# RestaurantDoctorAI > RestaurantDoctorAI helps independent restaurant owners see where their money goes and keep more of it. Its free profit check asks 12 multiple-choice questions about how the restaurant runs, scores the answers against RestaurantDoctorAI's targets for its type of restaurant, and shows a grade, the estimated money lost to avoidable costs a year and the costliest issue or biggest opportunity. A full report can be bought once on the website, with no subscription. It also sells an AI phone line that answers a restaurant's calls and takes pickup orders, table bookings as an add-on to that line, and a review reader; the phone line, table bookings and Review Watch are monthly and can be cancelled anytime. - The free check takes about 45 seconds and needs no account, name or email. - It needs no POS, bank or accounting login: the owner answers from what they already know, so it gives a first read in estimated ranges, not an audit of their transactions. - It is built for independent restaurants of every type: quick service and fast casual (counter pizzerias, takeout), cafés and coffee shops, full service (diners, sit-down pizzerias, casual dining), fine dining, and bars, pubs and taprooms. - Results are estimates from 12 multiple-choice answers and RestaurantDoctorAI's targets, not financial, tax or legal advice. - Money ranges are in US dollars, or in local currency for countries that use the euro, the pound sterling, the Canadian, Australian or New Zealand dollar, or the Swiss franc. Wording such as delivery apps and review sites follows the restaurant's country. - The full report adds estimates for each problem and opportunity, a 72-hour action plan, a 30-day roadmap, team cards and worksheets. It is only sold on restaurantdoctorai.com, and payment goes through Stripe. - RestaurantDoctorAI is provided by Gomorra Labs LLC. ## Free profit check - [Free profit check](https://restaurantdoctorai.com/): The 12-question check, its free result and the full report. - [How it works](https://restaurantdoctorai.com/how-it-works): The whole site in one page: the free 12-question profit check, the one-time report and plan, the free calculators and tools, and how they fit together. - [How we grade: our methodology](https://restaurantdoctorai.com/methodology): The 12 questions, RestaurantDoctorAI's targets for each type of restaurant, how the dollar estimates are worked out, what the AI does and the sources. - [Restaurant cost questions, answered with the math](https://restaurantdoctorai.com/faq): Owners' questions on prime cost, labor, food cost, break-even, delivery app prices, card and booking fees, the FICA tip credit and tip-outs, each answered by the free calculators for a stated example, plus the targets for each type of restaurant. - [Targets as data](https://restaurantdoctorai.com/targets.json): RestaurantDoctorAI's targets for each type of restaurant (prime cost, food and drink, labor, rent and utilities, as a % of net sales) in JSON, each marked as matching a named published figure or as RestaurantDoctorAI's own, with the National Restaurant Association's 2024 medians beside them and the sources. - [Sample result](https://restaurantdoctorai.com/sample-result): A full example of the free check's result and the paid report, for three made-up restaurants (a café, a quick-service pizza counter and a full-service restaurant), each scored by the same model, with one answer that can be changed to see the effect. - [Example report for How AI sees your website](https://restaurantdoctorai.com/online-check/example): A full example of the paid report, for a made-up pizzeria with made-up readings, scored by the same grading code as a real check: every check with its fix, the AI answers with their reasons and next steps, and the comparison with places nearby. - [Every check in How AI sees your website](https://restaurantdoctorai.com/online-check/checks): Each check on the Google listing, the website and the phone speed test, and how the score is worked out, plus the section on AI assistants, which is outside the score. - [The check-up](https://restaurantdoctorai.com/checkup): How the check-up in the premium report works: after the plan's first full month, the owner adds that month's numbers and sees then against now, with a next plan. - [Pricing](https://restaurantdoctorai.com/pricing): Everything that's free and everything on sale with its price: the reports, the AI phone line, table bookings, the review reader, what an AI agent can buy, and how payment and refunds work. - [Report codes for advisors](https://restaurantdoctorai.com/advisors): Prepaid premium report codes, in packs of 10 or 50, for consultants, accountants and associations who run the check for restaurants they advise. Codes never expire. - [Use RestaurantDoctorAI in your AI assistant](https://restaurantdoctorai.com/ai-assistants): How to add the free check, recipe costing, the calculators, the three printable sheets (cash-out, tip-out, prep and par), the website check, the deadline calendar and the library to Claude or another assistant that supports remote MCP servers, and what is sent. - [How our AI tools work](https://restaurantdoctorai.com/ai-assistants/how-it-works): What an AI assistant reads and receives when it uses RestaurantDoctorAI's MCP server: how it finds the server, the server's greeting, every tool's description and inputs word for word, one tool's answer from start to finish, what the tools say when a figure is missing, and what is kept. ## AI phone line, table bookings and reviews - [AI phone ordering](https://restaurantdoctorai.com/phone-ordering): An AI that answers the calls a restaurant's team can't get to, says it's an AI and that the call is recorded, takes pickup orders from the restaurant's menu, reads them back and sends them to a kitchen screen on any tablet. A person is one word away, and orders are paid at pickup. It doesn't connect to a POS yet: staff ring the orders in. Monthly, after a short free trial that starts with the AI's first call. No contract or setup fee; cancel anytime. The price is on https://restaurantdoctorai.com/pricing#phone-ordering. - [AI phone line setup](https://restaurantdoctorai.com/phone-ordering/setup): The six setup steps, to click through with an example restaurant: the menu, hours, staff number and address, the AI number, call forwarding, and the kitchen tablet with a test order. - [AI table bookings](https://restaurantdoctorai.com/table-bookings): An add-on to the AI phone line, not sold on its own: on the same calls and AI number, the AI books tables from the restaurant's own table book, by its hours, tables and rules, offers the nearest free times when it's full, sends big parties to staff, and texts a reminder to guests who said yes. The host stand gets a table book on any phone or tablet. Monthly on top of the phone line, with texts included each month and more to buy; the price is on https://restaurantdoctorai.com/pricing#table-bookings. An example table book is at https://restaurantdoctorai.com/bookings/demo. - [How AI sees your website](https://restaurantdoctorai.com/online-check): A check of what a basic AI reader gets from a restaurant's home page next to what a guest sees, with a grade of what guests see on Google and on its website and the first three things to change, free. The full report, with every check and what to fix, how it compares nearby and what to do next, is paid once, under a private link for 90 days. - [Review reader](https://restaurantdoctorai.com/reviews): Reads a restaurant's newest Google reviews, up to 100: the stars, a chart of what guests praise and complain about, and what to do about each complaint, from official rules and RestaurantDoctorAI's own tools. Paid once. Review Watch, a monthly plan, adds a full reading every month and a reply drafted for each new unhappy review. The prices are on https://restaurantdoctorai.com/pricing#review-reader. - [AI restaurant expert](https://restaurantdoctorai.com/ask): Answers questions about costs, staff, pricing and the rules that apply to a restaurant, with its sources. 5 questions free, then a pass paid once for 30 days of questions; it never renews. ## MCP server for AI assistants - [MCP endpoint](https://restaurantdoctorai.com/api/mcp): A public, read-only MCP server over Streamable HTTP. No sign-in. It takes only the 12 answers and a two-letter country code for the free check, a recipe and the owner's prices for a recipe draft, the owner's own figures for a calculator, a restaurant's website address for the website check, a US state and a few answers about the business for the deadline calendar, or a query, a kind or an entry's id for the library; it takes no payment and sends no email. The answers aren't stored; only an anonymous daily count of checks is kept. The recipe isn't stored; only an anonymous daily count of drafts is kept. The figures aren't stored; only an anonymous daily count of calculator runs is kept. The website isn't stored; only an anonymous daily count of checks is kept. The deadline calendar's answers aren't stored; only an anonymous daily count of calendars is kept. The query isn't stored; only an anonymous daily count of searches and reads is kept. A list of the library counts as a search. - [restaurantdoctor_get_questions](https://restaurantdoctorai.com/ai-assistants#tools): A tool that lists the 12 questions in quiz order, with the answer values each one accepts, worded for the restaurant's country (default US). It shows the exact money ranges and currency for a country. - [restaurantdoctor_run_free_check](https://restaurantdoctorai.com/ai-assistants#tools): A tool that scores the 12 answers and returns the free result the website shows: the grade, grades by area, the estimated money lost to avoidable costs, all of them together, a month and a year, the costliest issue or biggest opportunity with its estimated money a month, what one point of prime cost is worth, and the first fix, plus a link that opens the same result on restaurantdoctorai.com. - [restaurantdoctor_cost_recipe](https://restaurantdoctorai.com/ai-assistants#tools): A tool that costs a recipe's ingredient lines as a draft, in US dollars: each line matched to a USDA food and weighed, priced with the owner's prices or dated US public prices (BLS and USDA ERS store averages, USDA wholesale, and USDA weekly store-ad sale prices, which run low), what each line still needs, the cost per plate as a range, and a link that opens the draft on the plate cost calculator, where the owner checks each line. It never gives one plate cost or a menu price, and doesn't check allergens or nutrition. - [restaurantdoctor_check_website](https://restaurantdoctorai.com/ai-assistants#tools): A tool that checks a restaurant's own website the way How AI sees your website does: https and its own address, made for phones, a menu written as text, online orders on its own terms or only through delivery apps, the address, hours and phone written on the page, ways to get in touch, sign up and buy gift cards, what Google reads, the shared-link preview, the page's content, and Google's phone speed test. It returns the score and grade, the score for each area, the first three things to fix and each AI question's state, plus a link to the full check on the website (every check with what to fix, the reasons, the Google listing and places nearby). It reads only the public website, never the Google listing. - [restaurantdoctor_deadlines](https://restaurantdoctorai.com/ai-assistants#tools): A tool that lists a US restaurant's coming tax, payroll, license and permit deadlines from its state and yes/no answers, as the deadline calendar does: federal and state dates moved for weekends and holidays by each agency's rule, each with the rule, what a miss costs and the official source; a state date only where the state's own page confirms it, otherwise what to check with the agency; the rules for each hire; and a link to the same calendar on the website, where the owner gets free email reminders. Reminders, not legal or tax advice. - [restaurantdoctor_search_knowledge](https://restaurantdoctorai.com/ai-assistants#tools): A tool that searches RestaurantDoctorAI's sourced library, the same one its own AI restaurant expert answers from: the guides and blog posts listed below, RestaurantDoctorAI's step-by-step fixes, RestaurantDoctorAI's targets by type of restaurant, sourced industry facts, and an expert knowledge base of sourced how-tos, rules and checklists. It returns the best passages, each with its id, kind, title, link and source, and can be narrowed to one kind (guide, post, play, targets, fact or knowledge). It is keyword search: "break-even", "breakeven" and "break even" are one term, and plurals match. Figures are to be quoted with their source; RestaurantDoctorAI's targets are its own, not industry benchmarks. - [restaurantdoctor_list_knowledge](https://restaurantdoctorai.com/ai-assistants#tools): A tool that lists the library's entries, all or one kind, up to 50 at a time with an offset for the next page: each with its id, kind, title, link and source, and for guides and blog posts the day it was published and last updated. Only posts that are out are listed. - [restaurantdoctor_read_knowledge](https://restaurantdoctorai.com/ai-assistants#tools): A tool that reads one library entry in full, by the id restaurantdoctor_search_knowledge or restaurantdoctor_list_knowledge gave, with its title, link and source. A long entry comes 8,000 characters at a time, with its total length and the offset of the next part. - [restaurantdoctor_prime_cost](https://restaurantdoctorai.com/ai-assistants#tools): A tool that works out prime cost from the owner's figures for a week or a month, as the prime cost calculator does: food and drink plus labor as a share of net sales, against the target for the type of restaurant, with what each point is worth. - [restaurantdoctor_labor_cost](https://restaurantdoctorai.com/ai-assistants#tools): A tool that works out labor cost as the labor cost calculator does, in one of its three modes: labor as a share of net sales against the range for the type of restaurant, the hourly staff hours next week's sales can pay for, or the sales per labor hour below which the first person goes home. - [restaurantdoctor_break_even](https://restaurantdoctorai.com/ai-assistants#tools): A tool that works out the sales a month needs to pay every bill, also as a week, a day open and guests a day, as the break-even calculator does. - [restaurantdoctor_restaurant_budget](https://restaurantdoctorai.com/ai-assistants#tools): A tool that splits a month's expected sales by RestaurantDoctorAI's targets for the type of restaurant, as the restaurant budget calculator does: food and drink, labor, prime cost and rent with property charges and utilities as money ranges, other operating costs at the owner's share or 20% of sales (a working assumption, not a target or a published figure), what's left after every bill, and break-even sales with labor counted both ways. It needs only the expected sales and the type; rent and the owner's own pay are optional. - [restaurantdoctor_plate_cost](https://restaurantdoctorai.com/ai-assistants#tools): A tool that works out one plate's cost from the owner's own pack prices and amounts, or takes a plate cost they know, then the food cost at the menu price and the price for a target food cost, as the plate cost calculator does. For a recipe as written, without the owner's prices, restaurantdoctor_cost_recipe makes a draft instead. - [restaurantdoctor_delivery_price](https://restaurantdoctorai.com/ai-assistants#tools): A tool that works out a dish's break-even price on a delivery app from its in-house price and the app's real commission rate, as the delivery app price calculator does. - [restaurantdoctor_menu_calories](https://restaurantdoctorai.com/ai-assistants#tools): A tool that works out a dish's calories from its recipe, with USDA FoodData Central values or pack labels, rounded the way menus in the US, Ontario, or the UK and EU show them, as the menu calorie counter does. It's an estimate, not a lab test. - [restaurantdoctor_fica_tip_credit](https://restaurantdoctorai.com/ai-assistants#tools): A tool that estimates a US restaurant's FICA tip credit (IRS Form 8846, the 45B credit) for a year, as the FICA tip credit check does, from its state, a year's tips reported through payroll (the four Forms 941 line 5b, W-2 box 7 totals or the payroll tips report; never service charges) and, where the cash wage is under 5.15 dollars an hour, tipped staff's hours. It gives the reasons the real answer may be zero first, then the credit as a range, its worth after the deduction it replaces, the years that may still be open and the sources. It files nothing and isn't tax advice: an estimate to take to an accountant. - [restaurantdoctor_card_fees](https://restaurantdoctorai.com/ai-assistants#tools): A tool that checks a US restaurant's card fees for one month, as the card fees check does: the cost per 100 dollars of card sales (the effective rate) against a dated fair range for the type of restaurant, the money above it a month and a year, the processor's published rate where there is one, what's realistic for that processor, what to ask it and the fees to question on the statement. It doesn't read statements or contact a processor. - [restaurantdoctor_booking_fees](https://restaurantdoctorai.com/ai-assistants#tools): A tool that checks a US restaurant's booking app costs for one month, as the booking fee check does: what OpenTable, Resy, Tock, Yelp Guest Manager, Toast Tables, SpotOn, SevenRooms or another reservation app costs a month, a year and per guest booked through it, and as a share of the average spend per guest, from the plan, the owner's invoice and the guests booked. Every published price is dated and sourced, and the owner's invoice comes first. It needs only the plan, and on OpenTable the guests booked through its own app or website (on SevenRooms or another app, which publish no price here, the invoice's monthly fee); it never assumes a spend per guest or a number of guests. It shows cost, not which app to choose, doesn't read invoices or contact a platform, and RestaurantDoctorAI isn't affiliated with any of them. - [restaurantdoctor_cash_out](https://restaurantdoctorai.com/ai-assistants#tools): A tool that works out a restaurant's nightly cash close for one drawer and one day or shift, as the free cash-out sheet does: expected cash (starting bank + cash sales - cash tips paid out - paid outs), over or short against the cash counted, the deposit, the card difference between the processor's batch and the POS's card sales, and comps, voids and discounts as a share of net sales. It sets no normal for over or short, and its link goes to the free cash-out sheet. - [restaurantdoctor_tip_out](https://restaurantdoctorai.com/ai-assistants#tools): A tool that works out how tips are shared, as the free tip-out sheet does: a server's tip-out to support roles as a percentage of their net sales or tips, with what the server keeps, or a tip pool shared by hours and optional points, rounded to the cent. It returns the U.S. Department of Labor's published points on tips with their sources; it works out the split from the numbers sent and doesn't say whether a tip-out or a tip pool is allowed. Its link goes to the free tip-out sheet. - [restaurantdoctor_par_levels](https://restaurantdoctorai.com/ai-assistants#tools): A tool that works out par levels and what to order for up to 50 items, as the free prep and par sheet does: par = use per day × days covered + safety stock, and order = par minus what's on hand, rounded up to whole packs when a pack size is given. Days covered runs from the day of the count and the order to the day the next delivery after this one arrives. An item missing a figure is returned as incomplete, never counted as 0. Its link goes to the free prep and par sheet. - [restaurantdoctor_prime_cost_consult](https://restaurantdoctorai.com/ai-assistants#tools): A tool that runs the guided consult "Fix my prime cost" one step at a time from the answers so far, as the AI restaurant expert does on the website: set questions, a check on any figure that looks like a slip, the free prime cost calculator's result, a few more questions on the half that comes first, then one finding: the biggest problem at these figures, the step-by-step fix to start with, its sources and at most two other things worth a look. Fixed rules and the calculators produce every figure and sentence, with no AI in between. Nothing is kept between calls: the answers so far come in on every call, and a dish is only its price, plate cost and number sold, with no name. - [Calculator tools](https://restaurantdoctorai.com/ai-assistants#tools): The ten calculator tools take the owner's figures in any currency (the FICA tip credit, card fees and booking fee tools: a US restaurant's, in US dollars) and never count a missing figure as 0. Each returns the calculator's result, how it was worked out, and a link that opens the same calculator on restaurantdoctorai.com with the figures filled in. - [Web API](https://restaurantdoctorai.com/api/tools): The same free tools, except the website check, without MCP. GET /api/tools lists each tool with its input JSON Schema and an example body; POST a JSON body to /api/tools/ to run it, and the answer is JSON with the result and the same text the MCP tool returns. No key; 60 calls a minute from one address; what is sent isn't stored, only the same anonymous daily counts. Described at https://restaurantdoctorai.com/ai-assistants#web-api and, as OpenAPI 3.1, at https://restaurantdoctorai.com/openapi.json. - [A2A endpoint](https://restaurantdoctorai.com/api/a2a): For agents that speak the A2A protocol (version 1.0, JSON-RPC, send the header A2A-Version: 1.0). One skill: SendMessage searches the same library as restaurantdoctor_search_knowledge, with the message's text as the query, and answers with the passages as text and as data. It keeps no tasks, doesn't stream and sends no push notifications. No key; 60 calls a minute from one address; what is sent isn't stored, only an anonymous daily count of searches made this way. Its Agent Card is at https://restaurantdoctorai.com/.well-known/agent-card.json. - [API catalog](https://restaurantdoctorai.com/.well-known/api-catalog): An RFC 9727 linkset naming the web API, the MCP server, the A2A endpoint and the targets data, with where each is described. The MCP server's card is at https://restaurantdoctorai.com/.well-known/mcp/server-card.json. - [Pages as Markdown](https://restaurantdoctorai.com/faq): The main pages, restaurant type pages, guides, blog posts, calculator pages, their lists and the questions page answer in Markdown to a request with "Accept: text/markdown". - [Paid MCP endpoint for agents](https://restaurantdoctorai.com/api/mcp/agent): A separate MCP server for agents that buy for a restaurant owner. restaurantdoctor_buy_report prepares the full report from the 12 answers, the country, the restaurant's name and the owner's email; with tier "premium" it is the premium report instead (the owner is emailed the report page's link and adds last month's real numbers there within 24 hours, and it includes one check-up), and with fixKit true it adds the Restaurant Fix Kit (food cost, labor, delivery and waste workbooks for Excel, with printable PDFs). restaurantdoctor_buy_online_check finds the restaurant on Google Maps and prepares a restaurant's full check (How AI sees your website). restaurantdoctor_buy_report_codes prepares a pack of 10 prepaid premium report codes for an advisor: once paid, the codes are returned to the agent and emailed to the address it gives, and they never expire (the pack of 50 is sold only on https://restaurantdoctorai.com/advisors). All three return a payment link and charge nothing themselves. A link is paid by an agent with Stripe's machine payments (MPP: an HTTP 402 with the price, then the agent's payment credential), or by a person in a browser on the website's checkout. restaurantdoctor_use_code spends a prepaid report code, from a pack bought on the website or by an agent. Agents should ask the owner before paying. The prices are on https://restaurantdoctorai.com/pricing#ai-agents. ## Guides - [Why is my restaurant busy but not making money?](https://restaurantdoctorai.com/guides/restaurant-busy-but-not-making-money): Why a busy restaurant can still lose money, the six places the money usually goes, and the order to check your numbers in. - [How do I know if my restaurant is profitable?](https://restaurantdoctorai.com/guides/is-my-restaurant-profitable): A 15-minute test with last month's numbers: profit against cash, your own pay, prime cost and rent against targets, margin and break-even, with an example. - [What is prime cost in a restaurant? Formula and targets](https://restaurantdoctorai.com/guides/restaurant-prime-cost): What restaurant prime cost is, the formula, what counts in it, why rent doesn't, targets by restaurant type (bars included), and a weekly routine. - [Food cost percentage: formula and targets by restaurant type](https://restaurantdoctorai.com/guides/food-cost-percentage): How to calculate food cost percentage and plate cost, how to compare it with the targets for your format, and how to lower a high one. - [Restaurant labor cost percentage: what's good + formula](https://restaurantdoctorai.com/guides/restaurant-labor-cost): What labor cost percentage a restaurant should aim for by type, how to calculate it, what counts as labor, payroll percentage, and an SPLH cut line. - [Tipped minimum wage by state: 2026 rates and 2027 changes](https://restaurantdoctorai.com/guides/tipped-minimum-wage): Today's tipped cash wage and tip credit in every state, the 2027 changes from each state's own page, and a calculator for what a raise costs. - [What is a restaurant service charge, and should you add one?](https://restaurantdoctorai.com/guides/restaurant-service-charge): What a service charge is, the IRS's tests that tell it from a tip, what it changes for payroll tax and the FICA tip credit, what to show guests, and the trade-offs. - [Sales per labor hour (SPLH): formula and your target](https://restaurantdoctorai.com/guides/sales-per-labor-hour): What sales per labor hour means, how to work it out, and the cut line that tells a manager when to send someone home. - [Bar pour cost: how to calculate and fix it](https://restaurantdoctorai.com/guides/bar-pour-cost): How to calculate bar pour cost for a period or a single drink, and the fixes to try when it runs high. - [Delivery app commission: what it takes and how to cover it](https://restaurantdoctorai.com/guides/delivery-app-commissions): What a delivery app commission does to the profit on one order, how to price a delivery menu to cover it, and how to move regulars to direct ordering. - [How much rent can a restaurant afford?](https://restaurantdoctorai.com/guides/restaurant-rent-percentage): How to work out occupancy cost as a share of sales, the targets by restaurant type, and how much you need to sell to carry your rent. - [How many covers a night do I need to break even?](https://restaurantdoctorai.com/guides/restaurant-covers-to-break-even): The covers a night that pay every bill: the formula, a worked example from the break-even calculator's math, what your own pay adds, and a check against your seats. - [Restaurant formulas: food cost, labor, prime cost and more](https://restaurantdoctorai.com/guides/restaurant-formulas): Every formula a restaurant owner uses, as plain text you can copy, with what each tells you, what to aim for and the free calculator or guide for it. - [Do restaurants have to list calories on the menu?](https://restaurantdoctorai.com/guides/restaurant-menu-calories): Which restaurants have to put calories on the menu in the US, the UK, Canada and the EU, what the menu must show, and a free counter that works out a dish's calories. - [How to donate your restaurant's surplus food safely](https://restaurantdoctorai.com/guides/restaurant-food-donation): How to give surplus food to people who need it: cut over-prep first, follow the safety rules, and find a food rescue group that picks up near you. - [What does a restaurant have to post and keep?](https://restaurantdoctorai.com/guides/restaurant-required-posters-and-records): The federal labor posters, what a restaurant displays for guests, and the pay, hiring, tip and food safety records it keeps, each from its free official source. - [Can I check restaurant numbers without POS or bank access?](https://restaurantdoctorai.com/guides/check-restaurant-numbers-without-pos-or-bank): How to get a first read of your restaurant's costs with no POS or bank login, what such a check can't see, when connecting a tool is worth it, and five safety questions to ask first. ## Restaurant costs by type - [Restaurant cost targets by type of restaurant](https://restaurantdoctorai.com/restaurant-types): RestaurantDoctorAI's prime cost, food and drink, labor and rent targets for quick service, fast casual, full service, fine dining, bars and cafés, the National Restaurant Association's 2024 medians beside them, how to tell which type a restaurant is, and the closest type for pizzerias and food trucks. - [Full-service restaurant costs: targets and 2024 medians](https://restaurantdoctorai.com/restaurant-types/full-service): Targets for full service and casual dining beside the NRA's 2024 medians (labor split by profitable and loss-making restaurants, pretax income by sales volume), an example month, where the money goes, and tipped wages by state. - [Quick-service restaurant costs: targets and 2024 medians](https://restaurantdoctorai.com/restaurant-types/quick-service): Targets for quick service and, in its own section, fast casual, beside the NRA's 2024 limited-service medians, an example month, where the money goes, and who California's fast food wage covers. - [Bar, pub and taproom costs: targets and where money goes](https://restaurantdoctorai.com/restaurant-types/bars): RestaurantDoctorAI's own targets for bars, pubs and taprooms (no independent source publishes any), an example month, where money goes behind the bar (pours, comps and shrink), bartender pay and taprooms versus brewpubs. - [Café and coffee shop costs: targets and where money goes](https://restaurantdoctorai.com/restaurant-types/cafes): RestaurantDoctorAI's own targets for cafés and coffee shops (no café-only survey is public), the NRA's 2024 limited-service medians and a drive-thru coffee chain's results beside them, an example month, and four places to check: cups and milk, card fees on small sales, pastry waste and idle hours. ## Blog - [Can guests order from you in ChatGPT, Claude or Google Maps?](https://restaurantdoctorai.com/blog/order-through-chatgpt-claude-google-maps): Which restaurant systems take orders through ChatGPT, Claude and Google's Ask Maps today, the steps on Square and Toast, and what to do if you're on neither. Checked November 1, 2026. - [Can AI take your restaurant's phone orders? What goes wrong](https://restaurantdoctorai.com/blog/ai-phone-ordering-for-restaurants): How to count your own missed calls, what went wrong with AI ordering at three chains, the safeguards a phone AI needs, what ours won't do, and when it isn't worth it. - [Restaurant review analysis: what your reviews tell you](https://restaurantdoctorai.com/blog/restaurant-review-analysis): How to read your restaurant's reviews by topic, month by month and against places nearby, with a made-up worked example, the research, and the rules for asking guests. - [How to respond to bad restaurant reviews](https://restaurantdoctorai.com/blog/how-to-respond-to-bad-restaurant-reviews): How to reply to a bad restaurant review: a short sample reply for each hard case, what never to write, when to flag a review to Google, and how to use AI drafts safely. - [Who pays for that “best POS” list? Referral fees, explained](https://restaurantdoctorai.com/blog/restaurant-software-referral-fees): How restaurant software referral programs and partner networks pay for recommendations, what the FTC says about disclosing paid links, and how to tell whether a software list was paid for. - [The restaurant audit is moving into your AI assistant](https://restaurantdoctorai.com/blog/ai-agents-restaurant-audits): How AI assistants that can call tools are changing the restaurant audit, why fixed formulas beat a chat model's guesses, and how to connect RestaurantDoctorAI's free tools. - [How to price catering: one 40-person order, costed](https://restaurantdoctorai.com/blog/how-to-price-catering): How to price catering per person from everything an order uses: plate costs, packaging, staff hours at their true cost, delivery and rentals, worked out for one 40-guest order. - [Menu engineering: one 10-dish menu, worked through](https://restaurantdoctorai.com/blog/menu-engineering-worked-example): Menu engineering on one made-up ten-dish menu: the two cut-off lines as the method sets them, each dish sorted into star, plowhorse, puzzle or dog, and what to test with each. - [How much do restaurant owners make? Find your own number](https://restaurantdoctorai.com/blog/how-much-do-restaurant-owners-make): Why salary-site figures aren't what a restaurant owner keeps, how to work out your own pay from last month's numbers, an example at three sales levels, and draw vs salary per the IRS. - [Actual vs theoretical food cost: find your gap in a week](https://restaurantdoctorai.com/blog/actual-vs-theoretical-food-cost): How to compare what your dishes should have used with what left your stock, using a one-week count of a few top dishes, and the causes to check when the two numbers disagree. - [What one hour of restaurant labor really costs](https://restaurantdoctorai.com/blog/true-cost-of-an-hour-of-restaurant-labor): Price one hour of restaurant labor and an overtime hour: the wage plus employer taxes, workers' comp and benefits, from federal sources, with a method to repeat on your own rates and a note on tips. - [Restaurant lease extension: the math to do before you sign](https://restaurantdoctorai.com/blog/restaurant-lease-extension): Before signing a lease extension, check occupancy cost as a share of sales for each year of the term, see how far break-even moves, and read the clauses that raise the cost later. - [How to cut labor cost on slow shifts, not your best people](https://restaurantdoctorai.com/blog/cut-labor-cost-slow-shifts): Find the slow hours in your POS, set a cut line from your own wages, then trim hours with later starts, earlier finishes and cross-training, after checking your state and city scheduling rules. - [Café food cost: does the 30% rule work for resold pastries?](https://restaurantdoctorai.com/blog/cafe-food-cost-resold-items): Why one food cost percentage misleads a café that resells pastries and bottled drinks, and how to set a target per category and price resold items from their unit cost. - [The hidden costs of delivery apps, beyond the commission](https://restaurantdoctorai.com/blog/hidden-costs-of-delivery-apps): What DoorDash, Uber Eats and Grubhub cost a restaurant beyond the commission: guests you can't reach, charges on the payout, paid placement, app prices, changing terms, and what to do. - [Restaurant consultant cost, free help and when to hire](https://restaurantdoctorai.com/blog/restaurant-consultant-vs-profit-report): What restaurant consultants do and charge, the free mentors at SCORE and Small Business Development Centers, and when to start with an automated report instead. - [Free restaurant calculators compared: 6 brands checked](https://restaurantdoctorai.com/blog/free-restaurant-calculators-compared): Six restaurant software companies offer free calculators. We checked which ask for your details first, which show the formula and which say where their targets come from. - [Restaurant POS reports: what they show and what they miss](https://restaurantdoctorai.com/blog/pos-reports-vs-profit-check): What Toast, Square, Clover and Lightspeed say their reports show, the five to pull every week, what a POS can't know on its own, and where a free profit check fits beside it. - [Restaurant software cost: what subscriptions add up to](https://restaurantdoctorai.com/blog/restaurant-software-cost-per-year): Scheduling, invoice and POS subscriptions each look small by the month. Here are their listed starting prices, the yearly total, what they do well, and how to audit yours. - [ChatGPT for restaurants: how to check its numbers](https://restaurantdoctorai.com/blog/chatgpt-restaurant-numbers): AI chatbots help owners draft and brainstorm, but their makers say they can be wrong while sounding sure. Where they slip on food cost and pay rules, and how to check an answer. - [Do 9 in 10 restaurants fail? What the data says](https://restaurantdoctorai.com/blog/restaurant-failure-rate): Federal counts show most new food and lodging businesses are open after one year and about half after five. Closings are not always failures. Then: what to check this week, break-even and prime cost. - [What food waste costs a restaurant, and what to do](https://restaurantdoctorai.com/blog/restaurant-food-waste-cost): Waste is food you already paid for. What ReFED and WRAP say restaurants waste, where it happens, how to cost yours in a week with a log, and the order to fix it: measure, prep less, then donate. - [Restaurant break-even point: how to work out yours](https://restaurantdoctorai.com/blog/restaurant-break-even-point): Break-even sales are the bills that stay the same divided by what each sale leaves after the costs that rise with it. Find yours by month, week and day, and see which cost moves it most. - [Credit card fees for restaurants: what you pay and why](https://restaurantdoctorai.com/blog/restaurant-credit-card-processing-fees): The three layers of a restaurant's card fees, the Federal Reserve's debit cap, where surcharges are limited, the status of the Visa and Mastercard settlement, and what to ask your processor. - [Average restaurant profit margin: what to expect](https://restaurantdoctorai.com/blog/average-restaurant-profit-margin): What the National Restaurant Association's survey says restaurants keep before tax, why other published margins differ, and how a point of food or labor cost moves what you keep. - [No tax on tips: what it means for restaurant owners](https://restaurantdoctorai.com/blog/no-tax-on-tips-restaurants): The tips deduction is a federal income tax break for workers, so an owner's payroll tax on tips stays the same. What the IRS pages say, the 2026 W-2 changes, and what to ask your payroll provider. - [Coffee prices in 2026: what cafés should do about the cup](https://restaurantdoctorai.com/blog/coffee-prices-cafes-2026): What US government and coffee-market data say about green and shelf coffee prices, and how a café can cost a cup, see what a bag price rise does, and choose how to respond. - [Delivery app fees in 2026: what restaurants really pay](https://restaurantdoctorai.com/blog/delivery-app-fees-restaurants-2026): The plans DoorDash, Uber Eats and Grubhub publish, what city fee caps say, and the arithmetic for what an order leaves you and the app price that restores your margin. - [How much should you raise menu prices? Work it out per dish](https://restaurantdoctorai.com/blog/how-much-to-raise-menu-prices): Work out each dish's raise from its own plate cost and your target food cost, then use one formula to see how many guests you can lose before a price rise leaves you worse off. - [Rising beef prices: what to do with your beef dishes](https://restaurantdoctorai.com/blog/rising-beef-prices-restaurants): What USDA's latest data says about beef prices, and how to re-cost each beef dish and decide between a higher price, a smaller portion, a different cut or blend, or pushing other dishes. - [Minimum wage increase 2027: what it costs your restaurant](https://restaurantdoctorai.com/blog/minimum-wage-increase-2027-restaurants): The January 1, 2027 minimum wage increases confirmed on official pages, and how to budget them: the extra weekly cost, the new labor percentage, and when to reprice, trim hours or absorb it. - [Restaurant sales down? What to check before you cut](https://restaurantdoctorai.com/blog/restaurant-sales-down): Before you cut, find which number fell: guests or average check, on which days and channels, against the same weeks last year. Then check break-even and cut only what guests won't miss. ## Calculators - [Prime cost calculator](https://restaurantdoctorai.com/calculators/prime-cost): Food, drink and labor as a share of net sales for a week or a month, against the prime cost target for the restaurant's type, with what each point is worth. - [Restaurant labor cost calculator](https://restaurantdoctorai.com/calculators/labor-cost): Labor cost percentage (also called payroll percentage) for a week or a month, with payroll taxes and benefits in, against the range for the restaurant's type, plus the hours next week's sales can pay for and the sales per labor hour to cut staff at. - [Restaurant break-even calculator](https://restaurantdoctorai.com/calculators/break-even): The sales a restaurant needs each month, week and day open to pay every bill, and how many guests that takes. - [Restaurant budget calculator](https://restaurantdoctorai.com/calculators/restaurant-budget): A month's expected sales split into food and drink, labor, rent and other operating costs by RestaurantDoctorAI's targets for the type of restaurant, with what's left after every bill and break-even sales, and a spreadsheet (CSV) of the same lines to download. - [Food cost calculator: plate cost and menu price](https://restaurantdoctorai.com/calculators/recipe-cost): What one plate costs to make from its ingredients, its food cost percentage and the menu price that hits a target food cost. It can fill itself in from a pasted recipe: the owner checks each line, and every line is weighed with USDA data and priced. - [Delivery app price calculator](https://restaurantdoctorai.com/calculators/delivery-app-price): The break-even app price for a dish once the app takes its cut, and the markup on the in-house price it needs. - [Restaurant deadline calendar](https://restaurantdoctorai.com/deadline-calendar): Every federal and state tax, payroll, license and permit date for a US restaurant through 2027, from its state and a few yes/no answers, each with the rule, the weekend or holiday move and the official source. A state date is shown only where the state's own page confirms it. Free, with email reminders and a calendar link; no account. - [Menu calorie counter](https://restaurantdoctorai.com/guides/restaurant-menu-calories#calorie-counter): A dish's calories from its recipe, rounded the way menus in the US, the UK, the EU and Ontario show them, at the top of the guide on who has to put calories on a menu. - [FICA tip credit calculator](https://restaurantdoctorai.com/calculators/fica-tip-credit): An estimate, from a year's payroll totals, of the FICA tip credit a US restaurant may have generated (IRS Form 8846), with the reasons the real answer may be zero first, its worth after tax, the IRS web page's more cautious figure and the tax years that may still be open. An estimate to take to an accountant; it files nothing. - [Credit card fees calculator](https://restaurantdoctorai.com/calculators/credit-card-fees): A US restaurant's real cost per 100 dollars of card sales (the effective rate) for a month, against a dated fair range for its type, with what's realistic for its processor, the words to ask it and the fees to question on the statement. - [Reservation fees calculator: OpenTable, Resy and more](https://restaurantdoctorai.com/calculators/booking-fees): What a US restaurant's booking app (OpenTable, Resy, Tock, Yelp Guest Manager, Toast Tables, SpotOn, SevenRooms or another) really costs a month, a year and per guest booked through it, from its plan and, on OpenTable, the guests booked through OpenTable's own app or website, with each platform's published price dated and sourced, what the owner can do and what other apps publish, as information. Not affiliated with any of the platforms. ## Free templates - [Free restaurant templates](https://restaurantdoctorai.com/templates): Printable PDF and .xlsx templates for a restaurant's nightly cash-out, tip-out, prep list and par sheet, free to download with no email asked. - [Restaurant cash-out sheet](https://restaurantdoctorai.com/templates/restaurant-cash-out-sheet): An end-of-day sheet for one cash drawer: expected cash from the POS end-of-day (Z) report, over/short against the cash counted, the deposit, and optional card, comps, voids and discounts lines, with a calculator. - [Tip-out sheet and calculator](https://restaurantdoctorai.com/templates/tip-out-sheet): A server's tip-out to support roles by percentage of net sales or tips, or a tip pool split by hours and optional points, to the cent, with the U.S. Department of Labor's published points on tips and their sources. - [Prep list and par sheet](https://restaurantdoctorai.com/templates/prep-list-and-par-sheet): Each item's par (use per day × days covered + safety stock) and what to order or prep, rounded up to whole packs, with a calculator for up to 50 items. ## Help and policies - [About](https://restaurantdoctorai.com/about): Who builds RestaurantDoctorAI (Felix S. Pavel, Gomorra Labs LLC), why, and how its numbers are worked out and checked. - [Why I build with AI](https://restaurantdoctorai.com/ai): The founder on why RestaurantDoctorAI uses AI, the jobs it does well, the one it should never do, and how it already helps restaurant owners. - [Help & support](https://restaurantdoctorai.com/support): Missing reports, receipts, refunds and data requests. Email support@restaurantdoctorai.com. - [Privacy Policy](https://restaurantdoctorai.com/privacy): What is collected, including when an AI assistant runs the free check, costs a recipe, runs a calculator, checks a website, lists a restaurant's deadlines or searches, lists or reads the library, and how long it is kept. - [Terms of Service](https://restaurantdoctorai.com/terms) - [Refund and Return Policy](https://restaurantdoctorai.com/refunds) - [Service providers](https://restaurantdoctorai.com/subprocessors): The companies RestaurantDoctorAI uses to run the site and what each one handles. - [Paid links](https://restaurantdoctorai.com/partner-links): What a link marked "Paid link" is: a company we may be paid a fee by if a visitor signs up through it. A fee never changes a grade or what a report says. ## Optional - [Everything in one file](https://restaurantdoctorai.com/llms-full.txt): This list, then the FAQ, the calculators, every guide, every blog post that's out, the main pages and the restaurant type pages, in full, as plain text, for a tool that reads a whole site at once. The home, pricing and how it works pages are left out, because they're built with a live price. - [All guides](https://restaurantdoctorai.com/guides): The guides above, on one page. - [Blog](https://restaurantdoctorai.com/blog): Short, dated posts on what's changing in restaurant costs, prices, labor and fees, with their sources. RSS: https://restaurantdoctorai.com/blog/rss.xml - [All calculators](https://restaurantdoctorai.com/calculators): The calculators above, on one page. They're free, work in any currency (the FICA tip credit, card fees and booking fee calculators are for US restaurants), and the figures typed into them stay in the browser. # The pages in full The FAQ, the calculators, the guides, the blog posts that are out, the main pages and the restaurant type pages, each after a short block with its title and address. Left out: the home, pricing and how it works pages, which are built with a live price. The list above describes what they sell. --- title: "Restaurant cost questions, answered with the math" canonical: "https://restaurantdoctorai.com/faq" --- # Restaurant cost questions, answered with the math Each answer with a figure is the free calculator's own reply for the example in the question. ## My full-service restaurant took $80,000 last month and spent $26,000 on food and drink and $25,500 on labor. Is my prime cost OK? Prime cost is 64.4% of net sales. Food, drink and labor: 51,500 of 80,000 in net sales. That's inside the 65% target for full service, with 0.6 points to spare. Each point of prime cost (1% of your sales) is worth about 800 a month, or 9,600 a year, at these sales. [Prime cost calculator](https://restaurantdoctorai.com/calculators/prime-cost) ## A quick-service restaurant pays $3,900 in wages and $450 in payroll taxes on $15,000 of sales a week. Is its labor cost too high? Labor cost is 29% of net sales. Pay, taxes and benefits: 4,350 of 15,000 in net sales. That's inside the 26–31% range for quick service. Each point of labor cost (1% of your sales) is worth about 150 a week, or 7,800 a year, at these sales. Wages alone are 26% of sales; payroll taxes and benefits add 3 points. [Labor cost calculator](https://restaurantdoctorai.com/calculators/labor-cost) ## A dish costs $5.40 in ingredients and sells for $18 at a full-service restaurant. Is that a good food cost? One plate costs 5.40 in ingredients. Food cost: 30.0% of your 18.00 price. That's the share of the price that goes on the food. Full-service restaurants aim for 30–34% food and drink cost across the whole menu. This dish sits inside that range. [Food cost calculator](https://restaurantdoctorai.com/calculators/recipe-cost) ## How much does a restaurant need to sell to break even, with $60,000 a month in sales, $19,000 of food and drink, $14,000 of hourly pay, $3,000 of other costs that rise with sales and $16,000 of fixed bills? You break even at 40,000–47,369 a month. 40,000 if your hourly staff time goes up and down with sales, 47,369 if it stays about the same. Your real schedule sits somewhere in between. At 60,000 a month you're 12,632–20,000 above break-even (21.1–33.3% of your sales): your cushion before you start losing money. You had about 8,000 left over after every bill, before income tax. [Break-even calculator](https://restaurantdoctorai.com/calculators/break-even) ## How much can a café keep from $40,000 a month in sales, with $4,000 in rent and utilities? At 40,000 a month, RestaurantDoctorAI's targets for a café leave 4,000–7,200 a month (10–18% of sales) after every bill, before income tax. With the budget at the top of the targets, you break even at 30,000–34,595 a month: 30,000 if labor goes up and down with sales, 34,595 if it stays about the same. Your expected 40,000 is above it. [Restaurant budget calculator](https://restaurantdoctorai.com/calculators/restaurant-budget) ## What should a $15 dish cost on a delivery app that takes 30%? Price it at 21.43 or more on the app. That's 43% above your in-house price. At that price the app keeps 6.43 and you keep the same 15.00 as a dining room sale. [Delivery app price calculator](https://restaurantdoctorai.com/calculators/delivery-app-price) ## Are $1,600 in Square fees on $50,000 of card sales too much for a quick-service restaurant? Your card fees came to $1,600 on $50,000 of card sales: $3.20 per $100, an effective rate of 3.20%. A fair range for a quick-service restaurant is about $2.35 to $3.95 per $100, wide because the average card sale isn't known: it covers average card sales of $10 to $20. The number of card sales on the statement narrows it. That's inside the fair range. That's about the price of Square's published in-person rates (2.4% + 15¢ to 2.6% + 15¢, by plan), about $3.15 to $4.10 per $100 with average card sales of $10 to $20: the price of that plan, not a fee to question. Paying less means another plan or another processor. [Card fees calculator](https://restaurantdoctorai.com/calculators/credit-card-fees) ## What does OpenTable Core really cost with 400 guests a month from OpenTable and 300 from my own website? OpenTable Core costs you at least $699 a month, $8,388 a year. Counting the guests given, that's about $1.00 for every guest who booked through it. Each guest from OpenTable's own app and website cost $1.75 all in: the plan and the $1.00 fee per guest together. If 1 in 4 of the guests from OpenTable's own app and website booked on your website instead, about $100 a month would stay with you ($1,200 a year), as long as they'd have come anyway. [Booking fees calculator](https://restaurantdoctorai.com/calculators/booking-fees) ## Can a Texas restaurant, an S corporation, claim the FICA tip credit on $300,000 of reported tips over 12,000 tipped hours? Your business may have generated about $20,177 to $20,262 of FICA tip credit in a year like this. The check lists first why the real answer may be $0, such as a credit already claimed on Form 8846 or no federal income tax owed. [FICA tip credit calculator](https://restaurantdoctorai.com/calculators/fica-tip-credit) ## How much does a server tip out on $2,000 of sales and $360 in tips, at 2% to the busser and 1% to the bartender? The total tip-out is 60.00 (3% of net sales). The server keeps 300.00. [Tip-out sheet](https://restaurantdoctorai.com/templates/tip-out-sheet) ## What are good prime cost, food cost and labor percentages for each type of restaurant? As a share of net sales, RestaurantDoctorAI's targets are: Quick service: prime cost 60% or less, food and drink 26–32%, labor 26–31%. Fast casual: prime cost 60% or less, food and drink 28–33%, labor 27–32%. Full service / casual dining: prime cost 65% or less, food and drink 30–34%, labor 30–35%. Fine dining: prime cost 70% or less, food and drink 32–38%, labor 34–38%. Bar, taproom or lounge: prime cost 57% or less, food and drink 20–26%, labor 25–32%. Café or coffee shop: prime cost 60% or less, food and drink 24–32%, labor 28–34%. The prime cost lines for quick service and full service match RestaurantOwner.com's; the rest are RestaurantDoctorAI's own. Many restaurants run over them. The same targets are published as data at restaurantdoctorai.com/targets.json. [The targets and where they come from](https://restaurantdoctorai.com/methodology#targets) ## What should a coffee shop's food cost and labor be? RestaurantDoctorAI's targets for a café or coffee shop are food and drink 24–32% of sales, labor 28–34% and prime cost 60% or less, with rent and utilities at 10% or less. No public survey covers cafés on their own, so these are our own, set between what coffee roasters and listed coffee chains report and the National Restaurant Association's counter-service medians. [Targets for cafés and coffee shops](https://restaurantdoctorai.com/restaurant-types/cafes) ## Where do these numbers come from? Fixed formulas work out every figure, against RestaurantDoctorAI's targets for each type of restaurant. The free check and the calculators use no AI; in the paid report, AI writes the words around the figures, never the figures. The methodology page lists each target, says which match a published figure, and names the sources, checked against the National Restaurant Association's latest medians. The targets are also published as data at restaurantdoctorai.com/targets.json. [How we grade](https://restaurantdoctorai.com/methodology) ## What does RestaurantDoctorAI cost? The profit check, the calculators, the templates and the free AI tools cost nothing and need no account. The full report is $9.99, or $19.99 with your real numbers and a check-up, paid once, never a subscription, with a 14-day refund for any reason. [Every price, on one page](https://restaurantdoctorai.com/pricing) ## Can I use these calculators from ChatGPT or Claude? Yes. The same calculators are free tools any AI assistant that supports MCP can use, with no account or key: add restaurantdoctorai.com/api/mcp as a connector. The assistant sends the figures you give it and gets back the answer with its working and a link that opens the calculator with them filled in. [Our free AI tools](https://restaurantdoctorai.com/ai-assistants) --- title: "Free restaurant calculators" description: "Free restaurant calculators: prime cost, labor, break-even, plate cost, delivery prices, card and booking fees, FICA tip credit, calories and deadlines." canonical: "https://restaurantdoctorai.com/calculators" --- # Free restaurant calculators - [Prime cost calculator](https://restaurantdoctorai.com/calculators/prime-cost): Food, drink and labor as a share of your sales, against the target for your type of restaurant, so you know whether your prime cost needs fixing. By API: https://restaurantdoctorai.com/api/tools/restaurantdoctor_prime_cost - [Restaurant labor cost calculator](https://restaurantdoctorai.com/calculators/labor-cost): Your labor cost %, the hours next week's sales can pay for, and the sales per labor hour to cut staff at. By API: https://restaurantdoctorai.com/api/tools/restaurantdoctor_labor_cost - [Restaurant break-even calculator](https://restaurantdoctorai.com/calculators/break-even): How much you need to sell each month, week and day, and how many guests that is, so you know what a day has to bring in just to cover the bills. By API: https://restaurantdoctorai.com/api/tools/restaurantdoctor_break_even - [Restaurant budget calculator](https://restaurantdoctorai.com/calculators/restaurant-budget): Your expected sales for a month, split into food and drink, labor, rent and other costs by RestaurantDoctorAI's targets for your type of restaurant, with what's left and your break-even, so you can plan the month before you spend it. By API: https://restaurantdoctorai.com/api/tools/restaurantdoctor_restaurant_budget - [Food cost calculator: plate cost and menu price](https://restaurantdoctorai.com/calculators/recipe-cost): What one plate costs you to make, its food cost, and the price that hits your target. By API: https://restaurantdoctorai.com/api/tools/restaurantdoctor_plate_cost - [Delivery app price calculator](https://restaurantdoctorai.com/calculators/delivery-app-price): The break-even app price for a dish once the app takes its cut, so you can set app prices that leave you what the same dish earns in-house. By API: https://restaurantdoctorai.com/api/tools/restaurantdoctor_delivery_price - [Restaurant deadline calendar](https://restaurantdoctorai.com/deadline-calendar): Every tax, payroll, license and permit date for your restaurant through 2027, worked out for your state, so you can file and renew before the deadline, not after a late notice. By API: https://restaurantdoctorai.com/api/tools/restaurantdoctor_deadlines - [FICA tip credit calculator](https://restaurantdoctorai.com/calculators/fica-tip-credit): Enter a few numbers from your payroll so you can see whether your FICA tip credit is worth a call to your accountant. By API: https://restaurantdoctorai.com/api/tools/restaurantdoctor_fica_tip_credit - [Credit card fees calculator](https://restaurantdoctorai.com/calculators/credit-card-fees): Your real cost per $100 of card sales against a fair range for your type of restaurant, so you can see whether your fees are worth questioning and what to ask. By API: https://restaurantdoctorai.com/api/tools/restaurantdoctor_card_fees - [Reservation fees calculator: OpenTable, Resy and more](https://restaurantdoctorai.com/calculators/booking-fees): For restaurant owners: what your booking app really costs a month and a year, from OpenTable's fee per guest to a flat monthly price, so you can see what each guest it books really costs you. By API: https://restaurantdoctorai.com/api/tools/restaurantdoctor_booking_fees Every tool as a web API: https://restaurantdoctorai.com/api/tools --- title: "Prime cost calculator" description: "Free prime cost calculator. Enter net sales, food and drink purchases and labor for a week or month to see your prime cost against your format's target." canonical: "https://restaurantdoctorai.com/calculators/prime-cost" updated: "2026-10-10" --- # Prime cost calculator Food, drink and labor as a share of your sales, against the target for your type of restaurant, so you know whether your prime cost needs fixing. What to have to hand: Net sales, food and drink invoices and payroll, all for the same week or month. Use it on the page: https://restaurantdoctorai.com/calculators/prime-cost Or by API: POST JSON to https://restaurantdoctorai.com/api/tools/restaurantdoctor_prime_cost (no key). Its input schema and an example are at the same address. The same tool is on the MCP server at https://restaurantdoctorai.com/api/mcp as restaurantdoctor_prime_cost. The guide that explains the number: [What is prime cost in a restaurant? Formula and targets](https://restaurantdoctorai.com/guides/restaurant-prime-cost) The targets it uses, as data: https://restaurantdoctorai.com/targets.json --- title: "Restaurant labor cost calculator" description: "Free labor cost calculator for restaurants, no sign-up: labor % against your format's range, the hours next week's sales pay for, and when to cut staff." canonical: "https://restaurantdoctorai.com/calculators/labor-cost" updated: "2026-10-09" --- # Restaurant labor cost calculator Your labor cost %, the hours next week's sales can pay for, and the sales per labor hour to cut staff at. What to have to hand: Net sales, payroll and time clock hours, all for the same week or month. Use it on the page: https://restaurantdoctorai.com/calculators/labor-cost Or by API: POST JSON to https://restaurantdoctorai.com/api/tools/restaurantdoctor_labor_cost (no key). Its input schema and an example are at the same address. The same tool is on the MCP server at https://restaurantdoctorai.com/api/mcp as restaurantdoctor_labor_cost. The guide that explains the number: [Restaurant labor cost percentage: what's good + formula](https://restaurantdoctorai.com/guides/restaurant-labor-cost) The targets it uses, as data: https://restaurantdoctorai.com/targets.json --- title: "Restaurant break-even calculator" description: "Free restaurant break-even calculator. See the sales you need each month, week and day open to cover your costs, and how many guests that takes." canonical: "https://restaurantdoctorai.com/calculators/break-even" updated: "2026-10-14" --- # Restaurant break-even calculator How much you need to sell each month, week and day, and how many guests that is, so you know what a day has to bring in just to cover the bills. What to have to hand: A typical month's sales, the bills that stay the same, and the costs that rise with every sale. Use it on the page: https://restaurantdoctorai.com/calculators/break-even Or by API: POST JSON to https://restaurantdoctorai.com/api/tools/restaurantdoctor_break_even (no key). Its input schema and an example are at the same address. The same tool is on the MCP server at https://restaurantdoctorai.com/api/mcp as restaurantdoctor_break_even. The guide that explains the number: [Why is my restaurant busy but not making money?](https://restaurantdoctorai.com/guides/restaurant-busy-but-not-making-money) The targets it uses, as data: https://restaurantdoctorai.com/targets.json --- title: "Restaurant budget calculator" description: "Free restaurant budget calculator: split a month's sales into food, labor, rent and other costs by type of restaurant, with what's left and break-even." canonical: "https://restaurantdoctorai.com/calculators/restaurant-budget" updated: "2026-10-14" --- # Restaurant budget calculator Your expected sales for a month, split into food and drink, labor, rent and other costs by RestaurantDoctorAI's targets for your type of restaurant, with what's left and your break-even, so you can plan the month before you spend it. What to have to hand: Your expected sales for a month and your type of restaurant. Your rent and your own pay, if you want them in. Use it on the page: https://restaurantdoctorai.com/calculators/restaurant-budget Or by API: POST JSON to https://restaurantdoctorai.com/api/tools/restaurantdoctor_restaurant_budget (no key). Its input schema and an example are at the same address. The same tool is on the MCP server at https://restaurantdoctorai.com/api/mcp as restaurantdoctor_restaurant_budget. The guide that explains the number: [What is prime cost in a restaurant? Formula and targets](https://restaurantdoctorai.com/guides/restaurant-prime-cost) The targets it uses, as data: https://restaurantdoctorai.com/targets.json --- title: "Food cost calculator: plate cost and menu price" description: "Free food cost calculator for restaurants: add a recipe's ingredients to get its plate cost, its food cost % and the menu price that hits your target." canonical: "https://restaurantdoctorai.com/calculators/recipe-cost" updated: "2026-10-10" --- # Food cost calculator: plate cost and menu price What one plate costs you to make, its food cost, and the price that hits your target. What to have to hand: What you paid for each ingredient, and how much of it goes on one plate. Use it on the page: https://restaurantdoctorai.com/calculators/recipe-cost Or by API: POST JSON to https://restaurantdoctorai.com/api/tools/restaurantdoctor_plate_cost (no key). Its input schema and an example are at the same address. The same tool is on the MCP server at https://restaurantdoctorai.com/api/mcp as restaurantdoctor_plate_cost. The guide that explains the number: [Food cost percentage: formula and targets by restaurant type](https://restaurantdoctorai.com/guides/food-cost-percentage) The targets it uses, as data: https://restaurantdoctorai.com/targets.json --- title: "Delivery app price calculator" description: "Free delivery app markup calculator: enter a dish's in-house price and your real commission rate to see its break-even app price and the markup it needs." canonical: "https://restaurantdoctorai.com/calculators/delivery-app-price" updated: "2026-10-04" --- # Delivery app price calculator The break-even app price for a dish once the app takes its cut, so you can set app prices that leave you what the same dish earns in-house. What to have to hand: A dish's in-house menu price, and the real commission rate from your last payout statement. Use it on the page: https://restaurantdoctorai.com/calculators/delivery-app-price Or by API: POST JSON to https://restaurantdoctorai.com/api/tools/restaurantdoctor_delivery_price (no key). Its input schema and an example are at the same address. The same tool is on the MCP server at https://restaurantdoctorai.com/api/mcp as restaurantdoctor_delivery_price. The guide that explains the number: [Delivery app commission: what it takes and how to cover it](https://restaurantdoctorai.com/guides/delivery-app-commissions) The targets it uses, as data: https://restaurantdoctorai.com/targets.json --- title: "FICA tip credit calculator" description: "Free FICA tip credit calculator for restaurants: estimate the credit on your staff's reported tips, see why it may be $0, and the tax years still open." canonical: "https://restaurantdoctorai.com/calculators/fica-tip-credit" updated: "2026-10-11" --- # FICA tip credit calculator Enter a few numbers from your payroll so you can see whether your FICA tip credit is worth a call to your accountant. What to have to hand: A year's tips reported through payroll (the four quarterly Forms 941 your payroll files, or your payroll tips report), the hours tipped staff worked that year, and your state. Use it on the page: https://restaurantdoctorai.com/calculators/fica-tip-credit Or by API: POST JSON to https://restaurantdoctorai.com/api/tools/restaurantdoctor_fica_tip_credit (no key). Its input schema and an example are at the same address. The same tool is on the MCP server at https://restaurantdoctorai.com/api/mcp as restaurantdoctor_fica_tip_credit. The guide that explains the number: [Tipped minimum wage by state: 2026 rates and 2027 changes](https://restaurantdoctorai.com/guides/tipped-minimum-wage) The targets it uses, as data: https://restaurantdoctorai.com/targets.json --- title: "Credit card fees calculator" description: "Free credit card fees calculator for restaurants: your real cost per $100 of card sales against a fair range, the fees to question and what to ask." canonical: "https://restaurantdoctorai.com/calculators/credit-card-fees" updated: "2026-10-11" --- # Credit card fees calculator Your real cost per $100 of card sales against a fair range for your type of restaurant, so you can see whether your fees are worth questioning and what to ask. What to have to hand: One month's card statement: the month's card sales, the total fees taken that month, and the number of card sales if the statement shows it. Use it on the page: https://restaurantdoctorai.com/calculators/credit-card-fees Or by API: POST JSON to https://restaurantdoctorai.com/api/tools/restaurantdoctor_card_fees (no key). Its input schema and an example are at the same address. The same tool is on the MCP server at https://restaurantdoctorai.com/api/mcp as restaurantdoctor_card_fees. The targets it uses, as data: https://restaurantdoctorai.com/targets.json --- title: "Reservation fees calculator: OpenTable, Resy and more" description: "Free calculator for restaurant owners: what OpenTable, Resy, Tock, Yelp or Toast Tables really costs you a month, a year and per guest it books." canonical: "https://restaurantdoctorai.com/calculators/booking-fees" updated: "2026-10-11" --- # Reservation fees calculator: OpenTable, Resy and more For restaurant owners: what your booking app really costs a month and a year, from OpenTable's fee per guest to a flat monthly price, so you can see what each guest it books really costs you. What to have to hand: Your booking app's plan (its name is on the invoice) and, on OpenTable, how many guests booked through its own app or website last month; on SevenRooms or another app, the monthly fee on its invoice instead. Nothing else is required. Use it on the page: https://restaurantdoctorai.com/calculators/booking-fees Or by API: POST JSON to https://restaurantdoctorai.com/api/tools/restaurantdoctor_booking_fees (no key). Its input schema and an example are at the same address. The same tool is on the MCP server at https://restaurantdoctorai.com/api/mcp as restaurantdoctor_booking_fees. The targets it uses, as data: https://restaurantdoctorai.com/targets.json --- title: "Why is my restaurant busy but not making money?" description: "A full dining room can still lose money. See why busy restaurants lose profit, the six usual places the money goes, and the order to check your numbers in." canonical: "https://restaurantdoctorai.com/guides/restaurant-busy-but-not-making-money" published: "2026-09-28" updated: "2026-10-15" --- # Why is my restaurant busy but not making money? A busy restaurant loses money when it keeps too little of each sale. More covers bring in more sales, but each one leaves only a thin slice, and a month of thin slices may not cover the rent. The usual places the money goes are food and drink cost, labor that doesn't flex, delivery commissions and rent, and you can check each one with a number you already have. The short version - Busy means sales. Profit is what's left after costs, and most restaurant costs grow with every sale. - Start with prime cost: food and drink cost plus labor, as a share of sales. The ceilings RestaurantDoctorAI uses run from 57% for a bar to 70% for fine dining. - The usual problems in a busy restaurant: recipes not re-costed, waste nobody measures, free pours, shifts that don't flex, app orders at regular prices, and rent above your format's range. - Check the same numbers every Monday, in the same order, and fix the biggest gap first. ## Busy isn't the same as profitable Sales tell you how many guests came. Profit tells you what each of them left behind. When food, drink and labor take too much of each sale, a packed dining room adds only a sliver of profit per order, and the slivers don't add up to enough to pay for the building. Most restaurant costs move with sales: the food and drink you buy, the hours you schedule, the commission on app orders. Rent doesn't, but it gets paid out of whatever those costs leave. So the first question isn't how to get busier. It's how much of each sale you keep. [Free calculatorWork out yours: break-even calculatorThe sales you need each month to pay every bill, from what you keep of each sale.](https://restaurantdoctorai.com/calculators/break-even) ## A busy month in numbers Take a full-service restaurant with $80,000 of net sales in a month. Its food, drink and labor come to $55,200, a prime cost of 69%, against the 65% ceiling RestaurantDoctorAI uses for full service. Rent, property charges and utilities come to $10,400, or 13% of sales, against a 12% top of range. | Cost | This month | At the target | Gap a month | | --- | --- | --- | --- | | Prime cost (food, drink and labor) | $55,200 (69%) | $52,000 (65%) | $3,200 | | Rent, property charges and utilities | $10,400 (13%) | $9,600 (12%) | $800 | | Total | | | $4,000 a month, $48,000 a year | *An example month for a full-service restaurant, against RestaurantDoctorAI's targets. Round numbers, for illustration.* Every table can be full and this restaurant still ends the month about $4,000 short of where its targets would leave it. Each point of prime cost here is $800 a month, so the numbers below are where to look first. ## Start with prime cost Prime cost is your food and drink cost plus your total labor cost, as a share of sales. It's the biggest part of your costs you can control week to week, and the first number to check when a busy restaurant isn't making money. *Prime cost formula* Prime cost % = (food and drink used + total labor cost) ÷ net sales × 100 Use net sales, after discounts and comps, and the same week for costs and sales. | Restaurant type | Prime cost ceiling | | --- | --- | | Quick service | 60% or less | | Fast casual | 60% or less | | Full service / casual dining | 65% or less | | Fine dining | 70% or less | | Bar, taproom or lounge | 57% or less | | Café or coffee shop | 60% or less | *Prime cost ceilings RestaurantDoctorAI uses, as a share of net sales.* Over your ceiling, the problem is in food, drink or labor, and the next section shows where to look. Under it and still not making money, look at delivery and rent. For the full method, see [how restaurant prime cost works](https://restaurantdoctorai.com/guides/restaurant-prime-cost). ## Six places the money goes in a busy restaurant ### 1. Recipes that weren't re-costed Supplier prices move every week. Between updates, popular dishes quietly slip below their target margin, and a busy week sells more of them. Start with your best sellers: [work out their plate cost](https://restaurantdoctorai.com/guides/food-cost-percentage#plate-cost-for-a-single-dish) against today's invoices. ### 2. Waste nobody measures Trim, over-prep and plate waste are food you've already paid for. A busy kitchen preps more, so it can waste more, and none of it shows up until the month closes. A short daily waste log makes it visible while it's still fixable. ### 3. Free pours at the bar Over-pours, free drinks and spills go unrecorded on the highest-margin part of your sales. A busy bar pours more drinks, so it loses more that way. See [how to calculate and fix bar pour cost](https://restaurantdoctorai.com/guides/bar-pour-cost). ### 4. Shifts that don't flex A packed weekend can hide slow weekday hours. Paying a full crew through slow hours is a common labor cost, because every idle hour costs the same as a busy one. A [sales-per-labor-hour cut line](https://restaurantdoctorai.com/guides/restaurant-labor-cost#sales-per-labor-hour) tells your managers when to send people home. ### 5. App orders at regular prices The delivery apps take 15–30% of each delivered order, depending on the plan, and at regular menu prices that can wipe out your profit on the order. A busy app tablet can still mean a thin month. See [how much to mark up your app menu](https://restaurantdoctorai.com/guides/delivery-app-commissions#price-your-delivery-menu). ### 6. Rent that's too big a share of sales Rent and utilities don't grow with sales, so they squeeze profit when sales are low for the space. Above the top of your format's range, 10% of sales for quick service, fast casual and cafés or 12% for the rest, occupancy starts squeezing profit. Above 12%, it's a fixed cost no quick fix will change. See [how much rent a restaurant can afford](https://restaurantdoctorai.com/guides/restaurant-rent-percentage). ## Check your numbers in this order 1. Work out last week's prime cost and compare it with the ceiling for your type of restaurant. 2. If it's over, split it: food and drink cost against your range, then labor against yours (both are in the [targets by restaurant type](https://restaurantdoctorai.com/guides/restaurant-prime-cost#prime-cost-targets)). The half that's over is where to start. 3. Look at app orders on their own: what each app really took last month, from the payout statement. 4. Divide last month's rent, property charges and utilities by last month's sales. Above the top of your format's range is a warning sign: 10% for quick service, fast casual and cafés, 12% for the rest. 5. Pick the one biggest gap and fix it this week. Check the same numbers next Monday. Fix what each sale costs before you chase more sales: it raises what you keep on every order, busy night or quiet. Once prime cost is under your ceiling, each extra cover keeps a bigger share, and filling quiet days pays off more. No POS reports or bank login to hand? See how to [check your restaurant's numbers without POS or bank access](https://restaurantdoctorai.com/guides/check-restaurant-numbers-without-pos-or-bank). 45-second profit check Find where your restaurant is losing money Answer 12 quick questions about how your restaurant runs and get its grade and its biggest estimated loss, free, in about 45 seconds, so you know what to fix first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Common questions ### Why is my restaurant losing money? Usually because each sale costs too much to make, or the space costs too much for the sales it brings in. Check prime cost first, then delivery commissions and rent, in the order above. ### What is a healthy prime cost? At or under the ceiling for your type of restaurant: 57% of sales or less for a bar, 60% for quick service and fast casual, 65% for full service and 70% for fine dining. Under the ceiling is a good sign, not a guarantee of profit, since rent and every other cost still come out of what's left. ### Why is there no money left at the end of the month? Profit and cash aren't the same. A month can show a profit on paper and still leave the bank account empty, because some payments aren't costs on the profit and loss statement: the part of a loan payment that pays back the loan, the owner's own draws, extra stock bought ahead, and sales tax collected for the state. If your costs look in range, list last month's money out of the bank next to your costs and see what's on one list and not the other. For the full test, see [how to know if your restaurant is profitable](https://restaurantdoctorai.com/guides/is-my-restaurant-profitable). ### Should I get busier or cut costs first? Fix what each sale costs first, because that raises what you keep on every order, not just the extra ones. After that, quiet weekdays are the cheapest place to find extra sales, because rent, staff and utilities cost the same on a quiet Tuesday. --- title: "How do I know if my restaurant is profitable?" description: "A 15-minute test with last month's numbers: count your own pay, check prime cost and rent against targets, then work out your margin and break-even." canonical: "https://restaurantdoctorai.com/guides/is-my-restaurant-profitable" published: "2026-10-14" updated: "2026-10-14" --- # How do I know if my restaurant is profitable? Your restaurant is profitable when a month's sales cover every cost, a fair wage for your own hours included, and something is left over. You can test it in about 15 minutes with last month's numbers: count your own pay, compare prime cost and rent with the targets for your type of restaurant, then work out your margin and your break-even sales. The short version - Money in the bank isn't profit. Loan payments, draws, stock and taxes pull the two apart. - Count a fair wage for your own hours before you call anything profit. - Check prime cost and rent against the ceilings for your type of restaurant. - Then work out your margin and how far your sales sit above break-even. ## What you need One recent month, the same month for every figure: net sales (after discounts and comps, without sales tax), food and drink purchases, payroll with its taxes and benefits, rent and utilities, every other bill, your loan payments and the hours you worked yourself. Your P&L has most of these. Bank statements and your payroll report fill the gaps. ## The test, step by step ### 1. Separate profit from cash in the bank Profit is sales minus the month's costs. Cash is what's in the account. They drift apart because: - Only a loan's interest is a cost on the P&L. The part that pays the loan down leaves the bank uncounted. - Money you take out for yourself isn't a cost on the P&L either. - If your books count food used rather than food bought, a full walk-in is cash spent that isn't a cost yet. - Sales tax you collected sits in your account but belongs to the state. - A new oven leaves the bank at once but reaches the P&L a little each year. So a healthy balance can hide a loss, and a thin one can hide a profit. The steps below use the P&L view. Step 6 brings loan payments back in. ### 2. Pay yourself first in the math If you work in the restaurant and aren't on payroll, your P&L counts your hours as free. Put a number on them: what you would pay a manager to do your job, for the hours you work. Subtract it from the profit before you judge the result. If you're already on payroll at a fair wage, it's in labor, so don't count it twice. A restaurant that shows a profit only while you work for nothing is paying you less than the job is worth. ### 3. Check prime cost against your ceiling Prime cost is food and drink cost plus total labor, as a share of net sales. It's the biggest block of costs, and the one you can change week to week. *Prime cost formula* Prime cost % = (food and drink cost + total labor cost) ÷ net sales × 100 Use the same month for costs and sales. | Restaurant type | Prime cost ceiling | Occupancy ceiling | | --- | --- | --- | | Quick service | 60% or less | 10% | | Fast casual | 60% or less | 10% | | Full service / casual dining | 65% or less | 12% | | Fine dining | 70% or less | 12% | | Bar, taproom or lounge | 57% or less | 12% | | Café or coffee shop | 60% or less | 10% | *RestaurantDoctorAI's targets, the same ones its free profit check grades against, as a share of net sales. Occupancy is rent, property charges and utilities together.* Over your ceiling, the problem is in food, drink or labor. The [prime cost guide](https://restaurantdoctorai.com/guides/restaurant-prime-cost) shows how to find which half is over. [Free calculatorWork out yours: prime cost calculatorYour sales, food and drink and payroll, against the ceiling for your type.](https://restaurantdoctorai.com/calculators/prime-cost) ### 4. Check rent against its ceiling Add rent, the property charges your lease passes on, and utilities. Divide by net sales. Above the top of your format's range, 10% of sales for quick service, fast casual and cafés or 12% for the rest, occupancy starts squeezing profit. It changes least of all your costs, so know it before a lease renews. See [how much rent a restaurant can afford](https://restaurantdoctorai.com/guides/restaurant-rent-percentage). ### 5. Work out your margin Margin is what's left of each dollar after every cost, your fair wage from step 2 included. *Profit margin* Profit margin % = (net sales − every cost − your fair wage) ÷ net sales × 100 Before income tax, so it compares with the figures below. Above zero means the restaurant is profitable. For scale: in the National Restaurant Association's [survey of more than 900 restaurants](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/new-association-report-helps-operators-gauge-their-restaurant-performance), the median income before taxes in 2024 was 2.8% of sales for full-service restaurants and 4.0% for limited-service ones. The Association says these aren't standards or goals for any one restaurant, and its article doesn't say how each counted its owner's pay. Read them as where the middle restaurant landed, not as a pass mark. For why published margins differ so much, see [average restaurant profit margin](https://restaurantdoctorai.com/blog/average-restaurant-profit-margin). ### 6. Find your break-even sales Break-even is the monthly sales at which nothing is left. Split your costs into the ones that rise with every sale, like food and drink, and the bills that stay about the same, like rent, salaries, insurance, loan payments and your own pay. If a loan's interest is already in your bills, add only the part that pays the loan down. *Break-even sales* Break-even sales = monthly bills ÷ (1 − costs that rise with sales ÷ net sales) Write the share as a decimal: 32% is 0.32. The further your sales sit above break-even, the more room you have for a slow month. [Free calculatorWork out yours: break-even calculatorThe sales you need each month, week and day open, and how many guests that takes.](https://restaurantdoctorai.com/calculators/break-even) ## What each cost comes to per $100,000 of sales Steps 3 to 5 in dollars. Here is how RestaurantDoctorAI's targets split each $100,000 of sales for two types of restaurant, with the National Restaurant Association's medians beside them. | Per $100,000 of sales | Full service / casual dining: RestaurantDoctorAI's target | NRA median, full service | | --- | --- | --- | | Food and drink | 30–34% $30,000–$34,000 | 32.0% $32,000 | | Labor, taxes and benefits | 30–35% $30,000–$35,000 | 36.5% $36,500 | | Prime cost (the two together) | 65% or less $65,000 or less | Not published free | | Rent and utilities (occupancy) | Up to 12% Up to $12,000 | 5.7% $5,700 | | Left for everything else and profit | At least 23% $23,000 or more | Not published free | *Full service per $100,000 of net sales: RestaurantDoctorAI's targets, the ones its free check grades against, with what's left when prime cost and occupancy both sit at their ceilings, next to the National Restaurant Association's medians for 2024 ([food](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/), [labor](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/), [occupancy](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-occupancy-costs-were-more-than-5-of-sales-in-2024/), checked on October 10, 2026): the food figure leaves alcohol out, and occupancy counts rent, property taxes and insurance without utilities.* | Per $100,000 of sales | Quick service: RestaurantDoctorAI's target | NRA median, limited service | | --- | --- | --- | | Food and drink | 26–32% $26,000–$32,000 | 32.4% $32,400 | | Labor, taxes and benefits | 26–31% $26,000–$31,000 | 31.7% $31,700 | | Prime cost (the two together) | 60% or less $60,000 or less | 65% $65,000 | | Rent and utilities (occupancy) | Up to 10% Up to $10,000 | 5.2% $5,200 | | Left for everything else and profit | At least 30% $30,000 or more | Not published free | *Quick service per $100,000 of net sales, the same way. The National Restaurant Association's medians for 2024 ([food](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/), [labor](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/), [prime cost](https://restaurant.org/research-and-media/media/press-releases/new-resource-from-national-restaurant-association-provides-insights-into-operational-realities/), [occupancy](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-occupancy-costs-were-more-than-5-of-sales-in-2024/), checked on October 10, 2026): limited service counts quick service and fast casual together, the food figure leaves alcohol out, and occupancy counts rent, property taxes and insurance without utilities.* These are targets and typical figures, not results: your own numbers will differ, so work out yours from your own month before you compare. ## A worked example A full-service restaurant with $80,000 in net sales last month. The owner works the floor about 60 hours a week and isn't on payroll. All the figures are round examples. *Example: a full-service restaurant, one month* - Net sales: $80,000 - − Food and drink (32%): $25,600 - − Payroll, taxes and benefits (34%): $27,200 - − Rent, charges and utilities (9%): $7,200 - − Every other bill (18%): $14,400 - = Profit on the P&L (7%): $5,600 - − The owner's fair wage: $4,000 **Profit after the owner's wage (2%):** $1,600 $1,600 ÷ $80,000 × 100 = 2%. The $4,000 is an example: use what you'd pay someone to do your job. 1. Prime cost is 32% + 34% = 66%, one point over the 65% ceiling for full service. One point of $80,000 is $800 a month. 2. Occupancy is $7,200 ÷ $80,000 = 9%, under the 12% ceiling. 3. The margin is 7% on the P&L, but 2% once the owner's hours are paid for: under the Association's 2.8% median for full service. 4. The loan's interest is in the other bills. The $1,500 a month that pays the loan down leaves $100 of cash. 5. Break-even: food and drink rise with sales (32%). Everything else, the owner's wage and the $1,500 included, is $54,300 a month. $54,300 ÷ (1 − 0.32) = about $79,900, just under the $80,000 sold. So the restaurant is profitable, but only just. The first place to look is the point of prime cost over the ceiling: $800 a month, half of what the owner keeps after a wage. This example keeps it simple It counts payroll as a bill that stays the same. In a real month, hourly pay and card fees rise with sales, and the break-even calculator lets you count them that way. ## What your result means - **Profitable after your wage, well above break-even:** run the test monthly and watch prime cost weekly. - **Profitable only before your wage:** the job pays less than it's worth. Start with the biggest gap over a ceiling. - **Below break-even:** check prime cost first, then rent, in the order of [the checks for a busy restaurant that isn't making money](https://restaurantdoctorai.com/guides/restaurant-busy-but-not-making-money#check-in-this-order). ## Common questions ### How often should I run this test? Once a month, when the month's bills are in. Prime cost needs only invoices, payroll and sales, so you can check it every week. 45-second profit check See where your restaurant stands Answer 12 quick questions about how your restaurant runs and get its grade and its biggest estimated loss, free, in about 45 seconds, so you know which number to look at first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) --- title: "What is prime cost in a restaurant? Formula and targets" description: "Prime cost is food and drink cost plus total labor, as a share of net sales: the formula, a worked example, targets by type, and why rent isn't in it." canonical: "https://restaurantdoctorai.com/guides/restaurant-prime-cost" published: "2026-09-27" updated: "2026-10-14" --- # What is prime cost in a restaurant? Formula and targets Prime cost is what your food, drink and team cost you: cost of goods sold plus total labor, payroll taxes and benefits included, divided by net sales. It's the biggest cost you control week to week. Rent isn't part of it, and neither are delivery app commissions: each needs its own check. The quick answer Keep prime cost at or under the ceiling for your type of restaurant. Prime cost % = (cost of goods sold + total labor cost) ÷ net sales × 100 | Restaurant type | Prime cost ceiling | | --- | --- | | Quick service | 60% or less | | Fast casual | 60% or less | | Full service / casual dining | 65% or less | | Fine dining | 70% or less | | Bar, taproom or lounge | 57% or less | | Café or coffee shop | 60% or less | Targets RestaurantDoctorAI uses, the same ones its free profit check grades against. [Work out yours with the free calculator](https://restaurantdoctorai.com/calculators/prime-cost), no sign-up. ## What is prime cost in a restaurant? Prime cost adds your two biggest controllable costs, what you spend on food and drink and what you spend on your team, and shows them as a share of what you sold. It covers what it takes to run the kitchen and the floor, before rent, marketing, insurance or anything else. Cost of goods sold is the food and drink you used in the period: opening stock plus purchases, minus closing stock. If stock levels hold roughly steady week to week, purchases alone are a fair stand-in. Total labor cost is gross pay for everyone on payroll, hourly and salaried, plus the employer's share of payroll taxes and the benefits you pay. Which taxes and benefits apply depends on your state and payroll setup, and your payroll report lists yours. It's the full cost of employing your team, not the number on a pay stub. *Prime cost formula* Prime cost % = (cost of goods sold + total labor cost) ÷ net sales × 100 Use the same period for every figure, usually a week or a month, and use net sales after discounts and comps. ## Does prime cost include rent? No. Prime cost is food and drink plus labor. Rent belongs to occupancy, with property charges and utilities: a separate line that stays about the same whatever you sell. The occupancy ranges RestaurantDoctorAI uses are 6–10% of sales for quick service, fast casual and cafés, and 7–12% for full service, fine dining and bar or lounge formats. For comparison, the National Restaurant Association's [figures for 2024](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-occupancy-costs-were-more-than-5-of-sales-in-2024/) put median occupancy at 5.7% of sales for full-service restaurants and 5.2% for limited service. Its figure counts rent, property taxes and insurance, with utilities left out, so it runs lower than ours. Check the two side by side: prime cost under your ceiling and occupancy inside its range. A good prime cost can't make up for rent that's too high for your sales. See [how much rent a restaurant can afford](https://restaurantdoctorai.com/guides/restaurant-rent-percentage). Delivery app commissions sit outside prime cost too: if apps bring in a real share of your orders, [check what those commissions cost you](https://restaurantdoctorai.com/guides/delivery-app-commissions) on their own. ## What counts in prime cost, and what doesn't | What | Prime cost? | | --- | --- | | Food and drink you used, alcohol included | Count it | | Hourly and salaried pay, managers included | Count it | | The employer's payroll taxes and the benefits you pay | Count it | | Your own pay, if you're on payroll | Count it | | Rent, property charges, utilities (occupancy) | Leave out | | Card fees, delivery app commissions | Leave out | | Tips guests give your staff | Leave out | | Packaging: the targets are food and drink only | Leave out | *How RestaurantDoctorAI counts prime cost, in its calculator and its free profit check.* Some sources count only hourly kitchen and floor staff as labor and leave managers' salaries out. That makes prime cost look lower than counting everyone on payroll, so check how a figure was worked out before you compare yours with it. ## Prime cost targets by restaurant type Prime cost ceilings run higher for service-heavy dining rooms, where both food and labor take a bigger share of each dollar. They run lower for bar-driven concepts, where drinks carry a lower cost of goods per dollar of sales. The figures below are the targets RestaurantDoctorAI uses for each format. Treat the ceiling as a line to aim under, not a hard pass-fail test. | Restaurant type | Prime cost ceiling | Cost of goods | Labor | | --- | --- | --- | --- | | Quick service | 60% or less | 26–32% | 26–31% | | Fast casual | 60% or less | 28–33% | 27–32% | | Full service / casual dining | 65% or less | 30–34% | 30–35% | | Fine dining | 70% or less | 32–38% | 34–38% | | Bar, taproom or lounge | 57% or less | 20–26% | 25–32% | | Café or coffee shop | 60% or less | 24–32% | 28–34% | *Targets RestaurantDoctorAI uses, as a percentage of net sales. Cost of goods covers food and drink together.* Add the top of the cost of goods range to the top of the labor range. For every format, the total is above the prime cost ceiling. That is why each half can look fine while the combined number runs over. Where these come from: the 60% and 65% ceilings match the lines [RestaurantOwner.com](https://www.restaurantowner.com/public/Why-Prime-Cost-Is-the-Most-Important-Number-That-Should-Be-on-Your-PL.cfm), an independent site for operators, gives for quick service and table service. No independent survey publishes a ceiling for fine dining, bars or cafés, so those three are RestaurantDoctorAI's own. The café ceiling sits at the quick-service line, 60%, and the café food, drink and labor ranges are our own too: no survey covers cafés on their own. Many restaurants run above their ceiling. In the National Restaurant Association's [2025 Restaurant Operations Data Abstract](https://restaurant.org/research-and-media/media/press-releases/new-resource-from-national-restaurant-association-provides-insights-into-operational-realities/), the median limited-service restaurant, meaning counter service, ran a prime cost of 65% of sales. That's 5 points over the 60% ceiling for quick service, fast casual and cafés. All of one type's targets, with an example month: [full service](https://restaurantdoctorai.com/restaurant-types/full-service), [quick service](https://restaurantdoctorai.com/restaurant-types/quick-service) and [fast casual](https://restaurantdoctorai.com/restaurant-types/quick-service#fast-casual), [bars](https://restaurantdoctorai.com/restaurant-types/bars), [cafés](https://restaurantdoctorai.com/restaurant-types/cafes), or [every type side by side](https://restaurantdoctorai.com/restaurant-types). ## What each cost comes to per $100,000 of sales The same targets in dollars, with what the typical restaurant reported beside them, so you can see the size of each line next to the others. | Per $100,000 of sales | Full service / casual dining: RestaurantDoctorAI's target | NRA median, full service | | --- | --- | --- | | Food and drink | 30–34% $30,000–$34,000 | 32.0% $32,000 | | Labor, taxes and benefits | 30–35% $30,000–$35,000 | 36.5% $36,500 | | Prime cost (the two together) | 65% or less $65,000 or less | Not published free | | Rent and utilities (occupancy) | Up to 12% Up to $12,000 | 5.7% $5,700 | | Left for everything else and profit | At least 23% $23,000 or more | Not published free | *Full service per $100,000 of net sales: RestaurantDoctorAI's targets, the ones its free check grades against, with what's left when prime cost and occupancy both sit at their ceilings, next to the National Restaurant Association's medians for 2024 ([food](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/), [labor](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/), [occupancy](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-occupancy-costs-were-more-than-5-of-sales-in-2024/), checked on October 10, 2026): the food figure leaves alcohol out, and occupancy counts rent, property taxes and insurance without utilities.* | Per $100,000 of sales | Quick service: RestaurantDoctorAI's target | NRA median, limited service | | --- | --- | --- | | Food and drink | 26–32% $26,000–$32,000 | 32.4% $32,400 | | Labor, taxes and benefits | 26–31% $26,000–$31,000 | 31.7% $31,700 | | Prime cost (the two together) | 60% or less $60,000 or less | 65% $65,000 | | Rent and utilities (occupancy) | Up to 10% Up to $10,000 | 5.2% $5,200 | | Left for everything else and profit | At least 30% $30,000 or more | Not published free | *Quick service per $100,000 of net sales, the same way. The National Restaurant Association's medians for 2024 ([food](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/), [labor](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/), [prime cost](https://restaurant.org/research-and-media/media/press-releases/new-resource-from-national-restaurant-association-provides-insights-into-operational-realities/), [occupancy](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-occupancy-costs-were-more-than-5-of-sales-in-2024/), checked on October 10, 2026): limited service counts quick service and fast casual together, the food figure leaves alcohol out, and occupancy counts rent, property taxes and insurance without utilities.* These are targets and typical figures, not results: your own numbers will differ, so work out yours from your own month before you compare. The NRA's free figures don't give what was left for other bills and profit as one line. [Its release](https://restaurant.org/research-and-media/media/press-releases/new-resource-from-national-restaurant-association-provides-insights-into-operational-realities/) says the median full-service restaurant kept 2.8% of sales before taxes, $2,800 per $100,000, and the median limited-service one 4.0%, or $4,000. 45-second profit check See where your restaurant stands Answer 12 quick questions, including one on prime cost, and get a free preliminary grade plus the costliest issue we find, if there is one, so you know where to start. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## How to calculate prime cost: a worked example Take an example full-service restaurant with $90,000 in monthly sales. Food cost and labor cost each sit inside their range on their own. Watch what happens once you add them together. *Full service, $90,000 in monthly sales* - Cost of goods sold (32%): $28,800 - + Total labor cost (34%): $30,600 - = Prime cost: $59,400 - ÷ Net sales: $90,000 **Prime cost %:** 66% 32% cost of goods sold and 34% labor both sit inside the 30–34% and 30–35% ranges for full service. Added together, they land at 66%, above the 65% ceiling for the format. Here the total is one point over. At these sales, one point is $900 a month. Taking that out of either half brings prime cost back to the ceiling. Work out yours here: your sales, food and drink and payroll for a week or a month, against the target for your type of restaurant. ## Prime cost vs food cost vs labor cost | Number | What it divides by net sales | | --- | --- | | Food cost % | Food and drink you used | | Labor cost % | Pay, payroll taxes and benefits | | Prime cost % | Both of them together | *Each one is a share of the same net sales, for the same dates.* Prime cost tells you whether the two together are too high; the halves tell you which one to fix. See [food cost percentage](https://restaurantdoctorai.com/guides/food-cost-percentage) and [restaurant labor cost percentage](https://restaurantdoctorai.com/guides/restaurant-labor-cost) for each half on its own, or [every restaurant formula on one page](https://restaurantdoctorai.com/guides/restaurant-formulas). ## What is the 30/30/30 rule for restaurants? A rule of thumb for splitting sales, often written 30/30/30/10: about 30% to food and drink, 30% to labor, 30% to everything else (rent, utilities, insurance, repairs, card fees), and the last 10% is profit. Food and labor together make 60%, the prime cost ceiling RestaurantDoctorAI uses for quick service, fast casual and cafés. We couldn't find a published source for it, and real restaurants rarely split that evenly. In the National Restaurant Association's [figures for 2024](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/), labor alone was 36.5% of sales at the median full-service restaurant, and the median counter-service restaurant ran a prime cost of 65%. Use the rule as a quick check on the shape of your costs, then judge your own numbers against the ceiling for your type of restaurant in [the targets above](#prime-cost-targets). ## Why a weekly number beats the monthly P&L A monthly P&L tells you what already happened, often weeks after the month ends, once invoices are entered and payroll is closed. By the time it lands, the month that hurt your margin is long over. A weekly prime cost number uses invoices you already have and a time clock you can pull today. It tells you on Monday whether last week ran hot. You still have time to correct it before the pattern repeats for a whole month. Food cost and labor cost also move on different clocks. A supplier price rise shows up the day it hits an invoice. A scheduling problem shows up the day a shift runs long. Waiting for the P&L means you see both a month too late to act on either. ## The Monday routine This is the same weekly check the RestaurantDoctorAI action plan gives owners who don't track prime cost yet. It needs no new software and no new hires, and the plan allows about 30 minutes a week for it. 1. Add up last week's food and drink invoices, Monday to Sunday. If stock levels stayed about the same, purchases are a fair stand-in for what you used. 2. Add up last week's labor: gross wages from payroll or the time clock, plus the employer taxes and benefits on your payroll report. 3. Divide each total by last week's net sales from your POS. That gives you food cost and labor cost as a share of sales. Add the two and you have your prime cost. Compare it with the ceiling for your format above. 4. Write all three numbers, food, labor and prime cost, on a sheet in the office. Repeat every Monday. Four weeks in a row show whether food or labor is the one creeping up. Count invoices by delivery date Use the date the goods arrived, not the date you paid. Paying a big bill late makes one week look great and the next look terrible. Last week's net sales are seven nightly closes added up. If you close each night with the free [restaurant cash-out sheet](https://restaurantdoctorai.com/templates/restaurant-cash-out-sheet), each night's net sales from the POS end-of-day report are already on paper, so you can add up the week on Monday. ## Fixing a prime cost that's over target When prime cost sits above your ceiling, split it back into its two halves before you touch anything. Food cost and labor cost need different fixes. Your Monday sheet shows which one has been rising. - If food has been creeping up, [work out and fix your food cost percentage](https://restaurantdoctorai.com/guides/food-cost-percentage). - If labor has been creeping up, [bring your labor cost back toward target](https://restaurantdoctorai.com/guides/restaurant-labor-cost). - If drinks are a meaningful share of sales, look at [bar pour cost](https://restaurantdoctorai.com/guides/bar-pour-cost) too. Sometimes neither half is over its own range and only the total runs over, as in the worked example. Then start with the half sitting nearer the top of its range, or the one that has risen over your last few Mondays. Don't cut across the board Trimming every shift and every dish by the same amount treats a specific problem like a general one. Fix the half that has been creeping up in your weekly numbers, using the guide for it, and leave what's working alone. ## Common questions ### What is a good prime cost for a restaurant? Anything at or under the ceiling for your format in the table above. A result just under the ceiling is fine, but it leaves little room for a supplier price rise or a slow month. Being under is a good sign, not a guarantee of profit. Rent, delivery commissions and every other cost still come out of what's left. ### Does prime cost include rent? No. Rent is occupancy, its own line with property charges and utilities. [Check the two side by side](#does-prime-cost-include-rent). ### Is prime cost the same as cost of goods sold (COGS)? No. Cost of goods sold is the food and drink you used. Prime cost is that plus your total labor cost, which is why it runs about twice as high. ### Is labor included in cost of goods sold? No. Cost of goods sold is the food and drink you used. Labor is the other half of prime cost, so you add them together rather than counting labor twice. ### What should prime cost be for a bar? 57% of sales or less for a bar, taproom or lounge, the lowest ceiling of any type, and one of RestaurantDoctorAI's own. Drinks are usually about half of a bar's sales or more, and they cost less per dollar of sales than food, so cost of goods runs 20–26% and labor 25–32%. Watch the drink side on its own with [pour cost](https://restaurantdoctorai.com/guides/bar-pour-cost). ### Does prime cost include payroll taxes, benefits and managers' pay? Yes. Count everyone on payroll, hourly and salaried, including kitchen and floor managers. Then add the employer's share of payroll taxes and the benefits you pay. Leaving any of it out makes prime cost look better than it is. ### Do I include my own salary in prime cost? If you're on payroll, it's already in. If you work in the restaurant without a salary, add what you'd pay someone to do your hours, or prime cost looks lower than the business really runs. ### Should I use gross or net sales? Use net sales: what you rang up for food and drink after discounts and comps, not counting sales tax. Gross sales are a bigger number, so they make both halves look smaller than they are. Use the same sales figure every week so the weeks compare fairly. --- title: "Food cost percentage: formula and targets by restaurant type" description: "Food cost percentage is the food you used divided by food sales. Get the formula, plate cost, targets by restaurant type and how to fix a high one." canonical: "https://restaurantdoctorai.com/guides/food-cost-percentage" published: "2026-09-27" updated: "2026-10-14" --- # Food cost percentage: formula and targets by restaurant type Food cost percentage is the cost of the food you used in a period, divided by your food sales, times 100. A kitchen that used $21,000 of food to sell $70,000 runs a 30% food cost. Work it out for the whole kitchen each week or month. Then do the same for each dish as plate cost, to see which dishes need a fix. The short version - Food cost % = (beginning inventory + purchases − ending inventory) ÷ food sales × 100, for a week, a month or any period you choose. - Plate cost % does the same job for one dish: ingredient cost of the plate ÷ menu price × 100. - The targets RestaurantDoctorAI uses are for food and drink cost combined, from 26–32% of sales at quick service to 32–38% at fine dining. Add your drink cost in before you compare. - Two blind spots often sit behind a high number: recipes that are rarely re-costed, and waste that's estimated instead of measured. If you sell drinks, check pours too. - Fixing it starts with your 10 best sellers and your 10 biggest ingredients, not the whole menu at once. ## How to calculate food cost percentage Food cost percentage is the cost of the food you used, divided by your food sales. You measure it over a period, usually a week or a month. Count what actually left your storeroom, not what's sitting on the shelf that day. It's the food part of your cost of goods sold. Check it on a set schedule, not only when something feels off. *Food cost % for a period* Food cost % = (Beginning inventory + Purchases − Ending inventory) ÷ Food sales × 100 Beginning and ending inventory come from a physical count of your food stock. Purchases are your food invoices for the same period. Leave drinks out of all three. *One month, one restaurant* - Beginning inventory: $8,000 - + Purchases: $22,000 - − Ending inventory: $9,000 - = Food used: $21,000 - ÷ Food sales: $70,000 **Food cost %:** 30% This restaurant used $21,000 of food to ring up $70,000 in food sales: a 30% food cost for the month. Try it with your own counts. The calculator below opens on the same example and compares your result with RestaurantDoctorAI's target for your type of restaurant. ## Plate cost for a single dish Your food cost percentage for the whole kitchen can look fine while some dishes underneath it earn far less than others. Plate cost percentage isolates one dish. It shows which dishes need a fix, instead of averaging the problem away. *Plate cost % for one dish* Plate cost % = Ingredient cost of the plate ÷ Menu price × 100 Ingredient cost means every component on the plate: protein, sides, garnish, sauce and any packaging that ships with it. Price each one by the usable amount after trimming, not the weight you bought. *One burger, priced at $14* - Patty: $2.10 - Bun: $0.35 - Cheese: $0.30 - Toppings and sauce: $0.35 - Fries side: $1.10 - = Total ingredient cost: $4.20 **Plate cost % (÷ $14.00 menu price):** 30% A 30% plate cost on a $14 burger leaves $9.80 to cover labor, rent and profit before the next ingredient price rise. [Free calculatorWork out a dish: plate cost calculatorAdd its ingredients to get the plate cost, its food cost and the price that hits your target.](https://restaurantdoctorai.com/calculators/recipe-cost) ## Food cost percentage by restaurant type The targets RestaurantDoctorAI uses are for your combined cost of goods. That's the food and drink you used, divided by your food and drink sales. It's the same figure that goes into prime cost. If you sell drinks, a food-only number isn't comparable to these targets, so add the drink side in first. *Combined cost of goods, to compare with the targets* Cost of goods % = (Food used + Drinks used) ÷ (Food sales + Drink sales) × 100 Work out drinks used the same way as food: beginning bar stock plus bar purchases, minus ending bar stock. Say the restaurant in the first example also used $6,000 of drinks to sell $30,000 of drinks. Its combined cost of goods is $27,000 ÷ $100,000, or 27%. That's the number to hold against the table below, not the 30% food cost. | Restaurant type | Cost of goods (% of sales) | | --- | --- | | Quick service | 26–32% | | Fast casual | 28–33% | | Full service / casual dining | 30–34% | | Fine dining | 32–38% | | Bar, taproom or lounge | 20–26% | | Café or coffee shop | 24–32% | *Combined food and drink cost of goods as a share of food and drink sales, by restaurant type.* The bar and lounge target is the lowest because drinks usually cost less per dollar of sales than food. Treat all of these as typical ranges, not a pass-or-fail line. One odd week, such as a slow week or a short-lived price spike, can push you outside your range. Watch the trend over several weeks instead. A supplier price rise that sticks is different: re-cost the dishes it touches that week, as in step 3 below. For comparison, the National Restaurant Association's [figures for 2024](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/) put the median cost of food and non-alcoholic drinks at 32.0% of sales for full-service restaurants and 32.4% for limited service, meaning counter service. Those figures leave alcohol out. If drinks make up half your sales or more, what you spend on drinks is a large part of that combined number. Track the drink side on its own with pour cost. See [how to calculate bar pour cost](https://restaurantdoctorai.com/guides/bar-pour-cost) for that version of the math. ## The two quiet costs behind a high number When food cost keeps drifting up, two blind spots are often behind it. Both are easy to miss, because neither looks like an emergency at the time. - **Recipes you cost rarely, or never.** You might set menu prices by feel, by what competitors charge, or by a plate cost you last checked when the menu launched. Every supplier price increase since then has gone straight into your margin instead of into your price. - **Waste you estimate instead of measure.** Trim loss, over-prepped batches and plates sent back leave your kitchen the same way spoiled food does. But if nobody weighs or logs them, they never turn into a number you can act on. Both share the same fix: measure the thing you're currently guessing at. If you sell drinks and your combined number is high, check pours as well. An unmeasured pour raises cost of goods the same way unmeasured waste does. ## How to lower a high food cost percentage Start with the dishes and ingredients that move the most money for you, not the whole menu at once. 1. Pull last month's item sales report from your POS and list your 10 best-selling dishes. 2. For each one, weigh the ingredients the way the line actually plates them. Price them from your latest invoices, using the usable amount after trimming rather than the weight you bought. That's your real plate cost. 3. List the 10 ingredients you spend the most on. When one of them rises more than 5%, re-cost the dishes that use it that same week, not just when the menu changes. 4. Tape a waste sheet by the bins with four columns: item, amount, reason and initials. The chef signs it at close. 5. For each dish, multiply its profit per plate (menu price minus plate cost) by the number sold, and sort the menu by that total. Put the top earners first on your specials board and online menu. None of these five steps needs new software or a big spend. They turn a number you're guessing at into one you're measuring. That's what makes the next price or portion decision worth trusting. Over-prepped batches often start with a prep list made by feel. Our free [prep list and par sheet](https://restaurantdoctorai.com/templates/prep-list-and-par-sheet) sets a par for each item from what you use, so you can prep and order to a number. Don't punish the log Never punish your staff for what they write on the waste log. If logging gets people in trouble, the waste stops being written down. It doesn't stop being thrown out. Good food that's still left at close can feed people instead of the bin: see [how to donate your restaurant's surplus food safely](https://restaurantdoctorai.com/guides/restaurant-food-donation). 45-second profit check See where your restaurant stands Answer 12 quick questions, including how you cost recipes and track waste. You'll get a free preliminary grade and your biggest estimated loss in about 45 seconds, so you know where to start. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Common questions ### What's a good food cost percentage? It depends on your format. The targets RestaurantDoctorAI uses are for food and drink cost combined. They run from 26–32% of sales at quick service to 32–38% at fine dining. A bar or lounge runs lower, at 20–26%. Add your drink cost in before you compare. A number that stays above your range for several weeks is worth digging into. ### Is 30% food cost high? Not for most restaurants. 30% of sales sits inside the ranges RestaurantDoctorAI uses for quick service (26–32%), fast casual (28–33%) and full service (30–34%), food and drink combined, and under the National Restaurant Association's [2024 medians](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/) of 32.0% for full service and 32.4% for counter service, which leave alcohol out. It's high for a bar or lounge, where the range is 20–26%. It's also the food share in the [30/30/30 rule of thumb](https://restaurantdoctorai.com/guides/restaurant-prime-cost#30-30-30-rule). ### Is my food cost too high? Compare your combined food and drink cost with the range for your type of restaurant in the table above. If it sits above the top of that range for several weeks in a row, yes. The usual causes are [the two quiet costs](#two-quiet-costs): recipes that weren't re-costed after a supplier price rise, and waste nobody measures. One high week on its own is usually noise. ### Is food cost percentage the same as prime cost? No. Food cost percentage covers only the food you used. Prime cost is your food and drink cost plus total labor cost, divided by sales. It's the biggest share of your costs you can control week to week, and the first number to check when profit slips. See [our guide to restaurant prime cost](https://restaurantdoctorai.com/guides/restaurant-prime-cost) for the full formula. ### How often should I calculate it? Calculate the period number weekly or monthly from your inventory and purchases. Re-cost your best sellers any time a top ingredient rises more than 5%, rather than waiting for your next menu change. --- title: "Restaurant labor cost percentage: what's good + formula" description: "What percentage should labor cost be in a restaurant? Targets by type, the NRA's 2024 medians, the formula with payroll taxes, and SPLH as a cut line." canonical: "https://restaurantdoctorai.com/guides/restaurant-labor-cost" published: "2026-09-27" updated: "2026-10-14" --- # Restaurant labor cost percentage: what's good + formula A restaurant's labor cost percentage is what it costs to staff the restaurant, wages plus payroll taxes and benefits, as a share of sales. What's good depends on how you serve: the targets RestaurantDoctorAI uses run from 26–31% of sales at quick service to 34–38% at fine dining. When yours runs high, look at the schedule before you look at pay rates. The quick answer Keep labor inside the range for your type of restaurant, with payroll taxes and benefits counted in. Labor cost % = (wages + payroll taxes + benefits) ÷ net sales × 100 | Restaurant type | Labor cost target | | --- | --- | | Quick service | 26–31% | | Fast casual | 27–32% | | Full service / casual dining | 30–35% | | Fine dining | 34–38% | | Bar, taproom or lounge | 25–32% | | Café or coffee shop | 28–34% | Targets RestaurantDoctorAI uses, the same ones its free profit check grades against. Treat the top of yours as a ceiling, not a goal. [Work out yours with the free calculator](https://restaurantdoctorai.com/calculators/labor-cost), no sign-up. ## How to calculate labor cost percentage Labor cost percentage compares what it cost to staff the restaurant with what it sold in the same period. Work it out weekly if you can. A monthly number arrives after the shifts that caused it are over. *Labor cost formula* Labor cost % = (gross wages + payroll taxes + benefits) ÷ net sales × 100 Use net sales after discounts and comps, and the same dates for labor and sales. *Fast casual, one week* - Gross wages from payroll, tips left out: $5,040 - + Payroll taxes and benefits: $630 - = Total labor cost: $5,670 - ÷ Net sales for the week: $18,000 **Labor cost %:** 31.5% Wages alone are 28% of sales. With payroll taxes and benefits, labor is 31.5%: inside the 27–32% range for fast casual, close to its top. Work out yours here: your net sales and payroll for a week or a month, against the range for your type of restaurant. ## What counts as labor cost Everything it costs you to have people on the payroll. Leave something out and the number looks better than it is, which is the most common reason two owners' figures don't compare. | What | Labor cost? | | --- | --- | | Hourly wages, overtime included | Count it | | Salaries, managers and kitchen leads included | Count it | | The employer's payroll taxes | Count it | | Benefits you pay: health insurance, retirement, workers' comp | Count it | | Your own pay, if you're on payroll | Count it | | Your hours, if you work without a salary: what you'd pay someone | Count it | | Tips guests give your staff, even through payroll | Leave out | | Outside services: cleaning, bookkeeping | Leave out | *How RestaurantDoctorAI counts labor, in its calculators and its free profit check.* For a week, count a 52nd of each year's salary; for a month, a 12th. Tips come from guests, not from your sales, so they stay out of both. Tip rules vary by state and city, so check yours. ## Labor cost percentage by restaurant type How much of your sales should go to labor depends on how you serve guests. A counter needs fewer paid hours per dollar of sales than a dining room with servers, runners and bussers. The quick answer at the top of this page has the target for each type. For comparison, the National Restaurant Association's [figures for 2024](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/) put the median labor cost of restaurants that made a profit at 30.0% of sales for limited service, meaning counter service, and 34.2% for full service. Across all full-service restaurants, profitable or not, the median was 36.5%. For bars and fine dining you'll find different ranges in different places, and no independent survey splits them out. Treat any one range, ours included, as a guide, and watch your own trend week to week. All of one type's targets, with an example month: [full service](https://restaurantdoctorai.com/restaurant-types/full-service), [quick service](https://restaurantdoctorai.com/restaurant-types/quick-service) and [fast casual](https://restaurantdoctorai.com/restaurant-types/quick-service#fast-casual), [bars](https://restaurantdoctorai.com/restaurant-types/bars), [cafés](https://restaurantdoctorai.com/restaurant-types/cafes), or [every type side by side](https://restaurantdoctorai.com/restaurant-types). Labor is one half of [restaurant prime cost](https://restaurantdoctorai.com/guides/restaurant-prime-cost). Cost of goods sold is the other. Each can sit inside its own range while the two together reach or pass your prime cost ceiling, so check them side by side. If food is the half running high, see [how to calculate and fix food cost percentage](https://restaurantdoctorai.com/guides/food-cost-percentage). 45-second profit check See if labor is your costliest issue Answer 12 quick questions, including how staffing changes when a shift gets slow. You'll get a free preliminary grade and your costliest issue or opportunity in about 45 seconds, so you know whether labor is where to start. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Payroll percentage vs labor cost percentage Restaurant payroll percentage is usually the same number under another name: payroll, with the employer's taxes and benefits, as a share of sales. Before you compare yours with a figure you've read, check what it counts. Some leave out payroll taxes and benefits, some leave out managers' salaries, and either makes labor look lower than it is. Outside the US the same number goes by other names. In the UK it's usually labour cost, and employer's National Insurance and pension count in it; in Ireland, employer's PRSI. In Australia owners call it wage percentage, or wages as a percentage of sales, and super counts in it. In New Zealand, KiwiSaver and ACC levies do. ## Sales per labor hour (SPLH): your cut line SPLH means sales per labor hour. Labor cost percentage tells you how last week went; sales per labor hour tells you how this hour is going, while a manager can still act on it. The short version is below; [the sales per labor hour guide](https://restaurantdoctorai.com/guides/sales-per-labor-hour) has the rest. *Sales per labor hour* Sales per labor hour = sales ÷ labor hours worked Labor hours means hours actually worked on the time clock, not hours scheduled. That figure becomes useful next to a cut line: the sales per labor hour below which labor runs over your ceiling. You set the line from your own wages and your format's target, written as a decimal. | Restaurant type | Ceiling as a decimal | | --- | --- | | Quick service | 0.31 | | Fast casual | 0.32 | | Full service / casual dining | 0.35 | | Fine dining | 0.38 | | Bar, taproom or lounge | 0.32 | | Café or coffee shop | 0.34 | *The top of each labor cost target, written as a decimal for the cut line.* *Your cut line* Cut line = average hourly labor cost ÷ labor ceiling as a decimal Average hourly labor cost is wages plus payroll taxes, averaged across the staff on the clock. It leaves out benefits and salaried pay, so aim a little above the line. When sales per labor hour fall below the line, labor is running above your ceiling. That's the moment to make the first cut. *Full service, one slow hour* - Average hourly labor cost: $17.50 - ÷ Labor ceiling as a decimal: 0.35 - = Cut line: $50.00 - Sales, 8–9 p.m.: $300 - ÷ Labor hours worked, 8 staff: 8 **Sales per labor hour:** $37.50 $37.50 is below the $50.00 line. Eight staff at $17.50 cost $140, about 47% of $300 in sales. With six on the clock, it's $300 ÷ 6 = $50.00, and $105 of hourly labor is 35%. [Work out your own cut line](https://restaurantdoctorai.com/calculators/labor-cost#cut-line) from last week's payroll and sales with the free labor cost calculator. It also counts benefits and salaried pay, which the formula above leaves out. Who goes home first depends on the format. In fine dining, cut runners, bussers or a prep cook before servers or captains. On a slow bar night, cut the barback and a bartender well before last call. At quick service and fast casual, stagger start times so extra staff arrive with the rush. ## Why labor cost runs high Before you look at pay rates, look at hours. Each of these patterns adds paid hours, or pricier ones, without adding sales. - **Fixed shifts that don't flex with traffic.** Everyone works their scheduled shift, busy or not. An idle hour costs the same as a packed one. - **Overtime above 8% of hours.** Overtime hours usually cost more than regular ones. RestaurantDoctorAI treats that share as a scheduling signal. The usual causes are doubles, covered call-outs, and clocking in early and out late. - **Start times ahead of the rush.** Staff who arrive an hour before guests, or stay long after tickets slow, are paid for idle time. - **Schedules written from habit.** Without a cover forecast, it's easy to staff every Friday for your best-ever Friday. To find your overtime share, divide overtime hours by total hours worked in the same pay period. If your team worked 600 hours over two weeks and 60 were overtime, that's 10%, above the 8% line. A low number can hide a problem too. When the owner covers a station, labor can look fine on paper while nobody watches costs, the schedule or the guest experience. ## How to reduce labor cost in a restaurant None of these steps needs new software or new hires. Steps 1, 2, 3 and 5 happen when you write next week's schedule. 1. Forecast covers for each day next week: average the same weekday over the last four weeks, then adjust for reservations, local events, holidays, paydays and the weather. Staff each shift to that forecast. 2. Pull four weeks of sales and staff hours by hour from your POS and time clock. Where staff outnumber the work at the start of a shift, move one or two start times 30 minutes later. 3. Set your cut line with the formula above. On every schedule, write the cut order: who goes home first, second and third, and who closes. 4. During the shift, the manager checks sales per labor hour on the hour and makes the first cut as soon as it drops below the line. 5. Total each person's hours before posting the schedule. Keep full-timers below the point where overtime starts, with room for a late close: where that's 40 hours a week, schedule 38 or fewer. Offer extra shifts to part-timers first. 6. Allow no clocking in more than five minutes early without a manager's OK, and check the time report every morning. Train a second person on each station so a call-out doesn't become a double. 7. If you cover a station yourself, train your most reliable team member on it over two slow shifts. Hand it over with a written checklist, or the work drifts back to you. Then block two hours off the floor every week for prime cost, next week's schedule and orders. Trim idle hours, not the side work that keeps the place clean and stocked. Redo the hour-by-hour review from step 2 every month. The sign it's working: weekly labor cost moves toward your range without longer ticket times or more complaints. Check the rules before you cut a shift short Overtime, break and scheduling rules vary by state and city, and some require minimum pay when a shift is cut short. Check yours before you change the schedule or send anyone home early. And never fix overtime with off-the-clock work. Every hour worked gets paid. ## Common questions ### What percentage should labor cost be in a restaurant? It depends on your format. The targets RestaurantDoctorAI uses are 26–31% of sales for quick service, 27–32% for fast casual, 30–35% for full service, 34–38% for fine dining and 25–32% for a bar or lounge, with payroll taxes and benefits counted in. Check it alongside food cost, since the two add up to prime cost. ### Is 25% labor cost high for a restaurant? No, it's low. It sits at the bottom of the bar and lounge range (25–32%) and under every other range RestaurantDoctorAI uses, which start at 26% for quick service. It's also under the National Restaurant Association's [2024 medians](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/): 30.0% for profitable counter-service restaurants and 34.2% for profitable full-service ones. Before you count on it, check that it includes payroll taxes, benefits, managers' salaries and any hours you work yourself unpaid. ### Is 20% labor cost good? It's below every range RestaurantDoctorAI uses and well below the medians above. When labor reads 20%, something is often left out of the count, such as payroll taxes, benefits, salaried managers or your own hours, or the floor is short-staffed. If the number holds up with everything counted, watch wait times and reviews to make sure guests aren't paying for it. ### What is the 30/30/30 rule for restaurants? A rule of thumb: about 30% of sales to food and drink, 30% to labor, 30% to everything else, and 10% left as profit. It has no published source, and labor alone often runs above 30% at full service. [How the rule compares with real restaurants](https://restaurantdoctorai.com/guides/restaurant-prime-cost#30-30-30-rule). ### Is payroll percentage the same as labor cost percentage? Usually, yes: it's the same share of sales under another name. Check what a payroll figure counts before you compare it with yours, since some leave out payroll taxes, benefits or managers' salaries. ### What does SPLH mean? Sales per labor hour: your sales for a period divided by the hours your staff worked in it. Compare it with your cut line, the point below which labor runs over your ceiling. See [sales per labor hour](https://restaurantdoctorai.com/guides/sales-per-labor-hour) for the full guide. ### What is a good sales per labor hour? It depends on your wages and your format's ceiling. Divide your average hourly labor cost, wages plus payroll taxes, by your ceiling as a decimal. At $17.50 an hour and a 35% ceiling, the line is $50. Below it, labor for that hour is running above 35% of sales. Benefits and salaried pay come on top, so aim a little above the line. ### Does labor cost include my own pay? If you're on payroll, yes, your pay is already in it. If you work in the restaurant without a salary, add what you'd pay someone to do your hours, plus the payroll taxes on it. Otherwise a restaurant that depends on unpaid owner hours looks healthier than it is. ### How much overtime is too much in a restaurant? RestaurantDoctorAI uses 8% of total hours as the line. Above it, overtime usually signals gaps in the schedule rather than a real need for more staff. When overtime starts, and whether it counts by the day or the week, depends on where you operate. Check your state and city rules. --- title: "Sales per labor hour (SPLH): formula and your target" description: "What SPLH means, how to calculate sales per labor hour, and how to set your restaurant's target from your own wages and your labor cost ceiling." canonical: "https://restaurantdoctorai.com/guides/sales-per-labor-hour" published: "2026-10-10" updated: "2026-10-10" --- # Sales per labor hour (SPLH): formula and your target Sales per labor hour (SPLH) is your sales divided by the hours your staff worked. It turns labor cost into a number a manager can check during a shift, not just on next month's P&L. There's no published “good” SPLH for restaurants: yours comes from your own wages and your labor cost ceiling. The quick answer Divide sales by the hours worked, then compare the result with your cut line: the SPLH below which labor runs over your ceiling. SPLH = net sales ÷ labor hours worked | Restaurant type | Cut line at $18 an hour | | --- | --- | | Quick service | $58.06 | | Fast casual | $56.25 | | Full service / casual dining | $51.43 | | Fine dining | $47.37 | | Bar, taproom or lounge | $56.25 | | Café or coffee shop | $52.94 | Worked out from the top of each labor cost target RestaurantDoctorAI uses, at an example cost of $18 an hour in wages and payroll taxes. Not a survey: put in your own hourly cost. [Work out yours with the free calculator](https://restaurantdoctorai.com/calculators/labor-cost#cut-line), no sign-up. ## What does SPLH mean? SPLH stands for sales per labor hour: how many dollars of sales each hour of paid staff time brought in. If the restaurant sold $1,200 between 6 and 7 p.m. with 20 staff hours on the clock, SPLH for that hour was $60. Labor cost percentage needs payroll, which you only see after the fact. SPLH needs only two numbers you can read mid-shift: sales from the point of sale and hours from the time clock. That's why managers use it to decide who goes home and when the next person should start. ## How to calculate sales per labor hour *Sales per labor hour* SPLH = net sales ÷ labor hours worked Net sales after discounts and comps, before tax and tips. Hours actually worked on the time clock, not hours scheduled. 1. Pick the period: an hour, a shift, a day or a week. Shorter periods tell you more about the schedule. 2. Take net sales for that period from your point of sale. 3. Take the hours your hourly staff actually worked in it from the time clock, overtime included. 4. Divide sales by hours. Count the same way every time, so this week compares with last week. *Quick service, one Tuesday* - Net sales for the day: $3,240 - ÷ Labor hours worked: 54 **Sales per labor hour:** $60.00 On its own, $60 says little. Next to the cut line below, it tells you whether Tuesday's schedule paid for itself. ## What is a good sales per labor hour? One at or above your cut line. We couldn't find a neutral published figure for a good restaurant SPLH, and one wouldn't fit anyway: the same $50 is healthy where staff earn $14 an hour and over budget where they earn $25. So work your line out from two of your own numbers. *Your cut line* Cut line = average hourly labor cost ÷ labor cost ceiling as a decimal Average hourly labor cost is hourly wages plus the payroll taxes on them, divided by the hours worked. It leaves out benefits and salaried pay, so aim a little above the line. *Full service: setting the line* - Hourly pay and payroll taxes, last week: $15,480 - ÷ Hours worked: 860 - = Average hourly labor cost: $18.00 - ÷ Full service ceiling as a decimal: 0.35 **Cut line:** $51.43 Any hour that brings in less than $51.43 per labor hour is running labor above 35% of sales. The ceiling is the top of the labor cost target for your type of restaurant, written as a decimal. The quick answer at the top of this page works the line out for each type at $18 an hour; see [labor cost percentage by restaurant type](https://restaurantdoctorai.com/guides/restaurant-labor-cost#labor-cost-targets) for the targets and the National Restaurant Association's medians beside them. All of one type's targets, with an example month: [full service](https://restaurantdoctorai.com/restaurant-types/full-service), [quick service](https://restaurantdoctorai.com/restaurant-types/quick-service) and [fast casual](https://restaurantdoctorai.com/restaurant-types/quick-service#fast-casual), [bars](https://restaurantdoctorai.com/restaurant-types/bars), [cafés](https://restaurantdoctorai.com/restaurant-types/cafes), or [every type side by side](https://restaurantdoctorai.com/restaurant-types). ## SPLH vs labor cost percentage They're the same thing seen from two sides. Labor cost percentage is your average hourly labor cost divided by your SPLH: at $18 an hour and $60 in sales per labor hour, labor is 30% of sales. That's why a cut line works. | | Labor cost % | Sales per labor hour | | --- | --- | --- | | What it divides | Pay, taxes and benefits by sales | Sales by hours worked | | What you need | Payroll and sales | Sales and the time clock | | When you see it | After payroll runs | During the shift | | Best for | Checking last week against your target | Deciding who goes home this hour | *Two views of the same labor cost.* 45-second profit check See if labor is your costliest issue Answer 12 quick questions, including what happens to staffing when a shift slows down, and get a free preliminary grade and your costliest issue in about 45 seconds, so you know whether the schedule is where to start. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Using SPLH during a shift Check it hour by hour against your line. One slow hour before a rush is normal. When an hour comes in under the line and the next hour's usual sales won't make up for it, that's when to make the first cut, starting with the role guests notice least. [Free calculatorWork out your cut line: labor cost calculatorLast week's payroll and sales give your line, salaried pay included, and how many hourly staff an hour's sales pay for.](https://restaurantdoctorai.com/calculators/labor-cost#cut-line) Check the rules before you send anyone home Some states and cities require a minimum amount of pay when a shift is cut short, and some require notice before a schedule changes. Check the rules where you operate before you use a cut line to end shifts early. ## How to raise sales per labor hour - Schedule to the hour, not the day: look at sales by hour for the last four of the same weekday and staff to that curve. - Stagger start times so people arrive with the rush, and plan cuts from the end of shifts rather than adding cover late. - Move prep, stocking and deep cleaning into the slowest hours, so those hours do work that would otherwise be paid at peak. - Cross-train so one person can cover two stations when it's quiet. - Lift the sales side too: a higher average check raises SPLH with the same hours. A very high SPLH isn't always good news. If it comes with long ticket times or slipping reviews, the floor is short, and guests are paying for the saving. ## Common questions ### What does SPLH stand for? Sales per labor hour: sales for a period divided by the hours staff worked in it. ### Do salaried managers' hours count in SPLH? Pick one way and keep it. Counting only hourly staff matches the decision a cut line is for, who goes home this hour. Salaried pay still costs money, so the free calculator's cut line takes it into account. ### Should SPLH use net or gross sales? Net sales, after discounts and comps, before tax and tips: the same sales your labor cost percentage uses, so the two numbers line up. ### What's the difference between SPLH and labor cost percentage? Labor cost percentage is labor cost as a share of sales, worked out after payroll. SPLH is sales per hour worked, which you can see during the shift. One is the other turned around: average hourly labor cost ÷ SPLH = labor cost percentage. See [restaurant labor cost percentage](https://restaurantdoctorai.com/guides/restaurant-labor-cost) for the full picture. --- title: "Tipped minimum wage by state: 2026 rates and 2027 changes" description: "The tipped minimum wage in every state: today's cash wage and tip credit, the January 1, 2027 changes from each state's own page, and what a raise costs." canonical: "https://restaurantdoctorai.com/guides/tipped-minimum-wage" published: "2026-10-10" updated: "2026-10-14" --- # Tipped minimum wage by state: 2026 rates and 2027 changes The tipped minimum wage is the cash wage an employer pays before tips. Under federal law it's $2.13 an hour, with tips making up the rest of $7.25. Many states set it higher, 7 require the full minimum before tips, and 16 states change their rates on January 1, 2027. The quick answer Federal law: $2.13 an hour in cash, with up to $5.12 counted from tips. These states change on January 1, 2027: Cash wage + tips ≥ minimum wage, or the employer pays the difference | State | Tipped cash wage, January 1, 2027 | | --- | --- | | Arizona | $12.65 | | California | Full minimum: $17.40 | | Colorado | $12.69 | | Connecticut | Not announced yet | | Maine | $7.85 | | Michigan | $6.30 | | Minnesota | Full minimum: $11.87 | | Montana | Full minimum: $11.20 | | Nebraska | Not announced yet | | New Jersey | $6.61 | | Ohio | $5.70 | | Rhode Island | $3.89 | | South Dakota | Not announced yet | | Vermont | Not announced yet | | Virginia | $2.13 | | Washington | Full minimum: $17.73 | From each state's own page, checked October 8, 2026. Today's rate in every state is in [the table below](#by-state). ## How the tip credit works Federal law lets an employer pay a tipped employee, someone who usually gets more than $30 a month in tips, a cash wage of $2.13 an hour and count up to $5.12 of their tips toward the $7.25 minimum. That $5.12 is the tip credit. If cash wage plus tips comes to less than the minimum for a workweek, the employer pays the difference. The employer has to tell staff about the tip credit in advance, and staff keep their tips, apart from a valid tip pool. See the Labor Department's [fact sheet on tipped employees](https://www.dol.gov/agencies/whd/fact-sheets/15-tipped-employees-flsa). States can set a higher cash wage or a smaller tip credit, and some allow none. Where the state's rule and the federal one differ, the one better for the employee applies. ## States where tipped staff get the full minimum before tips In these 7 states, tips don't count toward the minimum wage: the employer pays the full rate in cash, and tips come on top. - Alaska: $14.00 an hour. - California: $16.90 an hour. A higher city or county minimum comes first. - Minnesota: $11.41 an hour. - Montana: $10.85 an hour. For businesses with yearly sales over $110,000. Smaller businesses the federal law doesn't cover: $4.00. - Nevada: $12.00 an hour. - Oregon: $15.55 an hour. Standard rate. Portland metro: $16.80. Nonurban counties: $14.55. - Washington: $17.13 an hour. ## Tipped minimum wage by state | State | Cash wage | Tip credit | 2027 | | --- | --- | --- | --- | | Alabama | $2.13 | $5.12 | — | | Alaska | $14.00 | None | July 1 | | Arizona | $12.15 | $3.00 | $12.65 | | Arkansas | $2.63 | $8.37 | — | | California | $16.90 | None | $17.40 | | Colorado | $12.14 | $3.02 | $12.69 | | Connecticut | $6.38 | $10.56 | Not set | | Delaware | $2.23 | $12.77 | — | | District of Columbia | $10.30 | $8.10 | July 1 | | Florida | $11.98 | $3.02 | — | | Georgia | $2.13 | $5.12 | — | | Hawaii | $14.75 | $1.25 | — | | Idaho | $3.35 | $3.90 | — | | Illinois | $9.00 | $6.00 | — | | Indiana | $2.13 | $5.12 | — | | Iowa | $4.35 | $2.90 | — | | Kansas | $2.13 | $5.12 | — | | Kentucky | $2.13 | $5.12 | — | | Louisiana | $2.13 | $5.12 | — | | Maine | $7.55 | $7.55 | $7.85 | | Maryland | $3.63 | $11.37 | — | | Massachusetts | $6.75 | $8.25 | — | | Michigan | $5.49 | $8.24 | $6.30 | | Minnesota | $11.41 | None | $11.87 | | Mississippi | $2.13 | $5.12 | — | | Missouri | $7.50 | $7.50 | — | | Montana | $10.85 | None | $11.20 | | Nebraska | $2.13 | $12.87 | Not set | | Nevada | $12.00 | None | — | | New Hampshire | $3.27 | $3.98 | — | | New Jersey | $6.05 | $9.87 | $6.61 | | New Mexico | $3.00 | $9.00 | — | | New York | Varies | Varies | — | | North Carolina | $2.13 | $5.12 | — | | North Dakota | $4.86 | $2.39 | — | | Ohio | $5.50 | $5.50 | $5.70 | | Oklahoma | $2.13 | $5.12 | — | | Oregon | $15.55 | None | July 1 | | Pennsylvania | $2.83 | $4.42 | — | | Rhode Island | $3.89 | $12.11 | $3.89 | | South Carolina | $2.13 | $5.12 | — | | South Dakota | $5.93 | $5.92 | Not set | | Tennessee | $2.13 | $5.12 | — | | Texas | $2.13 | $5.12 | — | | Utah | $2.13 | $5.12 | — | | Vermont | $7.21 | $7.21 | Not set | | Virginia | $2.13 | $10.64 | $2.13 | | Washington | $17.13 | None | $17.73 | | West Virginia | $2.62 | $6.13 | — | | Wisconsin | $2.33 | $4.92 | — | | Wyoming | $2.13 | $5.12 | — | *Today's figures from the [US Department of Labor's table](https://www.dol.gov/agencies/whd/state/minimum-wage/tipped) (revised July 1, 2026), with Florida's September 30, 2026 raise from Florida's own page. The 2027 column is the cash wage from January 1, 2027 (the full minimum where there's no tip credit), from each state's own page, checked October 8, 2026. “Not set”: it changes January 1 at a rate not published yet. “July 1”: it changes then. A dash: we found no 2027 change.* ### Notes by state - Alabama: No state minimum wage law: the federal rule applies. - Connecticut: Bartenders who usually get tips: $8.23 cash wage, $8.71 tip credit. - District of Columbia: The cash wage is at least 56% of the standard minimum. - Florida: Since September 30, 2026: $15.00 less Florida's $3.02 tip credit. Set until December 31, 2027. - Georgia: Georgia's own minimum wage law doesn't cover tipped staff: the federal rule applies. - Hawaii: The tip credit is allowed only if wages plus tips come to at least $7.00 more than the minimum. - Illinois: The tip credit is 40% of the minimum. - Iowa: The tip credit is 40% of the minimum. - Kansas: Kansas law doesn't cover jobs the federal wage law covers. - Louisiana: No state minimum wage law: the federal rule applies. - Maine: The tip credit is 50% of the minimum. - Michigan: The tipped cash wage rises each year until it reaches 50% of the minimum in 2031. - Mississippi: No state minimum wage law: the federal rule applies. - Missouri: The tip credit is 50% of the minimum. - New Jersey: Seasonal employers and those with fewer than 6 staff: $15.23 minimum. The tip credit needs advance notice to staff. - New York: Tipped rates vary by region and type of employer: see New York's minimum wage overview. - North Carolina: The tip credit needs a signed record of tips from each employee and advance notice. - North Dakota: The tip credit is 33% of the minimum. - Ohio: For employers with yearly gross receipts of $405,000 or more. For smaller employers the minimum wage is $7.25 an hour; neither Ohio's poster nor the US Department of Labor's table says whether they may take a tip credit, so ask Ohio's Wage and Hour office. - Oklahoma: Oklahoma law doesn't cover jobs the federal wage law covers. - South Carolina: No state minimum wage law: the federal rule applies. - South Dakota: The tip credit is 50% of the minimum. - Tennessee: No state minimum wage law: the federal rule applies. - Texas: Texas law doesn't cover jobs the federal wage law covers. - Utah: Utah law doesn't cover jobs the federal wage law covers. - Vermont: The tip credit is 50% of the minimum. - West Virginia: The tip credit is 70% of the minimum. For employers with 6 or more staff. - Wyoming: Wyoming's own minimum is $5.15; staff covered by the federal law get the federal rate. Some states set lower rates for minors, students or new hires, and some cities set a higher minimum than their state. The Labor Department's table leaves those out, and so does this one. ## What changes in 2027 Each state's announced change, with the page it comes from: - **Arizona**, January 1: minimum $15.65, tipped cash wage $12.65. Employers can pay tipped staff up to $3.00 an hour less, if tips make up the difference. [Source](https://www.azica.gov/labor-department) - **California**, January 1: minimum $17.40, no tip credit. A higher city or county minimum comes first. [Source](https://www.dir.ca.gov/DIRNews/2026/2026-66.html) - **Colorado**, January 1: minimum $15.71, tipped cash wage $12.69. [Source](https://cdle.colorado.gov/press-releases/press-release-colorado-minimum-wage-will-increase-to-1571/hour-in-2027) - **Connecticut**, January 1: minimum $17.48, tipped wage not announced yet. The 2027 tipped cash wage isn't confirmed yet (in 2026 it was $6.38 for restaurant staff and $8.23 for bartenders, set in state law). [Source](https://portal.ct.gov/governor/news/press-releases/2026/08-2026/governor-lamont-announces-minimum-wage-will-increase) - **Maine**, January 1: minimum $15.70, tipped cash wage $7.85. To be paid the tipped wage, a service employee now needs $199 a month in tips (up from $191). [Source](https://www.maine.gov/labor/news_events/article.shtml?id=13362373) - **Michigan**, January 1: minimum $15.00, tipped cash wage $6.30. The tipped cash wage is 42% of the minimum in 2027 and rises 2 points a year to 50% in 2031. [Source](https://www.michigan.gov/leo/bureaus-agencies/ber/wage-and-hour/min-wage) - **Minnesota**, January 1: minimum $11.87, no tip credit. Training wage for workers under 20: $9.68. [Source](https://dli.mn.gov/news/minimum-wage-rate-adjusted-inflation-jan-1-2027) - **Montana**, January 1: minimum $11.20, no tip credit. Businesses not covered by federal wage law, with yearly sales of $110,000 or less: $4.00. [Source](https://erd.dli.mt.gov/labor-standards/wage-and-hour-payment-act/state-minimum-wage) - **Nebraska**, January 1: minimum not announced yet, tipped wage not announced yet. A 1.75% raise set by state law (LB 258). The Labor Department must publish the new rate by October 15, 2026. [Source](https://www.nebraskalegislature.gov/FloorDocs/109/PDF/Slip/LB258.pdf) - **New Jersey**, January 1: minimum $16.48, tipped cash wage $6.61. Seasonal employers and employers with fewer than 6 staff: $16.25. The most tip credit is $9.87 an hour. [Source](https://www.nj.gov/labor/lwdhome/press/2026/20260929_min_wage.shtml) - **Ohio**, January 1: minimum $11.40, tipped cash wage $5.70. For employers with $419,000 or more a year in gross receipts. Smaller employers pay the federal minimum, $7.25. [Source](https://dam.assets.ohio.gov/image/upload/com.ohio.gov/DICO/WageAndHour/2027_Minumum_Wage_Poster.pdf) - **Rhode Island**, January 1: minimum $17.00, tipped cash wage $3.89. [Source](https://webserver.rilegislature.gov/Statutes/TITLE28/28-12/28-12-3.htm) - **South Dakota**, January 1: minimum not announced yet, tipped wage not announced yet. Raised each year with the cost of living; the state posts the new rate by October 15, 2026. The tipped cash wage is half of it. [Source](https://dlr.sd.gov/employment_laws/minimum_wage.aspx) - **Vermont**, January 1: minimum not announced yet, tipped wage not announced yet. Raised at the start of each year under state law (21 V.S.A. § 384). The tipped wage is half of the new minimum. [Source](https://labor.vermont.gov/node/4865) - **Virginia**, January 1: minimum $13.75, tipped cash wage $2.13. Virginia sets no tipped wage of its own: the federal $2.13 cash wage applies to most employers, and tips must make up the rest. $15.00 from January 1, 2028. [Source](https://law.lis.virginia.gov/vacode/title40.1/chapter3/section40.1-28.10/) - **Washington**, January 1: minimum $17.73, no tip credit. [Source](https://lni.wa.gov/workers-rights/wages/minimum-wage/) - **Alaska**, July 1: minimum not announced yet, no tip credit. A step already set in state law; from January 1, 2028 the wage follows inflation. [Source](https://dol.alaska.gov/news/2025/news25-11.htm) - **District of Columbia**, July 1: minimum not announced yet, tipped wage not announced yet. Raised each July 1 with inflation. The tipped cash wage is 56% of the new minimum. [Source](https://code.dccouncil.gov/us/dc/council/code/sections/32-1003) - **Oregon**, July 1: minimum not announced yet, no tip credit. Raised each July 1 with inflation. Oregon has three rates: standard, Portland metro and nonurban counties. [Source](https://www.oregon.gov/boli/workers/pages/minimum-wage.aspx) [Put your state's and city's dates in your calendar](https://restaurantdoctorai.com/deadline-calendar) with the free deadline calendar, with a reminder before each one. The posters that carry these rates, and the pay and tip records to keep, are in [what a restaurant has to post and keep](https://restaurantdoctorai.com/guides/restaurant-required-posters-and-records). ## What the raise costs you The cost of a tipped wage raise is the rise per hour times your tipped staff's hours. Your share of Social Security and Medicare, 7.65% of pay (6.2% plus 1.45%, per the [IRS](https://www.irs.gov/taxtopics/tc751)), goes up with it. Free calculator ### What the raise costs you Nothing you type leaves this page. Example: Michigan on January 1, 2027. Replace the numbers with yours. Tipped staff People paid the tipped cash wage. Hours each works a week An average week, from your time clock. Cash wage now Per hour, before tips. Cash wage after the raise Per hour, before tips. #### With the example numbers **$181.44** more a week in pay, about $9,434.88 a year. With your share of Social Security and Medicare on top (7.65% of pay), it's $195.32 a week, or $10,156.65 a year. State unemployment tax isn't included. How we worked it out - Rise per hour: $6.30 − $5.49 = $0.81 - A week: 8 × 28 hours × $0.81 = $181.44 - A year: $181.44 × 52 = $9,434.88 - With 7.65% on top: $9,434.88 × 1.0765 = $10,156.65 To see what labor costs you as a share of sales after the raise, use the [free labor cost calculator](https://restaurantdoctorai.com/calculators/labor-cost), and see [restaurant labor cost percentage](https://restaurantdoctorai.com/guides/restaurant-labor-cost) for the targets by type. To share tips at the end of a shift, the free [tip-out sheet](https://restaurantdoctorai.com/templates/tip-out-sheet) works out a tip-out by percentage or a tip pool by hours, so you can show each person how their share was worked out. A charge you add to the bill isn't a tip, and the rules for it differ: see [what a restaurant service charge is, and whether to add one](https://restaurantdoctorai.com/guides/restaurant-service-charge). 45-second profit check Is labor your costliest issue after the raise? Answer 12 quick questions, including how staffing changes when a shift gets slow, and get your restaurant's grade and its biggest estimated loss, free, in about 45 seconds, so you know where to start. [Start the free check](https://restaurantdoctorai.com/#diagnostic) Check your state's rules before you change pay Notice rules, tip pools, service charges and overtime for tipped staff vary by state and city. This page is general information, not legal advice: check with your state's labor office before you change anyone's pay. ## Common questions ### What is the federal tipped minimum wage? $2.13 an hour in cash, with up to $5.12 counted from tips toward the $7.25 federal minimum. It applies where a state sets nothing higher. ### Which states don't allow a tip credit? Alaska, California, Minnesota, Montana, Nevada, Oregon, Washington. There, the full state minimum is owed before tips. ### What happens if tips don't make up the minimum? The employer pays the difference, so that cash wage plus tips reaches at least the minimum for the hours worked that week. ### Does “no tax on tips” change the tipped minimum wage? No. It's an income tax deduction for people who get tips; it doesn't change what an employer has to pay. [What it means for restaurant owners](https://restaurantdoctorai.com/blog/no-tax-on-tips-restaurants). ### Is later in 2027 covered? Where a state has said when its next change comes, yes: Alaska, District of Columbia, Oregon change on July 1, 2027, at rates not announced yet. We add rates as states publish them. Wage figures from the US Department of Labor and each state's own pages, as dated above. General information, not legal advice: check with your state's labor office. --- title: "What is a restaurant service charge, and should you add one?" description: "A service charge is a fee the restaurant sets, not a tip. What it means for payroll tax, the FICA tip credit, no tax on tips and guests, with an example." canonical: "https://restaurantdoctorai.com/guides/restaurant-service-charge" published: "2026-10-14" updated: "2026-10-15" --- # What is a restaurant service charge, and should you add one? A service charge is an amount the restaurant sets and adds to the bill, such as 18% on parties of six or more, while a tip is an amount the guest freely chooses. For tax purposes the charge is the restaurant's own revenue, and the part you pay out to staff is wages, not tips, which changes your payroll taxes, the FICA tip credit and what you have to show guests. The short version - A charge the guest can't refuse or change is a service charge, whatever the bill calls it. - Paid out to staff, it's wages: withholding, your 7.65% share of Social Security and Medicare, and no FICA tip credit. - It isn't a qualified tip for the federal "no tax on tips" deduction. - California lets restaurants add one if it's shown wherever prices are shown, and taxes it as part of the sale. ## Service charge or tip: the IRS's four tests The IRS doesn't go by the label on the bill. [Revenue Ruling 2012-18](https://www.irs.gov/irb/2012-26_IRB#RR-2012-18) says a payment may be a service charge, not a tip, when any of these is missing: - The guest pays it free from compulsion. - The guest has the unrestricted right to decide the amount. - The amount isn't negotiated or dictated by the restaurant's policy. - Generally, the guest decides who receives it. The ruling's own example is a menu that adds 18% to every bill for parties of six or more: not a tip, because the guest didn't choose it. The [IRS tip reporting page](https://www.irs.gov/businesses/small-businesses-self-employed/tip-recordkeeping-and-reporting) adds a required gratuity in an event contract, and a payment screen that makes the guest pick a tip above zero. Two things stay tips: suggested amounts printed on the receipt when the tip line is left blank, and anything a guest adds on top of a service charge. ## What a service charge changes for you ### It's your revenue The IRS says service charges are income to the restaurant whether you pay out all of them, part of them or none. For federal tax, you decide how they're split; with a tip, the guest decides. ### Paid out, it's wages What you pay staff from a service charge goes on their W-2 as wages, and [Publication 15](https://www.irs.gov/publications/p15) tells employers to withhold on service charges as on regular wages. The ruling makes it wages for Social Security and Medicare too, so you pay the employer's 6.2% plus 1.45%, 7.65% in all ([IRS Topic 751](https://www.irs.gov/taxtopics/tc751)). You pay the same 7.65% on tips your staff report. The difference is what comes back. ### No FICA tip credit on it The [FICA tip credit](https://www.irs.gov/businesses/small-businesses-self-employed/fica-tip-credit-for-employers) gives back the employer's 7.65% on reported tips as a credit against federal income tax, on Form 8846. The IRS page says distributed service charges and auto-gratuities are non-tip wages, left out of the credit. The credit only lowers income tax the business owes, and [the law](https://www.law.cornell.edu/uscode/text/26/45B) allows no deduction for the tax it covers. [Free calculatorWork out yours: FICA tip credit calculatorYour reported tips, hours and cash wage, and what the credit could be worth.](https://restaurantdoctorai.com/calculators/fica-tip-credit) ### Not a tip for “no tax on tips” The 2025 deduction lets workers deduct qualified tips on their own federal return. [The IRS's release on the final rules](https://www.irs.gov/newsroom/treasury-irs-issue-final-regulations-listing-occupations-where-workers-customarily-and-regularly-receive-tips-under-the-one-big-beautiful-bill) says qualified tips are paid voluntarily and don't include a service charge unless the guest can disregard or modify it; its example is an automatic 18% for large parties. So a server can't deduct pay that came from a service charge. The deduction doesn't change what you pay as the employer either way: [what no tax on tips means for restaurant owners](https://restaurantdoctorai.com/blog/no-tax-on-tips-restaurants). ### It counts toward minimum wage Under federal wage law, the Department of Labor's [fact sheet on tipped employees](https://www.dol.gov/agencies/whd/fact-sheets/15-tipped-employees-flsa) says a compulsory service charge isn't a tip, may be used toward the employer's minimum wage and overtime pay, and counts in the regular rate for overtime. Each state's tipped cash wage is in the [tipped minimum wage guide](https://restaurantdoctorai.com/guides/tipped-minimum-wage). ## A worked example: the same 18%, two ways A full-service restaurant sells $40,000 a month of dine-in food and drink, before tax. Guests either tip 18% or pay the same 18% as a service charge that all goes to staff. Its servers' cash wage is at least $7.25 an hour, so every reported tip counts toward the credit whether you read the base as the $5.15 in the credit's law or the $7.25 on the IRS's page. The figures are round examples. *Example: one month, as tips or as a service charge* - 18% of $40,000: $7,200 - Your Social Security and Medicare on it (7.65%), either way: $550.80 - FICA tip credit if it's tips: $550.80 - FICA tip credit if it's a service charge: $0 **Credit given up in a year: 12 × $550.80:** $6,609.60 The credit lowers federal income tax owed, so it's worth the full amount only to a business with that much income tax to offset. It moves your labor cost too. The [labor cost guide](https://restaurantdoctorai.com/guides/restaurant-labor-cost#what-counts-as-labor) leaves tips out of both sales and labor, as RestaurantDoctorAI's targets do, but a service charge is sales and its payout is payroll. Say payroll with its taxes is $12,000 before the 18%. As tips, labor is ($12,000 + $550.80) ÷ $40,000 = 31.4% of sales. As a service charge, it's ($12,000 + $7,200 + $550.80) ÷ $47,200 = 41.8%. Same staff, same hours, so compare months on the same basis. [Free calculatorCheck your labor cost percentageWages, payroll taxes and sales, against the target for your type of restaurant.](https://restaurantdoctorai.com/calculators/labor-cost) ## Telling guests **California.** Its honest pricing law, SB 478, in force since July 1, 2024, requires most listed prices to include every mandatory fee. The [Attorney General's page](https://oag.ca.gov/hiddenfees) says restaurants, bars and some other food sellers are exempt for their mandatory fees as long as the fee is clearly and conspicuously displayed wherever prices are shown, under SB 1524. Voluntary tips aren't affected. **Federal.** The FTC's rule on unfair or deceptive fees, in force since May 12, 2025, covers businesses that offer live-event tickets or short-term lodging, [its FAQ says](https://www.ftc.gov/business-guidance/resources/rule-unfair-or-deceptive-fees-frequently-asked-questions). A restaurant meal is neither. Rules differ by state and city, so read your state's own pages, its attorney general's and its labor department's, before you add a charge. ## Should your restaurant add a service charge? It fits where a set amount is expected, like large parties and event contracts, the IRS's own examples, or where you want to set how pay is shared across the team. Weigh it against: - **The FICA tip credit:** none on a service charge, as the example shows. - **Sales tax:** California's tax agency [says](https://cdtfa.ca.gov/formspubs/pub115/) an optional tip isn't taxed, but a mandatory one is part of taxable sales, even if you pay it all to staff. Check your own state's tax agency. - **Server pay:** the payout runs through payroll with withholding, follows your policy, and isn't a qualified tip for their deduction. Tell staff how it's split before you start. - **Guests:** anything a guest adds on top is a tip. For a month, note how often guests add one, and how often they ask about the charge. - **Your numbers:** labor rises as a share of sales, so mark the month you changed. If you'd rather keep tips, suggested amounts printed on the receipt, with the tip line left blank for the guest, stay tips, the IRS says. 45-second profit check See where labor sits in your costs Answer 12 quick questions about how your restaurant runs and get its grade and its biggest estimated loss, free, in about 45 seconds, so you know whether labor is where to start. [Start the free check](https://restaurantdoctorai.com/#diagnostic) --- title: "Bar pour cost: how to calculate and fix it" description: "Pour cost is the share of bar sales spent on what you pour. Get the formula for a week or one drink, the 18–24% US target, and how to fix a high one." canonical: "https://restaurantdoctorai.com/guides/bar-pour-cost" published: "2026-09-27" updated: "2026-10-14" --- # Bar pour cost: how to calculate and fix it Pour cost is the percentage of your bar sales that goes to the liquor, beer, and wine you pour. RestaurantDoctorAI uses 18–24% as the target for US bar programs, a range that comes from one software company's figures for its customers; no independent survey publishes one. Over-pours, free drinks, and unrung sales push that number up, and they hit the highest-margin part of your menu. The short version - Pour cost is the cost of what you pour, divided by what that pouring sells for. - In the US, RestaurantDoctorAI uses 18–24% of bar sales as the target. Outside the US, norms differ, so watch the weekly trend. - You can measure it for a whole week or for a single drink. - Over-pours, free drinks, and unrung sales are the usual reasons it climbs. - A jigger, a comp sheet, and a weekly bar count show where the gap is and whether it is shrinking. ## What pour cost means Pour cost does for the bar what food cost does for the kitchen. It shows how much of a sale goes to the product itself, before labor, rent, or anything else. You can work it out two ways. One covers a period, such as a week, using your bar inventory. The other covers a single drink, using what goes in the glass and its price on the menu. Tracking it regularly matters more than getting one count exactly right. A single number tells you where you stand today. A string of weekly numbers tells you whether the bar is drifting or holding steady. ## Pour cost formula for a week or month For a week or a month, compare what the bar used against what it sold. This shows the shape of the whole bar over time, not just one bottle. *Weekly or monthly formula* Pour cost % = (opening bar stock + bar purchases − closing bar stock) ÷ bar sales × 100 Use bar purchases and bar sales only. Food and drink often share a POS and a supplier, but this number looks at the drink side alone. *Example: one week at the bar* - Opening bar stock: $5,000 - + Bar purchases this week: $2,000 - − Closing bar stock: $4,000 - Product used: $3,000 - Bar sales this week: $15,000 **Pour cost:** 20% $3,000 ÷ $15,000 × 100 = 20%, inside the 18–24% range RestaurantDoctorAI uses for US bars. ## Pour cost for one drink For a single drink, compare the cost of everything in the glass against the menu price. Use it when you price a new cocktail, or to check whether an existing one still earns what it should. *Per-drink formula* Pours per bottle = bottle size ÷ standard pour Cost per pour = bottle cost ÷ pours per bottle Drink cost = each pour in the recipe + mixers + garnish Pour cost % = drink cost ÷ drink price × 100 Use your house standard pour, the one written on the card behind the bar and measured with a jigger. *Example: a single-spirit drink* - Bottle size: 750 ml, about 25 oz - ÷ Standard pour: 1 oz - Pours per bottle: 25 - Bottle cost: $25.00 - Cost per pour: $1.00 - Drink price: $5.00 **Pour cost:** 20% $1.00 ÷ $5.00 × 100 = 20%, the same 20% as the weekly example above. This drink has no mixer or garnish, so its cost is one pour. A cocktail takes one more step. Add up the cost of every spirit, liqueur, mixer, and garnish in the glass, then divide that total by the menu price. ## Where pour cost fits in your numbers Pour cost is part of a bigger number: your combined cost of goods sold, food and drink together. The targets RestaurantDoctorAI uses set that combined target by service format. | Service format | Combined COGS target | | --- | --- | | Quick service | 26–32% | | Fast casual | 28–33% | | Full service / casual dining | 30–34% | | Fine dining | 32–38% | | Bar, taproom or lounge | 20–26% | | Café or coffee shop | 24–32% | *The targets RestaurantDoctorAI uses for combined food and drink cost of goods sold, by service format.* Bars, taprooms, and lounges carry the lowest combined target, because drinks usually cost a smaller share of their price than food. Drinks are about half of a bar's sales or more. So the same slip in pour cost moves a bar's total cost of goods more than it would in a restaurant where drinks are a small share of sales. For every target RestaurantDoctorAI uses for a bar, see [bar costs and targets](https://restaurantdoctorai.com/restaurant-types/bars). Pour cost also feeds your [prime cost](https://restaurantdoctorai.com/guides/restaurant-prime-cost): food and drink cost plus labor, as a share of sales. Prime cost leaves out rent and other overhead, so track those separately. The bar number works the same way [food cost percentage](https://restaurantdoctorai.com/guides/food-cost-percentage) does in the kitchen: one number, tracked every week, that shows whether what you buy matches what you sell. ## Why pour cost drifts upward A few habits push the number up, often without anyone noticing. - **Free pouring.** A heavy hand with the bottle adds product you paid for but never billed. - **Comps and staff drinks.** A drink comped for a regular, a spill, or a staff drink uses product with no sale. Left unlogged, it looks like shrink instead of a choice someone can review. - **Unrung sales.** A drink poured but never rung into the POS uses stock with no sale to match it. - **Over-pours on premium bottles.** These cost more per ounce, so the same heavy hand costs more on a top-shelf pour than on a well drink. - **Inconsistent recipes.** When two bartenders build the same cocktail differently, the cost swings from shift to shift even without anyone pouring heavy on purpose. ## Fixing a high pour cost None of these fixes need new software. They need a jigger, a weekly count, and a place to write down every comp. 1. Measure every spirit pour with a jigger, and write the standard pour for each drink on a card behind the bar. 2. For wine by the glass, check pours against your glass size. For draft beer, compare kegs used with pints rung in. 3. Count your 10 best-selling bottles at open and close for a few nights, and compare what was used against what the POS says was sold. 4. Log every comp, spill, and staff drink on a sheet with a manager's initials, or ring it into the POS as a comp. 5. Pick a fixed day and person to count the whole bar every week, in the same order each time. 6. Work out the week's pour cost from that count. In the US, track it against the 18–24% range. Elsewhere, track the trend: a falling number means the gap is closing. Alcohol rules, including what you can pour, comp, or discount, vary by state and sometimes by county or city. Check yours before you set a house policy. Drinks for friends still count A jigger only works if everyone uses it, including managers and owners. Drinks for regulars and friends need to go through the same measure and the same comp sheet as every other pour. Once the count is a habit, pour cost becomes a number you check every Monday, not a mystery you solve once a quarter. 45-second profit check Check your whole restaurant, bar included Answer 12 quick questions about your kitchen, bar, and floor, and get your restaurant's grade and its biggest estimated loss, free, in about 45 seconds, so you know whether the bar is where to start. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Common questions ### What's a good pour cost for a bar? In the US, RestaurantDoctorAI uses 18–24%. That range comes from BevSpot, a bar inventory software company, which published it from its customers' data in 2016; no independent survey publishes a pour cost range, so treat it as a guide. Outside the US, norms differ, so track it weekly and watch the direction. Where you land depends on your drink mix and prices. Compare your bar with its own past weeks rather than with another bar. The trend matters more than any single week, since one busy night of comped rounds can push a count higher. ### How do you calculate pour cost on one drink? Divide the bottle's cost by how many standard pours it holds to get the cost of one pour. For a cocktail, add the cost of every pour, mixer, and garnish in the recipe. Divide that drink cost by the menu price and multiply by 100. Run the same math on a new cocktail before it goes on the menu, so the price is set from the recipe, not a guess. ### Is pour cost the same as liquor cost? Not always. Some people use liquor cost for spirits only. Pour cost in this guide covers spirits, beer, and wine together. Check which one a report uses before you compare it with a target. --- title: "Delivery app commission: what it takes and how to cover it" description: "What DoorDash, Uber Eats and other delivery apps really take, how much to mark up your app menu to cover it (with a free calculator) and how to go direct." canonical: "https://restaurantdoctorai.com/guides/delivery-app-commissions" published: "2026-09-27" updated: "2026-10-15" --- # Delivery app commission: what it takes and how to cover it A delivery app commission is the share of each order the app keeps. The big apps publish plans from 5% to 30%, and a delivered order costs at least 15% on every one of them. It comes off the full menu price, before you've paid for the food, the packaging or the people who made it. Price your app menu to cover it, and move your regulars to direct ordering. The short version - A commission comes off sales, not profit. At 30%, a $30 order hands the app $9.00 before any of your costs are paid. - The big apps publish plans from 5% to 30%, and a delivered order costs at least 15% on each. What an app really takes can also include marketing fees, promotions and refunds, and your payout statement shows it. - Break-even app price = your in-house price ÷ (1 − the rate the app really takes). - Apps are good at finding new guests. Keep them for that, and move repeat orders to your own phone line, website and pickup. - Some cities cap delivery commissions, and contracts differ. Check your agreement and your local rules before you change prices or add offers to app orders. ## How a delivery app commission works When a guest orders through a delivery app, the app charges the guest, keeps its share and pays you the rest. The commission is a percentage of what the order sells for. It grows with every dollar on the ticket, whether that dollar was profitable or not. The apps differ by country: DoorDash and Uber Eats in the US, plus SkipTheDishes in Canada; Deliveroo, Just Eat and Uber Eats in the UK and Ireland; Uber Eats and DoorDash in Australia and New Zealand. Rates and plans vary by app and by contract, but the math on this page works the same for all of them. The commission isn't money for nothing. Some apps offer more than one plan, and a lower rate can come with less included. Depending on your plan, the commission can pay for the driver, your listing and the marketing that puts your menu in front of strangers. That reach is real. The trouble is paying for it on every order, including orders from guests who already know you. The headline rate isn't always the whole story, either. Marketing fees, promotions you agreed to fund and refunds can come out of your payout too. So start from the payout statement, not the contract. *What an app really takes* Real commission rate % = (menu sales before tax and tips − your payout) ÷ menu sales × 100 Use a full month of payout statements, and work out each app on its own. If your payout includes sales tax or tips passed on to you, take them out first. Who collects sales tax on app orders varies by state, so check yours. *Example: one app, one month* - Menu sales before tax and tips: $10,000 - − Your payout: $7,000 - = What the app kept: $3,000 **Real commission rate:** 30% $3,000 ÷ $10,000 × 100 = 30%. If this restaurant's plan says 25%, the other 5 points are fees, promotions and refunds. ## What a 30% commission does to one order It's easy to read a commission as a slice of your profit. It's a slice of your sales, and it comes off first. Here is one $30 delivery order from a quick service restaurant at a 30% commission, with example costs. *Example: one $30 app order* - Order, at menu price: $30.00 - − App commission, 30%: $9.00 - − Food and drink, 30%: $9.00 - − Packaging: $1.50 - − Labor to cook and pack, 30%: $9.00 **Left for rent, other costs and profit:** $1.50 Example costs, not benchmarks. $30.00 − $9.00 − $9.00 − $1.50 − $9.00 = $1.50. Food and labor at 30% each sit inside the ranges RestaurantDoctorAI uses for quick service. Together they make a prime cost of 60%, which is that format's ceiling. The kitchen did its job. The commission and the packaging took what would have been the margin. That's how a restaurant can get busier on delivery and see nothing change at the bottom of the P&L. For the number behind all of this, see [how restaurant prime cost works](https://restaurantdoctorai.com/guides/restaurant-prime-cost). ## Check app orders against your own numbers Your format's prime cost target covers the whole restaurant, not one app order. A bar's ceiling is low because drinks are about half its sales. Full service and fine dining ceilings include the staff who serve your tables. An app order is mostly food, and nobody serves it at a table. So work out app orders from your own costs: the plate cost of your top app dishes, the packaging and the time to cook and pack them. Then take off the real rate from your payout statement. | Real commission rate | App keeps | Left from a $30 order | | --- | --- | --- | | No commission | $0.00 | $10.50 | | 25% | $7.50 | $3.00 | | 30% | $9.00 | $1.50 | | 35% | $10.50 | $0.00 | *Arithmetic on the example order above, with only the rate changing. What's left still has to pay for rent and everything else.* Every 5 points of real rate costs $1.50 on a $30 order, and here that's the difference between a thin margin and none. That's why you price from the rate on your payout statement, not the rate in your plan. ## How much to mark up menu prices on delivery apps An app order should leave you the same dollars as the same dishes sold in your dining room. To get there, price up from your in-house price, using the real rate from your payout statement. *Break-even app price* Break-even app price = in-house price ÷ (1 − real commission rate) Write the rate as a decimal: 30% is 0.30, so you divide by 0.70. *Example: a $14 dish at a 30% real rate* - In-house price: $14.00 - ÷ (1 − 0.30): 0.70 - = Break-even app price: $20.00 - − App keeps 30%: $6.00 **You keep:** $14.00 $14.00 ÷ 0.70 = $20.00, and $20.00 × 30% = $6.00. You keep the same $14.00 as a dining room sale. Treat it as the floor: packaging still comes out of your side. The markup is always bigger than the commission, because the app takes its cut from the higher price too. Here is the same $14 dish at other rates: | Real commission rate | Mark up by at least | Break-even app price | | --- | --- | --- | | 15% | 18% | $16.47 | | 20% | 25% | $17.50 | | 25% | 34% | $18.67 | | 30% | 43% | $20.00 | | 35% | 54% | $21.54 | *Break-even app price for a $14 in-house dish. Arithmetic only: in-house price ÷ (1 − rate).* Free calculator ### Your break-even app price Works in any currency. Nothing you type leaves this page. Example numbers. Replace them with yours. Your in-house price The dish's price on your own menu. Real commission rate % From a month of payout statements: (menu sales before tax and tips − your payout without tax and tips) ÷ menu sales × 100. Work out each app on its own. #### With the example numbers Price it at **20.00** or more on the app. That's **43%** above your in-house price. At that price the app keeps 6.00 and you keep the same 14.00 as a dining room sale. Packaging still comes out of your side, so treat it as the floor. How we worked it out - In-house price: 14.00 - ÷ (1 − 0.30): 0.70 - = Break-even app price: 20.00 - − App keeps 30%: 6.00 - = You keep: 14.00 - Markup on the in-house price, rounded up: 43% Round odd results up, not down. And know what a dish costs to make before you price it. If you're not sure, [work out its plate cost first](https://restaurantdoctorai.com/guides/food-cost-percentage#plate-cost-for-a-single-dish). 1. Pull last month's payout statement for each app and work out its real rate with the formula above. 2. Raise app prices to at least the break-even price. Read your agreement with each app first: it may have its own rules on app prices. 3. Take dishes that travel badly or earn little off the app menus. Put your highest-profit dishes and add-ons first. 4. Check whether each app offers a lower-commission plan. Some trade a smaller delivery area for a lower fee, which can pay off when most orders come from nearby. 5. Compare next month's payout statements with this month's. You want more kept per order while order counts hold roughly steady. Check the payout before you fund an app promotion such as buy-one-get-one. A promotion can cost more than the order earns. Before you raise app prices, read your app's pricing rules too. Some apps limit how much higher app prices can be if you want their badges or promotions, and higher prices can mean fewer app orders. Weigh that against a commission that takes your whole margin at regular prices. Check your contract and your city Some cities cap delivery app commissions, and rules vary by state and city. Contracts also differ on rates, fees, pricing and promotions. Read your agreement and check your local rules before you change prices, switch plans or add your own offers to app orders. ## Moving regulars to direct ordering Pricing covers the commission. It doesn't change who pays it. If your regulars order through an app every week, you pay a commission on guests who already know you. Direct ordering means orders placed with you: by phone, through your own ordering link or through pickup ordering in your POS. You still pay for card processing, but no app commission comes off the top. Check each app agreement before you start. Contracts differ on what you can put in the bag and how you can reach app customers. 1. Set app prices first, with the break-even formula above, so app orders cover their commission while you switch. 2. If your agreement allows it, put a card in every app order with your own phone number or ordering link and a reason to order direct next time, such as a free side. Your own printer is fine. (New York City's law requires apps to allow printed materials with any online order.) 3. Make sure your Google Business Profile and website send people to your own ordering link or phone number first. 4. If your POS already includes pickup ordering, turn it on and promote it. Pickup orders you take yourself pay no delivery app commission. 5. Track direct and pickup orders every week. You want them growing, and the apps' share of your sales slowly falling. Don't switch the apps off overnight. They're good at reaching people who have never heard of you, and that's worth paying for. Keep them for first orders, and move repeat orders to your own channels. 45-second profit check See if delivery is your costliest issue Answer 12 quick questions, including how much of your sales come through delivery apps, and get your restaurant's grade and its biggest estimated loss, free, in about 45 seconds, so you know if the apps are the cost to fix first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Common questions ### What is a typical delivery app commission? DoorDash publishes plans at 15%, 25% and 30%, Uber Eats at 20%, 25% and 30% (15% when you deliver yourself), and Grubhub at 5%, 15% and 20% plus a delivery fee from 10% when its drivers deliver (their merchant pricing pages, checked October 2026). The lower rates come with less included, such as a smaller delivery area. What you really pay can be higher once marketing fees, promotions and refunds come out. Your payout statement shows the real rate. ### Should restaurants charge more on delivery apps? If you want an app order to leave you what a dining room order does, yes. Divide your in-house price by 1 minus the real commission rate: at 30%, a $14 dish becomes $20.00. Check your agreement first, since contracts differ on pricing. ### Is DoorDash worth it for restaurants? DoorDash, or any delivery app, can be worth it as a way to reach new guests. The trouble starts when a large share of your sales pays full commission at regular menu prices, including orders from regulars. Work out the real rate from your payout statement, price the app menu to cover it, and move repeat guests to direct ordering. --- title: "How much rent can a restaurant afford?" description: "Judge restaurant rent as a share of sales. Get the occupancy cost formula, the targets by restaurant type, and the monthly sales your rent needs." canonical: "https://restaurantdoctorai.com/guides/restaurant-rent-percentage" published: "2026-09-28" updated: "2026-10-14" --- # How much rent can a restaurant afford? Judge rent as a share of sales, not as a dollar figure. Add up rent, property charges and utilities, divide by sales, and compare the result with the range for your type of restaurant: 6–10% for quick service and fast casual, and 7–12% for the rest. Above the top of your range, occupancy starts squeezing profit. The short version - Occupancy cost = rent + property charges + utilities, as a share of sales. - The ranges RestaurantDoctorAI uses: 6–10% for quick service, fast casual and cafés, and 7–12% for full service, fine dining and bars. - Above the top of your format's range, occupancy starts squeezing profit. Above 12%, it's over every format's range: a fixed cost no quick fix will change. - What your rent needs: monthly occupancy cost ÷ your target share = the monthly sales it takes. - The levers are sales per seat, starting with your quiet days, and lease terms at renewal. ## How to calculate occupancy cost percentage Occupancy cost is what it costs to be in the building: your rent, any charges the lease passes on to you, and utilities. Put it next to sales for the same month and you can compare it with any restaurant, whatever the rent in dollars. *Occupancy cost formula* Occupancy cost % = (rent + property charges + utilities) ÷ sales × 100 Use one month of each, and the same month for costs and sales. *Example: a full-service restaurant, one month* - Rent: $5,500 - + Property charges: $700 - + Utilities: $1,300 - = Occupancy cost: $7,500 - ÷ Sales for the month: $90,000 **Occupancy cost %:** 8.3% $7,500 ÷ $90,000 × 100 = 8.3%, inside the 7–12% range for full service. ## Rent as a percentage of sales, by restaurant type | Restaurant type | Occupancy cost (% of sales) | | --- | --- | | Quick service | 6–10% | | Fast casual | 6–10% | | Full service / casual dining | 7–12% | | Fine dining | 7–12% | | Bar, taproom or lounge | 7–12% | | Café or coffee shop | 6–10% | *Occupancy cost ranges RestaurantDoctorAI uses: rent, property charges and utilities together, as a share of sales.* Under 8% of sales, rent and utilities leave room for profit. Above the top of your format's range, 10% for quick service, fast casual and cafés or 12% for the rest, they start squeezing it. Above 12%, occupancy is over every format's range: a fixed cost that no quick fix will change, and it makes every other problem more urgent. A full-service restaurant at 12% occupancy with a [prime cost](https://restaurantdoctorai.com/guides/restaurant-prime-cost) of 68% keeps 20 cents of each sales dollar for every other cost and profit: 100 − 68 − 12 = 20. ## How much you need to sell to carry your rent Turn the formula around and it tells you the monthly sales a space needs. That's worth doing before you sign or renew a lease, not after. *Sales your occupancy cost needs* Monthly sales needed = monthly occupancy cost ÷ target share Write the target as a decimal: 8% is 0.08. | Monthly occupancy cost | At 6% | At 8% | At 10% | At 12% | | --- | --- | --- | --- | --- | | $5,000 | $83,333 | $62,500 | $50,000 | $41,667 | | $7,500 | $125,000 | $93,750 | $75,000 | $62,500 | | $10,000 | $166,667 | $125,000 | $100,000 | $83,333 | | $15,000 | $250,000 | $187,500 | $150,000 | $125,000 | *Monthly sales needed for a monthly occupancy cost to land at each share. Arithmetic only.* Find the column for the top of your format's range: 10% for quick service, fast casual and cafés, 12% for the rest. If your sales sit below it for your occupancy cost, the space costs more than your sales can comfortably carry. ## When occupancy is over your range Rent doesn't move week to week, so the fixes are on the sales side and at the lease. The levers are sales per seat and lease terms at renewal. 1. Fill your quiet days first. Rent, staff and utilities cost the same on a quiet Tuesday, so extra covers then are the cheapest sales you can add. Give guests one reason to come on your slowest night. 2. Start your own guest list, so you can invite regulars back on a slow night instead of paying to reach them again. 3. Get prime cost under its ceiling. A high occupancy share leaves no room for food, drink and labor running over. 4. Before the lease renews, work out your occupancy share on your real sales, and take the number to the conversation. Leases differ What counts as rent and property charges depends on your lease. Read yours, and talk to your accountant or a lawyer before you sign or renegotiate one. 45-second profit check See if rent is squeezing your profit Answer 12 quick questions, including your rent and utilities as a share of sales, and get your restaurant's grade and its biggest estimated loss, free, in about 45 seconds, so you know what to fix first while the lease stays as it is. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Common questions ### How much rent should a restaurant pay? Aim for rent, property charges and utilities together inside the range for your type: 6–10% of sales for quick service, fast casual and cafés, and 7–12% for full service, fine dining and bars. Above the top of your range, it squeezes profit. ### Does occupancy cost include utilities? In the ranges RestaurantDoctorAI uses, yes: rent, property charges and utilities together. If you compare with a rent-only figure from somewhere else, take the utilities out of your own figure before you compare. The National Restaurant Association, for one, counts rent, taxes and property insurance as occupancy, and lists utilities on their own line. On that basis, its [figures for 2024](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-occupancy-costs-were-more-than-5-of-sales-in-2024/) put median occupancy at 5.7% of sales for full-service restaurants and 5.2% for limited service, meaning counter service. [ENERGY STAR](https://www.energystar.gov/products/ask-the-experts/how-cut-utility-costs-your-commercial-kitchen) puts a restaurant's energy costs alone at about 3–5% of sales. That is why the ranges here, with utilities in, run higher than rent-only figures. ### What if I don't know my rent percentage? Divide last month's rent, property charges and utilities by last month's sales. Above the top of your format's range is a warning sign: 10% for quick service, fast casual and cafés, 12% for the rest. If your restaurant is busy but still not making money, rent is one of [the six places to check](https://restaurantdoctorai.com/guides/restaurant-busy-but-not-making-money). --- title: "How many covers a night do I need to break even?" description: "Break-even sales ÷ average spend per guest ÷ nights open: a worked example, what paying yourself adds, and a check of the covers against your seats." canonical: "https://restaurantdoctorai.com/guides/restaurant-covers-to-break-even" published: "2026-10-14" updated: "2026-10-14" --- # How many covers a night do I need to break even? Divide the sales you need each month to break even by what the average guest spends, then by the nights you're open. That's the covers you need each night to pay every bill. For the example restaurant below, it's 61 covers a night, or 70 once the owner pays themselves $4,000 a month. The quick answer Divide your break-even sales for a month by your average spend per guest, then by the nights you're open. Covers a night = break-even sales a month ÷ average spend per guest ÷ nights open a month | Example: $45 a guest, 6 nights a week | Covers a night | | --- | --- | | No pay for the owner | 61 | | With $4,000 a month for the owner | 70 | Example figures, worked out the way the calculator does. [Work out yours with the free break-even calculator](https://restaurantdoctorai.com/calculators/break-even), no sign-up. The short version - Work out break-even sales for a month first, then turn them into covers. - Use net sales ÷ covers from the same month as your average spend. - Count a fair wage for your own hours as a bill, or the number comes out too low. - Divide covers by seats to see how many times each seat has to fill a night. ## The formula, step by step A cover is one guest served. You need three numbers, all from the same typical month. 1. **Break-even sales a month.** Add up the bills that stay the same whatever you sell: rent and utilities, salaries, insurance, software and loan payments. Then work out the share of each sale that the costs rising with sales take: food and drink, card fees, app commissions, packaging, and hourly pay if you add and cut shifts with the rush. The [SBA's break-even page](https://www.sba.gov/business-guide/plan-your-business/calculate-your-startup-costs/break-even-point) divides the first by what's left of each sale, and calls the result an estimate. 2. **Average spend per guest.** Net sales for the month ÷ covers for the month, both from your POS. Net sales means after discounts and comps, without sales tax and tips. 3. **Nights open a month.** Nights open a week × 52 ÷ 12. Six nights a week is 26 nights a month. *Break-even sales a month* Break-even sales = bills that stay the same ÷ (1 − costs that rise with sales ÷ net sales) Write the share as a decimal: 62.5% is 0.625. *Covers you need a night* Covers a night = break-even sales a month ÷ average spend per guest ÷ nights open a month If you serve lunch too, use days open and the covers for the whole day. Round up. A night that falls one cover short of the number is a night under break-even. ## A worked example This is the break-even calculator's opening example, so you can check every figure there. A full-service restaurant sold $90,000 last month, is open 6 nights a week, and its guests spend $45 on average. Its hourly team is added and cut with the rush, so hourly pay counts with the costs that rise with sales. The figures are round examples, not industry data. *Example: one typical month* - Net sales: $90,000 - Costs that rise with sales (62.5%): $56,250 - Bills that stay the same: $26,400 - Break-even: $26,400 ÷ (1 − 0.625): $70,400 - ÷ Average spend per guest: $45 - ÷ Nights open a month (6 × 52 ÷ 12): 26 **Covers a night, rounded up:** 61 $70,400 ÷ $45 ÷ 26 = 60.2, so 61 covers, or $2,708 of sales a night. The costs that rise with sales are food and drink purchases of $28,800, hourly pay of $22,500 and $4,950 of card fees, app commissions and packaging. If about the same team works every shift, hourly pay is a bill that stays the same instead: break-even rises to $78,240 a month and the covers to 67 a night. The calculator works out both when you're not sure which fits. ## What paying yourself does to the number If you work in the restaurant without a wage, the bills that stay the same have nothing in them for your hours, and break-even is the point where you work for free. Put a number on your hours: what you'd pay a manager to do your job. Say it's $4,000 a month in this example. | Example | Without the owner's pay | With $4,000 for the owner | | --- | --- | --- | | Bills that stay the same | $26,400 | $30,400 | | Break-even a month | $70,400 | $81,067 | | Sales a night open | $2,708 | $3,118 | | Covers a night at $45 | 61 | 70 | *Example figures, worked out the way the break-even calculator does, with hourly pay rising with sales and 6 nights a week.* $4,000 more in monthly bills takes 9 more covers a night. If you're already on payroll at a fair wage, your pay is in the bills, so don't add it twice. The break-even calculator takes your own pay as its own line, so you can see the number with it and without it. The [profitability test](https://restaurantdoctorai.com/guides/is-my-restaurant-profitable#the-test) uses the same idea: pay yourself first in the math. ## Check it against your seats *Turns a night* Turns needed a night = covers a night ÷ seats A turn is one seating: every seat in the room filled once. Say the example restaurant has 50 seats. 61 covers is 1.2 turns a night: every seat filled once, and some a second time. With the owner paid, 70 covers is 1.4 turns. That has to happen on an average night, every night it's open. Before you compare it with your own room: - Take takeout and delivery covers out first. They don't need a seat. - Seats are rarely all full. A table for four with two guests leaves two seats empty, so your room seats fewer covers in a turn than it has seats. - Look at your busy nights and your slow ones, not only the average. The covers have to come on the nights you're open. Then work out what your room really does: covers a night from your POS for the last four weeks, divided by your seats. If break-even needs more turns than your room manages on a good night, more guests alone won't close the gap, and the costs, the average spend or the bills are where to look. ## What moves the number - **Bills that stay the same:** in the example, each $1,000 a month off them lowers break-even by $2,667 a month, about 2 covers a night. - **Costs that rise with sales:** one point of sales off them, such as food and drink from 32% to 31% of sales, lowers break-even by $1,829 a month. - **Average spend:** at the same cost shares, $5 more per guest takes the 61 covers down to 55. Each of these is an example, not a forecast: change one number in the calculator and it works out the rest. The [break-even post](https://restaurantdoctorai.com/blog/restaurant-break-even-point) goes through the same month in more detail. [Free calculatorWork out yours: break-even calculatorYour month's sales and costs, then break-even a month, a week, a day open and guests a day.](https://restaurantdoctorai.com/calculators/break-even) 45-second profit check See where your restaurant stands Answer 12 quick questions about how your restaurant runs and get its grade and its biggest estimated loss, free, in about 45 seconds, so you know which cost to look at first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Common questions ### Is a cover the same as a guest? Usually, yes: one cover is one guest served. For takeout, count orders instead, and use sales ÷ orders as the average spend. ### Should the average spend include tax and tips? No. Use net sales, after discounts and comps and without sales tax and tips: the same sales figure your break-even uses. ### What if my rent is the problem? Rent is one of the bills that stay the same, so it moves the covers you need. See [how much rent a restaurant can afford](https://restaurantdoctorai.com/guides/restaurant-rent-percentage). --- title: "Restaurant formulas: food cost, labor, prime cost and more" description: "Every formula a restaurant owner uses, in plain text you can copy: food cost, labor, prime cost, pour cost, rent, break-even and margin, each linked." canonical: "https://restaurantdoctorai.com/guides/restaurant-formulas" published: "2026-10-09" updated: "2026-10-15" --- # Restaurant formulas: food cost, labor, prime cost and more Every restaurant number below is a share of the same thing: net sales, after discounts and comps, before tax and tips, for the same dates as the costs. Each formula is plain text you can copy, with what it tells you and a link to the free calculator or guide that works it out with your numbers. ## The four to check every week If you track nothing else, track these on Monday for the week before. Together they show whether food, labor or the schedule is where the money is going. ### Prime cost percentage Prime cost % = (cost of goods sold + total labor cost) ÷ net sales × 100 Your two biggest controllable costs together: food and drink plus your team. Rent isn't part of it. Aim for: at or under the ceiling for your type of restaurant, from 57% or less for a bar to 70% or less for fine dining (targets RestaurantDoctorAI uses). [Prime cost calculator](https://restaurantdoctorai.com/calculators/prime-cost) ### Food cost percentage Food cost % = (beginning inventory + purchases − ending inventory) ÷ food sales × 100 The share of food sales spent on what you actually used, not what you bought. Count drinks separately. Aim for: the National Restaurant Association's [2024 medians](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/) were 32.0% of sales for full-service restaurants and 32.4% for limited service. [How to calculate and fix food cost percentage](https://restaurantdoctorai.com/guides/food-cost-percentage) ### Labor cost percentage Labor cost % = (wages + payroll taxes + benefits) ÷ net sales × 100 What your team costs as a share of sales, counted the full way: managers' salaries and payroll taxes in, tips out. Aim for: 26–31% at quick service up to 34–38% at fine dining (targets RestaurantDoctorAI uses). The NRA's [2024 median](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/) for profitable full-service restaurants was 34.2%. [Labor cost calculator](https://restaurantdoctorai.com/calculators/labor-cost) ### Sales per labor hour (SPLH) Sales per labor hour = net sales ÷ labor hours worked What each clocked hour brings in, the number a manager can check mid-shift. Aim for: above your cut line (below, under Labor). [Sales per labor hour (SPLH)](https://restaurantdoctorai.com/guides/sales-per-labor-hour) 45-second profit check Rather not work these out one by one? Answer 12 quick questions and get your restaurant's grade and its biggest estimated loss, free, in about 45 seconds, so you know which of these numbers to start with. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Food and drink formulas ### Cost of goods sold Cost of goods sold = opening stock + purchases − closing stock What you used in the period. If stock stays about the same week to week, purchases alone are a fair stand-in. [Food cost percentage guide](https://restaurantdoctorai.com/guides/food-cost-percentage) ### Plate cost percentage Plate cost % = ingredient cost of one plate ÷ its menu price × 100 Whether one dish is priced right, every component on the plate counted at its usable weight. [Food cost calculator](https://restaurantdoctorai.com/calculators/recipe-cost) ### Menu price from a target Menu price = plate cost ÷ target food cost (as a decimal) The price that puts a dish at your target. At a $4.20 plate and a 30% target, that's $4.20 ÷ 0.30 = $14.00. [Food cost calculator](https://restaurantdoctorai.com/calculators/recipe-cost) ### What each dish leaves you Contribution per dish = menu price − plate cost The dollars each sale leaves to pay for labor, rent and profit. A cheap plate isn't always the one that earns most. [Food cost calculator](https://restaurantdoctorai.com/calculators/recipe-cost) ### Pour cost percentage Pour cost % = (opening bar stock + bar purchases − closing bar stock) ÷ bar sales × 100 The bar's version of food cost: what you poured as a share of what the bar sold. Aim for: 18–24% of bar sales, the range RestaurantDoctorAI uses for US bars. It comes from BevSpot's 2016 figures for its customers (a software company); no independent survey publishes one. [Bar pour cost guide](https://restaurantdoctorai.com/guides/bar-pour-cost) ### Pour cost of one drink Pour cost % = cost of the pour ÷ drink price × 100 Whether one drink is priced right. Cost per pour = bottle cost ÷ pours per bottle. [Bar pour cost guide](https://restaurantdoctorai.com/guides/bar-pour-cost) ### What one point is worth One point = yearly net sales × 1% What one point of food cost, labor or waste is worth in dollars a year. On $900,000 of sales, one point is $9,000. [What food waste costs a restaurant](https://restaurantdoctorai.com/blog/restaurant-food-waste-cost) ## Labor formulas ### Total labor cost Total labor cost = gross pay + the employer's payroll taxes + benefits The full cost of the team, not the pay stub. Gross pay covers hourly and salaried staff, managers included; tips stay out. [What counts as labor cost](https://restaurantdoctorai.com/guides/restaurant-labor-cost#what-counts-as-labor) ### Cut line Cut line = average hourly labor cost ÷ labor ceiling (as a decimal) The sales per labor hour below which labor runs over your ceiling: the moment to send the first person home. At $17.50 an hour and a 35% ceiling, it's $50. [Work out your cut line](https://restaurantdoctorai.com/calculators/labor-cost#cut-line) ## Sales, profit and break-even ### Net sales Net sales = guests × average check Whether a slow month is fewer guests or smaller checks: they need different fixes. [Restaurant sales down? Where to look first](https://restaurantdoctorai.com/blog/restaurant-sales-down) ### Average check Average check = net sales ÷ guests What each guest spends. Count guests (covers) for a dining room, orders for a counter. [Restaurant break-even calculator](https://restaurantdoctorai.com/calculators/break-even) ### Break-even sales Break-even sales = monthly bills ÷ (1 − costs that rise with sales ÷ net sales) The sales you need just to cover the bills that stay the same each month, after the costs that rise with every sale. [Restaurant break-even calculator](https://restaurantdoctorai.com/calculators/break-even) ### Profit margin Profit margin % = (net sales − every cost) ÷ net sales × 100 What's left of each dollar after everything, rent and fees included. Aim for: no single target. In the National Restaurant Association's [2025 Restaurant Operations Data Abstract](https://restaurant.org/research-and-media/media/press-releases/new-resource-from-national-restaurant-association-provides-insights-into-operational-realities/), the 2024 median before tax was 2.8% of sales for full-service restaurants and 4.0% for limited service. [What's a normal restaurant profit margin?](https://restaurantdoctorai.com/blog/average-restaurant-profit-margin) ## Rent, delivery apps and card fees ### Occupancy cost percentage Occupancy cost % = (rent + property charges + utilities) ÷ net sales × 100 Whether your space is affordable at your sales. It's separate from prime cost. Aim for: 6–10% for quick service, fast casual and cafés, 7–12% for full service, fine dining and bars (targets RestaurantDoctorAI uses). [How much rent a restaurant can afford](https://restaurantdoctorai.com/guides/restaurant-rent-percentage) ### Sales your rent needs Monthly sales needed = monthly occupancy cost ÷ target share (as a decimal) The sales a lease asks of you before you sign it. [How much rent a restaurant can afford](https://restaurantdoctorai.com/guides/restaurant-rent-percentage) ### A delivery app's real commission Real commission rate % = (menu sales before tax and tips − your payout) ÷ menu sales × 100 What an app really keeps, fees and promotions included, from your payout report. Aim for: nothing to aim for: compare it with the plan you signed. A delivered order costs 15–30% or more on the apps' published US plans. [Delivery app commissions guide](https://restaurantdoctorai.com/guides/delivery-app-commissions) ### Break-even app price Break-even app price = in-house price ÷ (1 − real commission rate) The app price that leaves you what the same dish earns in-house. [Delivery app price calculator](https://restaurantdoctorai.com/calculators/delivery-app-price) ### Card fees as a rate Effective card rate % = total card fees ÷ card sales × 100 What card processing really costs you, every fee in, from your monthly statement. Aim for: no single target: GAO's review of industry estimates found 1.0% to 4.4% of sales. [Restaurant credit card processing fees](https://restaurantdoctorai.com/blog/restaurant-credit-card-processing-fees) ## Common questions ### Does prime cost include rent? No. Prime cost is food and drink plus labor; rent is occupancy, a line of its own. See [why rent isn't in prime cost](https://restaurantdoctorai.com/guides/restaurant-prime-cost#does-prime-cost-include-rent). ### What is the 30/30/30 rule for restaurants? A rule of thumb: about 30% of sales to food and drink, 30% to labor, 30% to everything else, and 10% left as profit. No published source backs it, so check your own numbers with the formulas above. [How it compares with real restaurants](https://restaurantdoctorai.com/guides/restaurant-prime-cost#30-30-30-rule). ### Do food cost and pour cost both go into prime cost? Yes. Prime cost counts all the food and drink you used, kitchen and bar together. Food cost and pour cost split it so you can see which side is running high. ### Should I use net or gross sales? Net sales, for every formula on this page: what you rang up after discounts and comps, before tax and tips. Gross sales are a bigger number, so every cost looks smaller against them than it is. --- title: "Can I check restaurant numbers without POS or bank access?" description: "Yes: a first read of your restaurant's costs needs no POS or bank login. What a no-login check can and can't see, and what to ask before connecting a tool." canonical: "https://restaurantdoctorai.com/guides/check-restaurant-numbers-without-pos-or-bank" published: "2026-10-14" updated: "2026-10-14" --- # Can I check restaurant numbers without POS or bank access? Yes. You can get a useful first read of your restaurant's numbers from figures you already know: your monthly sales, the share of them that food and drink, labor and rent take, and how you run a few things. No POS, bank or accounting login is needed. What you get is an estimate that shows where to look first, not an audit. The short version - A no-login check shows which cost is most likely out of line, as an estimated range. - It can't see item sales, voids, comps or clock-in hours. Connecting a POS or accounting tool adds that detail. - Before you give any tool access, ask what it reads, whether it's read-only, where the data is kept, how to disconnect and whether it's sold. ## Start with the numbers you already know You need four figures for the same month, from papers you already have. 1. Last month's sales, after discounts and comps, from your end-of-month sales summary. 2. What food and drink cost you that month, from supplier invoices, adjusted for stock on hand if you count it. 3. Total labor: wages, payroll taxes and benefits, from your payroll summary. 4. Rent, the charges on top of rent, and utilities, from the lease and the bills. *Share of sales* Cost share % = cost ÷ net sales × 100 Prime cost % = (food and drink + total labor) ÷ net sales × 100 Use the same month for every cost and for sales. ## A worked example at $60,000 a month A full-service restaurant sells $60,000 in a month. Food and drink cost $21,600 and total labor $19,800. *One month, full service* - Net sales: $60,000 - Food and drink (36%): $21,600 - Total labor (33%): $19,800 - Prime cost: $41,400 **Prime cost % of sales:** 69% RestaurantDoctorAI's targets for full service are food and drink at 30–34% of sales, labor at 30–35%, and a prime cost of 65% or less. This restaurant's prime cost is 4 points over that ceiling. At $60,000 in sales, one point is $600 a month, so 4 points are worth $2,400 a month, or $28,800 a year. Food and drink is the half to look at first: at 36%, it's 2 points over the top of its range, about $1,200 a month. Labor, at 33%, is inside its range. That took four figures and no login. What it can't tell you is why food and drink runs at 36%. [Free calculatorWork out yours: prime cost calculatorYour prime cost from your own sales, food and drink, and labor.](https://restaurantdoctorai.com/calculators/prime-cost) ## What a no-login check can see, and what it can't A check that never connects sees only what you tell it: your shares of sales, or the ranges they fall in, and how you run things. It has no transaction detail, so it can't see: - Your item mix: which dishes sell, and which carry a thin margin. - Voids, comps and discounts, by server or by shift. - Exact hours: who was clocked in while sales were slow. - Price changes on each line of each invoice. So its results are estimated ranges from your answers, not an audit of your books. If an answer is off, the estimate is off too. Use it to find where to look, then check that area against your own records. ## When connecting your POS or accounting tool is worth it Tools that connect to your POS, payroll or accounting software read the transactions themselves. They can show sales by item and by hour, voids and comps, labor against sales hour by hour, and a profit and loss statement that stays current. Connecting is usually worth it when: - You want to track food, labor and sales every week, not once. - You need item-level detail to reprice or trim a menu. - You suspect voids, comps or missing cash, and need the transaction log to see it. - Your bookkeeping is current, so the numbers a tool reads are right. Many owners do both, in order: a no-login read first to find the area, then a connected tool or their accountant for the detail in that area. ## Five questions to ask before you give any tool access These apply to any app or service that asks to connect to your POS, payroll, accounting or bank account. They follow the Federal Trade Commission's [cybersecurity advice for small businesses](https://www.ftc.gov/business-guidance/small-businesses/cybersecurity) on working with vendors. 1. **What data will it read?** Ask for the list. The FTC's advice is to limit a vendor's access to what it needs, and only for as long as it needs it. 2. **Is the access read-only?** Reading data and changing it are often separate permissions. Square, for example, [lists reading payments and writing them](https://developer.squareup.com/docs/oauth-api/refresh-revoke-limit-scope) as two permissions an app can ask for. A tool that only reports on your numbers shouldn't need to change them. 3. **Where is the data kept, and for how long?** Ask where it's stored, how long it's kept and how it's deleted when you leave. The FTC lists these among the terms to settle with a vendor. 4. **How do I disconnect it?** Find out before you connect. In Square, connected apps are under Settings, then App integrations, and each one has a [Disconnect App option](https://squareup.com/help/us/en/article/5437-manage-your-square-app-marketplace-subscriptions). In Toast, staff with the Manage Integrations permission can [remove a connected integration](https://doc.toasttab.com/doc/devguide/apiPartnersGettingAccessibleRestaurants.html) from the My Integrations page in Toast Web. 5. **Is my data sold or shared?** Look for the answer in the tool's privacy policy, or ask for it in writing. The FTC lists how a vendor can use, share or sell your data among the terms to decide on. A clear answer to all five is a good sign. A vague one is a reason to ask again before you connect. ## How RestaurantDoctorAI's free check works RestaurantDoctorAI's free check is a no-login check. It asks 12 multiple-choice questions: your type of restaurant, your monthly sales range, your prime cost as a range (or that you don't track it), the share of sales that goes to rent, property costs and utilities, and how you handle recipe costing, waste, drink pours, slow shifts, delivery apps, guest follow-up and reviews, and how full the dining room is across the week. None of them asks for a POS, bank or accounting login, or for your name or email. On our website, the free result is worked out in your browser. As our [privacy policy](https://restaurantdoctorai.com/privacy) says, your answers are sent to us only if you continue to checkout or use a prepaid code, and two of them (your type of restaurant and sales range) only if you ask to be introduced to a partner company. The result shows a grade, an estimate of the money lost to avoidable costs as a range, a month and a year, and the costliest issue or opportunity to look at first. [How it works](https://restaurantdoctorai.com/how-it-works) shows each step, and [our methodology](https://restaurantdoctorai.com/methodology) shows how the estimates are worked out. 45-second profit check Get a first read with no logins Answer 12 quick questions and see your restaurant's grade and its biggest estimated loss, free, in about 45 seconds. No account, and no POS or bank login. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Common questions ### Is it safe to use a restaurant audit tool that doesn't connect to my POS? A tool that never connects can't read or change your transactions, so what's left to check is what it keeps of the answers you type in. Read its privacy policy for that, and for whether anything is sold, as you would for any tool. Its limit is detail: it gives estimates, not an audit. ### Can I do a financial health check without bank access? Yes. Prime cost, food and drink cost, labor and rent as shares of sales come from your sales summary, invoices, payroll summary and lease. Compare each with the target for your type of restaurant, then look closer at the one furthest over. See [the targets by restaurant type](https://restaurantdoctorai.com/guides/restaurant-prime-cost#prime-cost-targets). --- title: "Do restaurants have to list calories on the menu?" description: "Most restaurants don't have to list calories. Who must in the US, England and Ontario, what the menu must show, and a free counter for a dish's calories." canonical: "https://restaurantdoctorai.com/guides/restaurant-menu-calories" published: "2026-10-02" updated: "2026-10-15" --- # Do restaurants have to list calories on the menu? Most independent restaurants don't have to put calories on the menu. The rules cover chains and large businesses: in the US, chains with 20 or more locations; in England, businesses with 250 or more employees; in Ontario, chains with 20 or more locations in the province. If you show calories anyway, they can't mislead your guests, and the free counter below works them out from your recipe. Free calculator ## Work out a dish's calories Add what goes into one dish and get its calories for the menu. Nothing you type leaves this page. Where's the menu?USCanada (Ontario)UK, Ireland or EUUS menus round to the nearest 5 up to 50 calories and to the nearest 10 above. We round once, at the end. Example numbers. Replace them with yours. Dish name (optional) 1. **Ingredient 1** Food (ingredient 1) USDA: Beef, ground, 80% lean meat / 20% fat, patty, cooked, pan-broiled · FDC 171798 Amount (ingredient 1) Unit (ingredient 1) Pan-broiled, a patty keeps about 73% of its raw weight (USDA cooking yield). Enter the weight after cooking. Weighed it raw? Pick “Ground beef 80/20, raw”. Calories in this ingredient**305.46** 2. **Ingredient 2** Food (ingredient 2) USDA: Rolls, hamburger or hotdog, plain · FDC 172796 Amount (ingredient 2) Unit (ingredient 2) Calories in this ingredient**122.76** 1 each = 44 g 3. **Ingredient 3** Food (ingredient 3) USDA: Cheese, cheddar (Includes foods for USDA's Food Distribution Program) · FDC 173414 Amount (ingredient 3) Unit (ingredient 3) Calories in this ingredient**112.84** 1 each = 28 g 4. **Ingredient 4** Food (ingredient 4) USDA: Catsup · FDC 168556 Amount (ingredient 4) Unit (ingredient 4) Calories in this ingredient**17.17** 1 tbsp (US) = 17 g 5. **Ingredient 5** Food (ingredient 5) USDA: Lettuce, iceberg (includes crisphead types), raw · FDC 169248 Amount (ingredient 5) Unit (ingredient 5) Calories in this ingredient**1.12** 1 each = 8 g 6. **Ingredient 6** Food (ingredient 6) USDA: Tomatoes, red, ripe, raw, year round average · FDC 170457 Amount (ingredient 6) Unit (ingredient 6) Calories in this ingredient**3.6** 1 each = 20 g - Enter what goes on the plate, after trimming. Weigh it if you can: cups of flour vary a lot. - Pick the item that matches how you weighed it, raw or cooked. A burger patty loses about a quarter of its weight in the pan, and some of that is fat. - Fried? Add the oil it soaks up as its own row. Weigh the fryer oil before and after a batch, and the difference is what the food took in. - Not on the list? Pick “Other: from the pack label” and type the kcal per 100 g from its nutrition panel. How do you serve it?By the plateBy weight (ladle or scoop)Making a batch, like a tray of 12 pancakes? Count the plates it makes. Makes how many plates? 1 for a single plate. ### With the example numbers **560** calories a plate Exactly 562.95 before rounding. US menus round to the nearest 10 above 50 calories. An estimate from USDA nutrition data and your recipe. It isn't a lab test. Nutrient values: USDA FoodData Central (public domain). FDC ID shown for each ingredient. How we worked it out - Dish: Burger plate - 1. Beef, ground, 80% lean meat / 20% fat, patty, cooked, pan-broiled (FDC 171798): 124.17 g × 246 ÷ 100 = 305.46 - 2. Rolls, hamburger or hotdog, plain (FDC 172796): 1 each = 44 g; × 279 ÷ 100 = 122.76 - 3. Cheese, cheddar (Includes foods for USDA's Food Distribution Program) (FDC 173414): 1 each = 28 g; × 403 ÷ 100 = 112.84 - 4. Catsup (FDC 168556): 1 tbsp (US) = 17 g; × 101 ÷ 100 = 17.17 - 5. Lettuce, iceberg (includes crisphead types), raw (FDC 169248): 1 each = 8 g; × 14 ÷ 100 = 1.12 - 6. Tomatoes, red, ripe, raw, year round average (FDC 170457): 1 each = 20 g; × 18 ÷ 100 = 3.6 - Whole recipe: added up = 562.95 calories - Rounded for a US menu: 562.95 → 560 - Nutrient data: USDA FoodData Central, SR Legacy (April 2018) ### Before you print - Most independent restaurants don't have to post calories. The US rule covers chains with 20 or more locations under the same name and substantially the same menu, and restaurants that register with FDA; New York City's covers chains with 15 or more. - A menu that has to post them rounds as above and carries two statements: “2,000 calories a day is used for general nutrition advice, but calorie needs vary.” and “Additional nutrition information available upon request.” - Keep your recipe and this breakdown, with the database and the FDC IDs. FDA can ask a covered chain for them. - Calories you show must not mislead, even if you post them by choice, and words like “low calorie” are claims with rules of their own. It opens on a burger plate, worked through in [how to work out a dish's calories](#work-out-calories). Change the rows to your recipe and pick where you trade: it rounds the way that country's menus do, and lists each ingredient with its USDA record for your files. Two things it leaves to you. It doesn't take off alcohol that cooks out of a sauce: USDA's [retention table](https://www.ars.usda.gov/ARSUserFiles/80400535/Data/retn/retn06.pdf) leaves 25% after an hour of simmering and 5% after two and a half hours, so a wine sauce comes out a little high, by 6% in our example. And it gives kcal only: [MenuCal](https://www.gov.uk/government/publications/menucal-calorie-and-allergen-tool), a free government tool, gives kJ too. The short version - US: chains with 20 or more locations have to post calories, and New York City's rule starts at 15. Any other restaurant can opt in by registering with FDA. - England: businesses with 250 or more employees. Ontario: chains with 20 or more locations there. Scotland, Wales, Northern Ireland, Ireland and the rest of the EU: voluntary. - Calories you choose to show still can't mislead. Keep the recipe and the sums behind every number. - Calories per plate = each ingredient's grams × its calories per 100 g ÷ 100, added up, ÷ the plates the recipe makes. - The big errors come from the kitchen, not the math: meat weighed raw, frying oil left out, a sauce divided by its raw weight. ## The short answer, country by country - **United States:** chains with 20 or more locations under the same name and substantially the same menu, plus any restaurant that [registers with FDA](https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-101/subpart-A/section-101.11). - **New York City:** chains with [15 or more locations nationally](https://www.nyc.gov/assets/doh/downloads/pdf/about/healthcode/health-code-article81.pdf). - **England:** businesses with [250 or more employees](https://www.legislation.gov.uk/uksi/2021/909/contents). - **Ontario:** chains with [20 or more locations in Ontario](https://www.ontario.ca/laws/statute/15h07). - **New South Wales:** chains with 20 or more outlets in the state or 50 nationally, and [in kilojoules](https://www.foodauthority.nsw.gov.au/retail/cafes-restaurants-and-retail-outlets/kJs-on-menus). - **Scotland, Wales, Northern Ireland, Ireland, the rest of the EU and Canada outside Ontario:** nobody. It's voluntary. Voluntary means we found no law requiring calories there. We haven't checked every EU country, Australia's other states or New Zealand, and in the US a state or city can add rules of its own. ## Who has to put calories on the menu ### United States FDA's [menu labeling rule](https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-101/subpart-A/section-101.11) has applied since [May 7, 2018](https://www.fda.gov/food/nutrition-food-labeling-and-critical-foods/menu-labeling-requirements) to restaurants in a chain of 20 or more locations “doing business under the same name” and selling substantially the same menu items. Franchised locations count toward the 20. Daily specials, dishes on the menu for less than 60 days a year and custom orders are left out. New York City's [health code](https://www.nyc.gov/assets/doh/downloads/pdf/about/healthcode/health-code-article81.pdf) draws the line at 15 locations nationally, so chains with 15 to 19 locations post calories there “even if they are not subject to federal requirements”, as the [city puts it](https://www.nyc.gov/assets/doh/downloads/pdf/permit/cal-label-faq.pdf). Below 20 locations, FDA leaves room for [states and cities](https://www.govinfo.gov/content/pkg/FR-2014-12-01/html/2014-27833.htm) to set rules of their own, so check with your local health department. ### United Kingdom In England, the [calorie labelling regulations](https://www.legislation.gov.uk/uksi/2021/909/contents) have covered businesses with 250 or more employees, part-time staff included, since April 6, 2022. The government's [guidance](https://www.gov.uk/government/publications/calorie-labelling-in-the-out-of-home-sector/calorie-labelling-in-the-out-of-home-sector-implementation-guidance) “encourages smaller food businesses” to label anyway, and says it will consider extending the rule to them. Scotland paused its plans in May 2023, and Wales was still reviewing the evidence in September 2024, [Food Standards Scotland reported](https://www.foodstandards.gov.scot/sites/default/files/migration/downloads/2_-_FSS_position_on_calorie_information_in_the_OOH_sector_-_2024_September_18_-_240902.pdf). In Northern Ireland, calories are [voluntary](https://assets.publishing.service.gov.uk/media/6a0f0aa32e20b377604ec7b6/calorie-wise-technical-guidance_july2021.pdf). ### Canada [Ontario's Healthy Menu Choices Act](https://www.ontario.ca/laws/statute/15h07) covers chains with 20 or more locations in the province. There's no federal menu rule, though the [Canadian Food Inspection Agency](https://inspection.canada.ca/en/food-labels/labelling/industry/nutrition-labelling/foods-sold-establishments) says nutrition values a restaurant does give “must be accurate and not false or misleading”. We found no other province with a rule. ### Ireland and the rest of the EU [EU law](https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:02011R1169-20250401) makes nutrition information on food sold loose optional “unless Member States adopt national measures”, and we found no country that has. Ireland drew up plans, then set them aside: in 2022 its Department of Health said there were “no plans to revisit the matter”, [TheJournal.ie reported](https://www.thejournal.ie/poll-support-calorie-labelling-menu-5769708-May2022/). ## If you post calories: what the menu has to show If a rule above covers you, it sets the format. If none does, it's a good model, and in Northern Ireland, Ireland and the rest of the EU part of it applies to voluntary calories too. ### US chains, and restaurants that register The [federal rule](https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-101/subpart-A/section-101.11) asks for: - Calories next to each item's name or price, in type at least as large as the name or price, with “Calories” or “Cal”. - Numbers rounded to the nearest 5 up to 50 calories and the nearest 10 above. Under 5 can be shown as 0. - On every menu page: “2,000 calories a day is used for general nutrition advice, but calorie needs vary.” - Also on the menu: “Additional nutrition information available upon request.” Behind it, a written sheet on site with 11 nutrients for each item, from fats and sodium to sugars and protein. - The same on your website: “menus on the Internet” count as menus. - Records FDA can ask for, due in about 4 to 6 weeks by its [guidance](https://www.fda.gov/media/108737/download): the database and its version, the recipe, what each ingredient adds, and two signed statements that the numbers are complete and the kitchen follows the recipe. Registering with FDA, in plain words A US restaurant outside the rule can opt in by [registering with FDA](https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-101/subpart-A/section-101.11) on Form 3757, renewed every other year. Everything above then applies to you. In return, state and local menu labeling rules that differ from FDA's stop applying, which can suit a small chain facing different rules in different cities. You don't need to register to put calories on your menu. ### England, at 250 or more employees The [regulations](https://www.legislation.gov.uk/uksi/2021/909/contents) and the government's [guidance](https://www.gov.uk/government/publications/calorie-labelling-in-the-out-of-home-sector/calorie-labelling-in-the-out-of-home-sector-implementation-guidance) ask for: - Calories per portion in kcal, next to the dish's description or price, with the portion size or how many people it serves. - The statement “adults need around 2000 kcal a day” on every menu page or board, word for word. Pages that only list children's food can leave it out. - The same on websites and delivery apps. - A menu without calories only for a guest who asks for one. - Calories within plus or minus 20% of the real figure, which the guidance calls “an acceptable margin”. It recommends keeping records of how you worked them out. ### Ontario, at 20 or more locations Ontario's [menu-labelling guide](https://www.ontario.ca/document/guide-menu-labelling-requirements/calorie-display-requirements) asks for: - Calories next to each item's name or price, in the same font and format and at least the same size, with “Calories” or “Cals”. - Numbers rounded to the nearest 10 above 50 calories, the nearest 5 from over 5 up to 50 and the nearest whole number below that, with 0 under half a calorie. - A set statement on daily needs, in English or French: adults and teens 13 and older “need an average of 2,000 calories a day”, children 4 to 12 an average of 1,500. Copy it word for word from the [regulation](https://www.ontario.ca/laws/regulation/160050). ### Everywhere else In Northern Ireland, Ireland and the rest of the EU, calories you show by choice must be in kJ as well as kcal ([FSA guidance](https://assets.publishing.service.gov.uk/media/6a0f0aa32e20b377604ec7b6/calorie-wise-technical-guidance_july2021.pdf), [EU rules](https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:02011R1169-20250401)). Northern Ireland also expects updates “at the next menu reprint, or within six months”, and suggests a line such as “The energy information is provided as a guide.” Wherever you are, the numbers can't mislead. In the US, food is misbranded if its labeling is “false or misleading in any particular” ([21 U.S.C. 343](https://www.govinfo.gov/content/pkg/USCODE-2023-title21/html/USCODE-2023-title21-chap9-subchapIV-sec343.htm)), and in Great Britain a misleading label is an offence ([Food Safety Act 1990](https://www.legislation.gov.uk/ukpga/1990/16/section/15)). Keep the recipe and the sums behind each number, and redo it when the recipe changes. Leave out words like “low calorie”. That's a nutrition claim, with rules of its own: in the US for any restaurant, chain or not ([21 CFR 101.10](https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-101/subpart-A/section-101.10)), and in Northern Ireland and the EU too ([FSA guidance](https://assets.publishing.service.gov.uk/media/6a0f0aa32e20b377604ec7b6/calorie-wise-technical-guidance_july2021.pdf)). ## How to work out a dish's calories Add up the calories in everything that goes into the dish, then divide by the plates it makes. You don't need a lab: FDA's [guidance](https://www.fda.gov/media/108737/download) says “You are not required to conduct laboratory analysis” and names USDA's nutrient database as a basis. England's [guidance](https://www.gov.uk/government/publications/calorie-labelling-in-the-out-of-home-sector/calorie-labelling-in-the-out-of-home-sector-implementation-guidance) points to the UK's own food tables for most food, and USDA's for food from the US. *Calories per plate* **Calories in an ingredient = grams × calories per 100 g ÷ 100:** Calories per plate = all the ingredients added up ÷ plates the recipe makes Weigh each ingredient as it goes into the dish, after trimming, and use the value for the same state: a raw value for a raw weight, a cooked value for a cooked weight. Round once, at the end. Take calories per 100 g from the pack's label where there is one. For fresh food, use [USDA FoodData Central](https://fdc.nal.usda.gov/api-guide), the US government's free food database, which is in the public domain. Its [SR Legacy](https://fdc.nal.usda.gov/download-datasets) data, the final edition of the database FDA's guidance names, is what the counter below uses. If you've worked out the dish's [plate cost](https://restaurantdoctorai.com/guides/food-cost-percentage#plate-cost-for-a-single-dish), you already have most of what you need: the same amounts, after trimming, and the same number of plates. *Example: a burger plate, from USDA values* - Beef patty, 80/20, pan-broiled, 124.2 g: 305.5 - Bun, 44 g: 122.8 - Cheddar, 1 slice, 28 g: 112.8 - Ketchup, 1 tbsp, 17 g: 17.2 - Lettuce, 1 leaf, 8 g: 1.1 - Tomato, 1 slice, 20 g: 3.6 **Calories per plate:** 563 Each line is grams × USDA's calories per 100 g ÷ 100: for the patty, 124.2 × 246 ÷ 100. The 562.95 total goes on a US menu as 560, and with UK and EU energy factors it's 554 kcal. The patty was 6 oz raw: USDA's cooking yield for a pan-broiled medium-fat patty is 73%. The [counter at the top of this page](#calorie-counter) opens on this plate. Change its rows to your recipe and pick where you trade: it rounds the way that country's menus do, and lists each ingredient with its USDA record for your files. ## The mistakes that throw the number off The sums are the easy part. These mistakes move the number far more, each measured on USDA data or our worked examples. - **Raw weight with a cooked value, or the other way round.** A 6 oz raw 80/20 beef patty is 432 calories by USDA's [raw value](https://fdc.nal.usda.gov/food-details/174036/nutrients) and 305 once pan-broiled, by its [cooked value](https://fdc.nal.usda.gov/food-details/171798/nutrients) and [cooking yield](https://www.ars.usda.gov/ARSUserFiles/80400535/Data/retn/USDA_CookingYields_MeatPoultry.pdf): it loses water and fat in the pan. Count the raw figure for the cooked burger and you overstate it by 41%. - **The wrong database item.** USDA's [fast-food fries](https://fdc.nal.usda.gov/food-details/170698/nutrients) are 312 calories per 100 g and its [fresh-cut fries](https://fdc.nal.usda.gov/food-details/2709458/nutrients) 198: 58% more for the same weight. Read the full USDA name and pick the one that matches how you cook. - **Frying oil left out.** In USDA's own recipe for fresh-cut fries, 59% of the calories come from the oil the potatoes soak up. To measure yours, weigh the fryer oil before and after frying a weighed batch, as [Health Canada](https://www.canada.ca/content/dam/hc-sc/migration/hc-sc/fn-an/alt_formats/hpfb-dgpsa/pdf/label-etiquet/guide-nutri_val_tc-tm-eng.pdf) describes, and add the difference as its own row. - **A sauce divided by its raw weight.** Simmering drives off water, not calories. If you serve by the ladle, weigh the pot after cooking. Our example sauce lost 22% of its weight, and dividing by the raw weight would have understated every ladle by the same 22%. - **Flour by the cup.** [Published weights](https://www.fao.org/4/ap815e/ap815e.pdf) for a cup of flour vary widely. In our example pancake recipe, they put one pancake anywhere from 133 to 160 calories, and USDA's cup weight gives 140. Weigh the flour once and write down the grams. - **Rounding as you go.** Round once, at the end. Each rounded line can be off by up to half a rounding step, and the errors add up. ## What professional nutrition analysis costs If you'd rather pay someone, or you're a chain that needs records FDA can check, these are the published prices. | Who does the work | Published price | | --- | --- | | [Food Label Maker](https://foodlabelmaker.com/restaurant-nutrition-analysis/): its nutritionists analyze your recipe | From $69 a recipe | | [Menu Analytics](https://www.menuanalytics.com/see-our-pricing/): recipe nutrition analysis | $150 a recipe | | [Food Lab, Inc.](https://foodlab.com/services-pricing/): restaurant menu items, from a database | $240 an item | | [Medallion Labs](https://www.medallionlabs.com/wp-content/uploads/2026/01/GMI_MedallionLabs_Catalog_2026_V1.pdf): lab test for fat, protein, moisture and ash, with calories worked out from them | $354 a sample | | Medallion Labs: full US nutrition panel | $1,656 a sample | | [MenuSano](https://www.menusano.com/pricing/): software you enter your recipes into | From $29 a month | | [NutriCalc](https://www.nutricalc.co.uk/) (UK): software you enter your recipes into | From £750 + VAT a year | | [MenuCal](https://www.gov.uk/government/publications/menucal-calorie-and-allergen-tool): government tool in Ireland, Northern Ireland and Scotland | Free | *Published prices, checked October 2, 2026. Menu Analytics' prices took effect October 1, 2024; Medallion Labs' run through December 2026.* When FDA costed its rule in 2014, it put a database analysis at $25 to $100 an item and a lab test at $650. With about 4 hours of record-keeping each, it estimated $280 to $1,030 an item, $660 on average, for an average of 117 items a chain ([FDA's cost analysis](https://www.fda.gov/media/116833/download)). For one restaurant posting calories by choice, a calculation from your own recipe, with the sums kept, is a method every rulebook in this guide allows. 45-second profit check See where your restaurant is losing money Calories tell guests what's in a dish. The free check looks at what your menu, kitchen and team leave you. Answer 12 quick questions and get your restaurant's grade and its biggest estimated loss, free, in about 45 seconds, so you know where to start. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Common questions ### Can restaurants lie about calories? No. Calories you post can't mislead, even when no rule made you post them. For chains, FDA starts by working with the restaurant and treats small slips, like extra pepperoni on a pizza, as minor ([FDA guidance](https://www.fda.gov/media/108737/download)). Beyond that, New York City can fine $200 a violation ([city FAQ](https://www.nyc.gov/assets/doh/downloads/pdf/permit/cal-label-faq.pdf)), England has a £2,500 fixed penalty ([regulations](https://www.legislation.gov.uk/uksi/2021/909/contents)), and Ontario's fines are set per day ([the Act](https://www.ontario.ca/laws/statute/15h07)). ### How accurate do menu calories have to be? The US sets no margin: FDA dropped a proposed 20% test in 2014 and judges each case on the restaurant's method and records ([final rule](https://www.govinfo.gov/content/pkg/FR-2014-12-01/html/2014-27833.htm)). England's guidance treats plus or minus 20% as acceptable. Either way, base the number on what your kitchen really serves. ### Do I need a lab test? No. FDA says you don't need one, and England, the EU and [Ontario](https://www.ontario.ca/document/guide-menu-labelling-requirements/determining-number-calories) accept a calculation from food data too. A lab test suits a dish you can't break into ingredients, and starts at $354 a sample at the lab above. ### Do calories have to be on delivery apps and online menus? Only if a rule covers you. Then yes: FDA counts menus on the internet, England's guidance includes websites and delivery apps, and [Ontario](https://www.ontario.ca/document/guide-menu-labelling-requirements/calorie-display-requirements) includes online menus. FDA's [2023 draft guidance](https://www.fda.gov/media/174528/download) also encourages chains to show calories on delivery apps, without requiring it. The rules are summarized from official sources as of the date above; check the current rules where you operate. General information, not legal, nutrition or medical advice. --- title: "How to donate your restaurant's surplus food safely" description: "How to donate restaurant surplus food: what you can give, the safety rules, what US law protects, and who can pick it up near you. Cut over-prep first." canonical: "https://restaurantdoctorai.com/guides/restaurant-food-donation" published: "2026-10-07" updated: "2026-10-07" --- # How to donate your restaurant's surplus food safely Cut over-prep first. Then, for the good food that's still left, book a regular pickup with a local food rescue group, where one collects near you. Donate only food that was never served to a guest and was kept cold or hot, or cooled quickly by the rules below. In the US, federal law protects restaurants that donate apparently wholesome food in good faith, free, to a nonprofit or straight to a person in need, unless they act with gross negligence or intentional misconduct. The finder below shows who may collect near you. The short version - Cut over-prep first. The EPA ranks preventing wasted food above donating it. - Donate only food that was never served to a guest and was held at 41°F or below or 135°F or above, or cooled within the Food Code's times. - In the US, the Bill Emerson Act protects donors who give apparently wholesome food in good faith, free, to a nonprofit that passes it on to people in need. Since January 5, 2023, it also covers restaurants giving food free, straight to a person in need. - Food rescue groups and food banks may collect your surplus when a driver is free. Call before your first donation: each one sets its own rules. - Food donations to qualifying charities may be tax-deductible. Ask your accountant. ## Cut over-prep first The cheapest surplus is the food you never make. On the EPA's [Wasted Food Scale](https://www.epa.gov/sustainable-management-food/wasted-food-scale), preventing wasted food is “the most environmentally beneficial option”. Donating good food comes next. Most restaurant surplus can't be donated anyway. [ReFED](https://refed.org/stakeholders/restaurants-and-foodservice/) counts nearly 70% of restaurant and foodservice surplus in 2024 as plate waste: food guests took or were served and didn't eat. So start with the food that never reaches a plate. 1. Log what you throw away for a week: the item, how much and why. Give over-prepped food its own reason. 2. Lower the prep amounts (pars) for the three items you over-prep most, and check next week's log. 3. Cook smaller batches in slow hours, so less is left in the hot well at close. 4. For good food that's still left, set up a regular pickup with a food rescue group (see [the finder below](#find-food-rescue)). Waste comes straight out of your food cost. To see how much it moves, read [how food cost percentage works](https://restaurantdoctorai.com/guides/food-cost-percentage#two-quiet-costs). ## What you can donate, and what you can't In the US, the [FDA Food Code](https://www.fda.gov/media/194741/download) is the model that states and counties base their restaurant rules on, often years later and with changes. It says food “stored, prepared, packaged, displayed, and labeled” by its rules “may be offered for donation.” The 2026 edition, [issued September 17, 2026](https://www.fda.gov/food/hfp-constituent-updates/fda-releases-2026-food-code), keeps that rule. In short: handle donated food exactly like food you sell. | Rule | What the Food Code says | | --- | --- | | Cold holding | 41°F (5°C) or below | | Hot holding | 135°F (57°C) or above | | Cooling cooked food | From 135°F to 70°F within 2 hours, and to 41°F within 6 hours in total | | Ready-to-eat food you made | Marked with a date if kept more than 24 hours, and kept no more than 7 days at 41°F or below. The day you make it is day 1 | | Food served to a guest | Never offered as food again | *The Food Code's numbers, 2026 edition (sections 3-501.16, 3-501.14, 3-501.17 and 3-306.14). Your health department enforces its own version.* - **Usually fine:** food that was never served and was held at safe temperatures or cooled within the Food Code's times, such as extra trays from the line, prepped items in the walk-in or soup you cooled properly. Sealed food in its original packaging. Whole fruit and vegetables. - **Never:** anything served to a guest, including half-eaten plates, returned dishes and bread baskets that went to tables. Food left out between 41°F and 135°F (5°C and 57°C) for 2 hours or more, other than while you cool it within the Food Code's times. Food past its use-by date, spoiled, moldy or in damaged packaging. - **Leave out:** alcohol, and food made with it. [Greater Pittsburgh's food bank](https://pittsburghfoodbank.org/wp-content/uploads/2025/12/1-INVTY-84-2025_Food-Recovery-Program-One-Sheets_Prepared-Food-Donation.pdf) and [Second Harvest](https://www.secondharvest.ca/get-involved/donate-food) in Canada don't take either, and [Felix](https://felix.org/get-involved/become-a-food-supplier) in the UK (formerly FareShare and The Felix Project) doesn't take alcohol. - **Ask first:** buffet or self-serve leftovers, reheated food, hot food straight from service, and raw items such as sushi. Many groups refuse them, and the Greater Pittsburgh food bank takes prepared food only if it's frozen at 0°F or below. Food banks set the label rules. The [Conference for Food Protection](https://www.ncdhhs.gov/comprehensive-resource-food-recovery-2016-versionpdf/open) suggests five things on each container: who it's going to, who it's from, what the food is, the date of donation, and an allergen statement. Check two sets of rules first Federal law doesn't override state or local health rules. Before your first donation, ask your health department and the group you'll donate to what they need. This is general information, not legal advice. ## What US law protects The [Bill Emerson Good Samaritan Food Donation Act](https://www.law.cornell.edu/uscode/text/42/1791) (42 U.S.C. 1791) says a donor isn't liable, civilly or criminally, for “the nature, age, packaging, or condition” of apparently wholesome food it donates in good faith to a nonprofit that passes it on to people in need. Apparently wholesome means food that meets all quality and labeling standards set by federal, state and local law, even if it's surplus or doesn't look good enough to sell. The protection ends at gross negligence or intentional misconduct. The law defines gross negligence as doing something, or failing to, while knowing at the time it was likely to harm someone's health or well-being. Since January 5, 2023 ([Public Law 117-362](https://www.congress.gov/117/plaws/publ362/PLAW-117publ362.htm)), a restaurant can also give food straight to a person in need and keep that protection, as long as the food is free. Before that, protection covered food given through a nonprofit such as a food bank. Selling surplus cheaply, for example through a discount app, is a sale, not a donation. The protection for food given straight to people covers food given at zero cost. An [NRDC legal fact sheet](https://www.nrdc.org/sites/default/files/2023-09/emerson-act-legal-fact-sheet-202306.pdf) from June 2023 says no court case related to food donation liability exists. Some states protect donors even more: ReFED's [Policy Finder](https://policyfinder.refed.org/) shows which, state by state. In California, restaurants with [250 or more seats, or 5,000 square feet or more](https://www.law.cornell.edu/regulations/california/14-CCR-18982), have had to donate the most edible food they can [since January 1, 2024](https://www.law.cornell.edu/regulations/california/14-CCR-18991.3). They need a written agreement with a food recovery group and must [keep records](https://calrecycle.ca.gov/organics/slcp/foodrecovery/donors/) (SB 1383). ## Taxes: ask your accountant Food donations to qualifying charities may be tax-deductible. Ask your accountant. One thing to know before you ask: [the IRS says](https://www.irs.gov/publications/p526) you can't deduct contributions to specific individuals, including people in need, so food handed out at the back door doesn't count. Whether a deduction helps you at all depends on how your business is taxed, and only your accountant can tell you that. Keep a record of every pickup: the date, what you gave, how much, the group, and any receipt it gives you, so your accountant has the records to look at. Outside the US, tax rules for donated food are set country by country, so ask your accountant there too. ## Find food rescue near you Food rescue groups and food banks send someone to collect surplus food and take it to pantries, shelters and meal programs. Many use volunteer drivers, as [Food Rescue US](https://foodrescue.us/about/faqs/) does, so pickup depends on who's free. Food rescue finder Pick where your restaurant is to see groups that may collect surplus food, so you know who to call first. Where is your restaurant?United StatesUnited KingdomCanadaAustraliaIrelandAnother country Pick a country to see who may collect near you. ### How to start 1. Pick one group from the finder and call it, or use its sign-up form. 2. Ask what it takes, the smallest amount it collects, its pickup days and times, and whether it gives a receipt. 3. Set a regular pickup on the same days each week, rather than calling only when you have extra. 4. Show the closing team what goes in the donation bin, how to cool it and how to label it. 5. Log every pickup apart from your waste, with the date, the amount and the group. If no group near you can collect, ask your local food bank whether you can drop food off, and how it wants it packed. Would you want your restaurant shown on a public “restaurants that donate” list? We're asking before we build anything. We count your answer, and nothing about who you are. ## Outside the US The US law above doesn't apply anywhere else, and the UK, Canada, Australia and Ireland have no single national law like it. Talk to your local food rescue charity and your food safety inspector first: between them, they know the local rules and what can be collected. - **UK:** there's [no special protection law](https://atlas.foodbanking.org/country/united-kingdom/) for food donors, so ordinary food law applies to donated food. It's [an offence](https://www.gov.uk/government/publications/food-safety-and-hygiene-guidance-for-food-banks-and-charities/food-safety-and-hygiene-guidance-for-food-banks-and-charities) to supply food after its use-by date. Food past its best before date can be passed on if its quality is still good. - **Ireland:** there's [no Good Samaritan law](https://atlas.foodbanking.org/country/ireland/) either. The [Food Safety Authority of Ireland](https://www.fsai.ie/business-advice/running-a-food-business/food-donations/food-businesses-donations/businesses-donating-food-q-and-a) says you can donate food within its use-by date if it's in acceptable condition and handled safely. It doesn't recommend donating hot food. Cooked food can be chilled or frozen to donate if it was held at 63°C or above, never reheated, and in the fridge within 2 hours of starting to cool. It also says to tell your food safety inspector before you start donating, and to keep a record of what you gave, to whom and when. - **Canada:** every province and territory has a [law protecting people who donate food](https://nzwc.ca/Documents/FoodDonation-LiabilityDoc.pdf), worded differently in each. Food sold “even for a nominal price” isn't a donation. - **Australia:** state civil liability laws [limit what a donor can be liable for](https://www.foodstandards.gov.au/business/charities), if the donor meets food safety conditions, such as making sure food is safe to eat when it leaves them. - **Other countries:** the [Global Food Donation Policy Atlas](https://atlas.foodbanking.org/), from Harvard Law School's Food Law and Policy Clinic and the Global FoodBanking Network, sums up the rules country by country. ## If someone needs food Share these with anyone who asks you for food: - In the US, call the [USDA National Hunger Hotline](https://www.fna.usda.gov/national-hunger-hotline) at 1-866-3-HUNGRY (1-866-348-6479), Monday to Friday, 8 a.m. to 8 p.m. ET, or text a question to its automated line at 914-342-7744. - Find a food bank by ZIP code with Feeding America's [food bank finder](https://www.feedingamerica.org/find-your-local-foodbank). - Call [211](https://211.org/) for local help with food and other needs. - In Canada, use Food Banks Canada's [map of food banks](https://foodbankscanada.ca/find-a-food-bank/), or call 211. - In Australia, use Foodbank's [Find Food](https://findfood.foodbank.org.au/). 45-second profit check See if waste is your costliest issue One of the free check's 12 quick questions asks how you track waste. Answer all 12 and get your restaurant's grade and its biggest estimated loss, free, in about 45 seconds, so you know whether waste is where to start. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Common questions ### Can restaurants donate leftover food? Yes, if it was never served to a guest and was kept at safe temperatures (41°F or below, or 135°F or above) or cooled within the Food Code's times. Food from guests' plates can never be donated. Check your health department's rules and the group's own rules first. ### Can a restaurant be sued for donating food? In the US, the Bill Emerson Act protects donors who give apparently wholesome food in good faith, free, to a nonprofit that passes it on to people in need, or straight to a person in need. It doesn't cover gross negligence or intentional misconduct, and it doesn't replace your local health rules. An NRDC fact sheet from June 2023 says no court case on food donation liability exists. This isn't legal advice. ### Is donating food tax-deductible? Food donations to qualifying charities may be tax-deductible. Ask your accountant. Food handed straight to people doesn't qualify: the IRS says you can't deduct contributions to specific individuals. ### Is selling surplus on a discount app the same as donating? No. A sale is a sale, even at a low price. In the US, the protection for food given straight to people covers food given free. In Canada, food sold “even for a nominal price” isn't a donation either. ### Can I donate alcohol? Leave it out. Greater Pittsburgh's food bank and Second Harvest in Canada don't take alcohol or food made with it, and Felix in the UK doesn't take alcohol. ## Sources 1. [Wasted Food Scale](https://www.epa.gov/sustainable-management-food/wasted-food-scale), US Environmental Protection Agency, September 16, 2026. 2. [Restaurants & Foodservice](https://refed.org/stakeholders/restaurants-and-foodservice/), ReFED, 2024 data. 3. [Food Code 2026](https://www.fda.gov/media/194741/download), US Food and Drug Administration, 2026 edition. 4. [The FDA Releases 2026 Food Code](https://www.fda.gov/food/hfp-constituent-updates/fda-releases-2026-food-code), US Food and Drug Administration, September 17, 2026. 5. [42 U.S.C. 1791, Bill Emerson Good Samaritan Food Donation Act](https://www.law.cornell.edu/uscode/text/42/1791), Legal Information Institute, Cornell Law School, current text, last amended January 5, 2023. 6. [Public Law 117-362](https://www.congress.gov/117/plaws/publ362/PLAW-117publ362.htm), US Congress, January 5, 2023. 7. [Federal Liability Protection for Food Donation: Legal Fact Sheet](https://www.nrdc.org/sites/default/files/2023-09/emerson-act-legal-fact-sheet-202306.pdf), NRDC, June 2023. 8. [U.S. Food Waste Policy Finder](https://policyfinder.refed.org/), ReFED. 9. [Cal. Code Regs. tit. 14, § 18982](https://www.law.cornell.edu/regulations/california/14-CCR-18982), Legal Information Institute, Cornell Law School. 10. [Cal. Code Regs. tit. 14, § 18991.3](https://www.law.cornell.edu/regulations/california/14-CCR-18991.3), Legal Information Institute, Cornell Law School. 11. [Food Donors](https://calrecycle.ca.gov/organics/slcp/foodrecovery/donors/), CalRecycle. 12. [Publication 526 (2025), Charitable Contributions](https://www.irs.gov/publications/p526), Internal Revenue Service, last reviewed April 30, 2026. 13. [Comprehensive Resource for Food Recovery Programs](https://www.ncdhhs.gov/comprehensive-resource-food-recovery-2016-versionpdf/open), Conference for Food Protection, April 2016. 14. [Prepared Food Donation Guidelines](https://pittsburghfoodbank.org/wp-content/uploads/2025/12/1-INVTY-84-2025_Food-Recovery-Program-One-Sheets_Prepared-Food-Donation.pdf), Greater Pittsburgh Community Food Bank, December 2025. 15. [Donate Food](https://www.secondharvest.ca/get-involved/donate-food), Second Harvest. 16. [Become a food supplier](https://felix.org/get-involved/become-a-food-supplier), Felix. 17. [FAQs](https://foodrescue.us/about/faqs/), Food Rescue US. 18. [Global Food Donation Policy Atlas](https://atlas.foodbanking.org/), Harvard Law School Food Law and Policy Clinic and the Global FoodBanking Network. 19. [United Kingdom](https://atlas.foodbanking.org/country/united-kingdom/), Global Food Donation Policy Atlas, April 2021. 20. [Ireland](https://atlas.foodbanking.org/country/ireland/), Global Food Donation Policy Atlas, June 2025. 21. [Food safety and hygiene guidance for food banks and charities](https://www.gov.uk/government/publications/food-safety-and-hygiene-guidance-for-food-banks-and-charities/food-safety-and-hygiene-guidance-for-food-banks-and-charities), Food Standards Agency, August 28, 2026. 22. [Businesses donating food: Q and A](https://www.fsai.ie/business-advice/running-a-food-business/food-donations/food-businesses-donations/businesses-donating-food-q-and-a), Food Safety Authority of Ireland. 23. [Food Donation and Civil Liability in Canada](https://nzwc.ca/Documents/FoodDonation-LiabilityDoc.pdf), National Zero Waste Council, April 2018. 24. [Information for charities and community groups](https://www.foodstandards.gov.au/business/charities), Food Standards Australia New Zealand, October 1, 2025. 25. [National Hunger Hotline](https://www.fna.usda.gov/national-hunger-hotline), US Department of Agriculture, September 17, 2025. 26. [Find Your Local Food Bank](https://www.feedingamerica.org/find-your-local-foodbank), Feeding America. 27. [211](https://211.org/), United Way. 28. [Find a Food Bank](https://foodbankscanada.ca/find-a-food-bank/), Food Banks Canada. 29. [Find Food](https://findfood.foodbank.org.au/), Foodbank Australia. Every rule and source on this page was checked on October 7, 2026. The rules are summarized from official sources as of the date above. Check the current rules where you operate, and with the group you donate to. General information, not legal or tax advice. --- title: "What does a restaurant have to post and keep?" description: "The federal posters, health permit and records a US restaurant has to post and keep, with free official sources and what your state and county add." canonical: "https://restaurantdoctorai.com/guides/restaurant-required-posters-and-records" published: "2026-10-14" updated: "2026-10-15" --- # What does a restaurant have to post and keep? A restaurant posts the federal labor law posters where employees can see them, its health permit where guests can see it, and a wash-hands sign at every staff hand sink. It keeps payroll and time records, Forms I-9, tip records and a short list of Food Code records for the lengths of time below. The posters come straight from the agencies that require them, and your state and county add their own rules. The quick answer Post the federal labor posters, your health permit and a wash-hands sign. Keep pay, hiring, tip and a few food safety records for the time each rule gives. Post it where people can see it. Keep it for as long as its rule says. | Post or keep | What the source says | | --- | --- | | Federal labor posters | From the agencies | | Health permit | Where guests can see it | | Payroll records | 3 years | | Time cards | 2 years | | Shellstock tags | 90 days | Federal and Food Code rules, read on the agencies' pages on October 14, 2026. Your state and county rules are what apply. The short version - The federal labor posters come straight from the Department of Labor, the EEOC and OSHA. DOL's and OSHA's are free, and OSHA says not to pay a third-party vendor for its poster. - Display your health permit where guests can see it, a wash-hands sign at each staff hand sink, and a consumer advisory on menus that list raw or undercooked animal foods. - Keep payroll records 3 years, time cards 2 years, and each Form I-9 for 3 years after hire or 1 year after the job ends, whichever is later. - Daily temperature, cooling, sanitizer, cleaning and pest logs are good practice under the model Food Code, not requirements of it. Your state code and county rules decide. ## Restaurant labor law posters: the federal ones The Department of Labor's [workplace posters page](https://www.dol.gov/general/topics/posters) lists each federal notice with its revision date. These are the ones that matter to most restaurants: | Notice | Who gives or posts it | | --- | --- | | Minimum wage (FLSA), with its tip credit part | Every employer covered by the FLSA's minimum wage rules | | Know Your Rights (EEOC) | Covered employers | | Job safety and health (OSHA) | Private employers, where workers can easily see it | | Family and Medical Leave (FMLA) | Employers with 50 or more employees in 20 or more workweeks | | Polygraph protection (EPPA) | Employers in or affecting commerce, not government | | Military service rights (USERRA) | Posted or given another way: see below | *From the agencies' own pages, read on October 14, 2026. Employers in states with their own OSHA plan use that state's version of the OSHA poster.* - **Minimum wage.** DOL says every employer of employees covered by the FLSA's minimum wage rules must post the [FLSA poster](https://www.dol.gov/agencies/whd/posters/flsa), in a place where employees can readily read it. Its current version is dated April 2023, and DOL says earlier versions, including the August 2016 one, no longer meet the requirement. The [poster itself](https://www.dol.gov/sites/dolgov/files/WHD/legacy/files/minwagebw.pdf) has a tip credit part: an employer that claims a tip credit pays a cash wage of at least $2.13 an hour, and makes up the difference if tips and that wage fall short of the minimum wage. - **Know Your Rights.** The [EEOC's page](https://www.eeoc.gov/poster) says the law requires covered employers to post it, and gives the penalty for not posting as $680, adjusted each year for inflation. The page also says its poster documents are “undergoing revision” under an executive order, so check it for a newer version before you print. - **Job safety and health.** [OSHA's page](https://www.osha.gov/publications/poster) calls it the “OSHA Cares That You Go Home Safe” poster (DOL's list still carries the older “It's the Law” title, and OSHA says an earlier version you already have is acceptable). A copy you reproduce must be at least 8.5 by 14 inches with 10-point type. - **FMLA.** Only employers with 50 or more employees in 20 or more workweeks need the FMLA poster. A smaller restaurant doesn't. - **EPPA.** DOL lists the Employee Polygraph Protection Act notice for employers engaged in or affecting commerce. It doesn't apply to federal, state or local governments. - **USERRA.** DOL says an employer may post the notice where employee notices are customarily placed, or give its full text another way, such as handing it out, mailing it or emailing it. It is the one federal notice that doesn't have to hang on a wall. You don't need to buy any of these. DOL provides its posters as free electronic copies, and its [Poster Advisor](https://webapps.dol.gov/elaws/posters.htm) helps employers meet the poster requirements of several DOL laws. OSHA says its poster “is free from OSHA” and tells owners not to pay a third-party vendor for it. The EEOC offers its poster as a download to print and post. ## State labor posters, and whether to pay a poster vendor The Poster Advisor covers federal DOL posters only. For state posters, such as a state minimum wage notice, DOL points employers to their state department of labor, and its [State Labor Offices page](https://www.dol.gov/agencies/whd/state/contacts) lists each state's labor office with its website. Posters change when rates change, so look at your state's page again when a new rate starts. The [free deadline calendar](https://restaurantdoctorai.com/deadline-calendar) has the dates for new wage rates. Companies sell laminated sets with the state and federal posters together, and some owners pay for the convenience of getting them in one order. That is a choice, not a requirement: each agency's own version is free. OSHA is the one that says so in plain words about its own poster. ## Restaurant posting requirements: what to display for guests The food rules in this section come from the FDA's Food Code. The section numbers below are the same in the [2026 edition](https://www.fda.gov/media/194741/download) and the [2022 edition](https://www.fda.gov/media/164194/download). The Food Code is a model: it binds a restaurant through the state or local code that adopts it (more on that below). - **The health permit (8-304.11).** The permit holder posts the permit in a place in the restaurant that is conspicuous to guests. - **A wash-hands sign (6-301.14).** A sign or poster telling food employees to wash their hands goes at every hand sink they use, clearly visible to them. - **The consumer advisory (3-603.11).** If an animal food such as beef, eggs, fish, lamb, milk, pork, poultry or shellfish is served raw or undercooked, tell guests about the added risk. A disclosure names or asterisks those foods, and a reminder is an asterisked footnote on the menu, a brochure, a table tent or a placard. - **An allergen notice (3-602.12).** The permit holder notifies guests in writing when unpackaged food contains a major food allergen as an ingredient. A line on the menu, a table tent or a placard can do it. - **Calorie labels, for chains.** [FDA's menu labeling rule](https://www.fda.gov/food/nutrition-food-labeling-and-critical-foods/menu-labeling-requirements) covers restaurants that are part of a chain with 20 or more locations under the same name and with substantially the same menu. Covered places show calories on menus and menu boards, two required statements, and written nutrition information on request. A restaurant that isn't covered may [register to follow the rule voluntarily](https://www.fda.gov/media/116000/download), with FDA Form 3757. See [do restaurants have to list calories on the menu](https://restaurantdoctorai.com/guides/restaurant-menu-calories). ## Required posters for restaurants vary by city and county Cities and counties add their own signs. New York City is one example. Its Health Department's [required signs page](https://www.nyc.gov/site/doh/business/food-operators/required-signs.page) lists a choking first aid poster, a poster on the availability of CPR equipment, an alcohol pregnancy warning sign, a no smoking sign and a wash-your-hands poster, and says you can get them free by calling 311. It sends owners to 311 to check whether the food allergy poster applies to them. NYC also [requires a letter grade card](https://www.nyc.gov/site/doh/business/food-operators/letter-grading-for-restaurants.page) to be posted where it can easily be seen by people passing by. That is one city. Another city or county has a different list, and some have none of these. Check your city or county health department's pages for the signs and notices it requires, and your state labor department for the labor ones. ## Where the food rules come from, and the 2026 Food Code The FDA says the Food Code is a model code that is not required. States, counties and cities adopt it, sometimes years after a new edition comes out. In its 2024 report on [adoption of the Food Code](https://www.fda.gov/food/fda-food-code/adoption-fda-food-code-state-and-territorial-agencies-responsible-oversight-restaurants-and-retail), FDA counted 11 state agencies in 7 states on the 2022 edition, and 30 state agencies in 24 states on the 2022 or the 2017 edition. What binds your restaurant is your state's food code and your county's rules, in the edition your state uses. The FDA [released the 2026 Food Code on September 17, 2026](https://www.fda.gov/food/hfp-constituent-updates/fda-releases-2026-food-code). Among its changes, it adds “a new requirement for written employee illness policies that are maintained and available,” an alternative cooling process for foods at retail, and an exception that allows double gloving in certain cases. In the 2026 text (2-501.11), the written policy covers reporting symptoms and diagnoses, and when an employee is excluded or restricted from work and when they return. It binds a restaurant once its state adopts it. ## Restaurant record keeping requirements: pay, hiring and tips | Record | How long | | --- | --- | | Payroll records (FLSA) | At least 3 years | | Time cards, wage rate tables, work schedules | 2 years | | Form I-9 | 3 years after the hire date, or 1 year after the job ends, whichever is later | | Daily tip record (employee) | Kept by the employee, to report tips each month | *From DOL, USCIS and the IRS, read on October 14, 2026. A state may ask for longer; check your state labor department.* - **Payroll and time records.** DOL's [Fact Sheet 21](https://www.dol.gov/agencies/whd/fact-sheets/21-flsa-recordkeeping) says employers keep payroll records for at least three years and the records used to work out wages, such as time cards and wage rate tables, for two. No particular form is required, and the records may be kept at the workplace or in a central records office. - **Tip credit records.** [Fact Sheet 15](https://www.dol.gov/agencies/whd/fact-sheets/15-tipped-employees-flsa) says an employer that takes a tip credit keeps records of the tips reported and the hours worked in tipped and non-tipped jobs, gives the tip credit notice before taking the credit (it can be oral or written), and pays out collected tips by the regular payday. - **Form I-9.** USCIS says to keep each hired employee's [Form I-9](https://www.uscis.gov/i-9-central/form-i-9-resources/handbook-for-employers-m-274/100-retaining-form-i-9) for three years after the date of hire, or one year after the date employment ends, whichever is later. - **Staff tip reports.** The IRS's [Publication 531](https://www.irs.gov/publications/p531), revised December 2024, says Form 4070 and Form 4070A are historical: the steps for keeping a daily tip record and reporting tips are now in the publication itself. An employee gives their employer a report for each month by the 10th of the next month, and doesn't report a month in which tips from one job total under $20. - **Form 8027, for large tipped places.** The IRS's [instructions for Form 8027](https://www.irs.gov/instructions/i8027) say you file it if you operate a large food or beverage establishment: in the 50 states or DC, where tipping is customary, and with more than 10 employees on a typical business day in the prior year. They count employees across all your food and beverage operations, by average hours per business day. It is filed yearly, and the instructions list the due dates. The free [deadline calendar](https://restaurantdoctorai.com/deadline-calendar) has the monthly staff tip reports and the yearly Form 8027 on it, for restaurants that answer that they take tips. ## Food safety logs and written procedures the Food Code requires Only a few records and written procedures are written into the Food Code itself. The rest of the daily paperwork in most kitchens is good practice, which the next section covers. The records and written procedures the Code requires include: - **A written clean-up procedure for vomiting and diarrhea, every restaurant (2-502.11 in the 2026 edition, 2-501.11 in 2022).** It tells staff what to do when someone vomits or has diarrhea onto a surface in the restaurant: the steps that keep it from spreading to staff, guests, food and surfaces. - **A written employee illness policy, in the 2026 edition (2-501.11).** Available when asked for, it covers the points in [the 2026 Food Code section above](#where-the-food-rules-come-from). It binds a restaurant once its state adopts the 2026 edition. - **Shellstock tags, 90 days (3-203.12).** The tag or label stays on the container until it is empty. The date the last shellfish from the container is sold or served goes on the tag, label or invoice. You keep the tags, labels or invoices for 90 calendar days from that date, in date order. - **Parasite destruction records, 90 days (3-402.12).** If you serve raw, raw-marinated, partially cooked or marinated-partially cooked fish in ready-to-eat form, the person in charge records the freezing temperature and time and keeps the record for 90 calendar days after the fish are served or sold. A written statement from the supplier that it froze the fish can stand in for your own record. - **Non-continuous cooking procedures (3-401.14).** If you partly cook raw meat, poultry, fish or eggs and finish cooking them later, you need written procedures that the health department approved in advance, kept in the restaurant. They say how the heating, cooling, holding and final cook are monitored and documented, what you do when a step is missed, and how the part-cooked food is marked and kept apart from ready-to-eat food. - **Time as a public health control (3-501.19).** If you hold food without temperature control for a limited time instead, you keep written procedures in the restaurant, ready for the health department, and you mark the food with the time it must be used or thrown out. The 4-hour version is the common one. - **HACCP records, for variances and reduced oxygen packaging (8-201.13, 8-201.14, 3-502.12).** A restaurant that cures food, smokes it to preserve it, uses additives to preserve it or sprouts seeds needs a variance from its health department first, and a HACCP plan comes with it. So does reduced oxygen packaging, as a variance or as a HACCP plan the health department has been given before the restaurant starts. The plan lists the records to keep. For cook-chill or sous vide, the Code says to keep the records that confirm cooling and cold holding times and temperatures for at least 6 months. ## Food safety logs that are good practice, not required Many kitchens keep daily temperature, cooling, receiving, sanitizer, cleaning and pest logs. The model Food Code doesn't require them. What it does require is this: - The person in charge makes sure employees monitor cooking, cooling and holding temperatures, through daily oversight (2-103.11). The Code doesn't say to write it down. - A test kit that measures the strength of the sanitizer must be provided (4-302.14). A log of the results isn't part of that rule. - Pests must be controlled by routine inspection of deliveries and of the premises (6-501.111). A record of the inspections isn't part of it. A log is how you show an inspector what you checked and when, and many owners keep them for that reason. A state or county may require some of them, and some health departments ask to see them. What applies to you is your state food code and your county's rules, so check those before you decide which logs to keep. ## Free templates, and the dates to keep Not every sheet in a restaurant is on a government list. For the daily paper that is, such as a prep list and par sheet, a cash-out sheet and a tip-out sheet, see our [free restaurant templates](https://restaurantdoctorai.com/templates). For the dates in this guide, such as new wage rates and tip reports, use the [deadline calendar](https://restaurantdoctorai.com/deadline-calendar). The tipped wage rules by state are in the [tipped minimum wage guide](https://restaurantdoctorai.com/guides/tipped-minimum-wage). 45-second profit check See where your restaurant stands Answer 12 quick questions about how your restaurant runs and get its grade and its biggest estimated loss, free, in about 45 seconds, so you know which cost to look at first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Common questions ### Do I have to pay for labor law posters? No. The Department of Labor offers its posters as free electronic copies, OSHA says its poster is free and not to pay a third-party vendor for it, and the EEOC offers its poster as a download to print and post. State posters come from your state labor department. Paying a company for a ready-made set is optional. ### Does an independent restaurant need calorie labels? The federal rule covers chains with 20 or more locations under the same name and substantially the same menu. A smaller restaurant isn't covered, and may register with FDA to follow it voluntarily. A state or city may have its own rule, so check your health department. ### Do I have to keep daily temperature logs? The model Food Code doesn't require them outside the records and procedures above. A state or county may. Your state food code and county rules are the ones that apply. ### How long do I keep a Form I-9? For three years after the hire date, or one year after the job ends, whichever is later. ## Sources 1. [Workplace Posters](https://www.dol.gov/general/topics/posters), US Department of Labor. 2. [Fair Labor Standards Act (FLSA) Minimum Wage Poster](https://www.dol.gov/agencies/whd/posters/flsa), US Department of Labor, revised April 2023. 3. [Employee Rights Under the Fair Labor Standards Act (poster, form WH1088)](https://www.dol.gov/sites/dolgov/files/WHD/legacy/files/minwagebw.pdf), US Department of Labor, REV 04/23. 4. ["Know Your Rights: Workplace Discrimination is Illegal" Poster](https://www.eeoc.gov/poster), US Equal Employment Opportunity Commission. 5. [OSHA Cares Job Safety and Health Workplace Poster](https://www.osha.gov/publications/poster), US Occupational Safety and Health Administration. 6. [FirstStep Poster Advisor](https://webapps.dol.gov/elaws/posters.htm), US Department of Labor (elaws). 7. [State Labor Offices](https://www.dol.gov/agencies/whd/state/contacts), US Department of Labor. 8. [Food Code 2026](https://www.fda.gov/media/194741/download), US Food and Drug Administration, 2026 edition. 9. [Food Code 2022](https://www.fda.gov/media/164194/download), US Food and Drug Administration, 2022 edition, January 18, 2023 version. 10. [The FDA Releases 2026 Food Code](https://www.fda.gov/food/hfp-constituent-updates/fda-releases-2026-food-code), US Food and Drug Administration, September 17, 2026. 11. [Adoption of the FDA Food Code by State and Territorial Agencies Responsible for Oversight of Restaurants and Retail Food Stores](https://www.fda.gov/food/fda-food-code/adoption-fda-food-code-state-and-territorial-agencies-responsible-oversight-restaurants-and-retail), US Food and Drug Administration, 2024 report. 12. [Menu Labeling Requirements](https://www.fda.gov/food/nutrition-food-labeling-and-critical-foods/menu-labeling-requirements), US Food and Drug Administration. 13. [Menu Labeling Rule: Key Facts for Industry](https://www.fda.gov/media/116000/download), US Food and Drug Administration. 14. [Required Signs](https://www.nyc.gov/site/doh/business/food-operators/required-signs.page), NYC Health Department. 15. [Letter Grading for Restaurants](https://www.nyc.gov/site/doh/business/food-operators/letter-grading-for-restaurants.page), NYC Health Department. 16. [Fact Sheet #21: Recordkeeping Requirements under the FLSA](https://www.dol.gov/agencies/whd/fact-sheets/21-flsa-recordkeeping), US Department of Labor, revised July 2008. 17. [Fact Sheet #15: Tipped Employees Under the FLSA](https://www.dol.gov/agencies/whd/fact-sheets/15-tipped-employees-flsa), US Department of Labor. 18. [Retaining Form I-9](https://www.uscis.gov/i-9-central/form-i-9-resources/handbook-for-employers-m-274/100-retaining-form-i-9), US Citizenship and Immigration Services. 19. [Publication 531, Reporting Tip Income](https://www.irs.gov/publications/p531), Internal Revenue Service, revised December 2024. 20. [Instructions for Form 8027](https://www.irs.gov/instructions/i8027), Internal Revenue Service, 2025 instructions. Every rule and source on this page was checked on October 14, 2026. The rules are summarized from the agencies' own pages, checked on the date above; your state, county and city add their own. General information, not legal advice. --- title: "Can guests order from you in ChatGPT, Claude or Google Maps?" description: "Which restaurant systems can take orders through ChatGPT, Claude and Google's Ask Maps today, what you do on each, what it costs and where guests pay." canonical: "https://restaurantdoctorai.com/blog/order-through-chatgpt-claude-google-maps" published: "2026-11-01" updated: "2026-11-01" --- # Can guests order from you in ChatGPT, Claude or Google Maps? ## Key takeaways - In the U.S., on Square with Online Ordering, guests may already reach you in ChatGPT and Claude. Google's Ask Maps also needs your profile connected and one setting. - Toast says restaurants using its online ordering need no extra work for Google's Ask Maps, in the U.S. We found no Toast announcement for ChatGPT or Claude. - On other systems we found no built-in way to take these orders, only a few small ChatGPT apps, and no page shows an assistant ordering from a plain website. - The only count we found is tiny, so it's no reason to change your register: DoorDash's CEO put outside agents at maybe several hundred orders. In the U.S., if you use Square's or Toast's online ordering, guests may already be able to order from you through an AI assistant, or you may be one setting away. On the other systems we checked, we found no way to do this yet. The only count we found is very small, so this isn't a reason to change your register or to worry. Checked November 1, 2026 This changes fast: Square's link to ChatGPT and Claude went live in July, and ordering in Google Maps started in August. If your screen shows something different from this post, trust your screen and the company's own page. In Google Maps, the part that takes orders is called Ask Maps: an AI you talk or type to. ## The one number we found On DoorDash's latest earnings call, its CEO said outside AI agents had sent “maybe several hundred orders in total,” [Magica reported](https://magica.com/news/doordash-third-party-ai-agent-orders-still-tiny). The same report puts DoorDash at 993 million orders in the third quarter. He gave no exact count, and the two figures don't cover the same stretch of time, but the gap is huge. We found no such numbers from Square or Toast. Choosing a restaurant is further along than ordering from one. In a [survey DoorDash commissioned](https://about.doordash.com/en-us/news/doordash-restaurant-industry-trends-report-2026) (Dynata asked 3,001 U.S. consumers in March 2026), 22% said they had used an AI tool to help choose a restaurant. This is worth one look at your settings, so you know where you stand. It isn't a project. ## Your system, one by one | Your system | Assistant | Live? | What you do | What it costs | Where the guest pays | | --- | --- | --- | --- | --- | --- | | Square | ChatGPT, Claude | Yes, since July 1, 2026. U.S. | Have an active [Square Online Ordering](https://squareup.com/us/en/ai/agentic-commerce) profile. Eligible sellers are opted in. | Your existing Square online processing rate. No platform fee, and [no marketplace commission](https://squareup.com/us/en/press/claude-chatgpt-integrations). | It varies: in the assistant, or on your Square ordering page. | | Square | Ask Maps | Rolling out since August 6, 2026. U.S. | The same, plus connect your Google Business Profile and enable Order with Google through Square. | Your existing Square online processing rate, no added fees. | The pages differ. [Square](https://squareup.com/us/en/press/ask-maps) says in Ask Maps, with Order by Cash App. [Google](https://support.google.com/maps/answer/17114780) says in the partner's app. | | Toast | Ask Maps | [Rolling out](https://blog.google/products-and-platforms/products/maps/order-food-in-ask-maps/) since August 6, 2026. U.S. | [Toast says](https://pos.toasttab.com/news/toast-google-ask-maps-integration) “no extra work required.” Orders run through its own online ordering, so that should be on. Toast doesn't spell this out for Ask Maps. | Toast calls it commission-free. That covers commissions only: the pages we read give no figure for Online Ordering or card processing. | Toast doesn't say. Google says in the partner's app. | | Toast | ChatGPT, Claude | No Toast announcement found. | Nothing yet. | Not applicable. | Not applicable. | | Clover | Any | Nothing live found. [Fiserv](https://www.fiserv.com/en/insights/articles-and-blogs/agentic-commerce-is-here-what-merchants-need-to-know-and-do.html), which owns Clover, said in February it was working on enabling all merchants and named Clover as one of its platforms. No date. | Nothing yet. | Not applicable. | Not applicable. | | DoorDash | ChatGPT | [Grocery only](https://about.doordash.com/en-us/news/openai). Restaurants aren't mentioned. | Nothing documented. | No fee stated. [Plans](https://merchants.doordash.com/en-us/products/pricing): 15%, 25% or 30% on delivery, 6% on pickup. | The DoorDash app. | | Uber Eats | ChatGPT | [Browsing only](https://help.uber.com/ubereats/restaurants/article/chatgpt-integration-for-eaters?nodeId=7e56650b-9482-4a55-a560-1902ba2dd70e): restaurants and menus. U.S. | Nothing documented. | No fee stated. [Plans](https://merchants.ubereats.com/us/en/pricing/): 20% (15% in some cities), 25% or 30% on delivery; 7% on pickup, or 10% without proof of in-store pricing. | The Uber Eats app. | | Uber Eats | Claude | [Browsing only](https://help.uber.com/en/ubereats/restaurants/article/claude-integration-for-eaters?nodeId=f076b231-b5c7-448e-bd52-b378506b3cb7): restaurants and menus. U.S. Since April 23, 2026. | Nothing documented. | No fee stated. Same plans. | The Uber Eats app. | | Uber Eats | Alexa+ | Rolling out since March 31, 2026, as of [Amazon's post](https://www.aboutamazon.com/news/devices/alexa-plus-grubhub-uber-eats-food-delivery): delivery, on newer Echo Show 8 devices and larger. | Nothing stated. The guest links an account in the Alexa app. | No fee stated. Same plans. | Amazon doesn't say. Alexa shows the total before the purchase. | | Uber Eats | Ask Maps | Google said “coming soon” in August, and its help page now names Uber Eats. We couldn't confirm it's live. | Nothing stated. | No fee stated. Same plans. | The Uber Eats app, Google says. | | Grubhub | ChatGPT | [Yes](https://chatgpt.com/plugins), browsing: restaurants and menus, reported since June 2026. | Nothing stated. | No fee stated. [Plans](https://get.grubhub.com/grubhub-pricing-and-fees/): 5%, 15% or 20% commission; its own delivery from 10%; processing extra. | On Grubhub, [FastCasual reports](https://www.fastcasual.com/news/wonder-owned-grubhub-launches-chatgpt-ordering-integration/). | | Grubhub | Alexa+ | Rolling out since March 31, 2026, as of Amazon's post: delivery, on newer Echo Show 8 devices and larger. | Nothing stated. The guest links an account in the Alexa app. | No fee stated. Same plans. | Amazon doesn't say. Alexa shows the total before the purchase. | | DoorDash, Uber Eats, Grubhub | Gemini | [Beta](https://support.google.com/pixelphone/answer/16940971), released gradually. Gemini works in the delivery app on newer phones: Pixel 10 and 11, Galaxy S26 and some others. Google lists the U.S., India, Japan, South Korea and the UK. | Nothing. Google's page names no food apps; [9to5Google](https://9to5google.com/2026/03/17/gemini-screen-automation-pixel/) named these three in March. | No fee stated. Same plans. | The delivery app, in Google's [DoorDash example](https://blog.google/innovation-and-ai/products/gemini-app/android-multi-step-tasks/). | | Bites | ChatGPT | [Yes](https://chatgpt.com/plugins). Its listing shows San Jose, Fremont and Santa Clara. | A [contact form](https://withbites.com/restaurants). Bites doesn't say which POS you need. | $1 per order. | Bites doesn't say. | *As published on November 1, 2026. Commission rates are each app's normal plans; none of these pages names a separate fee for assistants. A system or assistant that isn't listed means we found nothing from the company, not that nothing exists.* Square also says it plans to reach Alexa+. It isn't live yet, as far as Square's July release says. ## If you're on Square Square has the fullest instructions, on its [AI page](https://squareup.com/us/en/ai/agentic-commerce). They apply to U.S. food and beverage sellers. In ChatGPT and Claude, guests use an app called Order by Cash App. Both directories list it as something a guest adds, so a guest who hasn't added it may never see you there. 1. For ChatGPT and Claude, check that your Square Online Ordering profile is active. Square says eligible sellers are opted in with no setup, so you may already be on. 2. Keep your menu, hours and stock current in Square, so guests order what you really have. 3. For Ask Maps, Square lists two more things: connect your Google Business Profile, and enable Order with Google through Square. Square's August release mentions only the profile, so check your Dashboard to see where you stand. Your menu, hours and location then sync from Square. 4. Preview it, so you see what a guest sees. Add Order by Cash App from ChatGPT's app store or Claude's directory, and ask it to order from your business. In Google Maps, ask Ask Maps to order from your restaurant and pick Order by Cash App from the options. Stop before you pay, unless you want a real order on your kitchen screen. 5. Watch for them in your Dashboard, POS and Kitchen Display. Square says orders are tagged by source and shows the source in its reporting, so count them there. 6. To turn it off, go to Square Dashboard > Online Channels > AI assistants, and switch it off. Square doesn't say whether this covers each assistant, so check Ask Maps afterward. Square says these orders are billed at your existing Square online processing rate, with no platform fee. It doesn't say what fees a guest sees, or whether ChatGPT and Claude orders can be delivered as well as picked up. Google adds that ordering in Ask Maps is in English only, for guests 18 or over, in the U.S. ## If you're on Toast Toast says there's “no extra work required” from you for Ask Maps. Those orders run through Toast's own online ordering, so check that yours is on. To see what a guest sees, ask Ask Maps for your own restaurant and start an order. Stop before you pay. Toast does have a [setting for Google ordering](https://support.toasttab.com/en/article/Using-Order-with-Google): Toast Web > Takeout & delivery > Google Management > Google Listings, then Save to Google. You can also turn it off in your Google Business Profile, under Food ordering, with “Accept orders on your profile”, which overrides Toast Web. Changes can take up to 48 hours. Toast doesn't say whether this setting also controls Ask Maps: its help page, updated October 27, doesn't mention Ask Maps at all. We found no Toast announcement for ChatGPT or Claude. ## If you're on a delivery app The table shows what's live. None of the pages we read gives a restaurant anything to switch on, or a way to opt out. If you want to know what the percentages mean for your menu, our [delivery app commissions guide](https://restaurantdoctorai.com/guides/delivery-app-commissions) works through them. ## If you're on none of these Here's the honest answer. Beyond the big systems, a few small apps take orders inside ChatGPT, such as Bites and VoiceBit, each for restaurants on its own platform. We didn't test them, and we found nothing like this for Claude or Ask Maps. No page we read shows any assistant ordering from a plain website. So you may have to wait, and that's fine at today's numbers. Three things you can do now, none of which needs a new system: - **Put your own order link on Google.** In your Google Business Profile, go to [Food ordering](https://support.google.com/business/answer/10842217), choose “Add a link” at the bottom of the list, then “Set as preferred” for pickup or delivery, so a guest on your listing can order from you directly. Leave “Accept orders on your profile” on: turning it off hides every option. Google's page doesn't mention Ask Maps, and Ask Maps sends guests to partner apps, so this isn't an Ask Maps order. - Write your menu as text on your website, not only as a photo or PDF, so anyone reading the page gets the dishes and prices. - Keep your hours, address and menu current on Google and on your site, so guests get the right answer wherever they ask. What we don't know: where assistants get menus for restaurants that aren't connected to them, and whether a text menu changes what they say about you. We'd rather tell you that than guess. ## What's coming The [Universal Commerce Protocol](https://ucp.dev/) is an open standard for catalogs, carts, checkout and orders, and its food part is marked “coming soon.” Google's [page on it](https://developers.google.com/actions-center/verticals/ordering/ucp) offers only a waitlist for now, with no date. Meta's [help page](https://www.meta.com/help/superintelligence/1687253048996149/) on connectors for its Muse assistant names Facebook, Instagram, Threads, Gmail, Google Workspace, Apple Health and Android SMS, and no restaurant connector. That page isn't the full list, which sits in the app's settings, and we couldn't read it. Neither of these needs anything from you today. ## This week 1. Find your system in the table. 2. On Square: check that Online Ordering is active, decide about Ask Maps, and run one preview. 3. On Toast: ask Ask Maps for your restaurant and see what a guest sees. 4. On anything else: add your own order link to your Google Business Profile, and put your menu on your site as text. 5. Run our check, [How AI sees your website](https://restaurantdoctorai.com/online-check), to see where your restaurant stands. See it for your own restaurant Type your restaurant's name into our check, [How AI sees your website](https://restaurantdoctorai.com/online-check). It looks at your Google listing and your website, shows what a basic AI reader gets from your home page, and its paid full report has a section on ordering through assistants, so you know which steps above apply to you. 45-second profit check That's your online side. Now your numbers Answer 12 quick questions and get your restaurant's grade for each cost area and where to start, free, in about 45 seconds. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [DoorDash Reports Only Several Hundred AI-Agent Orders](https://magica.com/news/doordash-third-party-ai-agent-orders-still-tiny), Magica, October 30, 2026 - [From the Doorstep to the Dining Room: New DoorDash Survey Data Reveals the Full Picture of the Modern Restaurant Guest](https://about.doordash.com/en-us/news/doordash-restaurant-industry-trends-report-2026), DoorDash, May 18, 2026 - [Show up and sell through AI](https://squareup.com/us/en/ai/agentic-commerce), Square - [Square Introduces New ChatGPT and Claude Integrations, Helping Sellers Reach Customers Through AI-Powered Discovery](https://squareup.com/us/en/press/claude-chatgpt-integrations), Square, July 1, 2026 - [Square and Google Deepen Collaboration to Support the Future of Food Ordering](https://squareup.com/us/en/press/ask-maps), Square, August 6, 2026 - [Taking actions within Ask Maps with third-parties](https://support.google.com/maps/answer/17114780), Google Maps Help - [Google Maps' Ask Maps gets more helpful with food ordering](https://blog.google/products-and-platforms/products/maps/order-food-in-ask-maps/), Google, August 6, 2026 - [Arriving Now: AI-Powered Toast Ordering on Google Maps](https://pos.toasttab.com/news/toast-google-ask-maps-integration), Toast, August 6, 2026 - [Agentic Commerce Is Here: What Merchants Should Do Now](https://www.fiserv.com/en/insights/articles-and-blogs/agentic-commerce-is-here-what-merchants-need-to-know-and-do.html), Fiserv, February 19, 2026 - [DoorDash Launches Grocery Shopping App Within ChatGPT](https://about.doordash.com/en-us/news/openai), DoorDash, December 17, 2025 - [DoorDash Merchant Plans and Pricing](https://merchants.doordash.com/en-us/products/pricing), DoorDash - [ChatGPT Integration for Eaters](https://help.uber.com/ubereats/restaurants/article/chatgpt-integration-for-eaters?nodeId=7e56650b-9482-4a55-a560-1902ba2dd70e), Uber Help - [Uber Eats Pricing for Restaurants](https://merchants.ubereats.com/us/en/pricing/), Uber Eats - [Claude Integration for Eaters](https://help.uber.com/en/ubereats/restaurants/article/claude-integration-for-eaters?nodeId=f076b231-b5c7-448e-bd52-b378506b3cb7), Uber Help - [Alexa Plus orders food from Grubhub and Uber Eats naturally](https://www.aboutamazon.com/news/devices/alexa-plus-grubhub-uber-eats-food-delivery), Amazon, March 31, 2026 - [ChatGPT apps directory](https://chatgpt.com/plugins), OpenAI - [Pricing and Fees](https://get.grubhub.com/grubhub-pricing-and-fees/), Grubhub - [Wonder-owned Grubhub launches ChatGPT ordering integration](https://www.fastcasual.com/news/wonder-owned-grubhub-launches-chatgpt-ordering-integration/), FastCasual, June 19, 2026 - [Ask Gemini to handle your multi-step tasks in select Android apps](https://support.google.com/pixelphone/answer/16940971), Pixel Phone Help - [Gemini screen automation rolling out to Pixel 10 series](https://9to5google.com/2026/03/17/gemini-screen-automation-pixel/), 9to5Google, March 17, 2026 - [Gemini on Android lets you assign multi-step tasks](https://blog.google/innovation-and-ai/products/gemini-app/android-multi-step-tasks/), Google, February 25, 2026 - [Bites for Restaurants](https://withbites.com/restaurants), Bites - [Manage Order with Google](https://support.toasttab.com/en/article/Using-Order-with-Google), Toast Support, October 27, 2026 - [Manage online ordering options](https://support.google.com/business/answer/10842217), Google Business Profile Help - [Universal Commerce Protocol](https://ucp.dev/), Universal Commerce Protocol - [Google Universal Commerce Protocol (UCP) Guide](https://developers.google.com/actions-center/verticals/ordering/ucp), Google for Developers - [How Muse works with Connectors](https://www.meta.com/help/superintelligence/1687253048996149/), Meta Help Center --- title: "Can AI take your restaurant's phone orders? What goes wrong" description: "Count your missed calls, see why McDonald's, Taco Bell and Anthony's pulled back, and what an AI phone line needs before it takes your orders." canonical: "https://restaurantdoctorai.com/blog/ai-phone-ordering-for-restaurants" published: "2026-10-30" updated: "2026-10-30" --- # Can AI take your restaurant's phone orders? What goes wrong ## Key takeaways - Count missed calls in your own phone or POS call log before you buy: the popular missed-call figures come from companies that sell phone AI. - McDonald's, Taco Bell and Anthony's Coal Fired Pizza pulled back AI ordering after wrong orders, pranks, or guests who wanted a person. - A phone AI should say it's an AI, read the order back, send allergy questions to a person, and reach staff the moment a caller asks. - Ours has one monthly price after a 3-day trial capped at 30 calls; it takes pickup orders only, with no POS link and no card payment on the call. AI phone ordering for restaurants is sold with big numbers about missed calls and lost sales. Before you pay for any of it, count your own missed calls, look at where these systems have failed in public, and check that the one you try handles those failures. This post does all three, and shows what we built and what ours won't do. ## Count your own missed calls first Almost every missed-call figure you'll find online comes from a company that sells phone AI. Some can't be traced to any original study at all, so we don't repeat them. The better ones say how they were made. Maple, a phone AI seller, says restaurants miss about a third of calls at peak, and its own [method note](https://www.maple.inc/blog/the-state-of-restaurant-phone-communication-insights-from-1-million-calls) says that figure is what owners estimated when they signed up, not a count from call logs. Your own number is better than anyone's, and it's free to get: 1. Open your phone company's online account or your phone system's app and find the call log. Many business lines list answered, missed and abandoned calls with the time of each. 2. If your phone runs through your POS or a phone system with reports, look for its missed-call report instead. 3. Take one normal week. Count missed calls by hour, and note which hours they bunch in. 4. Count repeat calls from the same number within a few minutes once: that's one guest trying twice. If you miss a handful of calls a week, an AI is probably not your best fix. If you miss many at the same rush hours, and a good share of your calls are orders, it might be. ## What went wrong when big chains tried it AI ordering has had some very public trouble. These are the three cases owners mention most, and what the reports actually say. - **McDonald's** ended its drive-thru AI test with IBM, run at 100 U.S. locations, in June 2024. [CBS News](https://www.cbsnews.com/news/mcdonalds-ends-ai-drive-thru-ordering) reported mixed results, with customers saying it got even simple orders wrong, including one who said it added nine sweet teas to her order. McDonald's said it still expects voice ordering to be part of its future. This was a drive-thru, not a phone line, but the problem is the same: an order that grows by itself. - **Taco Bell** put voice AI in more than 500 drive-thrus. In August 2025, [PYMNTS](https://www.pymnts.com/news/artificial-intelligence/2025/taco-bell-reconsiders-voice-ai-after-seeing-mixed-results/), following the Wall Street Journal, reported glitches, customers who felt uncomfortable, and pranks such as an order for “18,000 cups of water.” Taco Bell said it was weighing which tasks need a person. - **Anthony's Coal Fired Pizza & Wings**, with 59 locations, dropped its AI phone system after less than a year, according to [PMQ Pizza](https://www.pmq.com/anthonys-ai-phone-system/) in September 2023. The CEO of its owner, BurgerFi, said guests told them they “want to speak with a human when placing orders.” Calls now go to a person. Three things go wrong again and again: wrong orders, orders nobody meant seriously, and guests who can't reach a person. Any system you try needs an answer for each. Most restaurants haven't tried any yet: [Restaurant Dive reported](https://www.restaurantdive.com/news/national-restaurant-assocation-operator-artificial-intelligence-adoption/812418/) in February 2026 that only 6% of restaurants use AI for customer orders. ## What we built for each problem Our [AI phone ordering](https://restaurantdoctorai.com/phone-ordering) line answers the calls your team can't get to, takes pickup orders from your menu and sends them to a kitchen screen, so you can keep taking phone orders during a rush without pulling someone off the line. Here is what it does about each failure above, exactly as the code does it today. | What went wrong | What we built for it | | --- | --- | | Orders that grow or come out wrong | Our code, not the AI, prices the order from your menu. At the end the AI reads the whole order back, then asks if that's right. Only a spoken yes places it, and only the order it read back. It repeats any number over five and adds only what the caller asked for. | | Prank and huge orders | More than 12 of one item, more than 24 items, or more than $150: the AI doesn't take it and offers the team instead. Paid at pickup, a fake order costs you the food. | | Guests who want a person | The first sentence says it's an AI and that the call is recorded and transcribed, then that the caller can say “staff” anytime. “I don't want a robot” gets “Of course, connecting you now.” | | A transfer that rings forever | A warm transfer: your staff phone rings for 25 seconds while the AI stays on the line and tells whoever answers why the guest is calling. If nobody picks up, it tries your backup number once, then tells the guest the team will call back. The kitchen screen keeps a call-back card, and you get one email. | | Allergy questions | It never says a dish is safe. Any allergy or dietary-safety question goes to a person. If the guest would rather just order, the allergy goes in the order note, the kitchen card shows a red allergy banner, and the email says a person should check it before it's made. | | A system error | If a tool fails, it never guesses a price or says “you're all set.” It tries once more, then hands the call to staff. | *What went wrong elsewhere, and the safeguard on our line. Prompt version 14, checked October 30, 2026.* One lesson from testing: a caller who said “oh, add a drink” after the order was placed heard it read back, said yes, and the drink was silently dropped. We found it in our own review of the code and fixed it before any restaurant used the line. Now one order is placed per call, and anything after that goes to your team, because the kitchen already has the ticket. ## How you'd know it's working After an order is placed, the AI asks once whether the caller would answer one quick question. Only a clear yes goes on: “How was it: poor, good or amazing?” After “poor,” it asks what could have been better, and never pushes. Your dashboard shows the month's counts, an average out of 3 and what unhappy callers said, next to every call's transcript. Read the ratings for what they are Only callers who finished an order and agreed to answer are asked, and the AI skips anyone in a hurry or upset. So the ratings lean happy. The notes after a “poor” are the useful part. Read some transcripts each week too. ## What ours doesn't do Say no to us early if you need any of these: - No POS connection and no printer. Orders arrive on a kitchen screen on any tablet, with an email copy, and your staff ring them into the POS. - No card payment on the call. Guests pay when they pick up, and the AI never takes a card number. - Pickup only, as soon as possible. No delivery, no orders for later or another day, and no changes once an order is placed: those go to your team. - No reservations. It offers to connect the caller to your team. - It answers only calls you forward to it. You keep your own number and send missed or busy calls to the AI's number, so calls your team picks up never reach it. - If nobody accepts a new order on the kitchen screen within 2 minutes, you get an alert by email, not a text. - English and Spanish only. It never calls or texts guests, and keeps no audio: transcripts and guest names and numbers are deleted after 90 days. ## What it costs, next to others Ours is one monthly price, with no contract, no setup fee and no charge per call; today's price is on the [AI phone ordering page](https://restaurantdoctorai.com/phone-ordering). The free trial lasts 3 days from your AI's first call and ends no later than 7 days after you sign up; it's capped at 30 calls, and after that calls go to your staff until it ends. You add a card when you sign up. Here is what other sellers publish, as we found it on October 30, 2026. | Product | Price | Takes the order on the call? | | --- | --- | --- | | RestaurantDoctorAI | One monthly price, on our AI phone ordering page | Yes: pickup, paid at pickup, to a kitchen screen; no POS | | [Maple Voice](https://www.maple.inc/pricing) | $150 a month, or $85 a month billed yearly | No: answering only | | Maple Pro | $350 a month, or $220 a month billed yearly | Yes: into the POS, card payment on the call | | [Slang.ai](https://www.slang.ai/pricing) | Starting at $399 per location (the page doesn't say per month) | Texts an ordering link | | [Loman](https://loman.ai/pricing) | Not public | Yes, on its Premium plan: takeout and delivery ordering, payment on the call | | [Kea](https://kea.ai) | Not public | Yes: its page says into the POS and the kitchen screen | | [Square AI voice ordering](https://squareup.com/help/us/en/article/8568-take-orders-with-ai-powered-voice-ordering) | Not stated; needs a Square Plus, Premium or Messages Plus plan | Yes, for pickup; every order paid first by a texted link | | [Owner.com AI Phone Ordering](https://www.owner.com/ai-phone-ordering) | Not public: a waitlist | Yes: its page says it sends the order to your kitchen | *AI phone answering prices for one location, as published on October 30, 2026. “Not public” means the page asks you to contact sales.* If you need orders in your POS and card payment on the call today, the dearer plans do that and ours doesn't. Any accuracy percentage you see on these sites is the seller's own, without a published method. And lists of the “best AI phone answering” are often ranked by who pays, as we explain in [restaurant software referral fees](https://restaurantdoctorai.com/blog/restaurant-software-referral-fees). ## Against a person on the phone The [Bureau of Labor Statistics](https://www.bls.gov/ooh/food-preparation-and-serving/food-and-beverage-serving-and-related-workers.htm) puts the median pay of hosts and hostesses, and of fast food and counter workers, at $15.00 an hour (May 2025). At that wage, Maple Voice's $150 a month, for example, is about 10 hours of someone's time, before payroll taxes. The [true cost of an hour of labor](https://restaurantdoctorai.com/blog/true-cost-of-an-hour-of-restaurant-labor) is higher than the wage. Do the sum with your own count. Say, as a made-up example, the phone takes 15 calls a day at about 2 minutes each: that's 30 minutes a day, or about 15 hours over 30 days, $225 at $15.00 an hour. That's the time the phone takes, not money you'd save. The AI answers only the calls you forward, so your team still answers the rest, and some of the guests it answers might have called back anyway. Our [labor cost guide](https://restaurantdoctorai.com/guides/restaurant-labor-cost) shows where phone time sits in your labor numbers. ## When it's not worth it - You miss few calls. Your call log is the test, not a seller's average. - Most of your calls are reservations. In Maple's own sample, from its customers (restaurants and other local businesses), 38% of calls were reservations and 12% were orders ([Maple](https://www.maple.inc/blog/the-state-of-restaurant-phone-communication-insights-from-1-million-calls), a seller). Ours takes no reservations. - You need delivery, orders for later, or orders straight into your POS. - Your callers are mostly regulars who like to chat with your staff, or ask about allergies, which ours hands to a person anyway. - Someone is already free to answer at your busy hours. ## Recording, transcripts and saying it's an AI Here's what the rules say, in plain words. Federal law lets a call be recorded with one side's consent, but a group of states want everyone on the call to agree. The [Reporters Committee for Freedom of the Press](https://www.rcfp.org/introduction-to-reporters-recording-guide/) lists California, Connecticut, Delaware, Florida, Illinois, Maryland, Massachusetts, Michigan, Montana, Nevada, New Hampshire, Pennsylvania and Washington for phone calls, and notes that courts disagree on whose law applies when a call crosses state lines. A guest can call you from anywhere, so the safe habit is to tell every caller at the start. Transcribing counts as well as recording. In February 2025, a federal court in California let a wiretap claim go ahead against Google over an AI that transcribed customer-service calls live, as [Goodwin](https://www.goodwinlaw.com/en/insights/publications/2025/02/alerts-practices-dpc-ftec-ai-voice-products-subject-to-california-invasion-of-privacy-claims) reported. This is still being argued in court, which is one more reason to say it up front. On saying it's an AI: Maine's [chatbot law](https://legislature.maine.gov/statutes/10/title10sec1500-DD-1.html) bars a business from using a chatbot in a way that could make a customer think they're talking to a person, unless the customer is clearly told. Our first sentence covers both points: “I'm their automated AI assistant, and this call is recorded and transcribed.” If a caller asks later, it says so again. Whatever system you choose, make sure its opening line does the same, and that a person is one word away. ## This week 1. Pull one week of your call log and count missed calls by hour, each guest once. 2. Keep a tally at the counter of what answered calls are about: orders, hours, reservations, questions. 3. If many missed calls are orders at the same rush hours, write down the four must-haves: says it's an AI, reads the order back, sends allergies to a person, reaches a person on request. 4. Call any phone AI's demo line and test each one: ask for staff, ask about nuts, order 30 of something. 5. If ours fits, see how [setup works](https://restaurantdoctorai.com/phone-ordering/setup) before you start the trial. 45-second profit check While you're looking at the phone, look at the rest Answer 12 quick questions and get your restaurant's grade for each cost area and where to start, free, in about 45 seconds. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [The State of Restaurant Phone Communication: Insights from 1 Million+ Calls](https://www.maple.inc/blog/the-state-of-restaurant-phone-communication-insights-from-1-million-calls), Maple (sells AI phone answering), December 7, 2025 - [McDonald's ends AI drive-thru orders — for now](https://www.cbsnews.com/news/mcdonalds-ends-ai-drive-thru-ordering), CBS News, June 17, 2024 - [Taco Bell Finds Voice AI Still Needs a Human Touch](https://www.pymnts.com/news/artificial-intelligence/2025/taco-bell-reconsiders-voice-ai-after-seeing-mixed-results/), PYMNTS, August 28, 2025 - [AI Fail: Anthony's Gives 'Becky' the Boot After Less Than a Year](https://www.pmq.com/anthonys-ai-phone-system/), PMQ Pizza, September 21, 2023 - [NRA: Over 25% of restaurant operators use AI](https://www.restaurantdive.com/news/national-restaurant-assocation-operator-artificial-intelligence-adoption/812418/), Restaurant Dive, February 18, 2026 - [Pricing](https://www.maple.inc/pricing), Maple - [Pricing](https://www.slang.ai/pricing), Slang.ai - [Pricing](https://loman.ai/pricing), Loman - [The Best Voice AI for Restaurants](https://kea.ai), Kea - [Take orders with AI-powered voice ordering](https://squareup.com/help/us/en/article/8568-take-orders-with-ai-powered-voice-ordering), Square Support Center - [AI Phone Ordering](https://www.owner.com/ai-phone-ordering), Owner.com - [Food and Beverage Serving and Related Workers: Occupational Outlook Handbook](https://www.bls.gov/ooh/food-preparation-and-serving/food-and-beverage-serving-and-related-workers.htm), U.S. Bureau of Labor Statistics, August 27, 2026 - [Introduction to the Reporter's Recording Guide](https://www.rcfp.org/introduction-to-reporters-recording-guide/), Reporters Committee for Freedom of the Press - [AI Voice Products Subject to California Invasion of Privacy Claims](https://www.goodwinlaw.com/en/insights/publications/2025/02/alerts-practices-dpc-ftec-ai-voice-products-subject-to-california-invasion-of-privacy-claims), Goodwin Procter, February 26, 2025 - [10 §1500-DD. Required disclosure of use of artificial intelligence chatbot to engage in trade and commerce](https://legislature.maine.gov/statutes/10/title10sec1500-DD-1.html), Maine Legislature, Maine Revised Statutes --- title: "Restaurant review analysis: what your reviews tell you" description: "How to read your restaurant's reviews by topic, month by month and against places nearby, what the research says, and how to ask within the rules." canonical: "https://restaurantdoctorai.com/blog/restaurant-review-analysis" published: "2026-10-30" updated: "2026-10-30" --- # Restaurant review analysis: what your reviews tell you ## Key takeaways - One study of Washington State Yelp data found a one-star rise went with 5 to 9% more revenue for independents, not chains; it is old data, not a promise. - No credible industry-wide list of top restaurant complaints exists, so count your own reviews by topic and by month. - Compare each complaint's share with places nearby: one the whole block shares is common, one only you have is yours to fix. - Google lets you ask every guest for a review but bars rewards, asking only happy guests and staff review quotas; Yelp asks businesses not to ask. Your star average is the slowest number on your listing: it is built from every review you have ever had, so a new problem can run for months before it moves. Reading your reviews by topic, month by month and next to the places nearby shows which complaint is growing now. Here is how to do it with a spreadsheet, worked on a made-up example, plus what research says reviews are worth and how to ask for more within the rules. ## What the research says reviews do to sales Two studies get quoted most. Neither is a promise about your restaurant. - **Revenue.** Michael Luca's [Harvard Business School working paper 12-016](https://econpapers.repec.org/RePEc:hbs:wpaper:12-016) found that a one-star rise in a restaurant's Yelp rating went with a 5 to 9% rise in revenue. The effect came from independent restaurants; chains showed none. The sales data came from the Washington State Department of Revenue, and the paper first came out in 2011, so the data is well over a decade old. - **Full tables.** Michael Anderson and Jeremy Magruder, in the Economic Journal in 2012, looked at San Francisco restaurants' 7 p.m. reservations. In [their own summary](https://giannini.ucop.edu/publications/are-update/issues/2013/16/5/does-yelp-affect/), an extra half-star on Yelp made a restaurant sell out 19 percentage points (49%) more often at peak hours, with larger effects where guests had less other information. For restaurants with more than 500 reviews they found no jump. Both are careful studies, but each covers one city or state, years ago, on Yelp, not Google. The fair reading: what guests see about you can move how many of them come, most of all when they know little else about you. ## There is no reliable list of top complaints Search for what restaurant guests complain about most and you will find neat percentages. We looked for where they come from and found no traceable source: no government survey, no large peer-reviewed count of US restaurant reviews. The academic studies read narrow samples, such as fine dining or one city, and the rest comes from companies that sell review software. What guests complain about differs from one restaurant to the next, which is the reason to count your own. ## Read by topic, not by stars A 3-star review can praise the food and complain about the wait in the same breath. Sorting by stars hides that; sorting by topic shows it. Eight topics cover most of what guests write about: food; service and staff; wait times; price and value; cleanliness; atmosphere; order accuracy; takeout and delivery. Open a spreadsheet with one row per review: the date, the stars, and two columns per topic, praise and complaint. Read each review once and mark every topic it touches. A hundred reviews take an evening. | Topic | Praise | Complaints | | --- | --- | --- | | Food | 46 | 9 | | Service and staff | 31 | 6 | | Wait times | 4 | 14 | | Price and value | 8 | 7 | | Cleanliness | 5 | 2 | | Atmosphere | 22 | 1 | | Order accuracy | 1 | 5 | | Takeout and delivery | 3 | 4 | *Example only, not real data: Corner Table Café, a made-up café, its newest 100 Google reviews by topic. One review can count in several topics.* The food is the café's strength. Wait times are the surprise: the most complaints of any topic, and almost no praise to balance them. ## Month by month: is it growing? A total doesn't say whether a problem is fading or growing. Split the same reviews by month. | Month | Reviews | Average stars | Wait complaints | | --- | --- | --- | --- | | June | 16 | 4.5 | 1 | | July | 18 | 4.5 | 1 | | August | 17 | 4.4 | 2 | | September | 22 | 4.2 | 4 | | October | 27 | 4.1 | 6 | | All five months | 100 | 4.31 | 14 | *Example only, not real data: Corner Table Café's newest 100 reviews by month, with the wait time complaints.* Wait complaints went from 1 review in 16 in June to more than 1 in 5 in October, while the month's average fell from 4.5 to 4.1. The café's overall rating, built from years of reviews, would hardly move. A second check: compare your newest 25 reviews with the 75 before them. A topic much worse in the newest 25 is recent. Then look for the cause in your own records: a schedule change, someone who left, busier lunches. The [sales per labor hour guide](https://restaurantdoctorai.com/guides/sales-per-labor-hour) shows whether a shift is short-staffed for its sales, and our [post on slow shifts](https://restaurantdoctorai.com/blog/cut-labor-cost-slow-shifts) covers moving hours to where the rush is. If value complaints grow after a menu change, see [how much to raise menu prices](https://restaurantdoctorai.com/blog/how-much-to-raise-menu-prices). Wrong orders often come down to one habit: reading the order back. ## Compare with the places nearby Some complaints belong to the whole block: when prices rise everywhere, every place gets more value complaints. Read three similar places near you the same way and compare the share of reviews, not the count. | Complaint | The café | Place A | Place B | Place C | | --- | --- | --- | --- | --- | | Wait times | 14% | 6% | 8% | 4% | | Price and value | 7% | 12% | 10% | 8% | *Example only, not real data: share of reviews with each complaint, Corner Table Café (100 reviews) and three made-up places nearby (50 reviews each).* Price complaints are common on this block, and the café has fewer than its neighbors. The wait is the café's own: more than twice the neighbors' average. Read what they are praised for, too. ## How many reviews make a pattern? There is no fixed number, and anyone who gives you one is guessing. Some rules of thumb: - One angry review is one guest's night. Answer it, but don't rebuild a menu around it. - The same complaint from different guests in different weeks is worth checking on the floor. - A complaint whose share rises month after month matters more than a large total that is fading. - With few reviews a month, one review moves the share a lot. Treat small numbers as a hint and check your records first. ## Asking for more reviews within the rules More reviews make your monthly counts steadier. The rules differ by who sets them. **The FTC.** The Federal Trade Commission's [rule on fake reviews](https://www.ftc.gov/news-events/news/press-releases/2024/08/federal-trade-commission-announces-final-rule-banning-fake-reviews-testimonials) bans fake reviews, including AI-generated fake ones, undisclosed reviews by owners and managers, and scaring guests into taking reviews down. Under [section 465.4](https://www.ecfr.gov/current/title-16/chapter-I/subchapter-D/part-465), a business may not offer anything for a review that must be positive or negative. The FTC's [questions and answers](https://www.ftc.gov/business-guidance/resources/consumer-reviews-testimonials-rule-questions-answers) say the rule took effect on October 21, 2024, and that asking only happy guests, while not banned by the rule itself, could still break the wider FTC Act. **Google is stricter.** Its [Maps content policy](https://support.google.com/contributionpolicy/answer/7400114?hl=en) does not let a business: - offer any reward, such as a discount or a free dessert, for posting, changing or removing a review; - discourage negative reviews, or ask only the guests it expects to be happy; - pressure guests to review while still in the restaurant, or ask for specific content; - ask staff to collect a certain number of reviews, or reviews that mention a staff member. Reviews from current or former employees are removed as a conflict of interest. What Google does suggest, in its [tips to get more reviews](https://support.google.com/business/answer/3474122?hl=en), is asking customers to visit a Google link or scan a QR code. Ask every guest the same way: a card with the check, a line on the receipt, a sign by the door. What Google says about reviews that name your staff Some marketing blogs say a Google update this spring banned reviews that name an employee. Google's policy page, read on October 30, 2026, says no such thing and shows no dated change. What it bars is the business asking staff to gather reviews that identify a staff member, and staff review quotas. A guest who thanks a server by name on their own is not against that page. Drop any “mention my name” cards or staff targets. **Yelp asks the opposite.** Its page [“Don't Ask for Reviews”](https://www.yelp-support.com/article/Don-t-Ask-for-Reviews?l=en_US) says its software may not recommend reviews that seem prompted. So point your sign to Google, not Yelp. ## If you'd rather not count by hand Our [review reader](https://restaurantdoctorai.com/reviews) reads your newest 100 Google reviews and counts praise and complaints across the same eight topics, sets each topic's share beside up to three places like yours nearby, and shows month by month how many reviews complain and their average stars, with your newest 25 reviews' stars against the rest when there are 50 or more. It lists staff guests praise by name (praise only) and drafts replies to unanswered 1 to 3 star reviews for you to edit and post. It reads Google only, never posts or asks for reviews, and you paste nothing. A “since your last reading” comparison comes only with Review Watch, which also checks your 20 newest reviews each week and emails a reply draft for each new, unanswered 1 or 2 star review that has words (up to five a week). See a [sample result](https://restaurantdoctorai.com/sample-result) first. ## This week 1. Mark your newest 50 to 100 Google reviews in a spreadsheet: eight topics, praise and complaint for each. 2. Split them by month and find the one topic whose complaints are rising. 3. Read three similar places nearby the same way, and compare shares, not counts. 4. Check the rising complaint against your own schedule, tickets and sales for those weeks before you change anything. 5. Put a Google review link or QR code where every guest sees it, and drop any reward, staff target or request to name a server. 45-second profit check Check the numbers behind the complaints The free check asks 12 quick questions and grades each cost area in about 45 seconds, so you can see whether a wait or value complaint has a cost problem behind it. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Reviews, Reputation, and Revenue: The Case of Yelp.com](https://econpapers.repec.org/RePEc:hbs:wpaper:12-016), Michael Luca, Harvard Business School Working Paper 12-016, September 2011, revised March 2016 - [Does Yelp Affect Restaurant Demand?](https://giannini.ucop.edu/publications/are-update/issues/2013/16/5/does-yelp-affect/), Michael L. Anderson and Jeremy Magruder, ARE Update, Giannini Foundation of Agricultural Economics, May/June 2013 - [Federal Trade Commission Announces Final Rule Banning Fake Reviews and Testimonials](https://www.ftc.gov/news-events/news/press-releases/2024/08/federal-trade-commission-announces-final-rule-banning-fake-reviews-testimonials), Federal Trade Commission, August 14, 2024 - [16 CFR Part 465: Rule on the Use of Consumer Reviews and Testimonials](https://www.ecfr.gov/current/title-16/chapter-I/subchapter-D/part-465), Electronic Code of Federal Regulations - [The Consumer Reviews and Testimonials Rule: Questions and Answers](https://www.ftc.gov/business-guidance/resources/consumer-reviews-testimonials-rule-questions-answers), Federal Trade Commission, November 2024 - [Prohibited & restricted content (Maps User Generated Content Policy)](https://support.google.com/contributionpolicy/answer/7400114?hl=en), Google - [Tips to get more reviews](https://support.google.com/business/answer/3474122?hl=en), Google Business Profile Help - [Don't Ask for Reviews](https://www.yelp-support.com/article/Don-t-Ask-for-Reviews?l=en_US), Yelp Support Center --- title: "How to respond to bad restaurant reviews" description: "Short, calm replies for the bad reviews owners actually get: wrong orders, rude staff, food poisoning, fake reviews, plus what Google and the FTC forbid." canonical: "https://restaurantdoctorai.com/blog/how-to-respond-to-bad-restaurant-reviews" published: "2026-10-30" updated: "2026-10-30" --- # How to respond to bad restaurant reviews ## Key takeaways - Write the reply for the next guest reading it: thank, name the problem, say what changed, and give a way to reach you. - Never offer anything in return for changing or removing a review; Google bans it and the FTC warns it could be unfair. - Move food-poisoning and allergy claims offline at once, and admit or deny nothing in public before you've checked. - AI drafts are a fine start, but read every one and make it sound like you before it goes up. Reply to a bad restaurant review in three or four calm lines: thank the guest, name the problem in plain words, say what you've changed or are checking, and give a way to reach you. Never argue, never share what you know about the guest, and never offer something in public in return for changing the review. Below is a short reply for each of the hard cases owners actually get, from the wrong order to the food-poisoning claim. ## Who your reply is really for The guest who wrote the review has mostly made up their mind. The person you're writing for is the next one, scrolling your listing tonight and deciding where to eat. They read your reply to see how you treat people when something goes wrong. In a 2026 [survey by BrightLocal](https://www.brightlocal.com/research/local-consumer-review-survey/), a company that sells review software, 89% of 1,002 US adults said they expect owners to respond to reviews, and 80% said they were more likely to use a business that responds to all of them. That's what people say, across all local businesses, not only restaurants. There's also [one study of hotels](https://www.eurekalert.org/news-releases/898223) on TripAdvisor: those that started replying saw their ratings rise 0.12 stars on average and got about 12% more reviews. The negative reviews that kept coming were fewer but longer. That's hotels, not restaurants, so read it as a hint, not a promise. ## Five rules before you type 1. **Wait until you're calm.** A reply written during service is the one owners regret. Write it, leave it an hour, then read it as a stranger would. 2. **Keep it short.** [Google's own advice](https://support.google.com/business/answer/3474122?hl=en) is to reply promptly, stay professional and polite, keep it short, and apologize when it's right. For anything complicated, it suggests asking the guest to call or email you. 3. **No deals in the reply.** The same page says to avoid offering deals or promotions in your response. Make it right in private, not as a public coupon. 4. **Never trade anything for an edit.** Under the [FTC's rule on reviews](https://www.ftc.gov/business-guidance/resources/consumer-reviews-testimonials-rule-questions-answers), in effect since October 21, 2024, you can't try to get a review taken down with groundless legal threats, intimidation or knowingly false accusations about the reviewer. The FTC also warns that offering something to change or remove a negative review could be an unfair practice. And [Google's content policy](https://support.google.com/contributionpolicy/answer/7400114?hl=en) bans rewards for revising or removing one. 5. **Nothing private.** No guest's name, order, visit date, what they drank or how often they come. Google's policy bars posting personal information without consent, and it reads badly even when it's true. ## Replies for the hard cases Each reply below is an example to make your own. Swap in what really happened and what you really changed. Sign it with your first name and role, so it reads like a person and not a brand. | The review | What to do | A short reply | | --- | --- | --- | | Wrong order (in house or delivery) | Own it without blaming the driver or the app. | Thanks for telling us, and sorry your order came out wrong. We've added a check of every bag before it leaves. If you'd like us to make it right, reach me at [phone or email]. Maria, owner | | Slow service | Don't explain the staffing. Say what changes. | Sorry about the wait, that's not the visit we want for anyone. We've changed how we staff Friday nights. I hope you'll give us another try. Maria | | Rude staff | Never name or defend the employee in public. | Thank you for telling us. That isn't how we want anyone treated here, and I'm talking it through with the team. Please reach me directly at [phone or email]. Maria | | Too expensive | Don't argue the price or list your costs. | Thanks for the honest feedback. We price from what our ingredients cost us, and we're always looking at value. Sorry this one didn't feel worth it. | | Food-poisoning claim | Take it seriously, admit nothing you haven't checked, move it offline at once. | We're sorry you were unwell, and we take this seriously. Please contact me at [phone or email] so I can get the details and look into it right away. Maria | | Allergy reaction or a missed allergy | Same as above. Don't debate what was in the dish in public. | I'm very sorry. Allergies are serious, and I want to understand exactly what happened. Please reach me at [phone or email]. Maria | | A review you think is fake or from a competitor | Flag it to Google first. If you reply, stay neutral and never accuse. | We can't find a visit that matches this, but we'd like to look into it. Please contact me at [phone or email]. Maria | | Something you've already fixed | Say what changed, once, without arguing the past. | Thanks for this. You were right about the patio heaters, and they've been replaced since. We'd love to have you back. | | One star, no words | A one-line reply is enough. Don't guess what went wrong. | Sorry we missed the mark. If you're willing to tell us what happened, I'm at [phone or email]. Maria | *Sample replies for the reviews owners find hardest to answer. All made up; edit before you post.* The fake-review reply leaves out any detail of who you think wrote it. “We can't find a visit” is fine. “You're the guy from across the street” is an accusation, and if you're wrong, it's exactly the kind the FTC rule names. Food safety and allergies: say less in public A public reply is not the place to admit fault or to deny it. Get the details in private: what they ate, when, and who else was at the table. Then check your own records for that day, such as temperature logs and who was on the line. Those details stay private on both sides. ## When to report a review instead You can [flag a review on Google](https://support.google.com/business/answer/4596773?hl=en) from your Business Profile or the Reviews Management Tool, but only when it breaks Google's rules. The page says plainly not to report a review just because you disagree with it. A review is removed for breaking the policy, never for being harsh. - It's fake or doesn't come from a real visit to your location. - It's from a conflict of interest, such as a competitor or a current or former employee. - It shares someone's personal information without their consent. - It's off-topic: a political rant or a complaint about a different business. - It's hateful, harassing or sexually explicit. Google says a decision typically takes several days, and you get one appeal per review through the same tool. Reply calmly in the meantime, as if it might stay. ## Should AI write your replies? It can write the first draft. Google itself has been [testing a reply-with-AI option](https://www.seroundtable.com/google-reply-to-reviews-with-al-41095.html) inside Business Profile for some businesses this year, and plenty of tools do the same. The risk isn't the draft. It's posting it without reading it: a reply that thanks a guest for a lovely meal they hated, or one that admits a food-safety fault nobody has checked. Treat any draft, ours included, the way you'd treat a [chatbot's numbers](https://restaurantdoctorai.com/blog/chatgpt-restaurant-numbers): useful, and yours to check. Read it out loud. If it doesn't sound like you, change the words until it does. Our [review reader](https://restaurantdoctorai.com/reviews) reads your newest Google reviews and drafts a reply for the unanswered one to three star reviews that have words, with a link to the review, so you can start from a calm draft instead of a blank box. You fill in what you've changed and how to reach you, read it, and post it yourself. Nothing is posted for you. It reads Google reviews only, not Yelp or delivery apps. ## The reply is half the job One review about a long wait is a bad night. Five in a month is a schedule. Before you write the tenth apology for the same thing, count how often each complaint comes up. If it's the wait, look at how many sales each hour of staff covers at the rush with the [sales per labor hour guide](https://restaurantdoctorai.com/guides/sales-per-labor-hour). If sales are down too, the [checks to run before you cut](https://restaurantdoctorai.com/blog/restaurant-sales-down) include reading your reviews month by month. On timing, there is no Google hour rule. In the same BrightLocal survey, 19% said they expect a reply the same day and 81% within a week. A habit that holds: one-star and two-star reviews the same day or the next, the rest in a weekly sweep. ## This week 1. Open your Google Business Profile and list every one to three star review from the last three months with no reply. 2. Answer the oldest first, three or four lines each, signed with your first name. 3. Flag any that break Google's rules, and reply calmly to those too. 4. Write down the complaint each one is about, and circle the one that comes up most. 5. Pick one change for that complaint and tell the team before Friday. 45-second profit check See where your restaurant stands Reviews show what guests feel. The free check shows the numbers behind it: 12 quick questions and a grade for each cost area in about 45 seconds, so you know which one to look at first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Local Consumer Review Survey 2026](https://www.brightlocal.com/research/local-consumer-review-survey/), BrightLocal (review software company), February 11, 2026 - [Online reputation management: hotels that respond to reviews get better ratings (Proserpio and Zervas, Marketing Science)](https://www.eurekalert.org/news-releases/898223), INFORMS via EurekAlert!, September 11, 2017 - [Tips to get more reviews](https://support.google.com/business/answer/3474122?hl=en), Google Business Profile Help, Read October 30, 2026 - [Consumer Reviews and Testimonials Rule: Questions and Answers](https://www.ftc.gov/business-guidance/resources/consumer-reviews-testimonials-rule-questions-answers), Federal Trade Commission, Read October 30, 2026 - [Maps user contributed content policy](https://support.google.com/contributionpolicy/answer/7400114?hl=en), Google, Read October 30, 2026 - [Report inappropriate reviews on your Business Profile](https://support.google.com/business/answer/4596773?hl=en), Google Business Profile Help, Read October 30, 2026 - [Google Tests Reply To Reviews With AI](https://www.seroundtable.com/google-reply-to-reviews-with-al-41095.html), Search Engine Roundtable, March 24, 2026 --- title: "Who pays for that “best POS” list? Referral fees, explained" description: "How restaurant software referral fees work, what Toast pays advocates, what the FTC says about paid links, and six signs a software list is paid for." canonical: "https://restaurantdoctorai.com/blog/restaurant-software-referral-fees" published: "2026-10-22" updated: "2026-10-22" --- # Who pays for that “best POS” list? Referral fees, explained ## Key takeaways - Many software recommendations are paid: referral programs pay sites, consultants and friends a fee when you sign up. - Toast's published terms pay advocates $1,000 for a first qualified referral and up to $1,500 for later ones. - The FTC says a paid connection should be disclosed clearly, and the closer to the recommendation the better. - Before you trust a list, ask whether they're paid, what the tool is bad at, and what you'll pay in a year, all in. Many “best POS for restaurants” lists are paid for. Not always by the list's writer being bribed, but by referral programs: the software company pays the site, the consultant or the friend a fee when you sign up. It's legal and common. What matters is whether you can tell, so here is how the money moves, what the rules say, and how to read a recommendation before you trust it. ## How the money moves Software companies grow by referral. The simplest version is a program that pays anyone who sends them a new customer. Toast publishes its terms. Under its [Toast Advocates Program](https://pos.toasttab.com/toast-advocates-terms), consultants, investors, friends and customers' staff in the US can refer a restaurant, including through a personal link shared on a website or blog. When the restaurant signs and has been live for at least seven days, Toast pays: | Referral | Paid to the advocate | | --- | --- | | 1st qualified referral | $1,000 | | 2nd and 3rd | $1,250 each | | 4th and every one after | $1,500 each | | Later locations of the same group (up to 14) | $500 each | *Toast Advocates Program bonus payouts, from Toast's own terms (effective March 1, 2026).* Toast is open about it, and that's to its credit: the terms are public. Its page also says that businesses with a partner agreement are paid under that agreement instead, so what a bigger partner earns isn't on the page. The other version runs through networks. Companies such as PartnerStack run marketplaces where software companies list their partner programs and approve the sites, consultants and agencies who promote them. Its own help center describes the network as the place partners go to [“join partner programs and earn commissions”](https://support.partnerstack.com/hc/en-us/articles/20574018677395). A site in that network earns when its readers sign up. None of this is wrong Getting paid for a referral is normal business. A recommendation can be paid and still be right. The only problem is a paid recommendation that looks like an independent one. ## What the rules say In the US, the Federal Trade Commission's endorsement guidance covers exactly this. Its plain-language [questions and answers](https://www.ftc.gov/business-guidance/resources/ftcs-endorsement-guides-what-people-are-asking) say that someone who earns a commission through links should disclose it clearly, so readers can decide how much weight to give the recommendation, and that [“the closer the disclosure is to your recommendation, the better”](https://www.ftc.gov/business-guidance/resources/ftcs-endorsement-guides-what-people-are-asking). - **“Paid link” next to the link** should be enough, the FTC says. - **“Affiliate link” on its own** may not be understood, and a “buy now” button isn't a disclosure at all. - **A note in the footer**, far from the recommendation, is the kind of placement the guidance warns about: a reader may not make the connection. The guidance is about telling readers, not about whether a site may earn. So the question for you is simple: when you read the list, could you see who pays? ## Six signs a list is paid None of these proves a list is bought. Two or three together are a reason to read it more carefully. | What you see | What it can mean | | --- | --- | | Every product has a “Get a demo” or “Get started” button | Each button may be a tracked referral link | | No prices, or “contact us for pricing” everywhere | The list is selling the sign-up, not helping you compare | | No downsides for any product | Nobody is paid to tell you what a tool is bad at | | The disclosure is only in the footer or on an About page | It's far from the recommendation, where readers miss it | | The same three brands top every list you find | Those may be the programs that pay best | | The writer sells software or services too | Their own product can quietly win | *What to look for on a software list, and why it matters.* ## Questions to ask before you trust a recommendation These work on a website, a consultant, an accountant, a friend in the business or a sales rep. People who are paid fairly don't mind them. 1. Do you get paid if I sign up? How much, roughly? 2. Which tool would you pick for a place like mine if nobody paid you? 3. What is this tool bad at, and who should not buy it? 4. What will I pay in a year, all in: software, hardware, card processing and add-ons? 5. How do I leave, and what does leaving cost? The fourth question matters most. Software is rarely the biggest number: card processing often is. Our post on [what restaurant software costs a year](https://restaurantdoctorai.com/blog/restaurant-software-cost-per-year) adds it all up. [Free calculatorBefore you switch: card fees calculatorSee what your card processing costs per 100 dollars of card sales, against a dated fair range, from your own statement.](https://restaurantdoctorai.com/calculators/credit-card-fees) ## How we do it We'd rather show our own rule than ask you to take it on faith. Our free profit check, calculators and AI tools recommend no product. If we ever show a company under a result, the link is marked “Paid link” right next to it, a fee never changes your grade or what we write about your restaurant, and the companies are listed on our [paid links page](https://restaurantdoctorai.com/partner-links). Right now we show none. ## This week 1. Find the list or person that recommended the software you use most, and check it for the six signs above. 2. Pull last month's statements and add up what that software really costs you: subscription, hardware, card processing, add-ons. 3. Put your card statement through the card fees calculator to see where you stand. 4. Next time a rep or a list recommends a tool, ask the five questions before you book the demo. 45-second profit check Want an answer nobody paid for? The free check asks 12 quick questions about how your restaurant runs and shows where money is going first, in about 45 seconds. It recommends no product. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [US Toast Advocates Program Terms & Conditions](https://pos.toasttab.com/toast-advocates-terms), Toast, Effective for referrals submitted on or after March 1, 2026 - [Joining the PartnerStack Network](https://support.partnerstack.com/hc/en-us/articles/20574018677395), PartnerStack - [FTC's Endorsement Guides: What People Are Asking](https://www.ftc.gov/business-guidance/resources/ftcs-endorsement-guides-what-people-are-asking), Federal Trade Commission --- title: "The restaurant audit is moving into your AI assistant" description: "Why the next restaurant audit starts in an AI assistant, why it needs real tools rather than guesses, and how to connect ours for free today." canonical: "https://restaurantdoctorai.com/blog/ai-agents-restaurant-audits" published: "2026-10-22" updated: "2026-10-22" --- # The restaurant audit is moving into your AI assistant ## Key takeaways - A restaurant audit used to mean a consultant and weeks of waiting; more owners now start by asking an AI assistant. - MCP, an open standard Anthropic introduced in November 2024, lets assistants call tools such as calculators. - A tool runs the same formula every time and links to a page you can check; a chat model on its own can guess wrong. - Our free MCP server has 21 read-only tools with no sign-in, listed in the Official MCP Registry and GitHub's MCP Registry. A restaurant audit used to mean a consultant, a stack of reports and a few weeks of waiting. More and more owners now start by asking an AI assistant. That only helps if the assistant can do the sums instead of guessing them, and that is what tools built for AI agents are for: the assistant calls the tool, the tool runs fixed formulas, and the owner gets a number they can check. ## What a restaurant audit used to look like For most independent owners, a real look at the numbers meant paying someone. A consultant or an accountant asks for the P&L, the invoices and the payroll, works through them, and comes back with what costs too much and what to change first. It is useful work, and it takes time and money that a busy restaurant often can't spare in the middle of a season. We compare the options in [a restaurant consultant or a profit report](https://restaurantdoctorai.com/blog/restaurant-consultant-vs-profit-report). So most owners check their numbers rarely, or never in one sitting. They see the bank balance, they feel the busy nights, and they find out at the end of the month that there wasn't much left. ## What changed: assistants that can use tools In November 2024, Anthropic [introduced the Model Context Protocol](https://www.anthropic.com/news/model-context-protocol) (MCP), an open standard for connecting AI assistants to the systems and tools where real work happens. A business puts its tools on an MCP server, and any assistant that speaks MCP can use them. The project's own [introduction](https://modelcontextprotocol.io/docs/getting-started/intro) names Claude and ChatGPT among the apps that connect this way, and calculators among the tools, and puts it in one line: [“Think of MCP like a USB-C port for AI applications”](https://modelcontextprotocol.io/docs/getting-started/intro). For an owner, that changes who does the work. Instead of opening five websites and typing the same numbers into each, you tell your assistant what you sold, what you bought and what you paid in wages, and it calls the tools it needs. The audit moves into the conversation you were already having. ## Why a tool beats a guess A chat model on its own writes the most likely next words. Ask it for your prime cost and it will usually get it right, but it can also get it wrong and sound just as sure. We wrote about that in [ChatGPT for restaurants: how to check its numbers](https://restaurantdoctorai.com/blog/chatgpt-restaurant-numbers). A tool is different in three ways that matter when money is on the line: - **It runs the same formula every time.** Food, drink and labor divided by sales is the same sum on a Tuesday and on a Saturday, whatever words you used to ask. - **It answers with its working.** Our tools return the figures they used and the steps, so the assistant can show you how it got there. - **It links back to a page you can check.** Every answer from our tools comes with a link that opens the same result on restaurantdoctorai.com, where you can change a number and see what moves. The assistant talks, the tool counts Let the assistant do what it's good at: understanding your question and explaining the answer. Let a tool do the arithmetic. ## What our free tools do for an assistant RestaurantDoctorAI's free MCP server has 21 tools, all read-only, with no account and no sign-in. Nothing you send is stored, apart from anonymous daily counts. The [how our AI tools work](https://restaurantdoctorai.com/ai-assistants/how-it-works) page lists every one, word for word. | Ask your assistant | What runs | | --- | --- | | “Run a profit check for my restaurant.” | The 12-question free check: a grade, grades by area and the biggest place money is going | | “What's my prime cost? We sold 24,000 and spent…” | The prime cost calculator, against the target for your type of restaurant | | “Cost this recipe for me.” | A recipe costing draft, with a link to check every line | | “What should this burger cost on the delivery app?” | The delivery app price calculator | | “How much do we need to sell to break even?” | The break-even calculator, with and without your own pay | | “What tax and permit dates are coming up in Texas?” | The deadline calendar for your state | | “Is my website doing its job?” | The website check: phones, menu, online orders, speed | | “How do I work out my food cost?” | A search of our guides and posts, with sources | *Ask your assistant this, and it can use these tools.* There are also calculators for labor cost, card fees, booking app fees, the FICA tip credit, a monthly budget, plate cost and menu calories, and the sheets for a nightly cash-out, a tip-out and par levels. When an owner wants the full report, an assistant can buy it for them, with their say-so, through the separate tools listed on our [AI assistants page](https://restaurantdoctorai.com/ai-assistants). [Free calculatorOr do it yourself: prime cost calculatorEvery tool is also a page on this site. Start with prime cost: food, drink and labor as a share of sales.](https://restaurantdoctorai.com/calculators/prime-cost) ## Where assistants find them An assistant can only use a tool it can find. That's why we list our server wherever assistants and the people who set them up look for tools: - The [Official MCP Registry](https://registry.modelcontextprotocol.io/v0.1/servers?search=com.restaurantdoctorai%2Fprofit-check), the MCP project's own list, which other directories copy from. Ours is com.restaurantdoctorai/profit-check. - GitHub's [MCP Registry](https://github.com/mcp), which VS Code searches too. GitHub approved ours on October 21, 2026. - Directories such as Glama and mcpservers.org, listed with links on our AI assistants page. We think this is where restaurant tools are heading. The owner's assistant becomes the front desk, and the businesses whose tools it can reach are the ones it can help with. A calculator that only works when a person finds the page and types into it will be found less and less often. ## What it can't do A tool is only as good as the numbers you give it. The free check works from your answers, so its figures are estimates, and a calculator works from the figures you type. Neither is an audit of your books. Use the result to decide where to look first, then check that area against your own invoices and payroll. ## This week 1. In Claude on the web or the desktop app, open Customize › Connectors › + › Add custom connector, name it RestaurantDoctorAI, paste https://restaurantdoctorai.com/api/mcp and choose no sign-in. 2. Ask it to run the free profit check for your restaurant. Answer its 12 questions in your own words. 3. Give it last week's sales, food and drink purchases and wages, and ask for your prime cost. 4. Open the link in each answer and check the numbers on the page. 5. Pick the one area the check puts first and look at that area's invoices this week. 45-second profit check Rather do it here? The same free check runs on this site: 12 quick questions about how your restaurant runs, a grade for each cost area and the biggest place money is going, in about 45 seconds. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Introducing the Model Context Protocol](https://www.anthropic.com/news/model-context-protocol), Anthropic, November 25, 2024 - [What is the Model Context Protocol (MCP)?](https://modelcontextprotocol.io/docs/getting-started/intro), Model Context Protocol - [Official MCP Registry: com.restaurantdoctorai/profit-check](https://registry.modelcontextprotocol.io/v0.1/servers?search=com.restaurantdoctorai%2Fprofit-check), Model Context Protocol - [MCP Registry](https://github.com/mcp), GitHub --- title: "How to price catering: one 40-person order, costed" description: "How to price catering per person: food at plate cost, packaging, staff hours at true cost, delivery and rentals, worked out for one order of 40." canonical: "https://restaurantdoctorai.com/blog/how-to-price-catering" published: "2026-10-21" updated: "2026-10-21" --- # How to price catering: one 40-person order, costed ## Key takeaways - A catering price set from food cost alone can come up short: in the example, a quote of $30 a guest for 40 guests misses the order's own costs by $7.28. - Count every staff hour at its true cost. BLS put employer cost in accommodation and food services at about $1.24 per $1.00 of wages in June 2026, tips not counted. - Price the order at the higher of two numbers: food cost at your target, and all the order's costs divided by what's left after your margin and card fees. - The FDA Food Code keeps hot food at 135°F or above and cold at 41°F or below; health departments adopt their own version, so check yours before you plan the drive. A catering order for 40 can look like your best night of December and still lose money. Price it from everything the order uses: the food at plate cost, the packaging, every staff hour at its true cost, the drive and any rentals, then add the margin you want and check the food cost. In the worked example below, a quote set from food cost alone comes out short before card fees, and the price that covers everything is 40% higher. If holiday party orders are starting to come in, this post works one order line by line, with every table adding up. Copy the method, not the numbers. ## One order, priced from food cost Here is an example, made up so you can follow the sums. Linden Avenue Kitchen, an example full-service restaurant, gets a call: an office holiday party for 40 people, a hot buffet, delivered 30 miles away, set up, and kept tidy for three hours. Every name and number here is an example. The owner starts with plate cost. Each dish has a costed recipe, and this is what one guest's share of the buffet costs to make. | Dish | Per guest | | --- | --- | | Roast chicken, 5 oz cooked | $2.40 | | Braised beef, 4 oz cooked | $3.00 | | Penne with roasted vegetables | $1.00 | | Roasted potatoes | $0.50 | | Green salad and dressing | $0.80 | | Rolls and butter | $0.40 | | Dessert bars | $0.90 | | Food per guest | $9.00 | *Linden Avenue Kitchen (example): the buffet menu, food cost per guest.* The restaurant aims for food cost at 30% of the price, so the owner divides: $9.00 ÷ 0.30 = $30.00 a guest. That is $1,200 for one order, with the food cost right on target. It looks like a great night. The trouble is that a catering order brings costs the dining room never has: pans and plates that leave the building, a van, hours of driving and setting up, staff at someone else's buffet. None of them is in the $9.00. ## What the order really uses **Food, with a cushion.** Guests serve themselves at a buffet, so exactly 40 portions can run out. Linden Avenue makes 4 extra, a 10% cushion from its own past orders: 44 × $9.00 = $396.00. You pay for the cushion, so it goes in the cost. [Free calculatorCost the whole order: plate cost calculatorEnter the ingredients for the full batch and how many portions it makes (“Makes how many plates?”), and it gives the cost per portion and the price at your target food cost.](https://restaurantdoctorai.com/calculators/recipe-cost) **Packaging and disposables.** Everything that leaves with the food and doesn't come back. Price each one from your own invoices. | Item | Working | Cost | | --- | --- | --- | | Foil pans with lids | 12 × $2.50 | $30.00 | | Chafing fuel cans | 8 × $2.00 | $16.00 | | Plate, cutlery and napkin sets | 50 × $1.20 | $60.00 | | Cups for water | 50 × $0.40 | $20.00 | | Serving spoons and tongs | 8 × $1.50 | $12.00 | | Foil, labels, bags, tape, gloves | | $12.00 | | Total | | $150.00 | *Example: packaging and disposables for the order, at made-up prices.* Fifty sets for 40 guests, because some people take a second plate for dessert. That's $3.75 a guest the $30.00 quote never saw. **The van.** Fuel, insurance and wear are real costs even with no bill for this one trip. If you don't track your own cost per mile, one stand-in is the IRS's [business standard mileage rate](https://www.irs.gov/tax-professionals/standard-mileage-rates), 76 cents a mile from July 1, 2026 (it was 72.5 cents from January 1 to June 30). The IRS sets it for working out a deduction for business driving, not as a measure of your van, so your own number may be higher or lower. Linden Avenue's round trip is 60 miles: 60 × $0.76 = $45.60. **Rentals.** Linden Avenue owns its hotel pans but not enough chafing dishes, so it rents 4 at $15.00 and 2 tablecloths at $10.00 from a local rental company (example prices): $80.00. ## Staff hours at their true cost In the example this is the biggest line. List every hour the order takes, not just the cooking. | Task | Hours | | --- | --- | | Taking the order: calls, quote, shopping list | 1.0 | | Extra prep and cooking, two cooks, 4 hours each | 8.0 | | Packing and loading the van | 2.0 | | Driving, two people, 45 minutes each way | 3.0 | | Setting up the buffet, two people | 2.0 | | Running the buffet and clearing, two people, 3 hours | 6.0 | | Unloading, washing and putting away | 2.0 | | Total | 24.0 | *Example: every staff hour the order takes at Linden Avenue Kitchen.* An hour costs more than the wage on the schedule. The Bureau of Labor Statistics' [Employer Costs for Employee Compensation release](https://www.bls.gov/news.release/ecec.nr0.htm) for June 2026, published September 9, 2026, gives the national picture. Its [Table 4](https://www.bls.gov/news.release/ecec.t04.htm) puts employer costs in accommodation and food services at $20.03 per hour worked, of which $16.18 is wages and $3.85 is benefits: about $1.24 of total cost for each $1.00 of wages. Know what that figure is before you use it. BLS [measures it per hour worked](https://www.bls.gov/opub/hom/ecec/concepts.htm), it's an average of every job in the industry, hotels included, and its wages leave out tips, which BLS counts as payments from someone else. It's a starting point until you have your own figure, which [our post on the true cost of an hour of labor](https://restaurantdoctorai.com/blog/true-cost-of-an-hour-of-restaurant-labor) shows how to work out from your payroll, taxes and insurance. *Example: the order's staff hours, at the BLS average ratio* - Wage, all staff on this order: $18.00 - Cost per hour: $18.00 × 1.24: $22.32 - Wages alone: 24 × $18.00: $432.00 - Taxes and benefits on top: $103.68 **Staff cost: 24 × $22.32:** $535.68 All figures are examples, and one wage for everyone keeps the sums simple. Use your own cost per hour for each person on the order. Three things change this line in December: - **Overtime.** If the order's hours push a cook past 40 in the week, those hours cost more than the regular ones. BLS counts the overtime premium in benefits, not wages, so the 1.24 already holds an industry-average share of it; for your own cook, the labor hour post works an overtime hour out from your rates. - **Tipped staff.** A server's hour is costed differently; the labor hour post shows how. - **Your own time.** If you drive the van yourself, the table doesn't pay you. Price your hours in, or know the margin is paying you. For staff cost as a share of sales across the month, use the [labor cost calculator](https://restaurantdoctorai.com/calculators/labor-cost). ## The full cost, and the price per guest | Cost | For the order | Per guest | | --- | --- | --- | | Food, 44 portions | $396.00 | $9.90 | | Packaging and disposables | $150.00 | $3.75 | | Staff, 24 hours | $535.68 | $13.39 | | Van, 60 miles | $45.60 | $1.14 | | Rentals | $80.00 | $2.00 | | All costs | $1,207.28 | $30.18 | | The $30.00 quote | $1,200.00 | $30.00 | | Short by | $7.28 | $0.18 | *Example: what the order costs Linden Avenue Kitchen, against the $1,200 quote.* The quote doesn't cover the order's own costs. If the client pays by card at an example rate of 3%, another $36.00 comes off, so the order ends $43.28 short, and pays nothing toward rent or insurance. Staff hours are 44% of the costs and food only 33%: the owner priced from the smaller line. *The price for the order* **Price from food = food cost ÷ target food cost:** Price from all costs = all costs ÷ (1 − margin − card fee) Use the higher of the two, then divide by the guests and round up. Your margin is the share you want left for rent, utilities, insurance, your own time and profit. Linden Avenue wants 25% of the price left after the order's own costs and card fees. That 25% is the example's own choice, not a benchmark, and so is its 30% food cost target. For the food cost targets RestaurantDoctorAI uses for each type of restaurant, which are our own and not industry figures, see [the food cost guide](https://restaurantdoctorai.com/guides/food-cost-percentage#targets-by-service-format). *Example: the price per guest, worked out* - All costs ÷ (1 − 0.25 − 0.03): $1,207.28 ÷ 0.72: $1,676.78 - Per guest: $1,676.78 ÷ 40: $41.92 - From food: $396.00 ÷ 0.30 ÷ 40: $33.00 - The higher, rounded up: $42.00 - The order: 40 × $42.00: $1,680.00 - Card fee at 3%: $50.40 - Left: $1,680.00 − $1,207.28 − $50.40: $422.32 **Price per guest:** $42.00, 25.1% left All figures are examples. At $42.00 the food cost is 23.6% of the price, under the 30% target, because the other costs, not the food, set the price for this order. You don't have to show it as one number. The same $1,680 can be $34.00 a guest plus a $320 delivery and setup charge, which tells the client what the service costs. How a charge is named can change how it's taxed, though, so read the next section before you split it. The style of service moves the price as much as the menu does. If the office picks the food up instead, the drive, setup, buffet hours, van, rentals and fuel drop out. | Per order | Delivered and served | Picked up | | --- | --- | --- | | Food, 44 portions | $396.00 | $396.00 | | Packaging | $150.00 | $134.00 | | Staff | 24 hours, $535.68 | 13 hours, $290.16 | | Van and rentals | $125.60 | $0.00 | | All costs | $1,207.28 | $820.16 | | Price per guest from all costs | $41.92 | $28.48 | | Price per guest from food | $33.00 | $33.00 | | Price per guest, the higher | $42.00 | $33.00 | *Example: the same menu for 40, delivered and served or picked up, same rates as above.* For the pickup order the food cost target sets the price, leaving $460.24 after costs and a 3% card fee. Neither number is always the higher, so check both. There is no official price per person We found no government source that publishes a typical catering price per person. The ranges in search results come mostly from software sellers and wedding sites, rarely say how they were measured, and don't agree with each other. Set your price from your own costs. ## Hot, cold and the clock: food safety as a planning cost Food that travels has to stay safe on the way and on the buffet, and that costs equipment and staff time. The rules come from your health department, and they build on the FDA's [Food Code](https://www.fda.gov/food/retail-food-protection/fda-food-code), which the FDA describes as a model that local, state and federal regulators use to write their own rules. The Code itself says it is neither federal law nor federal regulation, and that every state has adopted some edition of it or the model codes before it. The newest full edition is 2026, so ask your health department which version it follows. In both the [2022 Food Code](https://www.fda.gov/media/164194/download) and the [2026 Food Code](https://www.fda.gov/media/194741/download), section 3-501.16 and section 3-501.19 say the same thing on these points: - **Hot holding:** food that needs temperature control for safety is kept at 135°F (57°C) or above, with an exception for some roasts. - **Cold holding:** at 41°F (5°C) or below. - **Time instead of temperature, up to 4 hours:** the food starts, in most cases, at 41°F or below or 135°F or above, is marked with the time 4 hours after it leaves temperature control, and is served or thrown away by then. - **Time for cold food, up to 6 hours:** the food starts at 41°F or below, never goes above 70°F (21°C), is checked and marked, and is served or thrown away within 6 hours. Either way, written procedures have to be prepared in advance and kept on hand for the health department. For the price, that means chafing dishes and fuel, insulated carriers, cold packs, a probe thermometer, and someone who checks temperatures. If you plan to use time instead of temperature, the clock starts when the food leaves hot or cold holding, so if it travels without heat, the drive and the setup count. In the example, a 45-minute drive, an hour of setup and a three-hour party add up to more than 4 hours, so the plan has to keep the food hot all the way, not rely on the clock. Food thrown away at the end is a cost too, which is one more reason the cushion sits in the food line. ## Sales tax, service charges and deposits Sales tax on catering differs by state, and so does the tax on what you add to the bill. Delivery fees, setup charges and service charges can be taxable in some states and not in others. One example, California, as its tax agency publishes it: - [Regulation 1603](https://www.cdtfa.ca.gov/lawguides/vol1/sutr/1603.html) says tax applies to the entire charge a caterer makes for serving meals, food and drinks, including the use of dishes, silverware, glasses, chairs and tables. It also says an optional tip isn't taxed, but a mandatory tip, gratuity or service charge is part of taxable gross receipts. - [Regulation 1628](https://www.cdtfa.ca.gov/lawguides/vol1/sutr/1628.html) taxes charges for delivery to the buyer unless several conditions are met, one of them that the charge is stated separately. That's one state, shown as an example. Your state's tax agency publishes its own rules, and your accountant can say how they apply to your invoice. A service charge can also change your payroll and your staff's tips: our [guide to restaurant service charges](https://restaurantdoctorai.com/guides/restaurant-service-charge) covers what it changes and how to tell guests. **Deposits and cancellations** are business terms, not tax rules, and they're easier to hold to in writing. A catering order is food bought and hours scheduled days before you're paid. Write down, in the quote the client accepts: - The deposit, how much and by when. One way to set it: enough to cover the food you'll buy. - The final guest count and the date it's due, and that you bill for that count even if fewer people come. - What happens if the client cancels: by which date the deposit comes back, and after which date it doesn't. - What the price includes (delivery, setup, service hours, rentals), and when the rest is paid. This is a list of terms to decide and write down, not legal advice. Agree them before you buy the food, so nobody is surprised later. ## This week 1. Cost your catering menu per guest in the [plate cost calculator](https://restaurantdoctorai.com/calculators/recipe-cost), as a batch with the number of portions it makes, cushion included. 2. Price your disposables from your last invoices: pans and lids, plate sets, cups, utensils, fuel. Write the cost per guest. 3. List every hour one past order took, from the first call to the last pan put away, and price it at your cost per hour, not the wage. 4. Measure the round trip to a typical client and price it at your own cost per mile, or the IRS rate as a stand-in. 5. Work your next quote both ways, from food cost and from all costs, and quote the higher. 6. Look up your state tax agency's page on catering and delivery charges, ask your health department which Food Code it follows, and write your deposit and cancellation terms into the quote. 45-second profit check Know your costs before the next big order The free check asks 12 quick questions, including how you cost recipes and how staffing changes when a shift slows, and grades each cost area in about 45 seconds, so you can see which one to watch before you quote. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Employer Costs for Employee Compensation Summary](https://www.bls.gov/news.release/ecec.nr0.htm), U.S. Bureau of Labor Statistics, September 9, 2026 - [Table 4. Employer Costs for Employee Compensation for private industry workers by occupational and industry group, June 2026](https://www.bls.gov/news.release/ecec.t04.htm), U.S. Bureau of Labor Statistics, September 9, 2026 - [Employer Costs for Employee Compensation: Concepts](https://www.bls.gov/opub/hom/ecec/concepts.htm), U.S. Bureau of Labor Statistics, Handbook of Methods, September 30, 2025 - [Standard mileage rates](https://www.irs.gov/tax-professionals/standard-mileage-rates), Internal Revenue Service, July 28, 2026 - [FDA Food Code](https://www.fda.gov/food/retail-food-protection/fda-food-code), U.S. Food and Drug Administration, September 17, 2026 - [Food Code 2022](https://www.fda.gov/media/164194/download), U.S. Food and Drug Administration, 2022 edition, January 18, 2023 version - [Food Code 2026](https://www.fda.gov/media/194741/download), U.S. Food and Drug Administration, 2026 edition - [Regulation 1603. Taxable Sales of Food Products](https://www.cdtfa.ca.gov/lawguides/vol1/sutr/1603.html), California Department of Tax and Fee Administration, Amended March 22, 2023 - [Regulation 1628. Transportation Charges](https://www.cdtfa.ca.gov/lawguides/vol1/sutr/1628.html), California Department of Tax and Fee Administration, Amended effective July 19, 1995 --- title: "Menu engineering: one 10-dish menu, worked through" description: "Menu engineering worked on a 10-dish menu: the 70% popularity line, the average margin line, and what to test with stars, plowhorses, puzzles and dogs." canonical: "https://restaurantdoctorai.com/blog/menu-engineering-worked-example" published: "2026-10-21" updated: "2026-10-21" --- # Menu engineering: one 10-dish menu, worked through ## Key takeaways - Menu engineering sorts each dish by plates sold and by dollars left per plate after its plate cost, compared within one group, such as mains. - A dish counts as popular at or above 70% of an equal share of plates sold; with ten mains that's 7%. - The profit line is the weighted average margin per plate, so a dish with a 20% food cost can still earn too few dollars to clear it. - It uses one period's sales and counts only food, not labor, so treat each move as a test and run the sort again next month. Menu engineering sorts each dish by two numbers: how many plates you sell, and how many dollars each plate leaves after its food cost. Two lines split the menu into four groups: stars, plowhorses, puzzles and dogs. In the worked menu below, the best seller ties for the lowest food cost percentage on the list and earns the fewest dollars per plate of all ten dishes. ## What menu engineering measures The method comes from Michael L. Kasavana and Donald I. Smith. A 2007 review of menu analysis methods dates it to 1982, and a [library record](https://agris.fao.org/search/en/providers/122535/records/65de0be563b8185d9ca6a1f6) lists a revised 1990 edition of their book, *Menu Engineering: A Practical Guide to Menu Analysis*. We couldn't open the book itself, so every rule below is as three articles in a peer-reviewed hospitality journal, *FIU Hospitality Review*, describe it. It asks two questions about each dish, compared with the other dishes in the same group (mains with mains): - **Is it popular?** Its share of all the plates sold in the group. - **Does it earn its keep?** Its **contribution margin**: the menu price minus the plate cost. It's the dollars one plate leaves to pay for wages, rent and profit. Plate cost means the food on the plate. A 1995 article by three University of Central Florida hospitality faculty, [LeBruto, Quain and Ashley](https://digitalscholarship.fiu.edu/record/17973/files/V13n195e.pdf), adds that any other cost that goes out with each plate, such as paper goods, belongs in it too. So if a dish leaves in a to-go box, count the box. They also sum up why the method looks at dollars: [“operators deposit cash, not percentages”](https://digitalscholarship.fiu.edu/record/17973/files/V13n195e.pdf). ## The two cut-off lines, as the method sets them **The popularity line.** Take an equal share: 100% divided by the number of dishes in the group. A dish counts as popular when its share of plates sold is at or above 70% of that. LeBruto, Quain and Ashley work an eight-item example: 100% ÷ 8 × 70% is 8.75%, so a dish is popular at 8.75% of plates sold or more. A 2001 study by [Horton](https://digitalscholarship.fiu.edu/record/18086/files/V19n22001d.pdf) puts it the same way: at, or above, 70% of the average. **The profit line.** Work out each dish's contribution margin, multiply it by the plates sold, add those up, and divide by all the plates sold. That's the menu's weighted average contribution margin. A dish above it counts as high; below it, low. LeBruto, Quain and Ashley and a 2007 review by [Taylor and Brown](https://digitalscholarship.fiu.edu/record/18216/files/Fall2007_7.pdf) both describe it this way. Horton's own study used a plain average of the dishes' margins instead, which can draw the line in a different place. We use the weighted one. *The two lines* Popularity line = (100% ÷ number of dishes) × 70%Contribution margin = menu price − plate costProfit line = sum of (margin × plates sold) ÷ total plates sold Count only the dishes in one group, such as mains. Sales counts come from your POS item report for the same period as your plate costs. | Group | Popular? | Margin against the line | | --- | --- | --- | | Star | Yes | Above | | Plowhorse | Yes | Below | | Puzzle | No | Above | | Dog | No | Below | *The four groups, by where a dish sits against the two lines.* ## One menu, worked all the way through Here's Larkspur Kitchen, a made-up independent restaurant. Every name and number is an example. It has ten mains and sold 2,000 of them last month. The plate costs are from its recipe cards at this month's invoice prices. | Dish | Sold | Price | Plate cost | Food cost | Margin a plate | Margin, month | | --- | --- | --- | --- | --- | --- | --- | | Spaghetti pomodoro | 360 | $14.00 | $2.80 | 20% | $11.20 | $4,032 | | Cheeseburger | 300 | $17.50 | $5.95 | 34% | $11.55 | $3,465 | | Roast chicken | 280 | $24.00 | $7.20 | 30% | $16.80 | $4,704 | | Fish tacos | 240 | $17.00 | $5.10 | 30% | $11.90 | $2,856 | | Salmon | 220 | $28.00 | $8.40 | 30% | $19.60 | $4,312 | | Ribeye | 150 | $38.00 | $15.20 | 40% | $22.80 | $3,420 | | Mushroom risotto | 120 | $18.00 | $3.60 | 20% | $14.40 | $1,728 | | Grain bowl | 120 | $16.00 | $4.00 | 25% | $12.00 | $1,440 | | Braised short rib | 110 | $32.00 | $11.20 | 35% | $20.80 | $2,288 | | Pork chop | 100 | $25.00 | $7.45 | 29.8% | $17.55 | $1,755 | | All ten | 2,000 | | | 30.3% | | $30,000 | *Larkspur Kitchen (example): ten mains, one month. Margin = price − plate cost; month = margin × sold.* The month comes to $43,050 in sales of mains and $13,050 in plate costs, a food cost of 30.3% ($13,050 ÷ $43,050). Take one from the other and you get the same $30,000 as the last column adds up to. *Example: Larkspur Kitchen's two lines* - Dishes in the group: 10 - Equal share: 100% ÷ 10: 10% - Popularity line: 10% × 70%: 7% - In plates: 7% × 2,000: 140 plates - Total margin (last column): $30,000 - Profit line: $30,000 ÷ 2,000 plates: $15.00 a plate **Popular at 140 plates or more; high margin above:** $15.00 All figures are examples. No dish sits exactly on either line. | Dish | Share of plates | Popular? | Margin vs $15.00 | Group | | --- | --- | --- | --- | --- | | Roast chicken | 14.0% | Yes | +$1.80 | Star | | Salmon | 11.0% | Yes | +$4.60 | Star | | Ribeye | 7.5% | Yes | +$7.80 | Star | | Spaghetti pomodoro | 18.0% | Yes | −$3.80 | Plowhorse | | Cheeseburger | 15.0% | Yes | −$3.45 | Plowhorse | | Fish tacos | 12.0% | Yes | −$3.10 | Plowhorse | | Braised short rib | 5.5% | No | +$5.80 | Puzzle | | Pork chop | 5.0% | No | +$2.55 | Puzzle | | Mushroom risotto | 6.0% | No | −$0.60 | Dog | | Grain bowl | 6.0% | No | −$3.00 | Dog | *Each dish against the two lines (140 plates, $15.00 a plate).* | Group | Dishes | Plates | Margin, month | Share of margin | | --- | --- | --- | --- | --- | | Stars | 3 | 650 | $12,436 | 41.5% | | Plowhorses | 3 | 900 | $10,353 | 34.5% | | Puzzles | 2 | 210 | $4,043 | 13.5% | | Dogs | 2 | 240 | $3,168 | 10.6% | | All ten | 10 | 2,000 | $30,000 | 100% | *What each group earned in the month.* The shares are rounded, so they add to 100.1%. Look at the plowhorses: they're 900 of the 2,000 plates and more than a third of the margin. The spaghetti alone left $4,032 of margin, the third-largest total on the menu, behind only the roast chicken and the salmon. A dish can earn little per plate and still matter a lot in total, so a plowhorse is something to work on, not to drop. ## Why the dollars decide, not the food cost percentage On a food cost sheet, the spaghetti looks like the best dish on the menu: 20%, tied for the lowest, and the biggest seller. By dollars, it earns $11.20 a plate, the least of all ten. The ribeye has the highest food cost on the menu at 40%, and each plate leaves $22.80, more than twice as much. The mushroom risotto is the clearest case. Its food cost is 20% too, and it still lands in the dog group: $14.40 a plate is under the $15.00 line, and 120 plates is under 140. A low percentage on a low price is a small number of dollars. LeBruto, Quain and Ashley make the same point with a $10 dish that costs $5 (50%, $5 left) against a $6 dish that costs $2 (33%, $4 left). *Example: 30 spaghetti orders become ribeye orders* - Extra sales: 30 × ($38.00 − $14.00): +$720 - Extra plate cost: 30 × ($15.20 − $2.80): +$372 - Food cost: $13,422 ÷ $43,770: 30.7% (was 30.3%) - Margin: 30 × ($22.80 − $11.20): +$348 **Food cost % up, dollars up:** $30,348 for the month All figures are examples. The food cost percentage got worse and the restaurant kept $348 more. That's why the method ranks dishes by dollars. The [food cost guide](https://restaurantdoctorai.com/guides/food-cost-percentage#plate-cost-for-a-single-dish) shows how to cost one plate. The percentage is still worth watching for the kitchen as a whole. It just can't rank dishes against each other on its own. ## What to try with each group LeBruto, Quain and Ashley list the actions Kasavana and Smith suggest for each group. Treat each one as a test: change one thing, wait a month, and run the numbers again. None of them is sure to work, and guests decide. **Stars (roast chicken, salmon, ribeye).** The method suggests testing a higher price on them. Keep the recipe and the portion the same while you do. Watch how much room a star has: the ribeye sold 150 plates, only 10 above the 140 line. At $2 more it leaves $24.80 a plate, and it needs 138 plates to earn the month's $3,420 again. That's below the line, so it could slip into the puzzle group while earning the same. **Plowhorses (spaghetti, cheeseburger, fish tacos).** The method suggests raising the price toward the line, as long as the dish stays popular. LeBruto, Quain and Ashley warn about dishes far below the line: a raise big enough to get them there can cost them their popularity. The spaghetti sits $3.80 under today's line, and because a raise on 360 plates lifts the average too, it would take about $4.65 more (a 33% raise) at the same sales to clear it. Try a smaller step, or bring the plate cost down. Taylor and Brown list retooling the recipe to cut its cost as the other option. That might be a cheaper side, or bringing the portion back to what the card says. *Example: a $1 raise on the spaghetti* - New price, margin a plate: $15.00, $12.20 - Last month: 360 × $11.20: $4,032 - Plates needed to match it: $4,032 ÷ $12.20: 330.5, so 331 - Plates it can lose: 360 − 331: 29, about 8% **Still popular at 331 plates?:** Yes, the line is 140 All figures are examples. If a guest who skips the spaghetti orders another main instead, you lose less than the whole plate. How many guests a raise can lose before it costs you is worked out in [how much to raise menu prices](https://restaurantdoctorai.com/blog/how-much-to-raise-menu-prices). **Puzzles (short rib, pork chop).** The method suggests featuring them on the menu or lowering the price. Taylor and Brown mention where they sit on the menu and how you promote them. Move the short rib to the top of the mains, or have servers mention it, for one month. Count the gain honestly. If 30 more guests order the short rib, that isn't 30 × $20.80 = $624 extra if they'd otherwise have had the spaghetti. In that case it's 30 × ($20.80 − $11.20) = $288. **Dogs (mushroom risotto, grain bowl).** The method suggests taking them off or rotating them with similar dishes. Ask one question first: are they the only meat-free mains? If so, they may be why some tables come in, and the sort can't see that. The risotto is only $0.60 under the line, so $1 more would put it at $15.40 a plate, above the line. Re-cost it first; a dog with a wrong plate cost is just a mistake on paper. ## What the method can't see - **One period's sales.** A month is a snapshot. A summer salad and a winter braise won't look the same in July, so compare like months, or run it again each season. - **New dishes have no history.** A dish added mid-month, or a special, hasn't had the same chance to sell. Leave it out or mark it until it has a full month. - **Labor and kitchen time.** LeBruto, Quain and Ashley list this as a main weakness: the margin counts only food. When Horton timed the work on each dish in one restaurant, adding labor moved 6 of 52 dishes into another group. One salad took $1.16 of labor against $1.07 of food. Food and labor together are your prime cost (see the [prime cost guide](https://restaurantdoctorai.com/guides/restaurant-prime-cost)). - **The group you compare in.** In Horton's study, looking at starters, salads and sandwiches, and entrees separately changed the group of 32 of the 52 dishes. Horton's point is to group dishes the way your guests order them: if a starter or a salad often stands in for a main, sort it with the mains. - **The lines move.** Both lines come from your own menu's averages, so some dish always sits below the profit line, and moving one dish moves the lines. Change one dish and the others can change group too, as Horton and Taylor and Brown both note. - **Price isn't everything.** Taylor and Brown warn that chasing margin can favor higher-priced dishes, and those can lower demand over time. - **Plate costs have to be right.** A card costed at last year's prices, or a portion heavier than the card, moves a dish across the line on paper. See [actual vs theoretical food cost](https://restaurantdoctorai.com/blog/actual-vs-theoretical-food-cost) for checking a card against what really leaves your stock. A sort, not a verdict The four labels tell you where to look, not what to do. A star after food cost can still lose money after labor and rent, and a dog can be the reason a table books. Use the groups to pick next month's test. [Free calculatorCost each plate first: plate cost calculatorAdd a dish's ingredients to get its plate cost and food cost. Take the plate cost from the menu price for its margin, then put the ten margins into the tables above.](https://restaurantdoctorai.com/calculators/recipe-cost) ## This week 1. Pull last month's item report from your POS: how many of each main you sold. Mains only, and leave out dishes that weren't on for the whole month. 2. Get a current plate cost for each main, at this month's invoice prices. Add the to-go box for anything that leaves in one. 3. Work out each dish's margin (price minus plate cost) and its month (margin × sold). Add the month column and divide by the plates sold: that's your profit line. 4. Work out the popularity line: 100% ÷ the number of mains × 70%, then times the plates sold. Sort each dish into star, plowhorse, puzzle or dog. 5. Pick one test for one dish: a small raise on a plowhorse, a better spot for a puzzle, or a re-cost of a dog. Write the date down and run the same sort next month. 45-second profit check Is food cost where to start? Menu prices are one lever. The free check asks 12 quick questions, including how often you update recipe costs, and grades each cost area in about 45 seconds, so you can see which one to work on first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Menu engineering: a practical guide to menu analysis (revised edition), catalog record](https://agris.fao.org/search/en/providers/122535/records/65de0be563b8185d9ca6a1f6), AGRIS, Food and Agriculture Organization of the United Nations (record from the U.S. National Agricultural Library), Undated record of the 1990 revised edition - [Menu Engineering: A Model Including Labor (LeBruto, Quain and Ashley)](https://digitalscholarship.fiu.edu/record/17973/files/V13n195e.pdf), FIU Hospitality Review, Florida International University, Spring 1995 (Volume 13, Number 1) - [Labor and Menu Category: Effects on Analysis (Horton)](https://digitalscholarship.fiu.edu/record/18086/files/V19n22001d.pdf), FIU Hospitality Review, Florida International University, 2001 (Volume 19, Number 2) - [Menu Analysis: A Review of Techniques and Approaches (Taylor and Brown)](https://digitalscholarship.fiu.edu/record/18216/files/Fall2007_7.pdf), FIU Hospitality Review, Florida International University, 2007 (Volume 25, Number 2) --- title: "How much do restaurant owners make? Find your own number" description: "Salary sites don't show what a restaurant owner keeps. Work out your own pay from last month's numbers, with an example at three sales levels." canonical: "https://restaurantdoctorai.com/blog/how-much-do-restaurant-owners-make" published: "2026-10-21" updated: "2026-10-21" --- # How much do restaurant owners make? Find your own number ## Key takeaways - Salary sites count paychecks, and the BLS manager pay figures leave out the self-employed, so neither shows what an owner keeps after every cost. - Your pay is net sales minus every cost except your own wage, plus any wage you paid yourself: your pay before your own taxes. - In the example, one restaurant leaves its owner a loss at $50,000 a month, $5,000 at $75,000 and $13,500 at $100,000, because the monthly bills come first. - Sole proprietors and partners take money out, not wages; an S corporation must pay an owner who works in it reasonable compensation, the IRS says. The figure on the salary sites isn't your number. Those sites count paychecks, and they mix in the wages of managers who run restaurants for someone else. An owner's pay is what is left after every cost, plus any wage the owner pays themselves through payroll. You can work out yours from last month's numbers in about 10 minutes, and this post shows how, with one example restaurant at three levels of sales. If you want to know what share of sales a typical restaurant keeps, start with [our post on the average restaurant profit margin](https://restaurantdoctorai.com/blog/average-restaurant-profit-margin). This one turns that share into the money that reaches you. ## Why the salary-site number isn't yours A salary figure answers “what does this job pay?” An owner's question is different: what did the business leave me after it paid everyone else? Two public sources show why the first can't answer the second. The Bureau of Labor Statistics publishes what hired managers earn. Its [Occupational Outlook Handbook page on food service managers](https://www.bls.gov/ooh/management/food-service-managers.htm) puts their median pay at $69,390 a year, or $33.36 an hour, in May 2025. In food services and drinking places the median was $65,060. The lowest 10% earned less than $45,960 and the highest 10% more than $107,640. The same page counts 31% of food service managers as self-employed, and says its pay figures leave the self-employed out, because the survey behind them doesn't collect their pay. So many owners who manage their own place, the self-employed ones, aren't in that figure at all. The National Restaurant Association publishes what restaurants keep. In its [August 20, 2025 article on its operations survey](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/new-association-report-helps-operators-gauge-their-restaurant-performance) of more than 900 restaurants, the median income before taxes in 2024 was 2.8% of sales for full-service restaurants and 4.0% for limited-service ones. Its [October 16, 2025 article on sales volume](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/higher-volume-restaurants-reported-lower-food-cost-ratios-in-2024/) gives 4.3% for full-service restaurants with sales of $2 million a year or more and 1.1% for those below. The Association says the figures are not meant as standards or goals, and the articles don't say how each restaurant counted its owner's pay. A restaurant that pays its owner a salary and one that doesn't can report very different margins on the same month. Two numbers, two questions A manager's wage is what someone is paid for the hours. An owner's pay is the wage for the hours plus whatever the business makes on the money and risk the owner put in, and that second part can be large, small or below zero. ## What owners' own tax returns show The closest public record of what restaurant owners keep comes from the IRS itself. Its Statistics of Income division publishes [nonfarm sole proprietorship statistics](https://www.irs.gov/statistics/soi-tax-stats-nonfarm-sole-proprietorship-statistics): business receipts, deductions and net income reported on Schedule C of Form 1040, the form a sole proprietor files. A sole proprietor can't deduct a salary for themselves, so the net income on that form is the owner's pay before their own income tax and self-employment tax. The [table for tax year 2023](https://www.irs.gov/pub/irs-soi/23sp01br.xls), published in March 2026, has a row for restaurants (full and limited service) and drinking places. | Tax year 2023 | All returns | With net income | | --- | --- | --- | | Returns | 653,372 | 376,535 | | Business receipts | $77.2 billion | $58.7 billion | | Net income (all returns: less losses) | $2.6 billion | $8.0 billion | | Receipts per return, average | about $118,000 | about $156,000 | | Net income per return, average | about $4,000 | about $21,300 | *Restaurants (full and limited service) and drinking places, nonfarm sole proprietorships, tax year 2023. Source: IRS Statistics of Income, Table 1, estimates based on samples. The averages are our division of the table's totals.* Read it with care, for four reasons: - **About 58% of these returns showed net income.** The other 276,837 showed none or a loss, which is why the average across all of them is so low. - **They are mostly small.** Average receipts of about $118,000 a year are under $10,000 a month. A sit-down restaurant with $75,000 a month in sales is far bigger than the average return here. - **Only Schedule C.** A restaurant run as a partnership or a corporation files a different return and isn't in this table. - **Tax profit isn't cash.** Net income is worked out by tax rules, which can differ from the money an owner could take out in the same year. So the table tells you that many small restaurant owners keep little, and some lose money. It can't tell you what you keep. Only your own month can. ## Your number in about 10 minutes You need last month's sales report, your bank statement or bills for the month, and your payroll report. A typical month works best; if your months swing, average the last three. *What the restaurant paid you* Owner's pay = net sales − every cost except your own wage + any wage you paid yourself Net sales are what you rang up after discounts and comps, without the sales tax you collected (that belongs to the state). Every cost means what the month's bills really were, with loan payments counted in full. 1. Write down last month's net sales from your sales report. 2. Add up every cost of running the restaurant that month: food and drink, staff pay with the employer's payroll taxes, rent and utilities, card fees, supplies, insurance, repairs, software, marketing, accounting and loan payments. Divide yearly bills by 12. 3. If you are on payroll yourself, leave your own wage out of that total, but keep the employer's taxes on it in. 4. Subtract the costs from net sales. That is what the restaurant left you this month. Add any wage you paid yourself. The total is your pay before your own taxes. 5. Divide it by the hours you worked that month. That is what each of your hours earned. Step 5 answers “is this worth it?” Compare your hourly figure with what you would pay a manager to work your hours. The BLS median above is one public yardstick, and a hired manager would also cost you the employer's taxes on that wage, which [our post on what an hour of labor costs](https://restaurantdoctorai.com/blog/true-cost-of-an-hour-of-restaurant-labor) lists. This is a cash view, as the [profitability test](https://restaurantdoctorai.com/guides/is-my-restaurant-profitable) explains: a loan payment counts in full here, while your P&L counts only its interest. Both are right for their own question. This one answers “what could I take out?” ## One restaurant at three levels of sales Here is an example, made up so you can follow the sums. Birch Lane Kitchen, an example full-service neighborhood restaurant, is a sole proprietorship. The owner works 55 hours a week and isn't on payroll. Some costs rise with every sale: food and drink at 32% of sales, hourly staff with payroll taxes at 28%, and card fees, supplies and packaging at 6%. The rest are bills that stay the same each month. All the rates and bills are round numbers for the example, not benchmarks. | Per month | $50,000 | $75,000 | $100,000 | | --- | --- | --- | --- | | Food and drink, 32% | $16,000 | $24,000 | $32,000 | | Hourly staff with payroll taxes, 28% | $14,000 | $21,000 | $28,000 | | Card fees, supplies, packaging, 6% | $3,000 | $4,500 | $6,000 | | Kitchen lead's salary with payroll taxes | $5,500 | $5,500 | $5,500 | | Rent and utilities | $8,000 | $8,000 | $8,000 | | Insurance, repairs, software, marketing, accounting | $5,000 | $5,000 | $5,000 | | Loan payment | $2,000 | $2,000 | $2,000 | | Left for the owner | −$3,500 | $5,000 | $13,500 | | Left for the owner, a year | −$42,000 | $60,000 | $162,000 | | Per hour the owner worked | below $0 | about $21 | about $57 | *Example: Birch Lane Kitchen, one month at three levels of sales. Made-up round numbers. The owner's hours are 55 a week, 2,860 a year.* The same restaurant, run the same way, leaves its owner three very different years. At $50,000 a month the owner puts in $3,500 a month instead of taking anything out. At $75,000 the owner keeps $60,000 a year, about $21 for each hour worked. That $60,000 is less than the $65,060 median for a hired manager in food services and drinking places, for a job the BLS says is usually full time, and it leaves nothing for the money and risk the owner carries. At $100,000 the owner keeps $162,000: the manager-level wage, plus $96,940 that is the business's return to its owner. *Example: why each extra sale counts so much* - Costs that rise with sales (32% + 28% + 6%): 66% - Left of each extra dollar of sales: 34 cents - Bills that stay the same: $20,500 - Break-even, owner paid nothing (÷ 0.34): $60,294 **Break-even with $5,400 a month for the owner:** $76,176 $5,400 is about $65,060 ÷ 12, the BLS median for a manager in food services and drinking places, rounded. (20,500 + 5,400) ÷ 0.34 = $76,176. Each extra $25,000 of sales adds $8,500 to the owner's month (34 cents × 25,000). That is the shape of owner pay. The monthly bills are paid first, from the first dollars of the month, and the owner is paid last, from what is left. Below break-even the owner pays in. Above it, every extra dollar of sales leaves 34 cents, so owner pay climbs much faster than sales. One honest limit: the example lets hourly pay move exactly with sales. Real staffing doesn't, because a slow night still needs a minimum crew, so a $50,000 month usually costs more in labor than this table shows. Food and labor together, called prime cost, run from 71% of sales at $50,000 to 65.5% at $100,000 here. The [prime cost guide](https://restaurantdoctorai.com/guides/restaurant-prime-cost) gives the ceiling for each type of restaurant, and it is the first place to look when what's left is thin. [Free calculatorFind your break-even, with and without your payPut your own pay in its own box, and the calculator shows break-even with your pay and without it. Leave it blank and break-even assumes you work for free.](https://restaurantdoctorai.com/calculators/break-even) ## Draw or salary: what the IRS says How you take money out depends on how the business is set up for tax. This is what the IRS pages say, not advice on which setup or how much pay is right for you. That is a question for your accountant. | How it's set up | How the owner is paid | | --- | --- | | Sole proprietor | Not an employee of the business, so no wage to yourself. What you take out is a withdrawal, kept in a drawing account. You pay self-employment tax and income tax on the business's net earnings. | | Partnership, and by default an LLC with two or more members | Partners are not employees and get no W-2 for what the partnership pays them, distributions included. Those payments go on Schedule K-1. | | S corporation | An owner who works in the business must be paid reasonable compensation as wages, with employment taxes. After that, the part of the profit that comes from other staff's work and the business's equipment and money can be paid as distributions, without employment taxes. Managing the staff counts as the owner's work. | | Corporation | An officer is generally an employee, paid wages that should generally match the duties. | *How owners are paid, by the way the business is set up, as the IRS pages linked below describe it.* - **Sole proprietors.** The IRS's [Tax Guide for Small Business (Publication 334)](https://www.irs.gov/publications/p334) says you can't deduct your own salary or personal withdrawals, and that as a sole proprietor you are not an employee of the business. It describes a drawing account as a separate account for the business income you take out for personal and family costs. It puts the self-employment tax rate at 15.3% of net earnings (12.4% Social Security, 2.9% Medicare), and says you generally make estimated tax payments if you expect to owe $1,000 or more, self-employment tax included. With no paycheck, nothing is withheld, so the tax money has to come out of what you take. - **Partners and corporation officers.** The IRS page on [paying yourself](https://www.irs.gov/businesses/small-businesses-self-employed/paying-yourself) says partners are not employees and should not get a Form W-2 for what the partnership pays them, distributions included, and that a corporation's officer is generally an employee whose wages should generally match their duties. - **LLCs.** The IRS's [LLC page](https://www.irs.gov/businesses/small-businesses-self-employed/limited-liability-company-llc) says it treats an LLC as a corporation, a partnership or part of its owner's return, depending on its elections and number of members. By default, two or more members make a partnership, and one member is disregarded as separate from the owner. - **S corporations.** The IRS's [S corporation compensation page](https://www.irs.gov/businesses/small-businesses-self-employed/s-corporation-compensation-and-medical-insurance-issues) says “S corporations must pay reasonable compensation to a shareholder-employee in return for services” before any non-wage distributions are made. It names no dollar amount. Among the factors it lists are training and experience, duties, time and effort spent on the business, and what comparable businesses pay for similar services. For the method above, the setup changes only one line. A sole proprietor or partner has no wage of their own on payroll, so everything left is their pay. An S corporation owner on payroll adds that wage back to what's left. Either way, the total is the same question: what did the restaurant pay you, all in? ## This week 1. Pull last month's net sales, every bill paid that month (loan payments in full, yearly bills ÷ 12) and your payroll report. 2. Work out what was left after every cost except your own wage, and add any wage you paid yourself. That is your pay for the month. 3. Write down the hours you worked last month, from your schedule or your best honest count, and divide. That is what an hour of yours earned. 4. Compare it with the BLS manager figure, $65,060 a year in food services and drinking places, and see whether your restaurant pays you at least what you'd pay someone to do your job. 5. Put the same month into the [break-even calculator](https://restaurantdoctorai.com/calculators/break-even) with your own pay in its box, to see the sales a month needs to pay you. 6. Ask your accountant how your business is set up for tax, how you should be taking money out, and how much to set aside for your own taxes. 45-second profit check See what's taking the most from your month The free check asks 12 quick questions about your restaurant and shows the cost areas most likely costing you profit, in about 45 seconds, so you know where the money that should reach you is going. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Food Service Managers, Occupational Outlook Handbook](https://www.bls.gov/ooh/management/food-service-managers.htm), U.S. Bureau of Labor Statistics, Last modified September 14, 2026 (pay for May 2025) - [New Association report helps operators gauge their restaurant performance](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/new-association-report-helps-operators-gauge-their-restaurant-performance), National Restaurant Association, August 20, 2025 - [Higher-volume restaurants reported lower food cost ratios in 2024](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/higher-volume-restaurants-reported-lower-food-cost-ratios-in-2024/), National Restaurant Association, October 16, 2025 - [SOI Tax Stats - Nonfarm Sole Proprietorship Statistics](https://www.irs.gov/statistics/soi-tax-stats-nonfarm-sole-proprietorship-statistics), Internal Revenue Service, Statistics of Income, July 20, 2026 - [Table 1. Nonfarm Sole Proprietorships: Business Receipts, Selected Deductions, Payroll, and Net Income, by Industrial Sectors, Tax Year 2023](https://www.irs.gov/pub/irs-soi/23sp01br.xls), Internal Revenue Service, Statistics of Income, March 2026 - [Publication 334 (2025), Tax Guide for Small Business](https://www.irs.gov/publications/p334), Internal Revenue Service, April 30, 2026 - [Paying yourself](https://www.irs.gov/businesses/small-businesses-self-employed/paying-yourself), Internal Revenue Service, May 8, 2026 - [Limited liability company (LLC)](https://www.irs.gov/businesses/small-businesses-self-employed/limited-liability-company-llc), Internal Revenue Service, October 10, 2026 - [S corporation compensation and medical insurance issues](https://www.irs.gov/businesses/small-businesses-self-employed/s-corporation-compensation-and-medical-insurance-issues), Internal Revenue Service, March 3, 2026 --- title: "Actual vs theoretical food cost: find your gap in a week" description: "Compare what your dishes should have used with what left your stock. A one-week by-hand count for a few top dishes, and the causes to check." canonical: "https://restaurantdoctorai.com/blog/actual-vs-theoretical-food-cost" published: "2026-10-20" updated: "2026-10-20" --- # Actual vs theoretical food cost: find your gap in a week ## Key takeaways - Theoretical use is what the dishes you sold should have used; actual use is opening count plus purchases minus closing count. The difference is your gap. - Raw price per pound is not cost per cooked pound: the USDA's guide lists 0.74 lb cooked from 1 lb of ground beef with up to 20% fat, a meal-program average. - We found no official acceptable gap, so set your own: correct the cards, weigh portions, log waste and staff meals, then see what gap is left over a few weeks. - A gap alone never proves theft. Check yield, portions, waste, comps, price changes and counting errors one at a time before you suspect a person. Theoretical food cost is what the dishes you sold should have used, worked out from your recipes and your sales counts. Actual food cost is what really left your stock: the opening count, plus what you bought, minus the closing count. The difference is your gap. A one-week count of three to five top dishes shows how big it is and which ingredient it sits in. ## The two numbers, in plain words **Theoretical** is a number on paper. Take each dish you sold, multiply the sales count by the amount of each ingredient on the recipe card, and add it up. It's what the kitchen would have used if every plate matched its card and nothing was lost, given away or miscounted. **Actual** is a number from your shelves. You count your stock at the start of the week, add every invoice for the week, and subtract your count at the end. It's everything that left the shelves, whether it landed on a guest's plate or not. The [food cost guide](https://restaurantdoctorai.com/guides/food-cost-percentage#how-to-calculate-food-cost-percentage) uses the same sum for the month's food cost percentage. The gap is actual minus theoretical. It's the food your records can't account for. This post is the bridge between two things you may already have: a plate cost for each dish, and a food cost percentage for the month. It lets you compare the two on a few dishes, over seven days, with a pencil. It doesn't repeat how to log waste (see [what food waste costs a restaurant](https://restaurantdoctorai.com/blog/restaurant-food-waste-cost)) or how to work out the percentage. ## A one-week test you can do by hand Pick three to five dishes that sell the most and the costliest ingredient or two in each. If other dishes on your menu use the same ingredient, count what they use as well, or the gap will include their share. 1. **Write the recipe for each dish as a weight.** How many ounces of each picked ingredient go on one plate, as the line really portions it. Weigh a few plates for this rather than copying an old card. Where the card is in cooked ounces, turn it into raw ounces first (the yield section below shows how). 2. **Count the picked ingredients at close on the last night of the week before.** Count the same shelves in the same order each time, and count open packs by weight. The IRS's [small business tax guide](https://www.irs.gov/publications/p334) describes an inventory form with columns for the item, quantity, unit price and value, and a space for who counted, who priced and who checked the sums. It's written for tax records, but the layout works for a weekly count too. 3. **Keep every invoice for those ingredients** for the seven days, with the weight on each one. Count what was received, not what was ordered. 4. **Pull the sales counts** for the same seven days from your POS item report: how many of each dish were sold. 5. **Count again at close on the last night of the test week.** Same shelves, same order, same units. 6. **Work out both numbers in pounds first, then price them.** Use one invoice price per pound for both, so a price change doesn't blur the result. *The two numbers and the gap* Theoretical use = dishes sold × amount on the plate (as purchased)Actual use = opening count + purchases − closing countGap = actual use − theoretical use, in pounds, then in dollars Work each picked ingredient in pounds, then price it at one price per pound, from your latest invoice. Do the same for each ingredient and add the dollars. ## A worked week Here's Maple Street Kitchen, a made-up full-service restaurant. Every name and number here is an example. It tests four dishes that share three ingredients over one week. | Dish | Sold | Price | Picked ingredient on the plate | | --- | --- | --- | --- | | Cheeseburger | 200 | $14 | Ground beef, 6 oz raw patty | | Beef tacos | 80 | $13 | Ground beef, 5.4 oz raw (4 oz cooked) | | Caesar salad | 150 | $11 | Romaine, 4 oz as purchased | | Roast chicken plate | 120 | $19 | Chicken breast, 8 oz raw | *Maple Street Kitchen (example): the dishes sold in the test week and what each plate uses.* Sales of the four dishes come to $7,770 (200 × $14 is $2,800, 80 × $13 is $1,040, 150 × $11 is $1,650, and 120 × $19 is $2,280). The recipes say how much each ingredient should have gone. | Ingredient | Working | Theoretical use | Price per lb | Theoretical cost | | --- | --- | --- | --- | --- | | Ground beef, burgers | 200 × 6 oz = 1,200 oz | 75.0 lb | $5.00 | $375.00 | | Ground beef, tacos | 80 × 5.4 oz = 432 oz | 27.0 lb | $5.00 | $135.00 | | Ground beef, total | 75.0 lb + 27.0 lb | 102.0 lb | $5.00 | $510.00 | | Chicken breast | 120 × 8 oz = 960 oz | 60.0 lb | $3.50 | $210.00 | | Romaine | 150 × 4 oz = 600 oz | 37.5 lb | $2.40 | $90.00 | | All three | | | | $810.00 | *Theoretical use: dishes sold × amount per plate, at the example's invoice prices.* The counts and invoices for the same week give the actual use. | Ingredient | Actual use | Theoretical | Gap | Gap in dollars | | --- | --- | --- | --- | --- | | Ground beef | 40 + 100 − 30 = 110.0 lb | 102.0 lb | 8.0 lb | $40.00 | | Chicken breast | 20 + 60 − 15 = 65.0 lb | 60.0 lb | 5.0 lb | $17.50 | | Romaine | 10 + 40 − 6 = 44.0 lb | 37.5 lb | 6.5 lb | $15.60 | | All three | $883.10 | $810.00 | | $73.10 | *Actual use (opening + bought − closing) against theoretical use, then the gap.* *Example: the gap in dollars and in points of sales* - Actual: $550.00 + $227.50 + $105.60: $883.10 - Theoretical: $510.00 + $210.00 + $90.00: $810.00 - Gap: $883.10 − $810.00: $73.10 - Gap ÷ theoretical: $73.10 ÷ $810.00: 9.0% more than the recipes - Gap ÷ sales of the four dishes: $73.10 ÷ $7,770: 0.9 of a point **The week's gap on these three ingredients:** $73.10, 0.9 of a point All figures are examples. A point is 1% of sales. The sales are the four dishes' only, so this is not a share of the whole restaurant's sales, and one week alone can't say whether it's a typical week. The gap isn't spread evenly. Beef is $40.00 of the $73.10, so beef is where to look first. That's why you count by ingredient, not only in total. ## Where a gap comes from The number tells you how much, not why. Most gaps come from a handful of causes, and most of them can be measured one at a time. | Cause | How to check it | Measurable? | | --- | --- | --- | | Yield: cooked or trimmed weight is less than the weight you bought | Weigh an ingredient before and after cooking or trimming, three times | Yes | | Portions bigger than the card says | Weigh ten plates off the line at a busy time | Yes | | Waste: trim, spoilage, over-prep | Keep a waste log for the week | Yes, with a log | | Comps, remakes, staff meals, tasting, owner meals | Write each down as it leaves, with the dish | Yes, if written down | | Prices changed between the card and the invoice | Price both numbers at one invoice price per pound | Yes | | Counting and unit errors | Same shelves, same order, same units; have a second person count one shelf | Partly | | Food taken without being recorded | Rule out everything above first | No, only by elimination | *Causes to check, and how to check each one.* ## Yield: why a price per pound isn't a cost per cooked pound Raw price per pound is the price of what you bought. A cooked plate holds less than that, because meat loses fat and water, and bones and trim don't go on the plate. The USDA's Food Buying Guide for Child Nutrition Programs lists yields for this. In its [meats and meat alternates section](https://foodbuyingguide.fns.usda.gov/files/Reports/USDA_FBG_Section1_MeatsAndMeatAlternatesYieldTable.pdf), 1 pound as purchased of fresh or frozen ground beef with no more than 20% fat gives 0.74 pound of cooked, drained, lean meat. A whole chicken as purchased gives 0.36 pound of cooked, boned meat without skin, not counting the neck meat and giblets. Read those as proof that the two weights differ, not as numbers to copy. The guide is written for school meals and other child nutrition programs, not restaurants. On its [about page](https://foodbuyingguide.fns.usda.gov/Home/About), it describes its figures as average yields from good-quality food prepared with little waste, and says it generally rounds its servings down. Your cut, cooking method and trim will differ, so weigh your own before and after, three times, and use your average. *Example: beef tacos, 4 oz of cooked beef a plate, at $5.00 a pound raw* - Price per raw ounce: $5.00 ÷ 16: $0.3125 - Card as first written: 4 oz × $0.3125: $1.25 - Yield used as a stand-in: 0.74 cooked per raw: 0.74 - Raw beef needed for 4 oz cooked: 4 ÷ 0.74, about: 5.4 oz - Plate cost with yield: 5.4 oz × $0.3125, about: $1.69 - 80 plates, card as first written: 80 × 4 oz: 20.0 lb, $100.00 - 80 plates, with yield: 80 × 5.4 oz: 27.0 lb, $135.00 **Beef the first card leaves out, for 80 plates:** 7.0 lb, $35.00 All figures are examples. The 0.74 stands in for the restaurant's own weighed yield. A card written this way would show 7.0 lb of beef as a gap before anyone dropped anything. [Free calculatorCost a plate with its yield: plate cost calculatorEnter the raw price per pound as the price with a pack size of 1 lb, the cooked amount as the amount used, and your weighed yield as the usable share (under “Trims or cooks down?”), 74 in the example. The calculator divides by it, so the plate is costed at the weight you buy.](https://restaurantdoctorai.com/calculators/recipe-cost) ## Portions, comps and counts **Portions.** A card that says 6 oz doesn't mean every patty weighs 6 oz. Maple Street weighs ten patties off the line at the Friday rush and they average 6.5 oz. That's half an ounce over on 200 burgers: 100 oz, or 6.25 lb, about $31.25 of the 8.0 lb beef gap. That leaves 1.75 lb ($8.75) of beef still unexplained. One check turned part of the gap into a cause. **Comps and staff meals.** Food that leaves the shelves and never rings up still counts as used. A remade plate, a comped dessert, a staff meal or a taste for a new special all count as actual use and not as theoretical use. Unwritten, they're the part of the gap you can't explain. So is food the owner takes home: the IRS's guide says items taken for personal use come out of cost of goods sold, and how your books handle it is a question for your accountant. **Price changes.** If the card was costed at last spring's price and the invoice is higher now, theoretical cost in dollars runs low even if every ounce matches. Pricing both numbers at one invoice price takes this out of the weekly gap. Re-cost the cards at today's prices to see it on its own. **Counting and unit errors.** Cases against each, pounds against ounces, an open bag counted twice, a delivery counted but not logged. These make gaps that look real. The EPA tells businesses to [“compare purchasing inventory with customer ordering”](https://www.epa.gov/sustainable-management-food/how-prevent-wasted-food-through-source-reduction), which is the same comparison from the waste side: what came in against what was sold. A gap alone never proves theft A gap is a difference between two numbers. A recipe card out of date, a miscounted case or a heavier hand on the line makes the same number as food taken without being recorded, and the number can't tell you which. Rule out the measurable causes first. Treat a person as a possibility only when the process checks are clean. ## What gap is acceptable? Set your own We found no official threshold. In the government and trade-association sources we looked at, none publishes an acceptable gap between actual and theoretical food cost, so we don't print anyone's range. A range measured in another kitchen says little about yours. The ranges you'll find online come from software sellers, and they don't agree with each other. You can find your own range, from your own counts: 1. Correct the cards first: weigh yields, re-cost at today's invoice prices, and write each recipe in weights. 2. Weigh portions on the line for the picked dishes, and write down waste, comps, remakes and staff meals for the week. 3. Run the one-week count for three weeks in a row, on the same ingredients, with the same counting routine. 4. Look at the gap left in each of the three weeks. Their spread is a starting range for your kitchen: the gap you get when what you can control is under control. 5. After that, a week well above your range is the week to look into. Update the range as you add weeks. Say Maple Street's three clean weeks leave gaps of 0.4, 0.6 and 0.5 of a point on the four dishes (example figures). Its starting range is 0.4 to 0.6. The first week's 0.9, counted before the cards were corrected, sits above it. A range like this belongs to that kitchen and that menu. It's not a standard for yours. ## What to do with the result - **Start with the biggest line.** In the example, that's beef at $40.00 of $73.10. - **Change one thing at a time** (weigh patties, cost the taco card with its yield, write down staff meals), then run the count again. If you change four things at once, you won't know which one moved the number. - **Count and order to a par.** Our free [prep list and par sheet](https://restaurantdoctorai.com/templates/prep-list-and-par-sheet) gives you a sheet to count against and an amount to order or prep. - **Don't punish the number.** If writing down a comp gets someone in trouble, it stops being written down. It doesn't stop happening. The guide's section on [the two quiet costs behind a high number](https://restaurantdoctorai.com/guides/food-cost-percentage#two-quiet-costs) covers the same ground from the monthly number. ## This week 1. Pick three to five top-selling dishes and the one or two costly ingredients in each. Write each plate's amount in raw ounces. 2. Count those ingredients at close tonight, on an inventory sheet with the item, quantity, unit and who counted. Keep every invoice for them this week. 3. In seven days, count again, pull the sales counts for the same seven days, and work out theoretical use, actual use and the gap in pounds. 4. Price the gap at one invoice price per pound and write it in dollars and in points of those dishes' sales. 5. Take the biggest line, run one check from the table (weigh ten portions, or weigh a yield), change one thing and count the next week the same way. 45-second profit check Not sure where your food cost is coming from? A gap in the count is one place the money goes. The free check asks 12 quick questions, including how you cost recipes and track waste, and grades each cost area so you can see your biggest estimated loss in about 45 seconds. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [About the Food Buying Guide](https://foodbuyingguide.fns.usda.gov/Home/About), U.S. Department of Agriculture, Food and Nutrition Service, Last updated August 4, 2025 - [Food Buying Guide for Child Nutrition Programs, Section 1: Meats/Meat Alternates](https://foodbuyingguide.fns.usda.gov/files/Reports/USDA_FBG_Section1_MeatsAndMeatAlternatesYieldTable.pdf), U.S. Department of Agriculture, Food and Nutrition Service, Undated PDF - [Publication 334 (2025), Tax Guide for Small Business](https://www.irs.gov/publications/p334), Internal Revenue Service, 2025 edition, for 2025 returns - [Prevent Wasted Food Through Source Reduction](https://www.epa.gov/sustainable-management-food/how-prevent-wasted-food-through-source-reduction), U.S. Environmental Protection Agency, December 12, 2025 --- title: "What one hour of restaurant labor really costs" description: "What an hour of restaurant labor costs beyond the posted wage: employer taxes, workers' comp, benefits, overtime and tips, with a method for your rates." canonical: "https://restaurantdoctorai.com/blog/true-cost-of-an-hour-of-restaurant-labor" published: "2026-10-15" updated: "2026-10-15" --- # What one hour of restaurant labor really costs ## Key takeaways - An hour costs more than the posted wage: employer Social Security and Medicare alone are 7.65%, before unemployment tax, workers' comp, paid leave and benefits. - BLS puts employer cost in accommodation and food services at $20.03 per hour worked in June 2026, $16.18 of it wages: an industry average, not one cook's cost. - Federal overtime is at least 1.5 times the regular rate over 40 hours in a workweek. Only costs that follow wages grow with it, not ones like health insurance. - Tips are outside the BLS wage figure, and cash wages and overtime rules differ by state, so check your state labor office before you set a schedule. An hour costs you more than the wage on the schedule. Add the employer's Social Security and Medicare, the unemployment taxes, workers' compensation, and any paid leave, insurance or retirement pay you offer, then divide by the hours people actually worked. In June 2026 the U.S. average for restaurants and hotels was about $1.24 for every $1.00 of wages, tips not counted, but your own rates decide yours. This post gives you the method, a worked hour and a worked overtime hour to repeat with your numbers, and what changes for tipped staff. It's about dollars per hour, not percent of sales. For what share of sales labor should be, and the targets for each type of restaurant, see [the labor cost guide](https://restaurantdoctorai.com/guides/restaurant-labor-cost). ## The wage on the schedule is where the cost starts The national picture comes from the Bureau of Labor Statistics. Its [Employer Costs for Employee Compensation release](https://www.bls.gov/news.release/ecec.nr0.htm) says total employer costs for private industry averaged $46.89 per hour worked in June 2026, published September 9, 2026. The [release's Table 4](https://www.bls.gov/news.release/ecec.t04.htm) splits that by industry. | Group | Total | Wages | Benefits | Wages share | | --- | --- | --- | --- | --- | | All private industry | $46.89 | $32.82 | $14.07 | 70.0% | | Leisure and hospitality | $21.33 | $17.25 | $4.08 | 80.9% | | Accommodation and food services | $20.03 | $16.18 | $3.85 | 80.8% | *Dollars per hour worked, private industry, June 2026. Source: BLS Employer Costs for Employee Compensation, Table 4.* In accommodation and food services, the $3.85 of benefits is what sits on top of $16.18 of wages: about $1.24 of total cost for each $1.00 of wages. The table breaks the $3.85 into legally required benefits ($1.82), insurance ($0.86), paid leave ($0.63), supplemental pay ($0.28) and retirement and savings ($0.25). The lines are rounded, so they don't add to the cent. BLS counts [Social Security, Medicare, federal and state unemployment insurance and workers' compensation](https://www.bls.gov/opub/hom/ecec/concepts.htm) as legally required benefits. Four things to know before you use that number: - **It is per hour worked.** That is how BLS states the unit, so the cost of paid leave is spread over the hours people were on the job. - **It is an average of every job in the industry.** The group is accommodation and food services, so it includes hotels, managers, cooks and dishwashers alike. It isn't the cost of one server or one cook. - **Its wages leave out tips.** BLS lists tips, which come from guests, among the payments not counted in wages and salaries. - **The overtime premium sits in benefits, not wages.** BLS counts premium pay for overtime as supplemental pay, so the wage figure is closer to straight-time pay. ## What sits on top of the wage Some of these costs come from federal law, some from your state, some from your insurer and some are your choice. That matters, because only some of them grow with every extra hour. - **Social Security and Medicare (federal law).** [IRS Topic 751](https://www.irs.gov/taxtopics/tc751) puts the employer's share at 6.2% for Social Security and 1.45% for Medicare, 7.65% together. The Social Security part applies up to a wage base, $184,500 of a person's wages in 2026, which an hourly cook is far from reaching. - **Federal unemployment tax (federal law).** [IRS Topic 759](https://www.irs.gov/taxtopics/tc759) says the rate is 6.0% on the first $7,000 you pay each employee in a year. With the maximum credit for state unemployment taxes paid, the rate is 0.6%, which is $42 per person per year. The credit is smaller in a state that has borrowed from the federal government and not repaid. The IRS page also says most employers pay both the federal and the state tax. - **State unemployment insurance (your state).** Your rate comes from your state, not the IRS. Ask your state's unemployment agency for your rate and the wages it applies to, and don't borrow a number from another state. - **Workers' compensation (state law and your insurer).** Each state has its own program and official: the Labor Department's [list of state workers' compensation officials](https://www.dol.gov/agencies/owcp/wc) links each state's page. The rate on your own policy is your number; ask your carrier how it is quoted and whether overtime pay counts in full. - **Insurance, retirement and paid leave (your choice, and sometimes your state's or city's).** What you offer is yours to price. If your state or city requires paid leave, your state labor office can say how it works. Not in that list: meals for staff, uniforms, training time and payroll service fees. Add them if you want a fuller number. ## One hour, worked out Work it over a year, so the costs that are charged per person, not per hour, get spread across the hours they cover. *The cost of an hour worked* Cost per hour worked = (wages paid + employer taxes + insurance + other benefits) ÷ hours worked Wages are everything you paid for the year, paid leave included. Hours are the hours actually worked, so paid leave hours aren't counted. Here is an example, made up so you can follow the sums. Harbor Street Grill, an example full-service restaurant, has a line cook at $18.00 an hour for 40 hours a week. That is 2,080 paid hours a year, 40 of them paid leave, so 2,040 hours worked. The workers' compensation and state unemployment rates, and the insurance, are round numbers for the example, not typical for anyone. Use the ones on your own notices and policies. | Cost | Per year | Per hour worked | Set by | | --- | --- | --- | --- | | Wages for 2,080 paid hours | $37,440 | $18.35 | You | | Social Security and Medicare, 7.65% | $2,864 | $1.40 | Federal law | | Federal unemployment, 0.6% of $7,000 | $42 | $0.02 | Federal law | | State unemployment, example 2% of $10,000 | $200 | $0.10 | Your state | | Workers' comp, example 3% of wages | $1,123 | $0.55 | State and insurer | | Health insurance, example $200 a month | $2,400 | $1.18 | Your choice | | Total | $44,069 | $21.60 | | *Example: one full-time cook at Harbor Street Grill, over a year. The unemployment, workers' comp and insurance figures are made-up round numbers.* The cook's $18.00 hour costs $21.60 per hour worked, 1.20 times the wage on the schedule. The wage line is $18.35 because the year's wages include 40 hours of paid leave, divided over 2,040 hours of work. The BLS industry average is about 1.24, built a different way and from every job in the industry. Neither is a target. Your own figure is the one to know. [Free calculatorFree labor cost calculatorIt takes your payroll taxes and benefits as one amount. The method above is how to work that amount out.](https://restaurantdoctorai.com/calculators/labor-cost) ## An overtime hour The federal rule, in the Labor Department's [overtime fact sheet](https://www.dol.gov/agencies/whd/fact-sheets/23-flsa-overtime-pay): covered employees, unless specifically exempted, get pay at “not less than time and one-half their regular rates of pay” for hours over 40 in a workweek. A workweek is a fixed, repeating period of 168 hours. It can start on any day, and the sheet says “Averaging of hours over two or more weeks is not permitted.” A 50-hour week followed by a 30-hour week is 10 hours of overtime, not zero. Say Harbor Street Grill's cook works 50 hours one week. The last 10 are overtime at 1.5 × $18.00 = $27.00. Which of the add-ons grow with those 10 hours? *Example: one overtime hour, same cook* - Wage, 1.5 × $18.00: $27.00 - Social Security and Medicare, 7.65%: $2.07 - Workers' comp, example 3%: $0.81 - Unemployment taxes, wage bases reached: $0.00 - Health insurance, a fixed monthly cost: $0.00 **Cost of the overtime hour:** $29.88 Taking 1.5 times the $21.60 regular hour would give $32.40, which is $2.52 too high, because the insurance, the unemployment taxes and the paid leave in that $21.60 don't grow with the hours. This assumes the cook is already past the unemployment wage bases for the year and that workers' comp counts overtime pay in full. Early in the year, before those wage bases are reached, add the two unemployment taxes. The same logic works with your numbers: the overtime wage, plus the costs that follow wages, and nothing for the costs that are fixed per person. Overtime rules are not only federal The federal rule is a floor. Some states add rules of their own. California, for one, counts overtime by the day: its [Labor Commissioner's overtime page](https://www.dir.ca.gov/dlse/faq_overtime.htm) gives one and one-half times the regular rate for hours over 8 in a workday and double time over 12. Check your own state with [your state labor office](https://www.dol.gov/agencies/whd/state/contacts). This is what the rules say, not advice on what you must do. ## More hours, or one more person Because some costs follow wages and others are charged per person, the same extra hours can cost differently depending on who works them. Say Harbor Street Grill needs 10 more cook hours a week. Either the cook works them as overtime, or a new part-timer is hired at the same $18.00. | Per week | Cook's overtime | New part-timer | | --- | --- | --- | | Wages | $270.00 | $180.00 | | Social Security and Medicare, 7.65% | $20.66 | $13.77 | | Workers' comp, example 3% | $8.10 | $5.40 | | Unemployment taxes | $0.00 | $4.41 | | Total | $298.76 | $203.58 | | Per hour | $29.88 | $20.36 | *Example: 10 extra cook hours a week at Harbor Street Grill, same example rates as above. Not a recommendation for either choice.* The part-timer's unemployment taxes are $42 of federal tax and $187.20 of the example state tax, spread over 52 weeks, because a new person starts their own wage bases. The overtime hours add none, because the cook already passed them. What the table leaves out matters as much: the time to hire and train, whether anyone is available, and whether the cook wants the hours. It also leaves out any paid leave or insurance you would offer a new hire, which are fixed per person and change the answer if they apply. The arithmetic doesn't say which choice is right. It says to compare costs hour by hour, not wage by wage. The same goes for a split shift, a lunch half and a dinner half. The costs that follow wages follow the dollars paid, so the cost per hour is the same. What can change is the rules: some cities and states require pay when a shift is cut short or changed late, and [our post on slow shifts](https://restaurantdoctorai.com/blog/cut-labor-cost-slow-shifts) lists the official pages we read. For choosing which hours to cut, see [the sales per labor hour guide](https://restaurantdoctorai.com/guides/sales-per-labor-hour). ## Tipped staff Don't use the BLS average for a server or bartender. Its wages leave out tips, and a tipped worker's cash wage can be far below the average. - **The cash wage is set by law.** The Labor Department's [fact sheet on tipped employees](https://www.dol.gov/agencies/whd/fact-sheets/15-tipped-employees-flsa) says the federal minimum cash wage is $2.13 an hour and the maximum tip credit is $5.12, against a $7.25 minimum. It also says some states require a higher cash wage and some don't allow a tip credit. Where state law differs, the sheet says the rule that protects the employee more applies. - **Overtime follows the full minimum wage.** When an employer takes a tip credit, the sheet says overtime is calculated on the full minimum wage, not on the cash wage. - **Reported tips carry employer tax.** [IRS Topic 761](https://www.irs.gov/taxtopics/tc761) treats the tips an employee reports as subject to Social Security and Medicare, and says the employer pays its portion. So the employer's 7.65% applies to cash wages and reported tips together. For a tipped hour, use the same method: the cash wage, the employer's tax on cash wages and reported tips, and the rest of the list above. Each state's cash wage is in [the tipped minimum wage guide](https://restaurantdoctorai.com/guides/tipped-minimum-wage). Whether a federal credit applies to the employer's tax on tips is a question for your accountant, and our [FICA tip credit calculator](https://restaurantdoctorai.com/calculators/fica-tip-credit) shows whether it is worth that call. ## This week 1. From your payroll report, get last year's wages and the employer taxes line, and from your time clock the hours actually worked, for one job or one group of people. 2. Pull your insurance invoices, your workers' comp policy and your state unemployment notice, so the rates in your table are yours. 3. Fill in the table above with your numbers: wages, employer taxes, insurance and other benefits, divided by hours worked. That is your cost per hour. 4. Work an overtime hour the same way: 1.5 times the regular wage, plus only the costs that follow wages. 5. Look up your state's overtime rules with [your state labor office](https://www.dol.gov/agencies/whd/state/contacts), including whether overtime counts by the day. 6. Put your payroll taxes and benefits into the [labor cost calculator](https://restaurantdoctorai.com/calculators/labor-cost) to see them as a share of sales. 45-second profit check See if labor is your costliest issue Answer 12 quick questions, including how staffing changes when a shift slows, and get a free preliminary grade and your costliest issue in about 45 seconds, so you know whether the cost of an hour is where to start. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Employer Costs for Employee Compensation Summary](https://www.bls.gov/news.release/ecec.nr0.htm), U.S. Bureau of Labor Statistics, September 9, 2026 - [Table 4. Employer Costs for Employee Compensation for private industry workers by occupational and industry group, June 2026](https://www.bls.gov/news.release/ecec.t04.htm), U.S. Bureau of Labor Statistics, September 9, 2026 - [Employer Costs for Employee Compensation: Concepts](https://www.bls.gov/opub/hom/ecec/concepts.htm), U.S. Bureau of Labor Statistics, Handbook of Methods, September 30, 2025 - [Topic no. 751, Social Security and Medicare withholding rates](https://www.irs.gov/taxtopics/tc751), Internal Revenue Service, September 24, 2026 - [Topic no. 759, Form 940, Employers Annual Federal Unemployment (FUTA) Tax Return – filing and deposit requirements](https://www.irs.gov/taxtopics/tc759), Internal Revenue Service, September 24, 2026 - [State Workers' Compensation Officials](https://www.dol.gov/agencies/owcp/wc), U.S. Department of Labor, Office of Workers' Compensation Programs - [Fact Sheet #23: Overtime Pay Requirements of the FLSA](https://www.dol.gov/agencies/whd/fact-sheets/23-flsa-overtime-pay), U.S. Department of Labor, Wage and Hour Division, Revised October 2019 - [Overtime](https://www.dir.ca.gov/dlse/faq_overtime.htm), California Labor Commissioner's Office (DLSE) - [State Labor Offices](https://www.dol.gov/agencies/whd/state/contacts), U.S. Department of Labor, Wage and Hour Division, January 1, 2026 - [Fact Sheet #15: Tipped Employees Under the Fair Labor Standards Act (FLSA)](https://www.dol.gov/agencies/whd/fact-sheets/15-tipped-employees-flsa), U.S. Department of Labor, Wage and Hour Division - [Topic no. 761, Tips – withholding and reporting](https://www.irs.gov/taxtopics/tc761), Internal Revenue Service, October 12, 2026 --- title: "Restaurant lease extension: the math to do before you sign" description: "Before you sign a restaurant lease extension: rent as a share of sales by restaurant type, the new break-even, and the lease terms to read first." canonical: "https://restaurantdoctorai.com/blog/restaurant-lease-extension" published: "2026-10-13" updated: "2026-10-14" --- # Restaurant lease extension: the math to do before you sign ## Key takeaways - Divide rent, property charges and utilities by sales. RestaurantDoctorAI's targets: 6–10% for quick service, fast casual and cafés, 7–12% for the rest. - Work out the share for every year of the new term, because a yearly increase stacks on itself while your sales may not. - A rise in occupancy cost lifts break-even by the rise divided by the share of each sale left after the costs that rise with sales. - Read the escalation, common-area charges, personal guaranty, relocation and option clauses, and act before the lease's notice deadline. Sign a lease extension only after three checks. First, divide rent, property charges and utilities by sales for every year of the new term, and compare the result with RestaurantDoctorAI's targets: 6–10% of sales for quick service, fast casual and cafés, 7–12% for full service, fine dining and bars. Second, see how far the new rent lifts the sales you need to break even. Third, read the clauses that change the cost later, and do all of it before the lease's deadline for giving notice. ## The rent math to check first A landlord's offer comes as a dollar figure: new base rent, maybe a new yearly increase. A dollar figure on its own can't tell you whether the space still works. Rent as a share of sales can, because you can put it next to a target for your kind of restaurant. Count everything it costs to be in the building, not base rent alone. Many leases add “additional rent” on top: the New York City Department of Small Business Services' [guide to commercial leasing](https://www.nyc.gov/assets/sbs/downloads/pdf/about/reports/commercial-lease-guide-accessible.pdf) lists a share of rising building operating costs, a share of rising real estate taxes, and a fixed yearly increase as common forms. Common-area charges (the shared parking lot, hallways, landscaping and building management) are another, covered below. *Occupancy cost as a share of sales* Occupancy cost % = (base rent + property charges + utilities) ÷ sales × 100 Use the same months for costs and sales. Property charges means everything the lease bills on top of base rent. | Restaurant type | Occupancy cost (% of sales) | | --- | --- | | Quick service (counter, takeout, delivery) | 6–10% | | Fast casual | 6–10% | | Full service / casual dining | 7–12% | | Fine dining | 7–12% | | Bar, taproom or lounge | 7–12% | *RestaurantDoctorAI's targets for occupancy cost: rent, property charges and utilities together, as a share of sales.* Above the top of your range, occupancy starts squeezing profit. Our [rent percentage guide](https://restaurantdoctorai.com/guides/restaurant-rent-percentage) explains the ranges and why they include utilities. Two habits make the check honest: - Use your real sales from the last 12 months, before tax and tips, not the sales you hope the next lease brings. - Work out the share for each year of the new term. A yearly increase stacks on itself, and your sales may not grow at the same pace. ## Example: a pizzeria's extension offer Northside Slice is an example pizzeria, and its numbers are invented for this post. It is a counter-service and takeout shop, so its target is the quick-service range, 6–10% of sales. It sells about $85,000 a month. Its landlord offers a 5-year extension: base rent up from $5,500 to $6,000 a month, then 3% more each year, with common-area and tax charges estimated at $1,000 a month instead of $900. *Example: Northside Slice, one month today* - Base rent: $5,500 - + Property charges (common areas, taxes, insurance): $900 - + Utilities: $1,600 - = Occupancy cost: $8,000 - ÷ Sales for the month: $85,000 **Occupancy cost %:** 9.4% Example figures. $8,000 ÷ $85,000 × 100 = 9.4%, inside the 6–10% range but near its top. Now the offer, year by year. The table holds sales at $85,000 and utilities and property charges flat, so it shows the effect of the rent terms alone. | Year of extension | Base rent a month | Occupancy cost a month | Share of sales | Sales needed for 10% | | --- | --- | --- | --- | --- | | Today | $5,500 | $8,000 | 9.4% | $80,000 | | Year 1 | $6,000 | $8,600 | 10.1% | $86,000 | | Year 2 | $6,180 | $8,780 | 10.3% | $87,800 | | Year 3 | $6,365 | $8,965 | 10.5% | $89,654 | | Year 4 | $6,556 | $9,156 | 10.8% | $91,564 | | Year 5 | $6,753 | $9,353 | 11.0% | $93,531 | *Example: Northside Slice under the offered extension, at $85,000 a month in sales. Arithmetic only.* Three things stand out. Day one of the extension already puts Northside Slice just over the top of its range. By year 5 it needs about $8,500 more in monthly sales than it makes today just to keep occupancy at 10%. And the 5 years of base rent add up to about $382,000, against $330,000 at today's rent: $52,000 more over the term, before any change in common-area charges. None of that means “don't sign.” Moving costs money too. It means the owner walks into the talk knowing which number matters. A first year at today's rent, then 2% increases, with property charges held at $900, would keep the share under 10% in all 5 years at today's sales. That is one counter-offer worth pricing out before the meeting, not a promise of what the landlord will accept. ## What the new rent does to break-even sales Rent is a bill that stays the same whatever you sell. The U.S. Small Business Administration lists rental lease payments among its examples of [fixed costs](https://www.sba.gov/business-guide/plan-your-business/calculate-your-startup-costs/break-even-point). So a rent rise moves your break-even point, the monthly sales that pay every bill with nothing left over. The rule of thumb from our [break-even post](https://restaurantdoctorai.com/blog/restaurant-break-even-point): divide the rise by the share of each sale left after the costs that rise with sales. *How far break-even moves* Change in break-even = change in monthly occupancy cost ÷ share of each sale left after the costs that rise with sales Write the share as a decimal: 40 cents of each dollar is 0.40. Say Northside Slice's food, packaging, card fees and hourly shifts take 60 cents of each sales dollar, so 40 cents are left to pay the bills that stay the same. Those bills come to $30,000 a month today, rent included. These are example figures too. | Example | Bills that stay the same | Break-even sales | Cushion at $85,000 of sales | | --- | --- | --- | --- | | Today | $30,000 | $75,000 | $10,000 | | Year 1 of the offer (+$600 a month) | $30,600 | $76,500 | $8,500 | | Year 5 of the offer (+$1,353 a month) | $31,353 | $78,383 | $6,617 | *Example: Northside Slice's monthly break-even under the offer (40 cents of each sale left to pay the bills that stay the same).* A $600 rise in monthly occupancy cost adds $1,500 to break-even, and by year 5 the rise of $1,353 adds about $3,383. The cushion above break-even shrinks by about a third over the term if sales stay where they are. The SBA also suggests adding a little extra, say 10%, for costs you can't predict, which makes a thin cushion thinner. [Free calculatorWork out yours: break-even calculatorEnter one typical month, then run it again with the new rent to see how many more sales a day it takes.](https://restaurantdoctorai.com/calculators/break-even) ## The terms to read before you sign The rent figure is only part of the cost. These clauses decide what you pay in year 3 and what happens if things go wrong. Lease law differs by state and even by city, and the SBA [suggests](https://www.sba.gov/business-guide/manage-your-business/buy-assets-equipment) having an attorney review a lease with you before you sign. ### Escalation: how the rent rises An escalation clause is the lease's rule for raising rent. The NYC guide describes three kinds: a fixed dollar amount each year, a fixed percentage each year, or your share of the landlord's actual increase in costs such as real estate taxes. It says fixed increases are typically 2–3% a year in that market. Some extensions instead reset rent to “fair market rent”; the same guide suggests asking for a limit, such as 90% or 95% of market, and a neutral broker or appraiser if you can't agree. Price each version for every year of the term, as in the table above. ### Common-area and other property charges A Seattle [lease checklist for small businesses](https://www.seattle.gov/documents/departments/economicdevelopment/restaurants/final%20full%20lease%20tool.pdf) lists what common-area maintenance (CAM) charges can include: maintenance and repair, administrative fees, utilities, landscaping, parking lot upkeep and management salaries, varying by building. It notes that exclusions, such as capital expenses, are highly negotiated, and that an audit right lets you check what was passed through; some leases make the landlord pay for the audit when the overcharge passes a set percentage. The city's [key lease terms](https://www.seattle.gov/documents/Departments/economicDevelopment/OISI/English%20FINAL%20Key%20Lease%20Terms.pdf) sheet adds a cap: your share can't rise more than a set percentage a year, often a fixed figure such as 3%. Ask the landlord for last year's statement and this year's estimate, in writing. ### Renewal option versus extension Read what you're being asked to sign. The Seattle key terms sheet explains that a renewal creates a new lease while an extension continues the old one, and that clauses personal to the tenant, such as a personal guaranty, don't carry into a renewal term unless the option sets them out again. Check how many further options the extension leaves you, and how rent will be set for them: the sheet names a consumer price index, appraisal or arbitration, or a set percentage over the last month's rent. ### Personal guaranty A personal guaranty is your promise to pay the lease yourself if the business can't. The NYC guide separates a full guaranty, covering everything the tenant owes, from a “good guy” guaranty, which holds you personally only while the business occupies the space; once you hand back the keys after a default, you stop owing rent personally for the time after. It warns that clauses making you pay the whole remaining rent, do extra repairs, or repay allowances, free rent or broker fees mean the guaranty isn't a true good guy guaranty. On an extension, read whether your existing guaranty now covers the new years and the higher rent. In Northside Slice's case, that is about $382,000 of base rent. ### Relocation clause A relocation clause lets the landlord move you to another space in the property. The Seattle key terms sheet says restaurants in particular should resist one, because the business depends on its location. If it stays, the sheet lists limits to ask for: at least 90 days' notice, not more than once in the term, not in the first 12 months or in November and December, comparable space, the landlord paying the moving and build-out costs, and the right to end the lease if the new space doesn't work. ### Percentage rent and a sales exit Some leases add percentage rent: a share of sales above a set level, the breakpoint, on top of base rent. If yours does, count it with the costs that rise with sales. The key terms sheet also describes a gross sales kick-out clause, which lets a tenant end the lease if sales stay under an agreed figure. Neither is in every lease, but both change the math above. Where the law differs Rules change from place to place. Seattle, for one, caps the personal guaranty a landlord can ask for on [new commercial leases](https://www.seattle.gov/documents/Departments/FAS/RegulatoryServices/Commercial-Lease-Requirements.pdf) signed after January 27, 2024, but the rule doesn't apply to lease extensions. Check your own state and city. ## Timing: start before the notice deadline Start well before the deadline, so you have time to ask and to answer. If your lease has a renewal or extension option, it sets when your notice must reach the landlord. The Seattle checklist explains why: landlords want notice early enough to market the space if you leave, but not so early that they can't see how you've performed. Find that date first and put it in your calendar with a reminder months ahead. If there's no option, the NYC guide says a landlord isn't required to renew, may ask any rent once the lease ends, and that you should contact the landlord well before the term ends. It also reminds tenants not to rely on spoken agreements or a letter of intent: what counts is the signed document. The same guide notes that landlords can be more willing to offer a reasonable rent to a tenant who pays on time and keeps to the lease. 1. Find the notice date and the method the lease requires (written notice, to which address). 2. Work out your occupancy share and break-even under the offer at least a month before that date. 3. Send your counter-offer in writing, with the numbers: your sales, the share, and the rent that keeps it in range. 4. Read the final draft against the offer before you sign: rent table, increases, charges, guaranty, relocation, options. ## This week 1. Find your lease and write down the end date and the notice deadline for any renewal or extension option. 2. Add up the last 12 months of base rent, property charges and utilities, and divide by sales for the same months. 3. Put the landlord's offer into a year-by-year table like Northside Slice's, and run the [break-even calculator](https://restaurantdoctorai.com/calculators/break-even) with today's rent and with year 1 and year 5 of the offer. 4. Mark the escalation, common-area charges, guaranty, relocation and option clauses, and list the questions to ask in writing. 45-second profit check See if rent is squeezing your profit Before the lease talk, answer 12 quick questions, including your rent and utilities as a share of sales, and see your restaurant's grade and its biggest estimated loss, free, in about 45 seconds. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Comprehensive Guide to Commercial Leasing in New York City](https://www.nyc.gov/assets/sbs/downloads/pdf/about/reports/commercial-lease-guide-accessible.pdf), NYC Department of Small Business Services - [Break-even point](https://www.sba.gov/business-guide/plan-your-business/calculate-your-startup-costs/break-even-point), U.S. Small Business Administration - [Buy assets and equipment](https://www.sba.gov/business-guide/manage-your-business/buy-assets-equipment), U.S. Small Business Administration - [Commercial Lease Checklist for Small Businesses](https://www.seattle.gov/documents/departments/economicdevelopment/restaurants/final%20full%20lease%20tool.pdf), City of Seattle Office of Economic Development and Wayfind - [Key Lease Terms](https://www.seattle.gov/documents/Departments/economicDevelopment/OISI/English%20FINAL%20Key%20Lease%20Terms.pdf), City of Seattle Office of Economic Development - [Summary of Commercial Lease Requirements](https://www.seattle.gov/documents/Departments/FAS/RegulatoryServices/Commercial-Lease-Requirements.pdf), City of Seattle Department of Finance and Administrative Services --- title: "How to cut labor cost on slow shifts, not your best people" description: "Read sales by hour, set a cut line from your own wages, stagger starts, cross-train, and check your state's rules before you send anyone home early." canonical: "https://restaurantdoctorai.com/blog/cut-labor-cost-slow-shifts" published: "2026-10-13" updated: "2026-10-13" --- # How to cut labor cost on slow shifts, not your best people ## Key takeaways - Cut hours, not people: find the slow hours in your POS sales by hour, trim those, and keep your best people on the busy ones. - A quick cut line is your average hourly labor cost divided by your labor ceiling as a decimal; counting salaries, as the calculator does, raises it. - Later starts, earlier finishes written into the schedule and cross-training trim slow hours with fewer same-day send-homes. - Some states and cities require pay when a shift is cut short or changed late, so check your own state or city labor office first. Cut hours, not people. Pull sales by hour from your POS, set a cut line from your own wages, and trim only the hours that fall below it: later starts, earlier finishes written into the schedule, and a cross-trained crew that can cover the lull. Your best people keep the busy hours. Before you send anyone home early, check your state and city rules, because some places require pay for a shift cut short. ## Start with sales by hour, not the week A weekly labor percentage tells you labor ran high. It doesn't tell you when. On a slow day, look first at three places: the gap between lunch and dinner, the first hour of a shift before guests arrive, and the last hour after the kitchen slows. You find those hours in your POS, not in payroll. If you're not sure what labor percentage you should be aiming for in the first place, start with [the labor cost guide](https://restaurantdoctorai.com/guides/restaurant-labor-cost): it has the targets for each type of restaurant and how to count labor. This post is about the next step, what to do with the slow hours. 1. In your POS, find the report that shows net sales by hour of the day. If yours only exports checks, export them with their close times and total them by hour in a spreadsheet. 2. Pull the same weekday for the last four weeks, say four Tuesdays, and average each hour. One week can be thrown off by a game, a storm or a party of 20. 3. From your time clock, count how many hourly staff were on the clock in each of those hours. Someone who worked 11 to 3 counts in four hours. 4. Put the two side by side: sales and staff on the clock, hour by hour. Leave out holidays and days with private events, or compare them only with each other. Use net sales, after discounts and comps, before tax and tips. Count hours actually worked, not hours scheduled: early clock-ins and late clock-outs are paid hours too, and they belong in the count. ## Set your cut line from your own numbers Sales per labor hour is sales divided by the hours your staff worked. On its own it says little. Next to a cut line, the sales per labor hour below which labor runs over your target, it tells a manager whether this hour is paying for the people on the clock. *The quick cut line* Cut line = average hourly labor cost ÷ labor ceiling as a decimal Average hourly labor cost is hourly pay plus its share of payroll taxes and benefits, divided by the hours worked. The ceiling is the top of your labor target, written as a decimal: 0.35 for 35%. Salaried managers and chefs are paid whatever the sales, so they use up part of the target before any hourly hour is worked. The free [labor cost calculator](https://restaurantdoctorai.com/calculators/labor-cost#cut-line) takes them out first, which raises the line: *The cut line with salaries counted* Cut line = average hourly labor cost ÷ (ceiling − salaries ÷ sales) Salaries include their share of payroll taxes and benefits. Sales are the week's net sales. The calculator rounds the hourly cost and the line up to the cent, so staying at or above the line keeps labor at or under the target. *Example: Juniper Table, a full-service restaurant, last week* - Net sales: $30,000 - Hourly pay, 600 hours worked: $9,000 - Salaries: $1,500 - Payroll taxes and benefits: $1,050 - Labor cost: $11,550 ÷ $30,000: 38.5% - Sales per labor hour: $30,000 ÷ 600: $50.00 - An hour of hourly staff: $9,900 ÷ 600: $16.50 - Salaries with taxes and benefits: $1,650 - Ceiling, the top of 30–35%: 0.35 **Cut line:** $55.94 Taxes and benefits are 10% of all pay ($1,050 ÷ $10,500), so an hour of hourly staff costs $15 × 1.10 = $16.50. Cut line: $16.50 ÷ (0.35 − $1,650 ÷ $30,000) = $55.94, rounded up to the cent. Without the salaries, the quick line would be $16.50 ÷ 0.35 = $47.15. Juniper Table's $50.00 sits between the two: above the quick line, below the real one. Juniper Table's labor ran at 38.5% against a 35% ceiling, the top of RestaurantDoctorAI's target for full service. At 35%, last week's sales paid for $10,500 of labor. Take out $1,650 of salaries and $8,850 is left for hourly staff: 536 hours at $16.50. The team worked 600. So the question for Juniper Table isn't who to let go. It's where 64 hours a week went that sales didn't pay for. [Free calculatorWork out your cut line: labor cost calculatorLast week's payroll and sales give your line with salaries counted, and how many hourly staff an hour's sales pay for.](https://restaurantdoctorai.com/calculators/labor-cost#cut-line) ## Find the hours that run under the line Divide each hour's sales by the cut line and round down. That's how many hourly staff that hour's sales pay for at your target. Here is Juniper Table's average Tuesday afternoon and late evening, from four Tuesdays of POS and time-clock reports. | Hour | Sales | On the clock | Pays for | | --- | --- | --- | --- | | 1–2 p.m. | $600 | 9 | 10 | | 2–3 p.m. | $250 | 8 | 4 | | 3–4 p.m. | $200 | 8 | 3 | | 4–5 p.m. | $350 | 8 | 6 | | 5–6 p.m. | $700 | 9 | 12 | | 8–9 p.m. | $450 | 10 | 8 | | 9–10 p.m. | $200 | 8 | 3 | *Example: Juniper Table's average Tuesday, last four Tuesdays. Pays for = sales ÷ $55.94, rounded down.* The busy hours pay for more people than are on the clock. The problem is five hours: the afternoon lull from 2 to 5 and the last two hours of the night. Two things keep this honest. - **There's a floor.** Juniper Table can't open with three people. Say it needs four to run safely: a cook, a dishwasher, a server and someone on the bar and the door. No hour goes below that, whatever the sales. - **One slow hour before a rush is normal.** The hour before dinner can run under the line while the crew sets up. The line is a reason to look, not an order to send people home. With a floor of four, the hours Juniper Table could trim on that Tuesday are 4 from 2 to 3, 4 from 3 to 4, 2 from 4 to 5, 2 from 8 to 9 and 4 from 9 to 10: 16 hours, about $264 at $16.50. If Monday to Thursday look like this Tuesday, that's 64 hours a week, the same gap the calculator showed. Your own days will differ; the point is that the hours sit in a few predictable places. ## Stagger the starts and plan the cuts Once you know which hours run under the line, change the schedule, not just the shift. A shorter shift written into next week's schedule gives your staff notice, where a long one cut short on the day doesn't, and in some states the two are treated differently in law too (see the rules section below). ### Later starts At Juniper Table, eight people are on from 4 to 5 and the hour pays for six. Two dinner servers who start at 4 could start at 5, when sales double. The same applies to the morning: if the first guests arrive at 11:30, a prep cook may need to start early, but a server probably doesn't. ### Earlier finishes, written in The lunch crew is the other half of the afternoon. If two lunch servers are scheduled until 4 and sales drop by more than half after 2, schedule them until 2:30 and let the dinner crew pick up the floor. Do the same at night: write the closing crew into the schedule by role, so the 9 to 10 hour runs on the floor of four rather than on whoever is still around. ### A written cut order For the days that come in slower than forecast, put the cut order on every schedule: who goes first, second and third, and who closes. Order it by role and by the hour, not by who's in favor. Ask who wants to go first; some people want the hours and some want the night. Then the manager on shift doesn't have to decide under pressure, and nobody reads an early cut as a verdict on their work. Give the lull useful work instead of empty time: prep for dinner, restocking and deep cleaning can move into 2 to 5, so those tasks aren't paid for during the rush. ## Cross-train so a small crew covers every station A floor like Juniper Table's is what keeps the lull expensive: you can't send the bartender home if nobody else can pour. Cross-training can lower the floor without lowering what guests get. - Train a server to make the drinks on the short list that sells in the afternoon, so the bar can open later or close earlier. - Train the host to run food and bus tables in the quiet hours, so one person covers the door and the floor. - Train a line cook on dish and prep, so the dishwasher can start with the dinner rush. - Train a second person on every station, so a call-out doesn't turn into a double shift. Train on slow shifts, a station at a time, with a short written checklist for each. Start with the people you most want to keep: someone who can cover two stations has more ways to stay on the schedule when it's slow. Where cross-training moves someone into a job paid at a different rate, check how your payroll handles the hours. ## The rules that differ by state and city Rules on cutting a shift short and changing a schedule differ by state and city. Some require pay when a shift is cut short. Others require schedules posted in advance and extra pay for late changes, and the ones we read are written for large employers and chains. Here is what the official pages we read say. It's a sample, not a list of every place with such a law. | Place | Who it covers | Shifts cut short or changed | | --- | --- | --- | | California | Employees who report for a scheduled day | Under half the day worked: paid for half the scheduled day, at least 2 hours, at most 4 | | New York State | Every employer in the hospitality industry | Call-in pay of at least 3 hours for one shift, or the regularly scheduled shift if shorter | | Oregon | Retail, hospitality and food service employers with 500+ employees worldwide | Half the regular rate for each scheduled hour cut | | Seattle | 500+ employees worldwide; full-service restaurants also need 40+ full-service locations | Pay for half the hours not worked when sent home early | | Chicago | Restaurants with 250 employees and 30 locations | 1 hour of pay for a change within 14 days | | Philadelphia | 250 employees and 30 locations worldwide | Schedules 14 days ahead and predictability pay | | New York City | Fast food chains with 30+ establishments nationally | $10 to $75 for each change under 14 days ahead | *What each official page says, read October 13, 2026. Check the page for the current rule.* ### California: reporting time pay The [California Labor Commissioner's reporting time pay FAQ](https://www.dir.ca.gov/dlse/faq_reportingtimepay.htm) says an employee who reports for a scheduled day and gets less than half of it is paid for half the scheduled day, never less than 2 hours or more than 4, at the regular rate. Its own example is a 6-hour shift sent home after 1 hour: 3 hours of pay. The FAQ lists exceptions, such as a utility failure or an employee leaving by choice, and says the extra pay doesn't count toward overtime. ### New York State: call-in pay New York's [Hospitality Industry Wage Order](https://forms.labor.ny.gov/WP/CR146.pdf) (section 146-1.5) says an employee who reports for duty at the employer's request or with its permission is paid for at least 3 hours for one shift, or the hours of the regularly scheduled shift if that's fewer. Time actually worked is paid at the usual rate less any tip credit; the rest of the 3 hours is paid at the basic minimum rate with no tip credit. ### Predictive scheduling and fair workweek laws These set notice periods and extra pay for changes after a schedule is posted. Oregon's [predictive scheduling law](https://www.oregon.gov/boli/workers/pages/predictive-scheduling.aspx) is statewide for employers with at least 500 employees worldwide, with schedules 14 days ahead. Seattle's [Secure Scheduling Ordinance](https://www.seattle.gov/laborstandards/ordinances/secure-scheduling) has the same 500-employee line, and a full-service restaurant also needs 40 or more full-service locations. Chicago's [Fair Workweek Ordinance](https://www.chicago.gov/city/en/depts/bacp/supp_info/fairworkweek.html) covers restaurants with 250 employees and 30 locations, and only staff at or under a pay ceiling the city sets. Philadelphia's [Fair Workweek law](https://www.phila.gov/services/working-jobs/worker-protections/wage-protections/report-a-fair-workweek-violation/) uses 250 employees and 30 locations worldwide and calls for schedules 14 days ahead. New York City's [Fair Workweek Law for fast food](https://www.nyc.gov/assets/dca/downloads/pdf/workers/FAQs-FairWorkweek-FastFood.pdf) covers chains with 30 or more establishments nationally, franchises included. A restaurant with one location is under the location lines in Chicago, Philadelphia and New York City, and under Seattle's for full service. The two state pages are different: California's FAQ sets no employer size, and New York's wage order covers every employer in the hospitality industry. Other states and cities have rules of their own, and rules change. Check your own state or city first This is what the rules say, not advice on what you have to do. Before you change start times or send anyone home early, check with [your state labor office](https://www.dol.gov/agencies/whd/state/contacts), and your city's labor office where your city has its own scheduling rules. Every hour worked gets paid. ## This week 1. Pull net sales by hour for the last four of one slow weekday from your POS, and staff on the clock by hour from the time clock. 2. Enter last week's sales and payroll in the [labor cost calculator](https://restaurantdoctorai.com/calculators/labor-cost#cut-line) and write your cut line at the top of the sheet. 3. Mark every hour where sales pay for fewer people than were on the clock. Write down the smallest crew that can run the room safely. 4. On next week's schedule, move one start later and one finish earlier in the slowest stretch, and write the cut order and the closing crew by role. 5. Pick one station to cross-train on a slow shift this week, with a written checklist. 6. Look up your state's and city's rules on reporting pay and schedule changes before the new schedule goes up. 45-second profit check See if labor is your costliest issue Answer 12 quick questions, including what happens to staffing when a shift slows down, and get a free preliminary grade and your costliest issue in about 45 seconds, so you know whether the schedule is where to start. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Reporting Time Pay (frequently asked questions)](https://www.dir.ca.gov/dlse/faq_reportingtimepay.htm), California Labor Commissioner's Office (DLSE), December 2016 - [Hospitality Industry Wage Order, 12 NYCRR Part 146](https://forms.labor.ny.gov/WP/CR146.pdf), New York State Department of Labor, September 2024 (CR 146, 09/24) - [Predictive Scheduling](https://www.oregon.gov/boli/workers/pages/predictive-scheduling.aspx), Oregon Bureau of Labor and Industries - [Secure Scheduling](https://www.seattle.gov/laborstandards/ordinances/secure-scheduling), City of Seattle, Office of Labor Standards - [Fair Workweek](https://www.chicago.gov/city/en/depts/bacp/supp_info/fairworkweek.html), City of Chicago, Department of Business Affairs and Consumer Protection - [Report a Fair Workweek violation](https://www.phila.gov/services/working-jobs/worker-protections/wage-protections/report-a-fair-workweek-violation/), City of Philadelphia, Department of Labor - [Fair Workweek Law in Fast Food: Frequently Asked Questions](https://www.nyc.gov/assets/dca/downloads/pdf/workers/FAQs-FairWorkweek-FastFood.pdf), NYC Department of Consumer and Worker Protection, September 15, 2023 - [State Labor Offices](https://www.dol.gov/agencies/whd/state/contacts), U.S. Department of Labor, Wage and Hour Division, January 1, 2026 --- title: "Café food cost: does the 30% rule work for resold pastries?" description: "The 30% food cost rule is a whole-menu target. How a café should cost and price bought-in pastries and bottled drinks next to what it makes itself." canonical: "https://restaurantdoctorai.com/blog/cafe-food-cost-resold-items" published: "2026-10-13" updated: "2026-10-14" --- # Café food cost: does the 30% rule work for resold pastries? ## Key takeaways - The 30% or 33% food cost line is a target for everything you sell together, not a markup every item has to hit. - A bought-in pastry carries the maker's labor in its price, so its cost % runs higher while your own payroll stays lower. - Price a resold item from its delivered unit cost, divided by the share you sell, plus packaging, then divided by its category's own target. - Check the whole café monthly: food and drink cost inside your range, and prime cost under the ceiling, 60% for a café in RestaurantDoctorAI's targets. Not item by item. The 30% (or 33%) food cost line is a target for everything you sell together, not a markup every item has to hit. A croissant you buy ready-made already carries the bakery's labor in its price, so its cost % runs higher while your own payroll stays lower. Give resold items their own target, price them from their delivered unit cost, and check that your blended food and drink cost and your prime cost still sit inside your range. ## Where the 30% and 33% come from Both numbers describe a whole business: all the food and drink you used in a month, divided by all your sales. The National Restaurant Association's [figures for 2024](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/) put the median cost of food and non-alcoholic drinks at 32.4% of sales for limited-service restaurants (counter service) and 32.0% for full service. For limited service, it notes that its 2010, 2013 and 2016 reports averaged about 33%. The Association also says plainly that these figures aren't meant as standards or goals for any one restaurant. They're medians: about half of the restaurants that answered ran higher, and half lower. RestaurantDoctorAI's targets work the same way. A café has targets of its own: our range for food and drink cost combined is 24–32% of sales, and prime cost (food and drink cost plus total labor, divided by sales) has a ceiling of 60%. Both are for the whole café, across everything on the menu. They're ours: no café-only survey is public, as [our café costs page](https://restaurantdoctorai.com/restaurant-types/cafes) explains. The [food cost guide](https://restaurantdoctorai.com/guides/food-cost-percentage#targets-by-service-format) has the ranges for every type of restaurant. So if you mark every item up three times, you're treating a whole-menu average as a rule for each item. Three times cost is a 33% cost on every item. That works for some items and prices others out of reach. ## What a made item costs compared with a bought one When you make something in-house, its food cost counts only the ingredients. The person who made it is paid out of your labor line. When you buy something ready-made, the price you pay already covers the maker's labor, ovens and margin. Nothing about it shows up in your payroll, but all of it shows up in your food cost. The IRS draws the same line for taxes. Its [Tax Guide for Small Business](https://www.irs.gov/publications/p334) says a merchant counts the cost of the goods it bought for sale, at the prices actually paid rather than the list price, and that freight on those goods is part of cost of goods sold. It adds that labor usually belongs in cost of goods sold only for manufacturers, and that small retailers usually have none to put there. How you file is a question for your accountant. For pricing, the takeaway is that a resold item's cost is what you paid to have it on your counter, delivery included. *Example: one croissant at $4.00, made or bought (Linden Street Café, a made-up café)* - Bought: $72.00 a case of 60, delivered: $1.20 - + Bag and napkin: $0.08 - = Bought, food cost: $1.28 ÷ $4.00: 32.0% - Made: butter, flour and the rest: $0.45 - + Bag and napkin: $0.08 - = Made, food cost: $0.53 ÷ $4.00: 13.3% - + Baker: 4 hours at $22.00 for 60: $1.47 - = Made, food and baker: $2.00 ÷ $4.00: 50.0% **Left per croissant after food and baker: bought, made:** $2.72, $2.00 All figures are examples. $22.00 an hour stands for pay plus the employer's payroll taxes. On food cost alone, the made croissant looks far better (13.3% against 32.0%). Once the baker's time is counted, the bought one leaves more per croissant. That flips if your baker is on the clock anyway with spare hours: then the made croissant's labor isn't extra, and it leaves $3.47. The point isn't that buying beats baking. It's that you can't compare the two on food cost %, because one has its labor inside the price and the other doesn't. A resold item isn't labor-free either. Someone receives the delivery, fills the case, warms, bags and throws out what didn't sell. It's usually seconds per item rather than the hours it takes to bake, but it's there. ## Why one blended number can mislead you Because each kind of item runs at its own cost %, your blended number moves when the mix moves, even if every price is right. Here's the example café's month, split by what it makes and what it resells. | Category | Sales | Cost % | Cost | | --- | --- | --- | --- | | Coffee and espresso drinks (made) | $16,000 | 22% | $3,520 | | Sandwiches and toast (made) | $6,000 | 32% | $1,920 | | Pastries from a bakery (resold) | $5,000 | 42% | $2,100 | | Bottled drinks and snacks (resold) | $3,000 | 50% | $1,500 | | Whole café | $30,000 | 30.1% | $9,040 | *Linden Street Café (example), one month. Cost % is each category's food and drink cost ÷ its sales.* The pastries run at 42% and the bottled drinks at 50%, well above any 30% rule. Yet the whole café sits at 30.1%, inside the 24–32% café range, because the coffee drinks run low. With $8,700 of labor (29% of sales), prime cost is $17,740 ÷ $30,000, or 59.1%, just under the 60% ceiling. Now say a new office nearby adds $3,000 a month of pastry sales, at the same 42% and with the same staff hours. *Example: the same café sells $3,000 more pastries* - Sales: $30,000 + $3,000: $33,000 - Food and drink cost: $9,040 + $1,260: $10,300 - Blended cost: $10,300 ÷ $33,000: 31.2%, up from 30.1% - Left after food and drink: $33,000 − $10,300: $22,700, up from $20,960 - Prime cost: ($10,300 + $8,700) ÷ $33,000: 57.6%, down from 59.1% **More left after food and drink each month:** $1,740 Example figures. The blended food cost got worse while the café took in more and its prime cost improved, because the extra sales came with no extra labor. Judged by the blended number alone, that month looks like a problem. It isn't one. The reverse happens too. Say sandwich cost creeps from 32% to 38% (an extra $360 a month) in a month when coffee sales grow by $2,000 at 22%. The whole café moves from 30.1% to 30.8%, less than a point, and nothing on the P&L tells you it's the sandwiches. A target per category catches it the month it starts. 33% is a median, not a rule The 33% number comes from a survey median of whole restaurants. A café that sells a lot of coffee and little resold food may sit well under it. One that resells a lot may sit near the top of its range and still be healthy. What matters is your prime cost and the dollars each item leaves. ## Set a target for each category Split your menu into the same groups your POS already reports, then sort them into made and resold. Give each group a target range of its own. The ranges below are the example café's own choices, not a standard. They follow its own costs. | Category | Made or resold | Target | Last month | | --- | --- | --- | --- | | Coffee and espresso drinks | Made | 20–25% | 22% | | Sandwiches and toast | Made | 30–34% | 32% | | Pastries from a bakery | Resold | 40–45% | 42% | | Bottled drinks and snacks | Resold | 45–50% | 50% | *Linden Street Café's targets (example). They're its own choices, set so that its sales mix blends into the 24–32% café range.* Then test the targets against your mix. If every category ran at the top of its range, the example café would spend $4,000 + $2,040 + $2,250 + $1,500 = $9,790 on $30,000 of sales: 32.6%, just over its range. So it keeps the coffee drinks near the middle of theirs, where most of its sales are. Your targets only work if, weighted by what you actually sell, they land inside your whole-café range and keep prime cost under your ceiling. To see each category's real cost, you need two splits each month: - **Sales by category** from your POS category report, before tax and tips. - **Purchases by category** from your invoices. Ask your bookkeeper to give resold goods their own accounts (bakery items, bottled drinks), apart from coffee, milk and sandwich food. Count each group in your month-end inventory too, so the cost is what you used, not just what you bought. ## How to price a resold item from its unit cost Start with what one item costs you to sell, not the case price. Three things go in: the delivered price of one, the share you throw out unsold, and the packaging that goes out with each one. *Price for a resold item* **Cost per one sold = Unit cost ÷ (1 − throw-out share) + packaging:** Price = Cost per one sold ÷ category target Unit cost is the case price plus any delivery charge, divided by the items in the case. Throw-out share is the share of what you buy that goes in the bin or gets given away at close, as a decimal: 1 in 10 is 0.10. Packaging (bag, napkin, sleeve) is added after it, because it only goes out with the items you sell. Enter the target as a decimal: 45% is 0.45. *Example: an almond croissant from a wholesale bakery* - Unit cost, delivered: $2.08 - ÷ 0.90 (1 in 10 thrown out at close): $2.31 - + Bag and napkin: $0.09 - = Cost per one sold: $2.40 - Price at 33%: $2.40 ÷ 0.33, about 3 × cost: $7.27 - Price at 45%: $2.40 ÷ 0.45: $5.33 **On the menu at 45%, rounded:** $5.25 (45.7%) All figures are examples. At $5.25 each one leaves $2.85 after its cost. At $7.25, the 33% price rounded, each one leaves $4.85. Dividing by the share you sell matters more for pastries than for bottled drinks. A sealed bottle can wait on the shelf until it sells. A croissant has a day, and each one you throw out raises the cost of the ones you sell. [Free calculatorCost a resold item: plate cost calculatorEnter the case price as the price, the items in the case as the pack size in each, and 1 as the amount used. Under trims, set usable to 90 for a 1-in-10 throw-out. Add the bag as a second ingredient, then enter your category's target food cost to get the price.](https://restaurantdoctorai.com/calculators/recipe-cost) ## When a high markup is costing you sales A higher markup leaves more per item, but only on the items that sell. If guests pass on a pastry at its price, the dollars per item don't help. To compare two prices, work out how many you'd need to sell at the lower one to take in the same dollars after cost: *Items to sell at the lower price to match* Items needed = Dollars left now ÷ Dollars left per item at the new price Dollars left per item = price − cost per one sold. Use a month of sales from your POS item report. The formula assumes your throw-out share stays the same, so if you sell more, check that you're not baking or ordering too many to keep up. Say the almond croissant sells 120 a month at $7.25. That leaves 120 × $4.85 = $582. At $5.25, each leaves $2.85, so the café needs $582 ÷ $2.85 = 204.2, or 205 a month, to take in as much. If it sells more than 205 at the lower price, the lower price leaves more. If it sells fewer, the higher one does. Nothing here tells you which way your guests will go. A test does. Change one item's price, keep everything else the same, and count it in your POS item report for four weeks against the four weeks before. For raising prices on the items you make, see [how much to raise menu prices](https://restaurantdoctorai.com/blog/how-much-to-raise-menu-prices), and for costing a cup, see [what cafés can do about coffee prices](https://restaurantdoctorai.com/blog/coffee-prices-cafes-2026). ## Check the whole café each month Category targets are how you price. The whole-café numbers are how you check. Once a month: - **Each category against its own target.** A category over its range for two months running is the one to look at: re-cost its best sellers on this month's invoices. - **Blended food and drink cost against your range.** If it moved but every category held, your mix moved. Check the dollars left before you act on it. - **Prime cost against your ceiling.** It's the one number that counts both the labor a made item takes and the labor a resold item doesn't. Many counter-service restaurants run above the 60% line. The National Restaurant Association's [2025 Restaurant Operations Data Abstract](https://restaurant.org/research-and-media/media/press-releases/new-resource-from-national-restaurant-association-provides-insights-into-operational-realities/) put the median limited-service restaurant's prime cost at 65 cents of every sales dollar. For how to work out yours, see the [prime cost guide](https://restaurantdoctorai.com/guides/restaurant-prime-cost). ## This week 1. **Split last month's sales.** From your POS category report, sort every category into made in-house or resold, and write down each one's sales. 2. **Split last month's purchases the same way.** Sort the invoices, or ask your bookkeeper for a separate account for resold goods. Work out each category's cost %. 3. **Cost your three best-selling resold items.** Delivered unit cost ÷ the share you sell, plus the bag, in the plate cost calculator. Write down the cost per one sold. 4. **Pick one item priced at three times cost that isn't selling.** Work out how many you'd need to sell at a lower price to match, then decide whether to test it for four weeks. 45-second profit check Is food cost your costliest issue, or labor? A blended food cost can hide where the money goes. The free check asks 12 quick questions and grades each cost area, so you can see your biggest estimated loss in about 45 seconds. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Restaurant operators kept food cost ratios in check in 2024](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/), National Restaurant Association, September 10, 2025 - [Publication 334 (2025), Tax Guide for Small Business](https://www.irs.gov/publications/p334), Internal Revenue Service, 2025 edition, for 2025 returns - [New Resource from National Restaurant Association Provides Insights into Operational Realities](https://restaurant.org/research-and-media/media/press-releases/new-resource-from-national-restaurant-association-provides-insights-into-operational-realities/), National Restaurant Association, August 20, 2025 --- title: "The hidden costs of delivery apps, beyond the commission" description: "What delivery apps cost a restaurant beyond the commission: the guest list, error charges, promotions, ads, app prices and the laws that now help." canonical: "https://restaurantdoctorai.com/blog/hidden-costs-of-delivery-apps" published: "2026-10-12" updated: "2026-10-12" --- # The hidden costs of delivery apps, beyond the commission ## Key takeaways - On the big apps' marketplaces, a restaurant generally gets a guest's first name, last initial and order, not their phone number or email, a federal court said in 2026. - Error charges, promotions you fund and ads can come out of a payout on top of the commission, so your statement's rate can be higher than your plan's. - California, Florida and New York City require a restaurant's agreement or consent before an app lists it or takes its orders, and two set removal deadlines. - Keep the apps for first orders, price the app menu from your statement's real rate, and move regulars to your own ordering, where their details stay with you. The commission is the cost you can see. A delivery app can also cost a restaurant in ways that never show up in the plan's rate: guests whose phone number and email stay with the app, error charges and promotions taken out of the payout, paying to be seen, an app price your guests may not know is higher, and an agreement that can change. They're hidden only in the sense that none of them is in the rate you signed up for. The apps give you something real. They can bring you guests who have never heard of you, and a California Senate committee's [2020 bill analysis](https://sjud.senate.ca.gov/sites/sjud.senate.ca.gov/files/ab_2149_gonzalez_-_senate_judiciary_committee_analysis.pdf) said they can help restaurants grow their sales and customer base. On a delivery order, the app finds the driver who picks up the food and takes it to the guest's door. What follows is the rest of the bill, from court records, city and state laws and the apps' own filings, as read on October 12, 2026. For the plans and rates themselves, see our [post on what the apps charge](https://restaurantdoctorai.com/blog/delivery-app-fees-restaurants-2026). ## The guest list stays with the app When a guest orders on an app's marketplace, they give the app their full name, phone number, email and delivery address. The restaurant that cooks the food generally gets the first name, the last initial and the order. That is how the federal appeals court in New York described it in its August 5, 2026 [decision in DoorDash v. City of New York](https://ww3.ca2.uscourts.gov/decisions/OPN/25-81_complete_opn.pdf). The court said the apps' contracts typically leave control of that data with the app: DoorDash's standard agreement lets restaurants use what they receive only to fill orders and says customer data belongs to DoorDash, and Grubhub allows wider use only in a few cases. New York City tried to change that. Its Customer Data Law of 2021 required the apps to give a restaurant that asked the name, phone number, email, address and order of every customer who had ordered from it, with each customer presumed to agree unless they opted out, order by order. DoorDash, Grubhub and Uber Eats sued. A federal district court held that the law violated the First Amendment and barred the city from enforcing it against the three apps' marketplaces, and the appeals court affirmed. It pointed to an opt-in as the less burdensome choice, and noted that Uber Eats already shares customer data with restaurants under certain agreements when customers opt in. Here is why that costs you. A guest you win through an app is a guest you can reach only through the app. The city told the court that a small share of a restaurant's customers, its loyal patrons, places an outsized share of its orders. While the law was being debated, the New York City Hospitality Alliance warned that a restaurant could not leave an app without “los[ing] access to their own customers.” Not every app product works this way. The same record says that with DoorDash's Storefront, which puts an ordering page on the restaurant's own website, what the guest enters stays with the restaurant. And since July 1, 2025, [Florida's law](https://www.leg.state.fl.us/statutes/index.cfm?App_mode=Display_Statute&URL=0500-0599/0509/Sections/0509.103.html) has required apps to give restaurants a way to contact a guest while the order is being made, on its way, and for up to two hours after pickup. That helps you fix an order. It doesn't give you a guest list. ## What comes off the payout besides the commission The plan rate is what the app charges on the order. Your payout is what's left after everything the app takes out, and three kinds of charges can sit between the two. - **Error charges.** When an order goes wrong, the app can charge part of the cost to you, depending on your app and your agreement. [California's law](https://leginfo.legislature.ca.gov/faces/codes_displayText.xhtml?lawCode=BPC&division=8.&title=&part=&chapter=22.4.&article=) calls these charges for customer order and delivery errors, and requires an app to tell a restaurant how they are calculated, how they are split between the app and the restaurant, and how to dispute them, including whether disputes may be resolved automatically. Outside California, ask your app the same three questions. - **Promotions you fund.** A discount you agree to pay for, such as money off or a free item, comes out of your side of the order. In April 2026, New York City's consumer agency [settled with the delivery app HungryPanda](https://www.nyc.gov/site/dca/news/030-26/mamdani-administration-acts-protect-immigrant-restaurant-owners-delivery-app-junk-fees) after its investigation found the app bundled fees into one line, often relabeled them, and called overcharges “promotion deductions”. HungryPanda agreed to pay more than $580,000 back to more than 380 restaurants. Check every promotion line on your statement against a promotion you agreed to. - **Ads.** Our fees post shows that all three big apps list ads among their optional extras, and for DoorDash, selling them is a growing business. [DoorDash's annual report for 2025](https://www.sec.gov/Archives/edgar/data/1792789/000179278926000013/dash-20251231.htm) says it offers advertising to merchants as a value-added service, and names a growing contribution from advertising among the reasons its 2025 revenue grew. Before you buy, work out what one extra order leaves you after the commission. Your statement is where all of this shows up. Under [Local Law 79 of 2025](https://www.nyc.gov/site/dca/about/new-laws-rules.page), New York City requires apps to send restaurants a monthly statement that itemizes the fees charged, and California requires an itemized cost breakdown of each order for the restaurant. Here is what the extra lines can do to one month. *Example: Main Street Noodle Bar, one app, one month* - Menu sales on the app, before tax and tips: $8,000 - − Commission on a 25% plan: $2,000 - − A promotion the restaurant funded: $300 - − Error charges: $120 - − Ads taken from the payout: $200 **Payout:** $5,380 Example numbers, not benchmarks. The app kept $2,620, or 32.75% of menu sales, on a 25% plan. If your ads are billed to a card instead of the payout, add them back in. That gap matters when you set app prices. At the 25% plan rate, a $20 dish needs $26.67 on the app to bring in $20. At the 32.75% on the statement, it needs $29.74. Price from the statement, not the plan. Our [delivery commission guide](https://restaurantdoctorai.com/guides/delivery-app-commissions#how-commissions-work) has the formula for your real rate. ## Listed on an app you never signed with For years, some apps listed restaurants that had never agreed to it. In December 2024 the FTC and the Illinois attorney general [alleged that Grubhub](https://www.ftc.gov/news-events/news/press-releases/2024/12/ftc-illinois-attorney-general-take-action-against-grubhub-harming-diners-workers-small-businesses) had as many as 325,000 restaurants on its app without their permission, more than half of the restaurants there. According to the FTC, those restaurants got orders from Grubhub drivers for food they didn't serve, drivers' Grubhub cards were sometimes declined and left food unpaid, and diners blamed the restaurants when the food came late or in poor condition. When restaurants asked to come off, the FTC said, Grubhub would try to sell them a paid partnership instead, and often removed them only after they threatened legal action. Grubhub settled, neither admitting nor denying the allegations, according to the [court's order](https://www.ftc.gov/system/files/ftc_gov/pdf/013_2024-12-31_grubhub_stipulated_order.pdf), and the FTC said the order requires it to list restaurants only with their consent. Chicago made a similar claim against DoorDash: the city's complaint said DoorDash listed restaurants without their consent. In November 2025, the [city announced a settlement](https://www.chicago.gov/city/en/depts/mayor/press_room/press_releases/2025/november/doordash-practices-settlement.html) in which DoorDash would pay $3.25 million to restaurants listed without consent that are no longer on the app, give $5.8 million in commission and marketing credits to eligible restaurants on it, and not list Chicago restaurants without consent again. A settlement ends the claims without a court ruling on them. DoorDash's annual report says it doesn't add new non-partner restaurants for delivery in the United States, and that many US places now require a restaurant's consent first. Three places now put it in writing, as their laws or agency pages read on October 12, 2026. Read each one for who is covered. - **California.** An app needs an agreement that expressly authorizes it before it arranges delivery of your orders, and must remove you within 3 business days of your request, existing contracts aside. - **Florida.** An app needs your express consent, written or electronic, and must remove your listing within 10 days of your request, unless an existing agreement says otherwise. - **New York City.** An app needs a written agreement before it lists you, says the city's [page for restaurants using delivery apps](https://www.nyc.gov/site/dca/businesses/Restaurants-Using-Delivery-Apps.page). The page sets no deadline for taking a listing down. Search for your restaurant's name on each app now and then. If you find a listing you never signed for, ask the app in writing to take it down, and keep the date you asked. ## The price guests see, and who they blame If you raise app prices to cover the commission, some guests will see a higher price than your menu without knowing why. Chicago's complaint said DoorDash did not disclose that menu prices on its app were often higher than buying directly from the restaurant. In April 2026 the [FTC asked for public comment](https://www.ftc.gov/news-events/news/press-releases/2026/04/ftc-seeks-public-comment-unfair-deceptive-fee-practices-online-food-grocery-delivery-services) on whether a rule is needed for delivery apps' fees and prices, including whether apps make clear when an item's price differs from the store's or restaurant's. That's no reason to leave app prices too low. New York City says restaurants can charge different prices on an app, with limited exceptions, and Florida bars an app from changing a restaurant's prices without an agreement. It is a reason to tell guests. California lets a restaurant direct the app to show customers the delivery fee and each fee or commission it charges the restaurant. Anywhere, a line on your own menu or website saying app prices are higher because of app fees is free to add, if your agreement allows it. A late or cold delivery also carries your name. The FTC said diners blamed the restaurants in its complaint for Grubhub's late and poor deliveries, and Florida now requires apps to give restaurants a way to respond to guests' ratings and reviews. Where your app has a reply tool, answer the guest, and say how to order from you directly next time. ## When the app changes the deal Your agreement can change. DoorDash's annual report says competition has led it, and may lead it in the future, to change its commission rates and fees. California requires an app to tell a restaurant about any material change to its agreement, and New York City's Local Law 79 requires apps to tell restaurant partners about new services, fee caps and fee changes. New York City has also been reviewing its fee caps (our fees post has the details), so check the current numbers. Leaving is allowed: the same report says partner merchants can generally end their agreements with 7 or 30 days' notice. What's hard to take with you is the guests. If your app regulars exist for you only as first names inside the app, a rate change leaves two choices: pay it, or leave and lose your easiest way to reach them. That is the cost of the guest list, and the reason to start building your own now. ## What to do: keep the apps for new guests None of this means switching the apps off. They're good at the first order. The aim is to stop paying a commission on the tenth and the fiftieth. *Example: one regular, one year* - 24 orders a year, 2 a month at $35: $840 - On the app at a 25% real rate, the app keeps: $210 - On your own ordering link for pickup, at an example 3% card fee: $25.20 **Difference for one guest in a year:** $184.80 Example numbers. $840 × 25% = $210 and $840 × 3% = $25.20. Delivery you run yourself has costs of its own; this example uses pickup. 1. **Put a card in every app bag** with your own ordering link or phone number and a reason to use it next time, so a guest who liked the food can find you without the app. Check your agreement first. New York City says restaurants may include menus and coupons in app orders. 2. **Build your own list, with permission.** A sign-up for news and offers on your website, receipts and bag cards lets you reach regulars yourself when an app changes its rates. Ask first: the appeals court faulted New York's law for building a marketing list no customer asked to join. Marketing emails and texts have legal rules of their own, so check them before you send. 3. **Make your own ordering the first thing guests find.** Check your Google listing's order button and phone number and your website's order links, so a guest who searches for you lands with you. Our check, [How AI sees your website](https://restaurantdoctorai.com/online-check), reads your website, and its paid full report says if its order links go only to the delivery apps. 4. **Price the app menu from your statement's real rate,** so each app order leaves you what a dining room sale of the same dishes does. 5. **Keep a log of error charges,** with the order number, the date and what the guest reported, so you can spot a pattern and have the details ready when you dispute one. [Free calculatorYour break-even app price: delivery price calculatorPut in your price and the real rate from your payout statement to get the app price that leaves you what a dining room sale does.](https://restaurantdoctorai.com/calculators/delivery-app-price) ## This week 1. **Pull last month's payout statement for each app** and list every line that isn't the commission: promotions, error charges, ads and adjustments. Add them up, so you can see your real rate next to your plan's. 2. **Ask each app three questions:** how error charges are calculated, how they're split between you, and how to dispute them. California makes apps answer these, and the answers tell you which charges are worth disputing anywhere. 3. **Search for your restaurant on each app and on Google,** so you can see whether the listing, the phone number and the order button lead to you. 4. **Print a bag card** with your own ordering link, if your agreement allows it. 5. **Re-price the app menu** with the real rate in the calculator, starting with the dishes that sell most on the apps. 45-second profit check Are the apps your costliest issue? The free check asks 12 quick questions, including how much of your sales come through delivery apps, and shows your biggest estimated loss in about 45 seconds. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [AB 2149 (Gonzalez), Food delivery platforms: Senate Judiciary Committee analysis](https://sjud.senate.ca.gov/sites/sjud.senate.ca.gov/files/ab_2149_gonzalez_-_senate_judiciary_committee_analysis.pdf), California State Senate, Judiciary Committee, August 2020 (hearing August 18, 2020) - [DoorDash, Inc. v. City of New York, No. 25-81 (opinion)](https://ww3.ca2.uscourts.gov/decisions/OPN/25-81_complete_opn.pdf), U.S. Court of Appeals for the Second Circuit, August 5, 2026 - [Section 509.103, Food delivery platforms](https://www.leg.state.fl.us/statutes/index.cfm?App_mode=Display_Statute&URL=0500-0599/0509/Sections/0509.103.html), The Florida Legislature, 2026 Florida Statutes - [Business and Professions Code, Chapter 22.4: Food Delivery Platforms](https://leginfo.legislature.ca.gov/faces/codes_displayText.xhtml?lawCode=BPC&division=8.&title=&part=&chapter=22.4.&article=), California Legislative Information, As amended through January 1, 2026 - [Mamdani Administration Acts to Protect Immigrant Restaurant Owners From Delivery App Junk Fees](https://www.nyc.gov/site/dca/news/030-26/mamdani-administration-acts-protect-immigrant-restaurant-owners-delivery-app-junk-fees), NYC Department of Consumer and Worker Protection, April 8, 2026 - [DoorDash, Inc. Annual Report (Form 10-K) for 2025](https://www.sec.gov/Archives/edgar/data/1792789/000179278926000013/dash-20251231.htm), DoorDash, Inc., filed with the U.S. Securities and Exchange Commission, February 18, 2026 - [New Laws & Rules](https://www.nyc.gov/site/dca/about/new-laws-rules.page), NYC Department of Consumer and Worker Protection - [FTC, Illinois Attorney General Take Action Against Grubhub for Harming Diners, Workers, and Small Businesses](https://www.ftc.gov/news-events/news/press-releases/2024/12/ftc-illinois-attorney-general-take-action-against-grubhub-harming-diners-workers-small-businesses), Federal Trade Commission, December 17, 2024 - [Stipulated Order for Permanent Injunction, Monetary Judgment, and Other Relief (FTC and Illinois v. Grubhub)](https://www.ftc.gov/system/files/ftc_gov/pdf/013_2024-12-31_grubhub_stipulated_order.pdf), U.S. District Court for the Northern District of Illinois, via the FTC, December 31, 2024 - [Johnson Administration Reaches $18 Million Settlement With DoorDash Over Deceptive Business Practices](https://www.chicago.gov/city/en/depts/mayor/press_room/press_releases/2025/november/doordash-practices-settlement.html), City of Chicago, Mayor's Press Office, November 14, 2025 - [Restaurants Using Delivery Apps](https://www.nyc.gov/site/dca/businesses/Restaurants-Using-Delivery-Apps.page), NYC Department of Consumer and Worker Protection, Updated January 2026 - [FTC Seeks Public Comment on Unfair and Deceptive Fee Practices in Online Food and Grocery Delivery Services](https://www.ftc.gov/news-events/news/press-releases/2026/04/ftc-seeks-public-comment-unfair-deceptive-fee-practices-online-food-grocery-delivery-services), Federal Trade Commission, April 14, 2026 --- title: "Restaurant consultant cost, free help and when to hire" description: "What a restaurant consultant costs, from two firms' published prices; free help from SCORE and SBDCs; and when an automated report is a better first step." canonical: "https://restaurantdoctorai.com/blog/restaurant-consultant-vs-profit-report" published: "2026-10-10" updated: "2026-10-10" --- # Restaurant consultant cost, free help and when to hire ## Key takeaways - Two consultants who publish prices charge $995 for a day, or from $1,250 to start, then $90 to $150 an hour. - SCORE mentors give advice at no cost, and nearly 1,000 Small Business Development Centers give no-cost consulting. - Our report is automated from 12 answers: no visit, no one doing the work with you and no weekly accountability. - Hire a consultant when someone has to see and fix how you run; start with the free check when you don't know where to look. Your sales look fine, but the bank account doesn't agree. So you wonder: is it time to hire a restaurant consultant? A good one can see things you can't and do the work with you. They also cost real money. Two consultants who publish their prices charge $995 for a day, or from $1,250 to start, and then $90 to $150 an hour. Before you pay anyone, there's free help from SCORE and Small Business Development Centers, and a 45-second free check that tells you where to look first. ## What a restaurant consultant does A restaurant consultant is someone you pay to look at how your place runs and help fix it. The two consultants below describe the work on their own pages, and it's hands-on: - They come and see. One visits first, usually without saying who he is, and orders a drink or an appetizer to watch the service. The other offers a day on site with a mini audit and a review of the kitchen and front-of-house workflow. - They find the problem and write a plan. Both start with a diagnosis and a list of what to fix first. - They do the work with you: menu costing and menu engineering, inventory systems, hiring steps, written procedures and staff training. - They can stay on. One bills further work by the hour. The other sells monthly retainers and a coaching program with weekly accountability. That last point is the one software can't copy. A person who checks in every week makes it harder to let the plan slide when the dinner rush hits. ## What a restaurant consultant costs Most consultants don't list prices, and we found no neutral survey of what they charge. So here are two who do, as each one's own page lists them on October 10, 2026: [5 Loaves Restaurant Consulting](https://www.5-loaves-marketing.com/marketing-tips/how-much-does-it-cost-to-hire-a-restaurant-consultant) in Connecticut, and [Chef Eric](https://911cheferic.com/restaurant-consultant-vancouver/pricing/) in Vancouver, Canada. They are examples, not a market average, and not a recommendation. | | 5 Loaves (Connecticut) | Chef Eric (Vancouver) | | --- | --- | --- | | How it starts | A visit and a talk about what he saw, then talk of working together | A free call, or a one-day session: $995 for 8 hours | | A first project | Full restaurant from $2,500; one area, like the bar, about $1,250 | 10 hours for $1,500, up to 100 hours for $10,000 | | By the hour | $150 for ongoing work | $150 down to $100 inside a package | | Ongoing help | Billed by the hour | $1,500 to $4,500 a month, for at least 4 months | | Expenses | Not listed on the page | Travel, stay, meals and supplies at cost plus 10% | *Two restaurant consultants' published prices, as listed on their own pages on October 10, 2026. Chef Eric's page shows a $ sign without saying US or Canadian dollars.* The shapes repeat: a starting fee or a day to diagnose, a package or project to fix, then hours or a monthly retainer. The total depends on how much you want done, and whether the consultant does it or your team does. ### How to ask for a quote Ask for it in writing, with answers to these: 1. What does the first fee include, and what is billed by the hour after it? 2. Is the work on site or online? Are travel and expenses extra, and how are they billed? 3. Is there a minimum term, a deposit or a notice period? One of the two examples above has all three. 4. What will I have in writing at the end, and who does the work: you or my team? 5. Which numbers do you need from me, and how will we both tell whether it worked? 6. Have you worked with a restaurant like mine: the same type, size and kind of service? ## Free help first: SCORE and Small Business Development Centers If you want a person but can't pay one yet, two national networks help small businesses, and the U.S. Small Business Administration lists both on its [resource partners page](https://www.sba.gov/counseling/local-assistance/resource-partners/). - SCORE is a nonprofit network of volunteer business mentors. The SBA says they give advice at no cost, by email, phone and video, on things like financing, staffing and business planning, and keep meeting with you over time. SCORE also runs workshops and webinars. - Small Business Development Centers (SBDCs) give one-on-one business advising and training: money management, planning, financing, marketing and more. Their association, [America's SBDC](https://americassbdc.org/about-us/), says nearly 1,000 local centers give no-cost consulting and low-cost training. They are hosted by universities, colleges, state agencies and private partners. The SBA page has a search by ZIP code for both. Two things to know. A mentor or adviser is matched by where you are, so ask early whether they have run or advised a restaurant. And they advise: they don't usually cost your menu or train your line cooks for you. For a question about money, like a loan, a lease or a cash squeeze, they are a strong first call. ## What our report does, and what it doesn't RestaurantDoctorAI is software. You answer 12 multiple-choice questions about how you run things: your type of restaurant, monthly sales, prime cost, how often you update recipe costs, waste, drink pours, slow-shift staffing, rent, delivery apps, how full the dining room is, repeat guests and your rating. Here's what you get and what you don't. ### What it does - The free check: your grade, a grade for each cost area, an estimate of the money lost to avoidable costs each year, your costliest issue and the first fix, in about 45 seconds. No account, name or email. - The full report, paid once with no subscription: dollar estimates for each problem, your prime cost against RestaurantDoctorAI's target for your type of restaurant, a 72-hour plan, a 30-day roadmap with the number to watch for each fix, and printable cards for your team. - It opens on screen, usually within a minute of paying, and comes by email. - The figures come from fixed formulas, not from the AI. The AI writes the words around them, and every fix comes from a playbook of standard restaurant fixes. ### What it doesn't do - No one visits. It can't see your menu, your prices, your lease, your team or your street. - It works from multiple-choice answers, so every figure is an estimate: a range, not a measurement of your books. - No one calls you next week to ask whether you did it. There's no accountability partner. The closest thing is a second report that adds last month's real numbers and one check-up after the plan's first full month, and that's still software. - It doesn't do the work. It hands you the plan and the team cards, and you and your staff carry them out. How every figure is worked out is on the [methodology page](https://restaurantdoctorai.com/methodology), and what each report costs is on the [pricing page](https://restaurantdoctorai.com/pricing). It is a one-time payment, smaller than one hour with either consultant above. ## Side by side | | A consultant | SCORE or an SBDC | Our check and report | | --- | --- | --- | --- | | Who looks | A person, on site or online | A volunteer mentor or a center's adviser | Software, from your 12 answers | | What it costs | $995 a day or from $1,250 to start, then $90 to $150 an hour | SCORE: no cost. SBDC: no-cost advising, low-cost training | Check free; report paid once | | Sees your restaurant | Yes | Only what you share | No | | Does the work with you | Yes, if you pay for it | Advises; you do the work | No: a plan and team cards | | Checks in later | Yes, on a retainer | Yes, if you keep meeting | One emailed check-up, on the second report only | *The three kinds of help side by side. Consultant prices are the two examples above, as listed on October 10, 2026, not an average.* ## Choose a consultant when… start with the report when… ### Choose a consultant when - The trouble is something a person has to see: slow tickets, a messy line, service that drags, a team that doesn't follow the recipe. - You need hands doing the work: costing every plate, building an inventory system, writing procedures, training staff. - You know you won't follow a plan alone, and you want someone to hold you to it each week. - You're opening, changing your concept or rebuilding the menu, and a mistake would cost far more than the fee. ### Start with free help when - You want a person to talk it through but can't pay one yet. - The question is about money: a loan, a lease, a business plan, cash flow. ### Start with the report when - The numbers feel off and you don't know where to look first. - You want an answer tonight, before you pay anyone by the hour. - You want a plan you and your team can start this week, with no new software or hires. - You'll hire a consultant anyway, and want to walk in knowing your costliest issue, so you can spend the paid hours on the fix, not the search. You can use more than one These aren't rivals. Take the free check tonight, book a free SCORE or SBDC session for the money questions, and hire a consultant for the part that needs hands in your kitchen. If you already work with a consultant or an accountant, they can run the check for you: [advisors can prepay report codes](https://restaurantdoctorai.com/advisors). ## This week 1. Take the free check. It takes about 45 seconds and shows your biggest estimated loss. 2. Pull one month's net sales, food and drink invoices and payroll, and put them in the [prime cost calculator](https://restaurantdoctorai.com/calculators/prime-cost). The [guide to the usual problems](https://restaurantdoctorai.com/guides/restaurant-busy-but-not-making-money) explains what the number means. 3. If the problem is money, like a loan, your lease or cash flow, find a SCORE mentor or an SBDC near you on the SBA's page and book a session. 4. If you call a consultant, send the six quote questions above and ask for the answers in writing. [Free calculatorWork out yours: prime cost calculatorFood, drink and labor as a share of your sales, against the target for your type of restaurant.](https://restaurantdoctorai.com/calculators/prime-cost) 45-second profit check Find out where to look before you pay anyone Answer 12 quick questions about prices, costs, staffing and guests. In about 45 seconds you get your grade and your biggest estimated loss, free, so you know which problem to bring to a consultant, a mentor or your own team. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [How Much Does It Cost to Hire a Restaurant Consultant?](https://www.5-loaves-marketing.com/marketing-tips/how-much-does-it-cost-to-hire-a-restaurant-consultant), 5 Loaves Restaurant Consulting, August 2, 2026; read October 10, 2026 - [Restaurant Consulting Packages & Pricing](https://911cheferic.com/restaurant-consultant-vancouver/pricing/), Chef Eric, 911cheferic, no date shown; read October 10, 2026 - [Resource Partners](https://www.sba.gov/counseling/local-assistance/resource-partners/), U.S. Small Business Administration, no date shown; read October 10, 2026 - [About Us](https://americassbdc.org/about-us/), America's SBDC, no date shown; read October 10, 2026 --- title: "Free restaurant calculators compared: 6 brands checked" description: "We checked free restaurant calculators from Toast, 7shifts, Homebase, Lightspeed, MarketMan and SpotOn: sign-up, formulas and target sources." canonical: "https://restaurantdoctorai.com/blog/free-restaurant-calculators-compared" published: "2026-10-10" updated: "2026-10-14" --- # Free restaurant calculators compared: 6 brands checked ## Key takeaways - We checked 8 free calculators from 6 software companies on October 10, 2026: 5 ask for at least your email before you see a result, and 3 don't. - On their own pages, only Homebase said where its targets come from; the others gave ranges with no source, or none. - Some do what ours don't: spreadsheet templates that recalculate as you type, labor built role by role with overtime, a P&L template and a printable break-even sheet. - RestaurantDoctorAI's calculators need no email, show each formula with your numbers in it, and say where their targets come from. Search for a free restaurant calculator and most of what you find comes from companies that sell restaurant software. Some tools work right on the page. Others are a spreadsheet you get after you hand over your name, email and phone number. We opened 8 free calculators from 6 well-known companies on October 10, 2026, and checked each one on four things an owner cares about. Then we held our own to the same four. ## How we checked We read each tool's own page in a browser. We didn't fill in or send any form, and we didn't download any file that sits behind one. So we only say what the page itself shows. If a spreadsheet may hold more inside, we say we couldn't see it. The four questions: - What it works out: food cost, labor cost, prime cost or break-even. - What it asks before you see a result: nothing, an email, or a full form. - Whether the formula is shown, so you can check the math by hand. - Whether it gives targets, and whether it says where they come from. ## Free restaurant calculators, side by side | Company and tool | Works out | Before you see a result | Formula shown | Targets and their source | | --- | --- | --- | --- | --- | | [Toast: Labor Cost Calculator](https://pos.toasttab.com/resources/restaurant-labor-cost-calculator) | Labor cost, hourly and salaried (Excel file) | A form: name, email, phone, business name and role; the link comes by email | In the file, the page says; not on the page | None on the page | | [Toast: COVID-19 Restaurant Financial Toolkit](https://pos.toasttab.com/resources/restaurant-financials-during-covid) | Prime cost as a share of gross sales, plus P&L, budget, balance sheet and overhead rate (Excel file) | The same five-field form | In the file; not on the page | None on the page | | [7shifts: labor cost calculator](https://www.7shifts.com/calculators/labor-cost-calculator/) | Labor cost and labor %, role by role, with overtime pay | Nothing: it works on the page | Yes, in its FAQ | 25–35% of revenue, called industry standards; no source named | | [7shifts: food cost template](https://www.7shifts.com/resources/templates/restaurant-food-cost-template/) | Food cost % (Excel file) | A form: first and last name, email, and whether you're a restaurant | Yes, in its FAQ | Ranges by type (for example 18–22% at quick service); no source named | | [Homebase: restaurant labor cost calculator](https://www.joinhomebase.com/labor-cost-calculator/restaurant) | Labor cost and labor %, with overtime and double time | No email box among its inputs | Yes, in its FAQ | 25–31% quick service, 34–38% full service; sources named, not linked | | [Lightspeed: food cost calculator](https://www.lightspeedhq.com/resources/retail/food-cost-calculator/) | A dish's cost, markup, profit per sale and price | The page asks for your business type and email first | Not on the page | None on the page | | [MarketMan: prime cost calculator](https://www.marketman.com/resources/restaurant-prime-cost-calculator) | Labor, cost of goods sold from your inventory counts, and prime cost (a sheet) | A form: name, email, cell phone, restaurant, job title, locations and country | The steps are described; the formula isn't written out | Checks you against benchmarks by type; no figures or source on the page | | [SpotOn: break-even calculator](https://www.spoton.com/Break-Even-Point-Calculator.pdf) | Break-even sales and guests by month, day and quarter (fillable PDF) | Nothing: the link opens the PDF | Yes, printed at the top | None | *Free restaurant calculators from 6 software companies, as each page showed them on October 10, 2026. A form means the tool is sent to you after you fill it in.* Three of the eight work without asking who you are: 7shifts' labor calculator, Homebase's and SpotOn's break-even sheet. The other five ask for at least an email first, and three of those ask for a phone number too. That isn't a trick: these are free tools from companies that sell software, and the form is how they hear from people who might buy. Toast's, for one, says sending it signs you up for its marketing emails. Just know before you start that the form is part of the deal. Targets are where the tools differ most. A number like “labor should be 30%” is only as good as where it came from. On these pages, Homebase is the only one that says where its ranges come from. The rest give ranges with no source, or none at all, at least on the page. ## What these tools do that ours don't Here's where each one does something RestaurantDoctorAI's calculators don't: - A spreadsheet you keep. Toast's, 7shifts' food cost and MarketMan's files are yours to keep, change and reuse week after week. Five of ours (prime cost, labor cost, plate cost, break-even and the restaurant budget) let you download your result as a spreadsheet (CSV), but it's a record of the figures you typed, not a template that works things out as you change them. - Labor built up role by role. 7shifts takes each role's hours and hourly rate and works out the overtime pay for you. Homebase adds double-time hours, which matter in California and other states with daily overtime. Ours starts from the hourly pay on your payroll report, overtime already in. - More than one statement. Toast's toolkit has a P&L, a budget and a balance sheet template. Our budget calculator's spreadsheet is a simple monthly budget; we have no P&L or balance sheet template. - A sheet to print. SpotOn's PDF has a line for every bill, from rent to natural gas, and gives a quarterly figure too. Ours works by month, week and day. - Sources named for labor targets. Homebase lists where its ranges come from, right in its FAQ. Compare like with like Calculators don't all count the same things. Toast's toolkit works prime cost out on gross sales, and 7shifts' labor tool leaves payroll taxes and benefits out, as it says itself. Ours use net sales and count payroll taxes and benefits in labor. The same restaurant can get different answers from two tools. Check which sales figure and which costs each one uses before you compare. ## RestaurantDoctorAI's calculators on the same four questions We read our own code to fill this in, not our memory. Our free [restaurant calculators](https://restaurantdoctorai.com/calculators) all work on the page, with no email, no account and no sign-up. Each page shows its formula, and a “How we worked it out” line opens the same formula with your own numbers in it, so you can check it by hand. | Calculator | Works out | Before you see a result | Formula shown | Targets and their source | | --- | --- | --- | --- | --- | | [Prime cost](https://restaurantdoctorai.com/calculators/prime-cost) | Prime cost %, food and labor halves, against your type of restaurant | Nothing | Yes, with your numbers | RestaurantDoctorAI's ceilings, 57–70% by type; the page names RestaurantOwner.com and the National Restaurant Association, and says which ceilings are our own | | [Labor cost](https://restaurantdoctorai.com/calculators/labor-cost) | Labor %, the hours next week's sales pay for, and sales per labor hour | Nothing | Yes, with your numbers | RestaurantDoctorAI's ranges, 25–38% by type, shown next to the National Restaurant Association's medians | | [Break-even](https://restaurantdoctorai.com/calculators/break-even) | Sales and guests needed by month, week and day; a range if you're not sure how hourly pay behaves | Nothing | Yes, with your numbers | None: break-even is your own number | | [Food cost](https://restaurantdoctorai.com/calculators/recipe-cost) | One plate's cost, its food cost % and the price that hits your target | Nothing | Yes, with your numbers | RestaurantDoctorAI's food cost ranges by type; sources on the methodology page | | [Delivery app price](https://restaurantdoctorai.com/calculators/delivery-app-price) | The app price that leaves you what the dish earns in-house | Nothing | Yes, with your numbers | None: it uses your real commission rate | *RestaurantDoctorAI's free calculators, from the site's own code, October 10, 2026.* Our targets are RestaurantDoctorAI's targets, not industry rules. Where an outside source publishes a line, as RestaurantOwner.com does for prime cost at quick service and table service, ours match it. Where none does, as for fine dining and bars, the page says the number is ours. The [methodology page](https://restaurantdoctorai.com/methodology) lists every source. In the prime cost, labor cost and break-even calculators you can tap “Remember on this device” to keep your numbers in your own browser, so they're there next time, and break-even offers to start from a month you saved in another one. Where ours falls short is in the list above: no spreadsheet template that recalculates as you type (five of ours give a CSV, a record of your figures, not a template), no role-by-role labor and no P&L template. ## Which one fits the job - To check one number now, without giving your email: 7shifts' or Homebase's labor tool, SpotOn's break-even sheet, or [ours](https://restaurantdoctorai.com/calculators). - To build next week's schedule role by role, with overtime pay worked out: 7shifts' labor calculator. - To track numbers week after week in your own file, and you don't mind giving your phone number: Toast's or MarketMan's spreadsheet. - To see whether your number is good for your type of restaurant, with the source of the target in view: our [prime cost](https://restaurantdoctorai.com/calculators/prime-cost) and [labor cost](https://restaurantdoctorai.com/calculators/labor-cost) calculators, or Homebase for labor. Whatever you pick, the math is the same. Every formula these tools use is in our [restaurant formulas guide](https://restaurantdoctorai.com/guides/restaurant-formulas), and the guides to [prime cost](https://restaurantdoctorai.com/guides/restaurant-prime-cost), [labor cost](https://restaurantdoctorai.com/guides/restaurant-labor-cost) and [food cost %](https://restaurantdoctorai.com/guides/food-cost-percentage) explain what counts in each. [Free calculatorStart with the biggest one: prime cost calculatorFood, drink and labor as a share of your sales, against the target for your type of restaurant. No sign-up.](https://restaurantdoctorai.com/calculators/prime-cost) ## This week 1. Pull one week's net sales, food and drink invoices and payroll from your POS and payroll reports. 2. Put them in a [prime cost calculator](https://restaurantdoctorai.com/calculators/prime-cost), ours or any of the others. Write down which sales figure and which costs it used. 3. If it gives a target, find out where the target comes from. If it can't tell you, treat it as a rule of thumb. 4. If labor is the half running high, build next week's schedule from the hours your sales can pay for. 45-second profit check A calculator gives you one number The free check asks 12 quick questions about prices, costs, staffing and guests, and shows your biggest estimated loss in about 45 seconds, so you know which number to work on first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Restaurant Labor Cost Calculator](https://pos.toasttab.com/resources/restaurant-labor-cost-calculator), Toast, No date shown; read October 10, 2026 - [The COVID-19 Restaurant Financial Toolkit](https://pos.toasttab.com/resources/restaurant-financials-during-covid), Toast, No date shown; read October 10, 2026 - [Restaurant labor cost calculator](https://www.7shifts.com/calculators/labor-cost-calculator/), 7shifts, No date shown; read October 10, 2026 - [Free restaurant food cost calculator for Excel](https://www.7shifts.com/resources/templates/restaurant-food-cost-template/), 7shifts, No date shown; read October 10, 2026 - [Restaurant Labor Cost Calculator + AI Schedule Optimizer](https://www.joinhomebase.com/labor-cost-calculator/restaurant), Homebase, No date shown; read October 10, 2026 - [Food Cost Calculator](https://www.lightspeedhq.com/resources/retail/food-cost-calculator/), Lightspeed, No date shown; read October 10, 2026 - [Restaurant Prime Cost Calculator](https://www.marketman.com/resources/restaurant-prime-cost-calculator), MarketMan, No date shown; read October 10, 2026 - [Restaurant Break-Even Point (BEP) Calculator](https://www.spoton.com/Break-Even-Point-Calculator.pdf), SpotOn, No date shown; read October 10, 2026 --- title: "Restaurant POS reports: what they show and what they miss" description: "What Toast, Square, Clover and Lightspeed reports show, the 5 to pull each week, what a POS can't know on its own, and where a profit check fits." canonical: "https://restaurantdoctorai.com/blog/pos-reports-vs-profit-check" published: "2026-10-10" updated: "2026-10-10" --- # Restaurant POS reports: what they show and what they miss ## Key takeaways - Every POS we checked reports net sales and what sold; most add a labor report built from staff clocking in on it. - Item cost columns use a set cost per item, so they show what a dish should cost, not what your counts say the kitchen used. - Without your invoices, counts, full payroll and rent added, a POS can't give real food cost, prime cost or rent share. - RestaurantDoctorAI's free check adds targets by restaurant type and your costliest issue; it doesn't connect to your POS. Your POS already knows what you sold, when you sold it and who rang it up. What it doesn't know on its own is what the food really cost you, what your whole payroll and rent come to, or whether your numbers are healthy for your kind of restaurant. Here's what Toast, Square, Clover and Lightspeed say their reports show, the five to pull every week, and where a profit check fits beside them. ## What your POS reports already show Every POS we checked offers the same three families of report: a sales summary, a report of what sold (often called the product mix), and, on most, a labor report built from staff clocking in on the system. This is what each maker's own help pages say, read on October 10, 2026. | POS | Sales | What sold | Labor | | --- | --- | --- | --- | | Toast | [Sales summary](https://support.toasttab.com/en/article/How-do-I-access-sales-reports): overall sales, revenue by category, tips and taxes, for the past 15 months | [Product mix](https://support.toasttab.com/en/article/Product-Mix-PMIX-Report-Overview): by menu, menu group, item and modifier | [Labor summary](https://support.toasttab.com/en/article/Labor-Reports-Overview): hours and pay, from clocked-in hours and each job's pay rate | | Square | [Sales summary](https://squareup.com/help/us/en/article/5381-in-app-summaries-and-reports): gross and net sales, discounts, comps, refunds, tips and taxes | [Item sales](https://squareup.com/help/us/en/article/8363-view-item-category-and-modifiers-sales-reports): quantity sold, gross and net sales for each item | [Labor vs sales](https://squareup.com/help/us/en/article/6140-employee-timecard-reporting): hourly labor cost against net sales, with a Square Shifts Plus plan and a wage on each profile | | Clover | [Sales report](https://www.clover.com/en-US/help/clover-reports-overview): gross sales, refunds and net sales, by daypart if you set them up | [Item sales](https://www.clover.com/en-US/help/understand-item-sales-report-information): quantity sold, net sales, and cost and gross profit for items that have a cost | Its reports overview lists an employee sales report (sales, tips and refunds by employee), not a labor cost report. Its [Time Clock app](https://www.clover.com/pos-systems/employees/m/time-and-attendance), powered by Homebase, records clock-ins and timecards | | Lightspeed (K-Series) | [Sales totals](https://k-series-support.lightspeedhq.com/hc/en-us/articles/18234292249883-Sales-Reports): gross or net sales, covers, orders and average order | [Products report](https://k-series-support.lightspeedhq.com/hc/en-us/articles/4403156003227-Products-report): sales by item, with cost and margin | [Labor report](https://k-series-support.lightspeedhq.com/hc/en-us/articles/4509139748251-Labor-Report): hours worked and average sales per hour, for each employee | *What four POS systems' help pages say their main reports show. Names are as each maker uses them; plans and menus change, so check yours.* These are good reports. They count every sale, down to the modifier, and they're there every morning. Nothing an owner types into a website can match that for what happened. The gap is on the other side of the ledger: costs. ## The 5 POS reports to pull every week Toast's own [reporting best practices](https://support.toasttab.com/en/article/Reporting-Best-Practices-Weekly-Monthly-Quarterly-Reports) suggest running the product mix weekly or monthly and the labor summary weekly or every two weeks for payroll, and point to the hourly sales report for building schedules. That's one maker's advice, not a rule. Here is a short weekly list built from it, using whatever your POS calls each report: 1. **Sales summary.** Write down net sales for the week. Square defines net sales as gross sales minus returns, discounts and comps. It's the number every cost share is divided by, so use the same figure every week. 2. **Product mix or item sales.** What sold and how many. It tells you which dishes carry the menu. Multiply a dish's count by its plate cost (the [recipe cost calculator](https://restaurantdoctorai.com/calculators/recipe-cost) works one out) to see what it costs you a week. 3. **Labor summary or labor vs sales.** Hours and pay against sales. It's what you check before next week's schedule goes out, and what payroll is built from. 4. **Hourly or daily sales.** Your busiest and slowest hours. Lightspeed's hourly sales report shows average labor hours for each hour beside the sales, and its daily sales report gives sales per labor hour. Use it to put shifts where the guests are (the [sales per labor hour guide](https://restaurantdoctorai.com/guides/sales-per-labor-hour) explains the number). 5. **Discounts, comps and removed items.** Money given away. Clover's reports overview describes a discounts report that shows the total impact on revenue, and a removed items report that records which employee took an item off an order. Toast suggests a weekly or monthly cash activity audit. ## What a POS can't know on its own A POS records sales. Most of your costs live elsewhere: supplier invoices, the payroll report, the lease. So four numbers that decide whether you make money don't come out of a POS on their own. ### Your real food cost Real food cost is what the kitchen used, against what it sold. The [food cost guide](https://restaurantdoctorai.com/guides/food-cost-percentage) works through it: *Food cost %* Food cost % = (beginning inventory + purchases − ending inventory) ÷ food sales × 100 Purchases come from invoices and inventory from a count on the shelf. A POS sees neither unless you add them. The cost columns in item reports run on a cost per item that you set up. Clover multiplies the quantity sold by the cost you give each item, and warns that gross profit reads higher than it should if some items have no cost. Lightspeed's [cost price](https://k-series-support.lightspeedhq.com/hc/en-us/articles/1260804656109-Importing-and-exporting-items-in-bulk) is the item's cost to the restaurant, used with its price to work out the margin. Toast's product mix shows cost of goods once items have a cost from Menu builder or from xtraCHEF recipes. That is what each dish should cost. Waste, over-portioning and a supplier's price rise only show up when invoices and counts come in. To be fair to the POS makers, some sell tools that close part of this gap. Toast's [xtraCHEF](https://support.toasttab.com/en/article/xtraCHEF-Recipe-Reporting) tracks how recipe ingredient costs change as you scan invoices. Square's [cost of goods sold report](https://squareup.com/help/us/en/article/8522-track-cost-of-goods-sold) comes with Square for Retail Plus or Premium, or Square Plus or Premium with advanced inventory added. If you run one of these and keep it fed, your food cost is measured, which beats any estimate. ### Your whole labor cost Toast works out its labor summary from clocked-in hours and each job's pay rate, and says it shows no labor cost for salaried staff. Square's labor vs sales compares hourly labor costs with net sales. Useful for the schedule, but the labor share can read lower than your full payroll. RestaurantDoctorAI's targets count total payroll: wages, salaries and your costs on them, such as payroll taxes. Those are on your payroll report. The [labor cost guide](https://restaurantdoctorai.com/guides/restaurant-labor-cost) shows how to add them up. ### Prime cost and rent share Prime cost is food and drink cost plus labor, as a share of sales. It needs both numbers above, so a POS's sales and labor reports alone give you only part of it. Rent, charges on top of rent and utilities aren't in POS sales or labor reports; they're in your lease and your bills. The [prime cost guide](https://restaurantdoctorai.com/guides/restaurant-prime-cost) and the [rent guide](https://restaurantdoctorai.com/guides/restaurant-rent-percentage) explain both. ### Whether a number is good for your kind of restaurant A report tells you what happened. Toast's [Benchmarking](https://support.toasttab.com/en/article/Toast-Benchmarking-Overview) compares net sales, sales per labor hour, orders and average check with similar restaurants near you, but none of the pages we read sets a food, labor or prime cost target for your type of restaurant. That matters, because the same number can be fine in one place and a problem in another. RestaurantDoctorAI's prime cost targets are ceilings by type: | Type of restaurant | Prime cost target | | --- | --- | | Quick service | 60% or less | | Fast casual | 60% or less | | Full service / casual dining | 65% or less | | Fine dining | 70% or less | | Bar, taproom or lounge | 57% or less | *RestaurantDoctorAI's prime cost targets, as a share of sales: at or under is healthy. Where they come from is on the [methodology page](https://restaurantdoctorai.com/methodology).* A prime cost of 63% is under the full-service target and over the quick-service one. ## What our free check and report add RestaurantDoctorAI starts where the POS stops: costs, targets and what to do first. The free check is 12 multiple-choice questions and takes about 45 seconds, with no account, name or email. The first question asks your type of restaurant, which picks the targets. The others cover your monthly sales range, your prime cost (“I don't track it regularly” is an answer), how often you update recipe costs, how you track waste and drink pours, slow-shift staffing, rent share, delivery apps, how full the dining room is, repeat guests and your rating. Your result shows a grade for the restaurant and for each cost area, an estimate of the money lost to avoidable costs each year, your costliest issue or opportunity, and the first fix. The free grade is worked out in your browser by fixed formulas, with no AI. The paid report, a one-time payment with no subscription (prices on the [pricing page](https://restaurantdoctorai.com/pricing)), puts a dollar estimate on every problem it finds and adds a 72-hour plan, a 30-day roadmap, team cards, calculators and printable worksheets. A second report lets you type in last month's totals: net sales and guests from your POS sales summary, food and drink purchases from invoices, total payroll, rent and utilities, and delivery app statements. Then it shows your real costs against the targets and what any gap costs, and compares then and now after your plan's first full month. [Free calculatorWork out yours: prime cost calculatorType net sales from your POS, purchases from your invoices and payroll from your payroll report, and see your prime cost against the target for your type.](https://restaurantdoctorai.com/calculators/prime-cost) ## What our tools don't do - No POS connection. We never ask for POS, bank or accounting logins, and nothing syncs. The check works from your answers, and the report and calculators from numbers you type. - Estimates, not records. From 12 answers, the dollar figures are estimated ranges, not a review of your books. - Not daily. Your POS updates with every sale; the check is a snapshot of the day you take it. - No item detail. We never see which dish sells, which server comps or which hour is slow. Your POS does. For what you can check with no login at all, see [checking restaurant numbers without POS or bank access](https://restaurantdoctorai.com/guides/check-restaurant-numbers-without-pos-or-bank). Keep your POS. Use both. For what sold, when and by whom, your POS is the better tool, every day. A profit check doesn't replace it. Use the check to see which cost to look at first and where it should sit for your type, then go back to your POS reports, invoices and payroll to measure it. ## This week 1. Pull last week's sales summary and write down net sales. 2. Pull the product mix. Take your 10 best sellers and check each has a current cost in your POS, if it shows cost columns. 3. Pull the labor summary or labor vs sales, then add salaried pay and payroll taxes from your payroll report. 4. Add up last week's food and drink invoices and put all three numbers in the [prime cost calculator](https://restaurantdoctorai.com/calculators/prime-cost). 5. Take the free check to see your biggest estimated loss and the targets for your type of restaurant. 45-second profit check Your POS shows sales. See where the profit goes. Answer 12 quick questions about prices, costs, staffing and guests. In about 45 seconds you get your grade and your biggest estimated loss, free, with no POS login, so you know which report to open first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [How do I access sales reports?](https://support.toasttab.com/en/article/How-do-I-access-sales-reports), Toast, Last updated May 1, 2026; read October 10, 2026 - [Product Mix (PMIX) Report Overview](https://support.toasttab.com/en/article/Product-Mix-PMIX-Report-Overview), Toast, Last updated May 28, 2026; read October 10, 2026 - [Labor Reports Overview](https://support.toasttab.com/en/article/Labor-Reports-Overview), Toast, Last updated August 13, 2026; read October 10, 2026 - [View sales summary, sales trends and payment methods reports](https://squareup.com/help/us/en/article/5381-in-app-summaries-and-reports), Square, No date shown; read October 10, 2026 - [View item, category and modifiers sales reports](https://squareup.com/help/us/en/article/8363-view-item-category-and-modifiers-sales-reports), Square, No date shown; read October 10, 2026 - [View team member timecard and scheduling reports](https://squareup.com/help/us/en/article/6140-employee-timecard-reporting), Square, No date shown; read October 10, 2026 - [Clover reports overview](https://www.clover.com/en-US/help/clover-reports-overview), Clover, No date shown; read October 10, 2026 - [Understand Item Sales report information](https://www.clover.com/en-US/help/understand-item-sales-report-information), Clover, No date shown; read October 10, 2026 - [Sales Reports (K-Series)](https://k-series-support.lightspeedhq.com/hc/en-us/articles/18234292249883-Sales-Reports), Lightspeed, No date shown; read October 10, 2026 - [Products report (K-Series)](https://k-series-support.lightspeedhq.com/hc/en-us/articles/4403156003227-Products-report), Lightspeed, No date shown; read October 10, 2026 - [Labor Report (K-Series)](https://k-series-support.lightspeedhq.com/hc/en-us/articles/4509139748251-Labor-Report), Lightspeed, No date shown; read October 10, 2026 - [Reporting Best Practices: Weekly, Monthly, Quarterly & Yearly](https://support.toasttab.com/en/article/Reporting-Best-Practices-Weekly-Monthly-Quarterly-Reports), Toast, Last updated July 7, 2026; read October 10, 2026 - [Importing and exporting items in bulk (K-Series)](https://k-series-support.lightspeedhq.com/hc/en-us/articles/1260804656109-Importing-and-exporting-items-in-bulk), Lightspeed, No date shown; read October 10, 2026 - [xtraCHEF: Variance Analysis and Product Mix Reports](https://support.toasttab.com/en/article/xtraCHEF-Recipe-Reporting), Toast, Last updated July 9, 2026; read October 10, 2026 - [Track cost of goods sold](https://squareup.com/help/us/en/article/8522-track-cost-of-goods-sold), Square, No date shown; read October 10, 2026 - [Time & Attendance](https://www.clover.com/pos-systems/employees/m/time-and-attendance), Clover, No date shown; read October 10, 2026 - [Toast Benchmarking Overview](https://support.toasttab.com/en/article/Toast-Benchmarking-Overview), Toast, Last updated March 25, 2026; read October 10, 2026 --- title: "Restaurant software cost: what subscriptions add up to" description: "Starting prices for 7shifts, Homebase, MarginEdge, Restaurant365, Toast and Square, what they add up to in a year, and a checklist to audit yours." canonical: "https://restaurantdoctorai.com/blog/restaurant-software-cost-per-year" published: "2026-10-10" updated: "2026-10-10" --- # Restaurant software cost: what subscriptions add up to ## Key takeaways - As listed on October 10, 2026, one location's starting paid plans ran from $30 a month (Homebase Essentials) to $350 (MarginEdge); Restaurant365 lists no price. - An example stack of Homebase Essentials, MarginEdge and Square Plus is $429 a month, or $5,148 a year, before hardware and card fees. - Subscriptions do daily work a one-time tool can't: taking payments, building schedules, processing invoices and connecting your systems. - Audit yours once a year: list every charge, multiply by 12, cut overlap and unused tools, and check free plans and cancellation terms. Restaurant software cost is easy to miss because it arrives in small monthly charges: one for scheduling, one for invoices, one for the point of sale. Below are the starting prices that six common tools list on their own pricing pages today, per location, and what they add up to over a year. Then a fair look at what those subscriptions do, and a checklist to audit your own. ## What each tool lists today These are the lowest paid plan prices each company shows for one location, as listed on October 10, 2026. All are software only. Hardware, card processing, payroll add-ons and sales tax are extra, and every company can change its prices at any time, so check the page before you decide. | Tool (what it does) | Starting paid plan, billed monthly | A year at that price | If billed yearly | | --- | --- | --- | --- | | 7shifts (scheduling) | Essentials, $49.99 a month per location (1 to 30 employees) | $599.88 | $44.99 a month: $539.88 | | Homebase (scheduling) | Essentials, $30 a month per location | $360 | $24 a month: $288 | | MarginEdge (invoices, inventory) | $350 a month per location | $4,200 | 10% off: $3,780 | | Restaurant365 (accounting, inventory) | Not listed (custom quote) | Not listed | Not listed | | Toast (POS software) | Point of Sale, from $69 a month | $828 | Not listed | | Square for Restaurants (POS software) | Square Plus, $49 a month per location | $588 | Not listed | *Starting software prices for one location, as listed on each company's pricing page on October 10, 2026. Yearly total = monthly price × 12.* A few details from each page that change the real number: - [7shifts](https://www.7shifts.com/pricing) shows different plans by team size. The $49.99 is with the team size set to 1 to 30 employees and billing set to monthly. It also has a free Comp plan for single locations with up to 15 employees. - [Homebase](https://www.joinhomebase.com/pricing) has a free Basic plan for 1 location with up to 10 employees. Its payroll add-on is listed at $49 a month plus $6 per employee paid, after a free period. - [MarginEdge](https://www.marginedge.com/pricing) lists one price with no tiers and no traditional setup fees. Its yearly billing saves 10%, but it commits you to 12 months: leave mid-year and the rest of the term is still due. - [Restaurant365](https://www.restaurant365.com/pricing/) shows no prices. Its pricing page asks you to request a custom quote. - [Toast](https://pos.toasttab.com/pricing) calls Point of Sale a monthly software subscription where you choose your own hardware. Its footnotes say the price applies to new customers and single locations only, and includes the first terminal; more devices cost more. It also lists a Starter Kit from $0 a month with a hardware kit and no upfront cost. - [Square](https://squareup.com/us/en/point-of-sale/restaurants/pricing) lists a free plan at $0 a month per location. The plan you pick also sets your card rate: in person it lists 2.6% + 15¢ on Free and 2.5% + 15¢ on Plus. Some apps cost extra per device, such as the kitchen display app at $30 a month per device on Plus. ## What a year adds up to Most owners don't buy one tool. They buy one of each kind. Here are two example stacks for a single location, built only from the prices above, billed monthly: | Example stack | A month | A year (× 12) | | --- | --- | --- | | Homebase Essentials + MarginEdge + Square Plus: $30 + $350 + $49 | $429 | $5,148 | | 7shifts Essentials + MarginEdge + Toast Point of Sale: $49.99 + $350 + $69 | $468.99 | $5,627.88 | | The same, plus the $50 a month Toast suite MarginEdge says Toast users need for the connection | $518.99 | $6,227.88 | *Example stacks for one location, software only, from the prices listed on October 10, 2026.* The first stack is about $14 a day ($5,148 ÷ 365). The third line comes from MarginEdge's own FAQ: Toast customers pay Toast $50 a month per location for its Restaurant Management Suite to use the connection, plus a pass-through fee that the page doesn't put a number on. So the real total for that stack is a bit higher than shown. At the other end, a small single location can start at $0 a month for software: Homebase Basic or 7shifts Comp for scheduling, and Square Free for the point of sale, within each free plan's limits. Card processing fees still apply. Software is a bill that stays the same Subscriptions don't rise and fall with sales, so they count with rent and insurance when you work out break-even. A $429 a month stack adds $429 of fixed bills, and it takes more than $429 of sales to cover it. The [break-even post](https://restaurantdoctorai.com/blog/restaurant-break-even-point) shows how to work out how much more. ## What a subscription does that a one-time tool can't To be fair to these tools: a monthly fee pays for work that happens every day, and no one-time report can do that. From their own pages: - They run the shift itself. Toast and Square take orders and payments. 7shifts and Homebase build and publish schedules and track time clocks. - They watch every day. MarginEdge processes invoices, tracks product prices with alerts, and lists a daily controllable P&L. 7shifts lists overtime alerts and labor targets set on projected sales. - They connect to each other. MarginEdge and 7shifts list POS integrations, and MarginEdge lists accounting integrations, so sales and invoices flow in without typing. - They keep records. Time punches, invoices and sales history sit in one place you can search later. If you need any of that, a subscription is the right kind of tool. A one-time check can't take a payment, schedule a shift or read Tuesday's invoice. ## When a subscription is worth it A rule of thumb, not a rule: a tool earns its fee when you can name what it does for you each month. It tends to fit when: - You or a manager open it every week and act on what it shows. - It replaces hours of typing invoices, building schedules or adding up tips that someone was paid to do. - It catches something you would miss, like a supplier price rise or a shift heading into overtime, early enough to act on. - It connects to tools you already pay for, so the numbers aren't typed twice. It tends not to fit when nobody has logged in for a month, when two tools do the same job, or when you pay for a higher plan for one feature you rarely use. Labor tools make the most sense once you know your labor numbers; the [labor cost guide](https://restaurantdoctorai.com/guides/restaurant-labor-cost) covers how to work them out. ## Audit your own subscriptions: a checklist One hour, once a year, or before any plan renews. 1. Pull three months of card and bank statements and list every software charge, small ones included. 2. For each one, write the plan name, the price, what it is charged on (per location, per device, per employee), whether it bills monthly or yearly, and the renewal date. 3. Add the extras: per-device apps, per-employee payroll fees, a second terminal, connection fees one tool charges to talk to another. 4. Multiply each monthly charge by 12 and add them up. That is your real yearly software cost. 5. Look for overlap. Two tools that both schedule, both track inventory or both send marketing emails is one too many. 6. Check who used each tool in the last 30 days. If nobody did, ask why before you pay again. 7. Check the free plans. 7shifts, Homebase and Square each list one; 7shifts and Homebase limit theirs by locations and employees, and Square Free charges only card fees. If you fit, try it. 8. Decide monthly or yearly billing. Yearly is cheaper on several of these pages, but on MarginEdge's it commits you to 12 months. Pay yearly only for tools you're sure you'll keep. 9. Read how each one cancels. 7shifts says it bills the current month and not again if you aren't under a contract; Homebase and Square say you can cancel anytime. Put each renewal date in your calendar. ## Where RestaurantDoctorAI fits RestaurantDoctorAI is not a subscription, and it doesn't replace any tool above. It doesn't schedule shifts, process invoices or take payments, and it doesn't watch your numbers every day. What it does: - A free profit check: 12 quick questions, about 45 seconds, no sign-up. It shows your restaurant's grade and your biggest estimated loss, so you know where to look first. - Free [calculators](https://restaurantdoctorai.com/calculators) for prime cost, labor, break-even, food cost per plate and delivery app prices, so you can check a number before you act on it. No account. - A free [deadline calendar](https://restaurantdoctorai.com/deadline-calendar) for tax, payroll, license and permit dates in your state. - A full report if you want one, in two versions, paid once. No subscription, no account, nothing renews. Prices are on the [pricing page](https://restaurantdoctorai.com/pricing). One honest way to use it next to the tools above: run the free check first, see which cost is losing you the most, and then decide whether a subscription aimed at that cost is worth its fee. If food cost is the problem, an invoice tool may earn its keep. If it isn't, maybe not yet. The guide to [why a busy restaurant can still lose money](https://restaurantdoctorai.com/guides/restaurant-busy-but-not-making-money) lists the usual causes. [Free calculatorAdd your software to your fixed bills: break-even calculatorPut your yearly software total in as one twelfth a month and see how much more you need to sell.](https://restaurantdoctorai.com/calculators/break-even) ## This week 1. List every software charge from your last three statements, with its plan, price and renewal date. 2. Multiply by 12 and write the yearly total next to your rent. 3. Mark any tool nobody opened last month, and any two that do the same job. 4. Before the next renewal, check that tool's pricing page for a free plan, a cheaper tier or yearly billing you'd actually use. 45-second profit check Know which cost to pay a tool for The free check asks 12 quick questions about prices, costs, staffing and guests, and shows your biggest estimated loss in about 45 seconds, so you can decide which problem is worth a monthly fee. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Plans & Pricing](https://www.7shifts.com/pricing), 7shifts, No date shown; read October 10, 2026 - [Official Homebase Pricing](https://www.joinhomebase.com/pricing), Homebase, No date shown; read October 10, 2026 - [Restaurant Invoicing Software Pricing](https://www.marginedge.com/pricing), MarginEdge, No date shown; read October 10, 2026 - [Restaurant Software Pricing](https://www.restaurant365.com/pricing/), Restaurant365, No date shown; read October 10, 2026 - [Official Toast Restaurant POS Pricing & Plans (2026)](https://pos.toasttab.com/pricing), Toast, Last updated August 20, 2026; read October 10, 2026 - [Restaurant POS Pricing & Plans](https://squareup.com/us/en/point-of-sale/restaurants/pricing), Square, No date shown; read October 10, 2026 --- title: "ChatGPT for restaurants: how to check its numbers" description: "Can ChatGPT work out your restaurant's food cost or pay rules? Where AI chatbots help, where they go wrong, and six steps to check an answer." canonical: "https://restaurantdoctorai.com/blog/chatgpt-restaurant-numbers" published: "2026-10-10" updated: "2026-10-10" --- # ChatGPT for restaurants: how to check its numbers ## Key takeaways - Chatbots help with drafts, ideas and explaining formulas, but their makers say they can be wrong while sounding sure, so check any figure or rule. - New York City's business chatbot told owners they could take a cut of workers' tips; federal law says employers can't keep any part of them. - Check an answer: ask for the source and open it, check the formula, run your numbers in a calculator that shows its math, and read the rule on the government page. - Pay rules differ by state: where state and federal minimum wages differ, the worker gets the higher one. Yes, ChatGPT and other AI chatbots can help you run a restaurant: they draft, brainstorm and explain a formula in plain words. Where they slip is the exact figure or rule. Their makers say so on their own help pages, and one big city's business chatbot told owners they could do things the law doesn't allow. So use the answer, then check it. This post shows where a chatbot helps, where it goes wrong, and six steps to check an answer about your numbers. ## Where a chatbot helps A chatbot is fast at the jobs where a first try is worth having and you can judge the result yourself. Good uses in a restaurant: - Drafting: a job post, menu descriptions, a note to staff about a new closing checklist, a polite email asking a supplier about a price change. - Brainstorming: specials that use up trim, ways to fill a slow Tuesday, questions to bring to your accountant. - Explaining: what prime cost means, why food cost is worked out from what you used and not what you bought, what a tip credit is. - Sorting your own numbers: paste in a week of sales by day and ask it to lay them out, then check the sums. In each of these, you are the check. You know if a menu description fits your food or if a schedule idea would work with your team. The trouble starts when the answer is a number or a rule you can't judge by reading it. ## Where it goes wrong, in its makers' own words OpenAI's help page, [Does ChatGPT tell the truth?](https://help.openai.com/en/articles/8313428-does-chatgpt-tell-the-truth), says ChatGPT can give incorrect or misleading answers and may sound confident even when it's wrong. Among the mistakes it lists are wrong facts and made-up citations or sources. It also says the models learn from data up to a cutoff date, so they don't know what happened after it unless a tool like search is used. Its advice is to treat ChatGPT as a first draft, not a final source, and to check important information. Google says much the same about its chatbot. Its [Gemini help page](https://support.google.com/gemini/answer/16279220?hl=en) says Gemini can hallucinate and present inaccurate information as fact, and that its answers shouldn't be relied on as legal, financial or other professional advice. That's not a reason to stop using them. It's the makers telling you how to use them: as a helper whose work you check. ### The New York City chatbot The clearest warning for restaurant owners came from a city government. New York City announced its MyCity business chatbot in October 2023, built on Microsoft's Azure AI services, to answer owners' questions about city rules. In March 2024, [The Markup reported](https://themarkup.org/artificial-intelligence/2024/03/29/nycs-ai-chatbot-tells-businesses-to-break-the-law) that it told owners they could take a cut of their workers' tips, that a restaurant could go cash-free, and that there were no rules on telling staff about schedule changes. All three were wrong, the article reports: a 2020 city law, for one, requires businesses to accept cash. The tips answer is wrong under federal law too. The US Department of Labor's [fact sheet on tipped employees](https://www.dol.gov/agencies/whd/fact-sheets/15-tipped-employees-flsa) says an employer can't keep any part of employees' tips, whether or not it takes a tip credit, and that managers and supervisors can't take employees' tips either. After the first story, the city added a note calling the bot a beta product and telling users not to treat its answers as legal or professional advice, but [the bot stayed online](https://themarkup.org/news/2024/04/02/malfunctioning-nyc-ai-chatbot-still-active-despite-widespread-evidence-its-encouraging-illegal-behavior). In January 2026, Mayor Zohran Mamdani named it as a place to save money and his office said the city planned to take it down, and [an update on February 4, 2026](https://themarkup.org/artificial-intelligence/2026/01/30/mamdani-to-kill-the-nyc-ai-chatbot-we-caught-telling-businesses-to-break-the-law) reports that the city has taken it down. MyCity isn't ChatGPT, and the chatbot you use may well get the tip question right. The lesson is the pattern: an answer from an official-looking tool, said with confidence, that would have put an owner on the wrong side of the law. A general chatbot can make the same kind of mistake, and it won't tell you when it has. ## Four places an answer about your numbers can slip These are the things to look for in any answer about food cost, labor, prices or pay rules, whoever gives it. - An old figure. Wage rates, tax rules and fees change. A model trained before the change can give last year's rate as if it were today's. - The wrong formula. A sensible-sounding method can still measure the wrong thing, as in the example below. - The wrong place. Federal law sets a floor, and many states and cities set their own rules on top of it. - A source that doesn't say it, or doesn't exist. OpenAI lists made-up citations among the mistakes ChatGPT can make. ### Example: the formula behind a food cost answer Say you tell a chatbot what you spent on food last month and what you sold, and ask for your food cost. An answer that divides purchases by sales looks right, but it measures what you bought, not what you used. Food cost for a period counts the change in what's on your shelves: *Food cost % for a week or month* Food cost % = (beginning inventory + purchases − ending inventory) ÷ food sales × 100 Inventory comes from a count of your food stock at the start and end of the period. Leave drinks out of all three. *Example: one month, two answers* - Food sales: $32,000 - Food purchases: $10,000 - Purchases ÷ sales: 31.3% - Beginning inventory: $8,000 - Ending inventory: $9,000 - Food used: $8,000 + $10,000 − $9,000: $9,000 **Food cost: $9,000 ÷ $32,000:** 28.1% Example figures, invented for this post. The month you stock up looks worse than it is; the month you run the shelves down looks better. Same month, about 3 points apart, and nothing in the first answer looks wrong. That's why the formula is the thing to check. The [food cost percentage guide](https://restaurantdoctorai.com/guides/food-cost-percentage) goes through it step by step. ## How to check an AI's answer about your numbers Six steps. The first three take a minute; the last three matter most when a rule or real money is involved. 1. **Ask for the source, and open it.** Ask the chatbot where the figure or rule comes from, then click the link yourself, as OpenAI's own help page suggests. If the page doesn't exist, or doesn't say what the answer says, drop the answer. 2. **Check the formula.** Ask it to write out the formula it used, step by step. Compare it with a formula you can trust: the [restaurant formulas guide](https://restaurantdoctorai.com/guides/restaurant-formulas) has food cost, labor, prime cost, break-even and more on one page. 3. **Run your own numbers in a calculator that shows its math.** RestaurantDoctorAI's free calculators for [food cost](https://restaurantdoctorai.com/calculators/recipe-cost), [labor cost](https://restaurantdoctorai.com/calculators/labor-cost), [prime cost](https://restaurantdoctorai.com/calculators/prime-cost) and [break-even](https://restaurantdoctorai.com/calculators/break-even) each have a “How we worked it out” line that puts your own numbers into the formula, so you can check every step by hand. If the chatbot's answer and the calculator's don't match, find the step where they part. 4. **Read the rule on the government page.** For federal pay and tip rules that's the Department of Labor: its [minimum wage page](https://www.dol.gov/general/topic/wages/minimumwage) and its fact sheet on tipped employees. Look at the date on the page, too. 5. **Check your state, and your city.** The Labor Department says that when state and federal minimum wages differ, the worker gets the higher one, and its [state minimum wage table](https://www.dol.gov/agencies/whd/minimum-wage/state) lists each state. Federal law allows a tipped cash wage of $2.13 an hour, but some states require more and some allow no tip credit at all. The [tipped minimum wage guide](https://restaurantdoctorai.com/guides/tipped-minimum-wage) has the rate by state, and the [deadline calendar](https://restaurantdoctorai.com/deadline-calendar) lists tax, payroll and license dates for your state. 6. Take legal and tax questions to a person. When the answer decides what you pay staff, what you owe in tax or whether you need a permit, use the chatbot's answer as your list of questions for your accountant, lawyer or the agency itself. A habit that makes checking easier Give the chatbot your own numbers and ask it to show each step and name the formula it used. A shown step is easy to check. A single figure on its own isn't. ## Rules of thumb are not facts Ask a chatbot what your food cost or labor cost should be and you'll get a range. It may be a fair rule of thumb, but it isn't a law, and it depends on what you sell. Ask where the range comes from and whether it fits your type of restaurant. The guides on this site name the source for each figure, and the targets the free check uses are listed as RestaurantDoctorAI's targets on the [methodology page](https://restaurantdoctorai.com/methodology), with where each one comes from. [Free calculatorCheck a plate's food cost: food cost calculatorAdd a dish's ingredients to get its plate cost, its food cost % and the price that hits your target, with the math shown.](https://restaurantdoctorai.com/calculators/recipe-cost) ## This week 1. Pick one number you've asked a chatbot about, or would like to: food cost, labor cost or a pay rate. 2. Ask it again, and this time ask for the formula and the source. Open the source. 3. Put the same numbers into the matching [free calculator](https://restaurantdoctorai.com/calculators) and open “How we worked it out”. Note any step that differs. 4. For any pay or tip rule, read your state's page and the Labor Department's page before you change a paycheck. 45-second profit check Want to know which number to check first? The free check asks 12 quick questions about prices, costs, staffing and guests, and shows your biggest estimated loss in about 45 seconds, so you know which number to look at first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Does ChatGPT tell the truth?](https://help.openai.com/en/articles/8313428-does-chatgpt-tell-the-truth), OpenAI Help Center - [Learn about responses from Gemini Apps](https://support.google.com/gemini/answer/16279220?hl=en), Google, Gemini Apps Help - [NYC’s AI Chatbot Tells Businesses to Break the Law](https://themarkup.org/artificial-intelligence/2024/03/29/nycs-ai-chatbot-tells-businesses-to-break-the-law), The Markup, March 29, 2024 - [Fact Sheet #15: Tipped Employees Under the Fair Labor Standards Act (FLSA)](https://www.dol.gov/agencies/whd/fact-sheets/15-tipped-employees-flsa), U.S. Department of Labor, Wage and Hour Division - [Malfunctioning NYC AI Chatbot Still Active Despite Widespread Evidence It’s Encouraging Illegal Behavior](https://themarkup.org/news/2024/04/02/malfunctioning-nyc-ai-chatbot-still-active-despite-widespread-evidence-its-encouraging-illegal-behavior), The Markup and THE CITY, April 2, 2024 - [Mamdani to kill the NYC AI chatbot we caught telling businesses to break the law](https://themarkup.org/artificial-intelligence/2026/01/30/mamdani-to-kill-the-nyc-ai-chatbot-we-caught-telling-businesses-to-break-the-law), The Markup and THE CITY, January 30, 2026 (updated February 4, 2026) - [Minimum Wage](https://www.dol.gov/general/topic/wages/minimumwage), U.S. Department of Labor - [State Minimum Wage Laws](https://www.dol.gov/agencies/whd/minimum-wage/state), U.S. Department of Labor, Wage and Hour Division, July 1, 2026 --- title: "Do 9 in 10 restaurants fail? What the data says" description: "Federal data show far fewer new food businesses close in year one than the 9-in-10 claim. See the 1, 5 and 10-year numbers and what to check this week." canonical: "https://restaurantdoctorai.com/blog/restaurant-failure-rate" published: "2026-10-08" updated: "2026-11-01" --- # Do 9 in 10 restaurants fail? What the data says ## Key takeaways - In federal data, 79 to 90 percent of new food and lodging establishments were still open after one year, and 49 to 60 percent after five. - That matches or beats all private industries in nearly every opening year, but a closing is not always a failure: owners sell, retire or leave. - Studies tie closings to the owner's life, a small start, and low earnings, debt and cash; the trade group put a typical pre-pandemic margin near 5 percent. - Check break-even and prime cost this week, because a thin margin leaves little room for a slow month. No. In the federal count of new food and lodging businesses, between 79 and 90 in every 100 were still open a year later, depending on the year they opened, and roughly half to 6 in 10 were open after five. That is as good as or better than all private-sector businesses, so the claim that 9 in 10 restaurants fail does not match the data. Restaurants do run on thin margins, though, so this week the useful checks are break-even and prime cost. ## Where the 90 percent claim came from Both a 2003 university release and a 2014 research draft tie it to television in 2003. Ohio State University's [release on a Columbus restaurant study](https://news.osu.edu/restaurant-failure-rate-much-lower-than-commonly-assumed-study-finds/), dated September 7, 2003, says the figure was repeated on every episode of an NBC reality show, and the study's author, H. G. Parsa, reports finding no evidence for it in the research. Tian Luo, then at BLS, and the Berkeley statistician Philip Stark [describe it as a television advertisement](https://arxiv.org/abs/1410.8603) from that year, and the two accounts differ in detail. Either way, neither source points to a count of anything behind it. Since then, owners have heard it from lenders, relatives and strangers. What follows is what government counts and published studies say instead, with who measured what, because the numbers differ by how they define “fail”. ## What the federal numbers say about years 1, 5 and 10 The Bureau of Labor Statistics follows every new private-sector establishment, meaning a single location with employees, and counts how many are still open each March after it opened. Its [survival table for accommodation and food services](https://www.bls.gov/bdm/us_age_naics_72_table7.txt) covers establishments that opened in each year ending in March from 1994 to 2025, with data through March 2025. Restaurants, bars and caterers sit in that group with hotels and other lodging, and BLS does not split restaurants out in these tables. Here is a selection of opening years, next to the [same count for every private-sector industry](https://www.bls.gov/bdm/us_age_naics_00_table7.txt). | Opened | After 1 year | After 5 years | After 10 years | | --- | --- | --- | --- | | 1994 | 79.9 / 79.6 | 49.4 / 49.6 | 34.6 / 33.6 | | 2000 | 82.5 / 78.4 | 53.6 / 48.2 | 39.3 / 32.9 | | 2005 | 81.5 / 80.1 | 50.7 / 46.8 | 36.7 / 33.8 | | 2010 | 81.5 / 78.6 | 52.8 / 51.1 | 37.1 / 35.3 | | 2015 | 88.2 / 79.6 | 58.5 / 50.2 | 41.6 / 34.7 | | 2020 | 85.5 / 80.9 | 59.3 / 51.4 | not yet | | 2024 | 85.3 / 77.9 | not yet | not yet | *Percent of new establishments still open, by the year they opened (year ended March). In each cell, the first figure is accommodation and food services and the second is all private industries. Source: BLS Business Employment Dynamics, Table 7.* Reading the whole table, not just these rows: - After one year, between 79.0 and 90.3 percent of food and lodging establishments were still open, depending on the opening year (31 opening years, 1994 to 2024). No opening year had more than 21 in 100 close in year one. - After five years, between 49.0 and 60.4 percent were open (27 opening years, 1994 to 2020), against 45.4 to 51.9 percent for all private industries. - After ten years, between 34.6 and 41.6 percent were open (22 opening years, 1994 to 2015), against 32.4 to 35.3 percent for all industries. - In 31 of 31 opening years the one-year figure beat the all-industry one, as did the five-year figure in 26 of 27 and the ten-year figure in 22 of 22. That count is mine, from reading the two tables side by side. Two cautions. First, by year ten, only about 4 in 10 are still open, and for the all-industry count it is about 3 in 10. Starting a business is risky in every industry, not only this one. Second, a closing is not always a failure, which is the next section. Opening years from 2016 to 2020 had their first five years run through 2020 and 2021, and their five-year figures, 57.1 to 60.4 percent, sit at the top of the range. This table cannot tell you why. ## A closing is not always a failure BLS [defines a closing](https://www.bls.gov/opub/hom/bdm/concepts.htm) by employment: a location that had paid employees and then has none, or goes inactive. It waits three quarters before treating a closing as permanent, so a seasonal restaurant that closes for winter is not counted as dead, and it tracks a business through a change of name or ownership. The data show that a place stopped employing anyone. They do not show whether the owner lost money, retired or sold up. Other studies count differently, which is why their numbers differ: | Study | What was counted | What it found | | --- | --- | --- | | BLS Business Employment Dynamics | New private establishments in accommodation and food services, nationwide, opened 1994 to 2024 | Between 79 and 90 percent open after one year (table above) | | [Luo and Stark](https://arxiv.org/abs/1410.8603) (draft, 2014) | Independent, single-location full-service restaurants in eight western states, born 1992 to 2011; a merger or spinoff ended the observation and was not counted as a closing | About 17 percent closed in year one, against 19 percent for other new service businesses; median life about 4.5 years | | [Parsa and co-authors](https://doi.org/10.1177/0010880405275598) (journal, 2005) | Restaurants in Columbus, Ohio, 1996 to 1999, using health department records; measured ownership turnover, not bankruptcy | 26.16 percent of all restaurants failed in year one (27.51 percent of independents, 24.06 percent of chains); across three years, 57.2 percent of chains and 61.4 percent of independents | | Dun & Bradstreet, as the Ohio State release reports it | Eating and drinking places that went bankrupt or closed owing money, in records it stopped publishing after 1998 | About 1.06 percent | *Four ways of counting, and what each found. These are different places, years and definitions, so they are not a single series.* Read down the table and the lesson is the definition. A count of bankruptcies gives a tiny number. A count of ownership changes in one Ohio city over three years gives a large one. Luo and Stark, who used non-public BLS records for the West, treat a merger or spinoff as the end of what they can observe, not a death. Their paper is a draft on arXiv and does not show a journal publication, and the authors say their opinions are not the Bureau's. The studies also disagree. Parsa's three-year figure of 59 percent for Columbus is well above what BLS shows for the nation: reading the same table, between 27 and 40 in every 100 new food and lodging establishments had closed by their third year. Different city, different decade and a different question could explain the gap, and no source here settles which one does. If you hear the claim from a lender The research above is about groups of restaurants, not yours. It is a fair reason to ask what number the lender is using and who counted it. It does not show that any one restaurant will last. ## How restaurants compare with other new businesses Luo and Stark ranked more than 500 kinds of single-location service startups in the West. Full-service restaurants were the most common, with about 81,500 in the data, and they landed in the middle. Their selected results are below, with the share still open after one year and the median life, the age by which half had closed. | Kind of business | Open after 1 year | Median life | | --- | --- | --- | | Offices of dentists | 93 percent | more than 19.5 years | | Offices of physicians | 90 percent | 10.75 years | | Drinking places (bars) | 83 percent | 4.5 years | | Full-service restaurants | 83 percent | 4.5 years | | Limited-service restaurants | 81 percent | 3.75 years | | Janitorial services | 76 percent | 3 years | *Western US, single-location startups born 1992 to 2011. Source: Luo and Stark, Tables 2 and 3.* Limited-service restaurants did somewhat worse than full-service ones on both measures. Size at the start matters too. The paper finds survival rises with the number of employees at opening, and restaurants that began with five or fewer employees had a median life of 3.75 years, a little shorter than the 4.0 years for other service businesses that began that small. ## Why restaurants close: what the research says No single cause shows up. Four things do: - **The owner's life.** Parsa interviewed Columbus owners. Many who closed named family issues such as divorce, poor health or wanting to retire, and all described how much time a restaurant takes. The Ohio State [release](https://news.osu.edu/restaurant-failure-rate-much-lower-than-commonly-assumed-study-finds/) says many restaurants close for personal reasons, not because they could not succeed financially. - **Starting small.** In Luo and Stark's data, survival rises with startup size. Larger startups may need more capital, they write, and still tended to last longer. - **Low earnings and heavy debt.** Luo and Stark's review of earlier research notes that restaurants with low earnings and high liabilities are more likely to go out of business. A 2024 study of publicly traded restaurant companies, from 2000 to 2019, [found](https://doi.org/10.3390/jrfm17090399) that the quick ratio (cash and near-cash against short-term bills), operating cash flow and working capital were the most crucial indicators of a bankruptcy filing. Those are large listed companies filing for bankruptcy, so treat it as a hint for independents, not a measure of them. - **Thin margins that costs can erase.** The [National Restaurant Association's July 2026 analysis](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-costs-continue-to-pressure-restaurant-profitability/) describes a typical independent restaurant before the pandemic as spending about 33 cents of each sales dollar on food, 33 on labor and 29 on everything else, leaving roughly 5 cents before tax. It estimates total expenses for an average restaurant rose 36 percent from 2019 to 2026, and reports that 42 percent of operators said their restaurant was not profitable in 2025. It is a trade group, and these are its estimates and its operator survey. Put that last bullet and the cash-flow finding together and you get the same message from two directions: when only a few cents of each dollar are left, a slow month or a jump in costs can use up the cushion before the bank balance does. ## What a thin margin looks like: an example This is an example, not a real restaurant. Rosa's Taqueria is a counter-service place with $80,000 in sales in an average month. Break-even is the sales a month needs to pay every bill; the [busy-but-not-making-money guide](https://restaurantdoctorai.com/guides/restaurant-busy-but-not-making-money#busy-vs-profitable) walks through why it is the first number to know. *Example: Rosa's Taqueria, an average month* - Sales: $80,000 - Costs that rise with each sale: 36% - Kept from each sales dollar: 64 cents - Bills that stay the same: $48,000 - Break-even: $48,000 ÷ 0.64: $75,000 - Cushion above break-even: $5,000 **Left over: $80,000 × 0.64 − $48,000:** $3,200 Costs that rise with each sale are food, drink, packaging and card fees. Bills that stay the same are rent, utilities, salaries, fixed shifts, insurance and loan payments, payroll taxes included. All numbers are invented round figures. Now two ordinary shocks, one at a time: - Sales dip 8 percent, to $73,600. Left over is $73,600 × 0.64 − $48,000, which is minus $896. The month loses money. - Costs per sale rise 3 points, from 36 to 39 percent, with sales unchanged. Left over is $80,000 × 0.61 − $48,000, which is $800. Break-even climbs to about $78,700. Neither shock needs a bad owner, a bad dish or a bad block. The cushion was only about 6 percent of sales. That is the pattern the research points to, and it is something you can see on your own numbers this week, before a bank balance shows it. [Free calculatorWork out yours: break-even calculatorOne typical month of sales and costs gives your break-even and your cushion.](https://restaurantdoctorai.com/calculators/break-even) ## What it means for an owner ### The average is not your restaurant The federal counts say new food and lodging establishments last as long as new businesses in general, or a bit longer. They say nothing about whether yours is healthy. Half of a group lasting five years tells you nothing about which half you are in. ### The pattern is money, not mystery The research that looks at money points to earnings, debt and cash. Those you can measure. Prime cost is food and drink cost plus all labor cost, as a share of sales; it is the part of your costs that moves with every shift you run, and the [prime cost guide](https://restaurantdoctorai.com/guides/restaurant-prime-cost) shows how to work it out and which ceilings to compare it to. Break-even tells you how much room you have if sales dip. ## This week 1. Enter last month's sales and costs in the [break-even calculator](https://restaurantdoctorai.com/calculators/break-even). Write down your cushion (sales minus break-even), and compare it with your worst recent month-to-month dip. 2. Work out last week's prime cost: food and drink used plus all labor cost, divided by net sales. The [prime cost guide](https://restaurantdoctorai.com/guides/restaurant-prime-cost) has the formula and the ceilings by type of restaurant. 3. Add up what leaves your bank account in a normal week: payroll, supplier invoices, rent and loan payments. Divide your balance by that. The result is how many weeks you could pay your bills with no sales at all. 4. If the numbers say you are busy but not keeping money, follow the order in the [busy-but-not-making-money guide](https://restaurantdoctorai.com/guides/restaurant-busy-but-not-making-money#check-in-this-order), and fix the biggest gap first. 45-second profit check See where the money goes Answer 12 quick questions about your food cost, labor, pricing and slow shifts, and you'll get a free grade and your biggest estimated loss in about 45 seconds, so you know what to look at first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Table 7. Survival of private sector establishments by opening year: Accommodation and Food Services](https://www.bls.gov/bdm/us_age_naics_72_table7.txt), U.S. Bureau of Labor Statistics, Business Employment Dynamics, Data through March 2025 - [Table 7. Survival of private sector establishments by opening year: Total Private](https://www.bls.gov/bdm/us_age_naics_00_table7.txt), U.S. Bureau of Labor Statistics, Business Employment Dynamics, Data through March 2025 - [Business Employment Dynamics: Concepts](https://www.bls.gov/opub/hom/bdm/concepts.htm), U.S. Bureau of Labor Statistics, October 4, 2018 - [Restaurant Failure Rate Much Lower Than Commonly Assumed, Study Finds](https://news.osu.edu/restaurant-failure-rate-much-lower-than-commonly-assumed-study-finds/), The Ohio State University, September 7, 2003 - [Only the Bad Die Young: Restaurant Mortality in the Western US](https://arxiv.org/abs/1410.8603), T. Luo (U.S. Bureau of Labor Statistics) and P. B. Stark (UC Berkeley), draft on arXiv, October 31, 2014 - [Why Restaurants Fail](https://doi.org/10.1177/0010880405275598), Cornell Hotel and Restaurant Administration Quarterly (H. G. Parsa and co-authors), August 2005 - [Bankruptcy Prediction for Restaurant Firms: A Comparative Analysis of Multiple Discriminant Analysis and Logistic Regression](https://doi.org/10.3390/jrfm17090399), Journal of Risk and Financial Management (Huo, Chan and Miller), September 6, 2024 - [Elevated costs continue to pressure restaurant profitability](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-costs-continue-to-pressure-restaurant-profitability/), National Restaurant Association, July 8, 2026 --- title: "What food waste costs a restaurant, and what to do" description: "Waste is food you already paid for. See what sources say restaurants waste, how to put a dollar figure on yours in a week, and what to fix first." canonical: "https://restaurantdoctorai.com/blog/restaurant-food-waste-cost" published: "2026-10-08" updated: "2026-10-08" --- # What food waste costs a restaurant, and what to do ## Key takeaways - Waste is food you already paid for, so log what you throw away for a week and cost it at what you paid for it. - ReFED counts nearly 70% of US restaurant and foodservice surplus as plate waste; your own log shows how much of yours is made-too-much, spoilage and trim. - A point of food cost is 1% of sales, so in the example restaurant it is worth $7,800 a year, and a week's waste of $180 is 1.2 points. - Work in this order: measure, cut over-prepping, then donate only food that was never served and was kept safe. Wasted food is food you already paid for, so every dollar you stop binning stays in your till. Put a number on it this week: write down what leaves your kitchen in the trash for seven days, cost it at what you paid, and see how many points of food cost it is. Then work in this order: measure, cut over-prepping, and only then donate what's still good. ## How much do restaurants waste? Nobody can give you your number, but two published sources show the shape. ReFED, a US nonprofit that researches food waste, says in its [page on restaurants and foodservice](https://refed.org/stakeholders/restaurants-and-foodservice/) that the sector generated 12.5 million tons of surplus food in 2024, and that more than 85% of it went to landfill or was incinerated. That's a weight, not a bill: the page gives no dollar figure for the sector. ReFED also splits that surplus by cause. It defines plate waste as customers not eating what they have taken or been served, and puts it at nearly 70% of the total. The [UK's WRAP, an organization that works on waste](https://www.wrap.ngo/resources/report/overview-waste-hospitality-and-food-service-sector), published a different split in a 2013 overview of UK hospitality and food service outlets (the page is now marked as archived): 920,000 tonnes of food waste a year, three quarters of it food that could have been eaten. | Cause | ReFED: US restaurants and foodservice, 2024 | WRAP: UK hospitality and food service, 2013 overview | | --- | --- | --- | | Plate waste | 69.6% of surplus | 34% (consumer plates) | | Food preparation | Not listed | 45% | | Spoilage | Not listed | 21% | | Overproduction | 11.9% of surplus | Not listed | | Catering overproduction | 9% of surplus | Not listed | *Where waste comes from, as each source counts it. Different countries, years and venues, so do not add them up or average them.* The two disagree a lot on plate waste, about 70% against a third. They count different countries, different years and different kinds of venue, so don't blend them into one number. What they tell you together is that waste comes from several places at once, and that only a count in your own kitchen shows which place is biggest in yours. ## Where it happens in your kitchen Sources group waste differently, but in a kitchen it comes down to four places. Each one needs a different fix. - **Prep trim.** Peels, ends, bones and fat. Some can't be avoided. The part you can fix is trimming more than the dish needs, or throwing out what could go in a stock, a staff meal or another dish. The EPA tells businesses to examine their production and handling practices to cut preparation waste, and the WRAP split above has preparation as the largest slice. - **Spoilage.** Food that goes off before you use it: wilted greens, fish past its day, milk, bread. It usually comes from ordering more than you sell, or from stock that sits at the back of the walk-in. - **Over-production.** Food you made and didn't sell: the soup batch, the extra tray of chicken, the rice left in the well at close. It comes from pars (the amounts you prep for the day) set for a busier day than you had. - **Plate waste.** Food a guest was served and left. It's the biggest slice in ReFED's figures, and the fixes are on the menu side: portion size, dishes guests send back, sides nobody finishes. The first three happen before a guest sees the food, and you can count them at the bin by the prep table. The EPA also advises businesses to [measure the type and amount of wasted food in a waste audit](https://www.epa.gov/sustainable-management-food/how-prevent-wasted-food-through-source-reduction) and to “Compare purchasing inventory with customer ordering.” If you buy 40 pounds of something a week and sell 30 pounds' worth, the gap is in the bin or in the walk-in. ## Put a dollar figure on your waste in a week You need a sheet, a bus tub and a scale if you have one. This follows the [EPA's waste logbook](https://archive.epa.gov/wastes/conserve/pubs/web/pdf/food-waste-log.pdf), an archived document, adapted for a small kitchen. 1. Put a sheet and a tub where food gets thrown away: the prep table, the line and the dish pit. Everyone who throws food away writes it down. 2. Write it down just before it goes in the bin, not at the end of the night. At close you won't remember. 3. Write two things for every entry: what it was and why it went. The EPA calls those the two most important details. Use a short list of reasons: trim, spoiled, made too much, dropped or burnt, past its date. 4. Weigh it. The EPA suggests putting the waste in an empty bus tub, weighing it and subtracting the tub's own weight later. With no scale, write down portions or volume, like half a pot or 2 gallons. 5. Write down food you give away on the same sheet, marked as donated, so it doesn't look like a loss and you can see how much good food is left over. 6. Read yesterday's sheet at the start of the next shift. At the end of the week, find the five items that cost the most and talk about them with the team. ### Cost it at what you paid Cost waste at what you paid for it, not at what it would have sold for. For each line, multiply the amount thrown away by the price per pound or per unit on your invoice. For trim, use the price of the whole item, because you paid for every pound of it. *One line of the log* Cost of waste = amount thrown away × price per unit Use the price per unit as you bought it, from your latest invoice. | Item | Why it went | Amount | Price | Cost | | --- | --- | --- | --- | --- | | Chicken thighs | Made too much | 20 lb | $3.00/lb | $60 | | Romaine | Spoiled | 10 heads | $2.50 each | $25 | | Tomatoes | Trim | 12 lb | $2.00/lb | $24 | | Tomato soup | Made too much | 4 qt | $2.50/qt | $10 | | Burger buns | Spoiled | 24 | $0.50 each | $12 | | Salmon | Spoiled | 3 lb | $12.00/lb | $36 | | Cooked rice | Made too much | 13 lb | $1.00/lb | $13 | | Week's total | | | | $180 | *Example only: one week of Corner Table's waste log, a full-service restaurant. The items, amounts and prices are made up for the example.* Group the cost by reason. In this example, food that was made too much cost $83, spoilage $73 and trim $24. The biggest reason points to the first fix: here, lower pars. Plate waste is a different count. The EPA's logbook says food coming back from the dining room is too varied to log item by item. It suggests tracking it about once a month, by the total weight of that trash or a count of bins or bags, on your busiest day, and repeating on the same weekday so the figures can be compared. Look at what's in the bin too. A side nobody finishes is a portion to change. What the log leaves out It counts what you threw away. It doesn't catch over-portioning, unrecorded comps or a stock count that's off. Those show up in your [food cost percentage](https://restaurantdoctorai.com/guides/food-cost-percentage#two-quiet-costs) as a gap you can't explain. ## What a point of food cost is worth A point of food cost is 1% of your sales. Your food cost percentage is the cost of the food you used divided by your sales, and it already includes everything you threw away: food that left the storeroom and earned nothing still counts as used. So waste is not an extra cost on top of food cost. It's part of it, and the log tells you how big a part. *What a point is worth* Value of one point = yearly sales × 1% Use a year of sales. Divide a week by 52 and you'll see what a week's waste is worth. *Example: Corner Table, a full-service restaurant* - Weekly sales: $15,000 - Yearly sales: × 52: $780,000 - Food cost: 30% of sales: $4,500 a week - Waste log, week one: $180 - Waste as share of sales: $180 ÷ $15,000: 1.2 points - Waste as share of food used: $180 ÷ $4,500: 4% - One point, a year: $7,800 - Three points, a year: $23,400 **That week's waste, over 52 weeks:** $9,360 A made-up restaurant with round numbers. $9,360 is $180 × 52 and assumes week one was a typical week, which one week alone can't tell you. Do the log for a second week, in a different part of the month, before you rely on it. Say Corner Table gets the made-too-much line down by a third, from $83 to about $55 a week, with the rest of the log unchanged. That's $28 a week, or about $1,440 a year: roughly 0.2 of a point. This is arithmetic on an example, not a forecast. Your first fix may be bigger or smaller, and the log will show you which. This is why food cost is worth watching alongside labor. Prime cost is your food cost and your labor cost added together, and waste moves the food half of it. [Free calculatorSee where your food cost sits: prime cost calculatorYour sales, food cost and labor cost give your prime cost, as a share of sales.](https://restaurantdoctorai.com/calculators/prime-cost) ## What to do first: measure, cut over-prepping, then donate ### 1. Measure for a week Don't change a thing in week one. The log only helps if it shows what really happens, and a kitchen that starts prepping less on day two hides its own numbers. Once you have a week, you know your biggest reason and your top five items. ### 2. Cut over-prepping The EPA's [Wasted Food Scale](https://www.epa.gov/sustainable-management-food/wasted-food-scale) ranks preventing wasted food as the most environmentally beneficial option, ahead of donating it. For you it's also the cheapest: food you never make costs you nothing, and food you make and give away still cost you the ingredients and the labor. - Lower the pars for the three items your log shows you over-prep most, and read next week's log to see what changed. - Cook smaller batches in slow hours, and a second batch if you need it, instead of one big one for the whole shift. - Compare what you bought with what you sold of each top-five item. If you buy 40 pounds and sell 30 pounds' worth, order less. - Use trim: stock, staff meal, a special. Move older stock to the front of the shelf. - Look at plate waste on your menu: the dishes that come back, and the sides nobody finishes. There is some evidence on how it turns out, with limits. A [February 2019 report from Champions 12.3](https://champions123.org/release-new-report-finds-restaurants-save-significant-money-fighting-food-waste), a coalition of leaders from government, business and civil society, looked at the costs and savings of 114 restaurants in 12 countries. On average, restaurants saved $7 in operating costs for every $1 they invested in cutting kitchen waste, and cut kitchen waste by 26% within a year. The investments included measuring and monitoring waste, training staff on handling and storage, and redesigning menus. The release doesn't say how the restaurants were chosen and covers restaurants in many countries, so it shows what happened for them, not what will happen for you. ### 3. Donate what's still good Once your pars are lower, less is left at close, but some good food will be. On the EPA's scale, donating comes right after preventing. Book a regular pickup with a local food rescue group, where one collects near you. Whether anyone does depends on where you are, so check before you plan around it. Donate only food that was never served to a guest and was kept cold or hot, or cooled quickly by the safety rules. Anything that went out to a table can't be donated, which rules out plate waste. Your health department and the group you donate to set their own rules, so ask both before your first donation. Our [food donation guide](https://restaurantdoctorai.com/guides/restaurant-food-donation) has the safety rules and a finder for who may collect near you. ## This week 1. Print or draw the log sheet: time, who, item, why, amount. Put a tub and a sheet at the prep table, the line and the dish pit, and tell the team why you're doing it. 2. Run it for seven days without changing your pars. On day eight, price every line from your invoices and add them up. 3. Work out what it is in points: the week's total ÷ your weekly sales. Then check your own [food cost percentage](https://restaurantdoctorai.com/guides/food-cost-percentage) and [prime cost](https://restaurantdoctorai.com/calculators/prime-cost) to see where waste sits in them. 4. Pick the top three items and cut their pars. Start the log again next Monday to see what changed. 5. When pars are down and some good food is still left, ask your health department and one local food rescue group what they need before you donate. 45-second profit check Find out what else is costing you Waste is one of the places a kitchen loses money. Answer 12 quick questions, including how you track waste, and the free check names your restaurant's biggest estimated loss so you know what to fix first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Restaurants & Foodservice](https://refed.org/stakeholders/restaurants-and-foodservice/), ReFED, a U.S. nonprofit that researches food waste, 2024 data (the page shows no publication date) - [Overview of waste in the hospitality and food service sector](https://www.wrap.ngo/resources/report/overview-waste-hospitality-and-food-service-sector), WRAP (UK), November 1, 2013 (archived page) - [Prevent Wasted Food Through Source Reduction](https://www.epa.gov/sustainable-management-food/how-prevent-wasted-food-through-source-reduction), U.S. Environmental Protection Agency, December 12, 2025 - [Food Waste Logbook](https://archive.epa.gov/wastes/conserve/pubs/web/pdf/food-waste-log.pdf), U.S. Environmental Protection Agency (archived document), undated; the file was last changed August 28, 2015 - [Wasted Food Scale](https://www.epa.gov/sustainable-management-food/wasted-food-scale), U.S. Environmental Protection Agency, September 16, 2026 - [RELEASE: New Report Finds Restaurants Save Significant Money From Fighting Food Waste](https://champions123.org/release-new-report-finds-restaurants-save-significant-money-fighting-food-waste), Champions 12.3, February 13, 2019 --- title: "Restaurant break-even point: how to work out yours" description: "Work out your restaurant's break-even point: what's fixed and variable, a worked example by month, week and day, and what moves the number most." canonical: "https://restaurantdoctorai.com/blog/restaurant-break-even-point" published: "2026-10-08" updated: "2026-10-08" --- # Restaurant break-even point: how to work out yours ## Key takeaways - Break-even sales a month are your bills that stay the same divided by the share of each sale left after the costs that rise with sales. - In the example, $26,400 of fixed bills and 62.5% of sales in rising costs give $70,400 a month, about $2,708 on each of 26 days open. - Hourly pay can sit on either side: counting it as a fixed bill raised the example's break-even to $78,240. - If you're below it, check the numbers, fix avoidable costs, match shifts to guests, review fixed bills, and add sales last. Your break-even point is the sales a month needs to pay every bill, with nothing left over and nothing missing. Work it out in three steps: add up the bills that stay the same, find how much of each sale is left after the costs that rise with it, and divide the first by the second. Then split it into a week and a day: in the example below, it comes to about $2,708 on each day open. ## What's fixed, what's variable, and what's in between Break-even needs every cost in one of two piles. The U.S. Small Business Administration (SBA) describes [fixed costs](https://www.sba.gov/business-guide/plan-your-business/calculate-your-startup-costs/break-even-point) as costs that stay the same when production or service levels rise or fall, and lists rent or lease payments, salaries, property taxes, insurance, interest and depreciation. Variable costs are the ones that move with sales. In a restaurant: ### Bills that stay the same - Rent. Utilities too, in our calculator: the SBA's calculator page counts utility bills as variable, because they change with usage. - Salaries, with the payroll taxes and benefits you pay on them. - Insurance, software, marketing, licenses. - Loan and lease payments, in full (the interest and the part that pays the loan down). - Your own pay, if you take it and it isn't on payroll. Leave it out and your break-even assumes you work for free. ### Costs that rise with every sale - Food and drink: the more you sell, the more you buy. - Card fees. - Delivery app commissions. - To-go packaging. - Percentage rent, royalties or ad fund fees, if your lease or franchise has them. ### In between: hourly labor, and a few bills Hourly pay is the cost that sits on either side. If you add shifts when it's busy and cut them when it's quiet, it rises and falls with sales, so count it with the costs that rise. If about the same team works every shift whatever the sales, it's a bill that stays the same. Yours may be a bit of both. The SBA says to split costs of this mixed kind (it names phone service, repairs, fuel and power) into their fixed and variable parts where you can. Our [break-even calculator](https://restaurantdoctorai.com/calculators/break-even) asks one question: does your hourly staff time go up and down with sales? Yes counts hourly pay with the costs that rise. No counts it as a bill that stays the same. Not sure works it out both ways and shows a range, because the answer moves break-even. It puts rent and utilities with the bills that stay the same, so do the same to get matching numbers. ## The math, in three steps The SBA gives the formula in sales dollars: fixed costs divided by the contribution margin. The contribution margin is the share of each sale left after the costs that rise with sales; the managerial accounting textbook from [OpenStax](https://openstax.org/books/principles-managerial-accounting/pages/3-summary) describes it as what's available to cover fixed expenses. In plain words, it's what each sales dollar leaves to pay the rent, the salaries and the rest. 1. Take one typical month. Use your POS net sales (before tax and tips, not a holiday rush) and use the same month for every other number. 2. Add up that month's bills that stay the same. Call it F. A yearly bill goes in as one twelfth and a quarterly one as one quarter; the [SBA's own break-even calculator](https://bec.www.sba.gov/) does the same. 3. Add up that month's costs that rise with sales. Call it V. Divide V by net sales, and take that away from 1. The rest is the share of each sale left to pay F. *Break-even sales a month* Break-even = F ÷ (1 − V ÷ net sales) Write shares as decimals: 62.5% is 0.625. The share left of each sale is 1 − V ÷ net sales. Two cautions. The SBA calls the result an estimate, and it suggests adding a little extra, say 10%, for costs you can't predict. And it assumes your prices, menu mix and cost shares stay as they were in that month. ## Worked example: Casa Lupe Casa Lupe is an example restaurant, and its figures are invented for this post: they are not industry data. It is open 6 days a week and its guests spend $45 on average. They are also the break-even calculator's opening example, so you can type them in and see the same answers. *Example: Casa Lupe, one typical month* - Net sales: $90,000 - Food and drink purchases: $28,800 - Hourly staff pay: $22,500 - Card fees, app commissions, packaging: $4,950 - V: costs that rise (hourly pay in): $56,250 - V ÷ net sales: 62.5% - Left of each sale: 1 − 0.625: 37.5% - F: bills that stay the same: $26,400 **Break-even: $26,400 ÷ 0.375:** $70,400 Example figures. F is rent and utilities, salaries and everything else that doesn't move with sales, payroll taxes included. Hourly pay is counted with the costs that rise, as if Casa Lupe adds and cuts shifts with the rush. Casa Lupe sold $90,000, so it had $19,600 of cushion above break-even (21.8% of its sales). OpenStax calls that cushion the margin of safety: the sales above break-even, where a higher one means a lower risk of a loss. After every bill, about $7,350 was left over, before income tax. Now the other way round. Suppose about the same team works every shift. Then the $22,500 is a bill that stays the same: V falls to $33,750 (37.5% of sales), F grows to $48,900, and each sale leaves 62.5 cents. Break-even is $48,900 ÷ 0.625, or $78,240. The same month, the same $7,350 left over, and a break-even $7,840 higher. | Example, break-even | Hourly pay rises with sales | Same team every shift | | --- | --- | --- | | A month | $70,400 | $78,240 | | A week (month × 12 ÷ 52) | $16,247 | $18,056 | | A day open (6 days a week = 26 days a month) | $2,708 | $3,010 | | Guests a day at $45 each | 61 | 67 | | Cushion at $90,000 of sales | $19,600 (21.8%) | $11,760 (13.1%) | *Example: Casa Lupe's break-even, worked out both ways. The calculator rounds each figure up to the next dollar.* Not sure which one you are? Use the range. Casa Lupe's real schedule sits somewhere between $70,400 and $78,240. If you add the SBA's 10% to the top of that range you get $86,064, close to the $90,000 it sold. That's a thin month, not a comfortable one. Break-even is not profit Sales at break-even leave nothing after every bill, and that includes your own pay and your loan payments. The calculator calls what's left above it “left over after every bill” and not profit, because depreciation and income tax aren't in it. Your accountant can tell you how those fit your books. ## Turn it into a number for a slow Tuesday A month is too big to feel. Divide it by the days you're open and you have a number you can see on the register at 3 p.m.: Casa Lupe needs $70,400 ÷ 26 days = about $2,708 a day if its hourly team flexes, or about $3,010 if it doesn't. At $45 a guest, that's 61 to 67 guests a day. It actually sold about $3,462 a day, or 77 guests, on average. An average day isn't a Tuesday. A slow day sits under the average and a busy one over it, and what matters is the month. So make the number work for you: - Pull the last 4 weeks of net sales by weekday from your POS. Put each Tuesday next to the daily break-even. - A day under the number isn't a loss by itself: busier days pay for it. But a weekday that runs far under, week after week, is a shift to look at. - Look at that day's hours. If it keeps running far under, the hourly shifts are the cost you can change in next week's schedule. - Total the week. If the 6 days add up to the weekly figure ($16,247 to $18,056 for Casa Lupe), you are at break-even for the week. For a night-by-night version, with turns and seats, see [how many covers a night you need to break even](https://restaurantdoctorai.com/guides/restaurant-covers-to-break-even). ## What moves it most Break-even moves when a cost moves, and the same dollar moves it by different amounts. A rule that works for bills that stay the same: divide the change by the share left of each sale. For Casa Lupe, a $1,000 a month rent rise adds $1,000 ÷ 0.375 = $2,667 to break-even. A dollar saved on a cost that rises helps because the share left of each sale grows. OpenStax makes the same point, that break-even may or may not change, depending on the type of cost affected. | Example, one change at a time | Hourly pay rises with sales | Same team every shift | | --- | --- | --- | | As it is today | $70,400 | $78,240 | | Food cost up 3 points (from 32% to 35%, $2,700 a month) | $76,522 | $82,185 | | Food cost down 1 point ($900 a month) | $68,572 | $77,008 | | Rent up $1,000 a month | $73,067 | $79,840 | | Hourly pay down $2,250 a month (2.5% of sales) | $66,000 | $74,640 | *Example: how Casa Lupe's monthly break-even changes (example figures; rounded up to the next dollar).* Three things stand out. First, a few points of food cost are expensive: 3 points added $6,122 to Casa Lupe's break-even in the first column. Any cost that rises with sales does the same, such as card fees or app commissions. Second, a rent rise looks small next to sales ($1,000 against $90,000) but adds $2,667. Third, every change moves the first column more than the second, because there each sale leaves only 37.5 cents to pay the bills against 62.5 cents in the second: the less a sale leaves, the more sales it takes to cover a cost. To see where yours sit, the guides cover each side: [food cost %](https://restaurantdoctorai.com/guides/food-cost-percentage), [labor cost](https://restaurantdoctorai.com/guides/restaurant-labor-cost), [rent as a share of sales](https://restaurantdoctorai.com/guides/restaurant-rent-percentage) and [delivery app commissions](https://restaurantdoctorai.com/guides/delivery-app-commissions). [Free calculatorWork out yours: break-even calculatorAdd one typical month and see your break-even by month, week and day, and how many guests that is.](https://restaurantdoctorai.com/calculators/break-even) ## If you're below break-even If your typical month sits under your number, you are short every month, so act in this order. The early steps are cheaper and quicker, and they raise the share of each sale you keep. 1. Check the numbers first. One typical month, every invoice in it, your own pay in F. A number from the wrong month, or a missing invoice, makes break-even look better or worse than it is. 2. Fix the costs that lose money at any level of sales: plates priced off old invoices, app orders at menu prices, waste. Each point you recover lowers break-even, as the table shows. Start with [the guide to the usual problems](https://restaurantdoctorai.com/guides/restaurant-busy-but-not-making-money). 3. Match hourly shifts to the guests you really have, starting with the daypart that runs furthest under its daily number. 4. Look at the bills that stay the same: rent, insurance, software, loan payments. They are slower to change and some can't change at all. If you ask, ask for something specific. 5. Add sales last. Each extra dollar leaves only 37.5 cents in Casa Lupe's first column, so the gap is wider than it looks: to close a $5,000 shortfall you would need $13,333 in extra sales ($5,000 ÷ 0.375). If sales have fallen and that's why you're under, begin with [what to check before you cut when sales are down](https://restaurantdoctorai.com/blog/restaurant-sales-down): it splits guests from average check before you decide what to change. ## This week 1. Pick one typical month and write down net sales, food and drink purchases, hourly pay, card and app fees, and every bill that stays the same, loan payments and your own pay included. 2. Enter them in the [break-even calculator](https://restaurantdoctorai.com/calculators/break-even). If you aren't sure about hourly pay, pick Not sure and keep both numbers. 3. Write the daily break-even on a card by the register or in your manager's log. Next to it, put what each weekday sold over the last 4 weeks. 4. Find the one cost that moves your number most in the table above, and make one change to it before the next schedule goes out. 45-second profit check Is break-even your real problem? Break-even tells you the floor. The free check asks 12 quick questions about prices, costs, staffing and guests, and shows your biggest estimated loss in about 45 seconds, so you know which cost to look at first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Break-even point](https://www.sba.gov/business-guide/plan-your-business/calculate-your-startup-costs/break-even-point), U.S. Small Business Administration - [Break Even Calculator](https://bec.www.sba.gov/), U.S. Small Business Administration - [Ch. 3 Summary, Principles of Accounting, Volume 2: Managerial Accounting](https://openstax.org/books/principles-managerial-accounting/pages/3-summary), OpenStax, Rice University, February 14, 2019 (book publication date) --- title: "Credit card fees for restaurants: what you pay and why" description: "Card fees have three layers: interchange, network fees and processor markup. See the debit cap, where surcharges are limited and how to read a statement." canonical: "https://restaurantdoctorai.com/blog/restaurant-credit-card-processing-fees" published: "2026-10-08" updated: "2026-10-08" --- # Credit card fees for restaurants: what you pay and why ## Key takeaways - Each card sale pays interchange to your customer's bank, set by the card network, plus a small network fee and your processor's markup. Interchange is usually biggest. - Debit from issuers with $10 billion or more in assets is capped at 21 cents plus 0.05% (plus 1 cent). A court set the cap aside but paused that while the Fed appeals. - Visa's rules allow a credit-only surcharge of up to 3% with notice and signs, but states such as Connecticut, Massachusetts and Maine limit it. - The Visa and Mastercard settlement has only preliminary approval. The court's final approval hearing is set for November 16, 2026. Every card sale carries three fees stacked together: interchange, which goes to your customer's bank and is set by the card network, a small network fee, and your processor's markup. Interchange is usually the biggest piece, and your processor doesn't set it. Debit cards from big banks have a federal cap, a surcharge on credit cards is allowed by Visa's rules but limited by some states, and the big Visa and Mastercard settlement is not final yet. ## What you pay on each swipe: three layers Three companies touch a card sale besides you. The customer's bank (the issuer) gave them the card. The card network (Visa, Mastercard and others) passes the information and the money between banks. Your acquirer, which sometimes uses a third-party processor, provides your card services and keeps your merchant account, according to the Federal Trade Commission's [guide for businesses](https://www.ftc.gov/business-guidance/resources/new-rules-electronic-payments-lower-costs-retailers). What comes off each sale is called the merchant discount, and it covers interchange and the other processing fees below. - **Interchange.** Paid to the issuer. The U.S. Government Accountability Office (GAO) says the networks set the rates, adjust them twice a year, and build each one from a fixed amount per sale plus a percentage of the purchase. Its [2025 report on what federal agencies pay to take cards](https://files.gao.gov/reports/GAO-25-107298/index.html) found interchange made up nearly 90% of their fees. It also says network officials generally don't negotiate interchange rates with individual merchants. - **Network fees.** GAO describes these as generally fixed amounts per sale, set by each network. On debit cards, the Federal Reserve found the average network fee was 12.9 cents a sale in [its report on 2023 data](https://www.federalreserve.gov/paymentsystems/2023-interchange-fee.htm), and acquirers and merchants paid 64.9% of all network fees; issuers paid the rest. - **Processor markup and other fees.** What's left. GAO says processors charge for routing sales to the networks and banks, and that other fees can cover things like software, fraud prevention and terminal rental. The Federal Reserve Bank of Kansas City puts the costs of a debit card sale in the same three groups in [its 2025 article on small merchants](https://www.kansascityfed.org/research/economic-review/debit-card-interchange-fees-charged-to-small-merchants-after-regulation-ii/). Your processor contract sets this layer. Visa publishes its interchange rates for anyone to read. Its [rate sheet effective April 18, 2026](https://usa.visa.com/dam/VCOM/download/merchants/visa-usa-interchange-reimbursement-fees.pdf) says merchants don't pay interchange directly: they pay the merchant discount to their bank or processor, which can bundle several services. For a card-present restaurant sale, the sheet lists 2.10% (with a 4-cent minimum) on basic consumer credit cards and 2.60% on Signature, Infinite and Signature Preferred cards. It also lists a higher “non-qualified” consumer credit rate of 3.15% plus 10 cents, which is worth asking your processor about. *Example: three layers on one $40 credit card sale* - Interchange: 2.10% of $40 (Visa's published rate, basic card): $0.84 - Network fee (example amount): $0.04 - Processor markup: 0.30% + $0.10 (example): $0.22 **Total on a $40 sale:** $1.10 That's 2.75% of the sale. Only the interchange line comes from a published rate sheet; the network fee and markup are made-up round numbers to show the layers. Yours are on your statement and in your contract. ## Debit: what the Federal Reserve capped Debit cards from big banks have a federal price limit. Under the Federal Reserve's Regulation II, the interchange on one of those debit sales may not go above 21 cents plus 0.05% of the sale, plus 1 cent if the issuer meets fraud-prevention standards, as the Board's [page of average fees](https://www.federalreserve.gov/paymentsystems/regii-average-interchange-fee.htm) states. The FTC's guide says the cap covers only issuers with $10 billion or more in assets, so cards from smaller issuers are exempt. It also says credit card fees and fees your processor adds aren't capped, and that merchants must be offered a choice of at least two networks for most debit sales. | Card | Visa's published rate | On a $40 sale | | --- | --- | --- | | Basic consumer credit | 2.10% (4-cent minimum) | $0.84 | | Signature or Infinite credit | 2.60% (4-cent minimum) | $1.04 | | Debit, large issuer (capped) | 0.05% + $0.21 | $0.23, plus $0.01 if the issuer qualifies | | Debit, smaller issuer (exempt) | 1.19% + $0.10 | $0.58 | *Example: Visa's published interchange rates for a $40 card-present restaurant sale (effective April 18, 2026). Interchange only; your processor adds its own fees.* The Board's numbers for 2024 show the gap. Capped debit sales averaged 23 cents, or 0.47% of the sale, while exempt debit sales averaged 51 cents, or 1.21%. All debit together averaged 34 cents, or 0.73%. The cap could move. In October 2023 the Board [proposed lowering it](https://www.federalreserve.gov/newsevents/pressreleases/bcreg20231025a.htm), from 24.5 cents to 17.7 cents on an average $50 debit sale. On the Board's average-fee page, updated December 19, 2025, the cap still reads 21 cents plus 0.05%. A court is also involved. In August 2025 a federal court in North Dakota found the Board's fee standard contrary to law and set it aside, but paused its own order while the Board appeals, to keep fees from becoming “a completely unregulated market,” per its [amended judgment](https://storage.courtlistener.com/recap/gov.uscourts.ndd.42550/gov.uscourts.ndd.42550.87.0.pdf). The appeal was argued on May 13, 2026, and the [Eighth Circuit docket](https://www.courtlistener.com/docket/71607569/corner-post-inc-v-board-of-governors/) showed no decision when I read it, though its copy may lag; the latest entry was July 14, 2026. For now the cap stands, but it could change. ## Can you add a surcharge? A surcharge is an extra fee added to the bill because the customer pays by card. A discount for paying another way is different. The FTC's guide says the card networks can't stop you from offering a discount or incentive for a payment method, as long as it is offered to everyone and clearly disclosed. It also says you may set a minimum for credit card purchases, if it's the same for all issuers and networks and no more than $10. For surcharges, start with the network's own rules. Visa's [public rules (April 18, 2026 edition)](https://usa.visa.com/dam/VCOM/download/about-visa/visa-rules-public.pdf) allow a U.S. merchant to add a surcharge to a credit card sale, “subject to applicable laws.” The rules also say: - The surcharge can be no more than 3.00%, and the sign at checkout must say it is not greater than your Visa credit card merchant discount rate. - It applies to credit only. Debit cards can't be charged it, including a debit card the customer runs as “credit.” - You tell your processor in writing at least 30 days before you start or announce it. - You disclose it at the entrance and at checkout, and the customer can cancel after seeing it on online and phone orders. - It is set the same for all Visa credit cards, or the same for each card product type, not both. - A clause about other card brands that limit surcharging can block it. Ask your processor how that applies to you. Mastercard and other networks have their own surcharge rules, so ask your processor which apply to your business. State law is a separate test, and it is where many owners get stopped. Connecticut's Department of Consumer Protection [says](https://portal.ct.gov/dcp/knowledge-base/articles/surcharge-faqs/what-is-the-connecticut-surcharge-law) its law bars businesses from adding a fee that raises the total because of the payment method. Massachusetts says [no seller may surcharge a customer](https://malegislature.gov/Laws/GeneralLaws/PartI/TitleXX/Chapter140d/Section28a) who uses a credit card instead of cash or a check, while a discount offered to all buyers and clearly disclosed is not a finance charge. Maine's [statute](https://mainelegislature.org/legis/statutes/9-A/title9-Asec8-509.html) bars extra charges for paying by credit or debit card, and says a discount is not a surcharge. Other states have different rules, and legislatures change them. For your state, check with its attorney general or consumer protection office, read your processor agreement and ask a lawyer. Not legal advice This post says what these pages say. It can't tell you what you may do in your state. Have the wording and the plan checked before any surcharge sign goes up. ## The big lawsuit: what is and isn't settled Since 2005, merchants' lawsuits against Visa and Mastercard over their rules have been combined in one federal case in Brooklyn (E.D.N.Y.). A class of more than 12 million merchants is covered by a settlement of the part that asks for changes to the networks' rules rather than money. A judge rejected an earlier version in June 2024. The merchants and networks signed an amended one in November 2025, and on June 9, 2026 Judge Cogan granted it [preliminary approval](https://storage.courtlistener.com/recap/gov.uscourts.nyed.249477/gov.uscourts.nyed.249477.9884.0.pdf), writing that this suggests final approval is likely. That is not final approval. According to the [court's scheduling order](https://storage.courtlistener.com/recap/gov.uscourts.nyed.249477/gov.uscourts.nyed.249477.9887.0_3.pdf), the final approval hearing is set for November 16, 2026, and the [docket](https://www.courtlistener.com/docket/4318960/in-re-payment-card-interchange-fee-and-merchant-discount-antitrust/) shows objections filed in September by merchants and trade groups, among them one filing for 978 merchants and trade associations. The latest entry I read was October 5, 2026. If it is approved as written, the court's order says it would: - Let merchants accept a network's standard consumer cards without its premium cards, or consumer cards without commercial cards. - Let merchants surcharge Visa and Mastercard credit at the brand or product level, up to the lesser of their cost of accepting the card or 3%, whether or not they accept other card brands. - Cut the combined average effective credit card interchange rate by 10 basis points (0.10%) for five years, with a proportional cut for large merchants on negotiated rates. - Cap posted interchange on standard consumer credit cards at 125 basis points (1.25%) for at least eight years. Nothing in that list is in force today. The 10 basis points is a system-wide average, so your statement may not move by that amount. The court also noted objectors' point that merchants in states that cap or ban surcharging couldn't use the surcharge change. ## How to read your statement Pricing comes in a few shapes. GAO lists three models common among processors: interchange-plus, tiered and flat-rate. The Kansas City Fed says a flat rate, such as 3% of every sale, covers interchange and all other fees for every card and brand, and that small merchants often choose it. Under interchange-plus, each fee is charged separately. Large merchants usually pick that, and the Bank says it is more transparent. With a flat rate you can't see how much of your percentage is interchange and how much is markup. *Your effective rate* Effective rate = Total card fees ÷ Card sales Use one full month: every fee on the statement (percentage fees, per-sale fees, monthly fees, terminal rental, any others), and your card sales for the same month from your POS or the statement. Compare it with a wide range. GAO's review of published industry estimates found of 1.0% to 4.4% of sales, which GAO did not verify. A rate near the bottom or the top of that range tells you less than the lines behind it, so read each line: - Interchange, by card type if the statement splits it out. Debit from a large bank should look like the capped figures above. - Network fees, usually a small amount per sale. - Your processor's markup, as a percentage plus cents per sale. - Everything else: monthly, statement, terminal, software and any fee for sales that don't qualify for a lower rate. On a flat-rate statement you may see one line. Ask your processor for the breakdown, or a quote on a pricing model that shows each part. ## What one point is worth: an example A percentage point sounds small until you multiply it by a month of card sales. Here is a made-up restaurant. *Example: Corner Table Café (made up)* - Monthly sales: $100,000 - Paid by card: 85%: $85,000 - Card fees at a 2.8% blended rate (example): $2,380 - Card fees at 1.8%: $1,530 **One point lower, per month:** $850 That's $10,200 a year. At an average card check of $34 (also an example), 2.8% is about 95 cents a check. The 2.8% is a round number inside GAO's 1.0% to 4.4% range, not a typical rate. In the example, card fees are 2.4% of all sales: $2,380 on $100,000. Your [prime cost](https://restaurantdoctorai.com/guides/restaurant-prime-cost), food and drink plus total labor as a share of sales, leaves that out. So a restaurant can hit its prime cost target and still lose a couple of points of sales to card fees. If you want to see where else the money goes, see [why a busy restaurant can still lose money](https://restaurantdoctorai.com/guides/restaurant-busy-but-not-making-money). ## This week 1. Pull your last three monthly statements. For each, divide total card fees by card sales to get your effective rate, and write the three numbers down. 2. Find your pricing model on the statement and mark each line: interchange, network fees, markup, everything else. If you can't tell, that is your first question. 3. Call or email your processor with four questions: What is my markup, as a percentage plus cents per sale? Which of my sales hit a non-qualified or other higher rate, and why? Is my debit routed to the lowest-cost network I'm offered? What does my contract say about price changes and ending early? 4. Before any surcharge or cash discount sign goes up, check your state's rule with its attorney general or consumer protection office, your processor agreement and, if you're unsure, a lawyer. Visa's rules also need 30 days' written notice to your processor. 5. Don't plan around the settlement or the debit cap until the courts rule. Write November 16 in your calendar and look at the docket afterward. 45-second profit check See where the money goes Card fees are one line on a long list of costs. Answer 12 quick questions and get a free grade and your biggest estimated loss in about 45 seconds, so you know whether fees are where to look first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Payment Cards: Costs and Benefits for Federal Entities (GAO-25-107298)](https://files.gao.gov/reports/GAO-25-107298/index.html), U.S. Government Accountability Office, April 30, 2025 - [New Rules on Electronic Payments Lower Costs for Retailers](https://www.ftc.gov/business-guidance/resources/new-rules-electronic-payments-lower-costs-retailers), Federal Trade Commission, September 2011 - [2023 Interchange Fee Revenue, Covered Issuer Costs, and Covered Issuer and Merchant Fraud Losses Related to Debit Card Transactions](https://www.federalreserve.gov/paymentsystems/2023-interchange-fee.htm), Board of Governors of the Federal Reserve System, December 19, 2025 - [Debit Card Interchange Fees Charged to Small Merchants After Regulation II](https://www.kansascityfed.org/research/economic-review/debit-card-interchange-fees-charged-to-small-merchants-after-regulation-ii/), Federal Reserve Bank of Kansas City, March 27, 2025 - [Visa USA Interchange Reimbursement Fees](https://usa.visa.com/dam/VCOM/download/merchants/visa-usa-interchange-reimbursement-fees.pdf), Visa, April 18, 2026 - [Regulation II: Average Debit Card Interchange Fee by Payment Card Network](https://www.federalreserve.gov/paymentsystems/regii-average-interchange-fee.htm), Board of Governors of the Federal Reserve System, December 19, 2025 - [Federal Reserve Board requests comment on a proposal to lower the maximum interchange fee that a large debit card issuer can receive](https://www.federalreserve.gov/newsevents/pressreleases/bcreg20231025a.htm), Board of Governors of the Federal Reserve System, October 25, 2023 - [Amended Judgment, Corner Post, Inc. v. Board of Governors of the Federal Reserve System, No. 1:21-cv-00095](https://storage.courtlistener.com/recap/gov.uscourts.ndd.42550/gov.uscourts.ndd.42550.87.0.pdf), U.S. District Court, District of North Dakota (copy on CourtListener), October 27, 2025 - [Docket, Corner Post, Inc. v. Board of Governors, No. 25-3000](https://www.courtlistener.com/docket/71607569/corner-post-inc-v-board-of-governors/), U.S. Court of Appeals for the Eighth Circuit (copy on CourtListener), July 14, 2026 - [Visa Core Rules and Visa Product and Service Rules](https://usa.visa.com/dam/VCOM/download/about-visa/visa-rules-public.pdf), Visa, April 18, 2026 - [What is the Connecticut Surcharge law](https://portal.ct.gov/dcp/knowledge-base/articles/surcharge-faqs/what-is-the-connecticut-surcharge-law), Connecticut Department of Consumer Protection, August 28, 2024 - [General Laws, Part I, Title XX, Chapter 140D, Section 28A](https://malegislature.gov/Laws/GeneralLaws/PartI/TitleXX/Chapter140d/Section28a), The Commonwealth of Massachusetts, General Court - [Maine Revised Statutes, Title 9-A, Section 8-509: Surcharge](https://mainelegislature.org/legis/statutes/9-A/title9-Asec8-509.html), Maine Legislature, Amended 2021 - [Memorandum Decision and Order granting preliminary approval, In re Payment Card Interchange Fee and Merchant Discount Antitrust Litigation, No. 05-md-1720](https://storage.courtlistener.com/recap/gov.uscourts.nyed.249477/gov.uscourts.nyed.249477.9884.0.pdf), U.S. District Court, Eastern District of New York (copy on CourtListener), June 9, 2026 - [Class Settlement Notice and Scheduling Order, No. 05-md-1720](https://storage.courtlistener.com/recap/gov.uscourts.nyed.249477/gov.uscourts.nyed.249477.9887.0_3.pdf), U.S. District Court, Eastern District of New York (copy on CourtListener), June 15, 2026 - [Docket, In re Payment Card Interchange Fee and Merchant Discount Antitrust Litigation, No. 05-md-1720](https://www.courtlistener.com/docket/4318960/in-re-payment-card-interchange-fee-and-merchant-discount-antitrust/), U.S. District Court, Eastern District of New York (copy on CourtListener), October 5, 2026 --- title: "Average restaurant profit margin: what to expect" description: "The latest published restaurant profit margins, why the figures differ, and how one point of food or labor cost moves what you keep, with an example." canonical: "https://restaurantdoctorai.com/blog/average-restaurant-profit-margin" published: "2026-10-08" updated: "2026-10-08" --- # Average restaurant profit margin: what to expect ## Key takeaways - The National Restaurant Association's 2024 medians are 2.8% of sales before tax for full-service restaurants and 4.0% for limited-service ones. - Published margins differ because they measure different things: before or after tax, a survey or a table of companies, and which year. - In the example, one point of food or labor cost is $900 a month on $90,000 of sales, and two more points on both turns a 3% margin into a loss. - Prime cost covers only food, drink and labor and can be checked weekly. Margin is what is left after every bill. The latest figure from the National Restaurant Association's survey of operators is a median profit before tax of 2.8% of sales for full-service restaurants and 4.0% for limited-service (counter-service) ones, in 2024. Other figures you will find are quite different because they measure different things: profit before or after tax, a survey of restaurants or a table of companies, one year or another. On $1,000,000 of sales, 2.8% is $28,000 for the year, so a point or two of food or labor cost decides whether you keep anything. ## What a typical restaurant keeps The best published picture of what ordinary restaurants keep comes from the National Restaurant Association's 2025 Restaurant Operations Data Abstract, a survey of more than 900 restaurants. In its [August 20, 2025 article on the report](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/new-association-report-helps-operators-gauge-their-restaurant-performance), the Association gives the median income before taxes in 2024: 2.8% of sales for full-service restaurants and 4.0% for limited-service. A median is the middle restaurant: half did better, half worse. Size matters in full service. In the Association's [October 16, 2025 article on sales volume](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/higher-volume-restaurants-reported-lower-food-cost-ratios-in-2024/), full-service restaurants with sales of $2 million a year or more had a median income before taxes of 4.3% of sales, and those below $2 million had 1.1%. The same article puts their median food cost (food and non-alcohol drinks) at 31.0% of sales for the larger group and 33.7% for the smaller. The Association also says its figures are “not intended to represent standards or goals” for individual restaurants. Read them as where respondents landed in 2024, not as a target. What a median can hide A median says nothing about how far apart restaurants were: some of the restaurants in the survey reported a loss, as the labor figures below show. Use the numbers to see roughly where you stand, then use your own books to decide what to change. ## Why the published margins don't match Search for the average restaurant profit margin and the figures run from under 3% to nearly 16% in the two sources below. Usually nobody is wrong. They are answering different questions. Ask these five of any number: - **Before or after tax?** The Association reports income before taxes. NYU Stern's Aswath Damodaran publishes a net margin, net income divided by total revenues, and a pre-tax operating margin, the profit from running the business before interest and tax, divided by sales (see his [variable definitions](https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/variable.htm)). - **Who is counted?** The Association's figures come from a survey of more than 900 restaurants. Damodaran's restaurant row covers 64 firms taken from data services, and his page doesn't say how big they are. - **Middle or total?** The Association reports a median. For a sector, Damodaran divides all the firms' net income by all their sales, so larger firms weigh more. - **Which year?** The Association's figures are for 2024. Damodaran's table says its data are as of January 2026, and it was last updated that month. - **Which kind of restaurant?** Full service and counter service differ, and so does a $1 million restaurant from a $3 million one. | Source | What it measures | Restaurants | | --- | --- | --- | | National Restaurant Association | Median income before taxes, % of sales | Full service 2.8%, limited service 4.0% | | Damodaran (NYU Stern), Restaurant/Dining row | Net margin, net income ÷ revenues | 9.37% (64 firms) | | Same row | Pre-tax unadjusted operating margin | 15.79% | *Two sources, side by side. Each row is as the source states it: the Association's from its 2025 Abstract (2024 results), Damodaran's from his Margins by Sector (US) table (data as of January 2026).* [Damodaran's table](https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/margin.html) is a table of firms from data services, not a survey of restaurant owners, so it can't tell you what a restaurant like yours keeps. The Association's survey is closer to your situation, but it is still a middle, and the pages we read don't spell out every cost line behind “income before taxes”. So when you compare your own margin with any figure, first make sure you are dividing the same profit by the same sales. ## What the sources say by type of restaurant There is no clean split of profit margin for every kind of place. The Association separates two kinds of service and, inside full service, restaurants above and below $2 million a year. We found no primary source that gives a margin for cafés, bars, pizzerias or fine dining on their own, so we won't give you one. | Group | Income before taxes | Where it comes from | | --- | --- | --- | | Full service, all | 2.8% | August 20, 2025 article | | Full service, sales $2 million or more | 4.3% | October 16, 2025 article | | Full service, sales below $2 million | 1.1% | October 16, 2025 article | | Limited service, all | 4.0% | August 20, 2025 article | *Median income before taxes, 2024, as a share of sales, from the National Restaurant Association's 2025 Operations Data Abstract. The Association's labels, not ours.* Pick the closest match, not a perfect one. A sit-down place with table service belongs next to full service. A counter-service café or pizzeria sits closest to limited service. A bar fits neither, and its costs run differently, so start from its [pour cost](https://restaurantdoctorai.com/guides/bar-pour-cost) and your prime cost target (57% or less for a bar) rather than a borrowed margin. ## Where each dollar of sales goes: an example Take an example full-service restaurant with $90,000 in sales a month, which is $1,080,000 a year. The sales, food and labor figures are the same example as in our [prime cost guide](https://restaurantdoctorai.com/guides/restaurant-prime-cost). The rent and “everything else” lines are made-up round numbers for this example: yours will differ. *Example: a full-service restaurant, one month* - Sales: $90,000 - − Food and drink (32 cents of each dollar): $28,800 - − Total labor (34 cents): $30,600 - = Prime cost (66 cents): $59,400 - − Rent, property charges and utilities (8 cents): $7,200 - − Everything else (23 cents): $20,700 **Left before tax (3 cents):** $2,700 Everything else stands for card fees, supplies, repairs, insurance, marketing, software and other bills, with no benchmark behind it: it is the part of this example that is simply assumed. $2,700 a month is $32,400 a year, which is 3% of $1,080,000. Food cost and labor sit inside the ranges our guides use for full service (30–34% and 30–35%), and occupancy inside 7–12%. Three cents of every dollar is left. That is close to the Association's 2.8% median for full service, but you can reach it with very different mixes. Two restaurants with the same sales and the same three cents can be spending their money in different places, and a change in any one line moves what they keep. Two things follow. First, margin is what is left after every bill, so it is small and moves a lot. Second, the two biggest lines, food and labor, are the ones you can change this week. ## Margin is not prime cost Prime cost is food and drink cost plus total labor cost, divided by sales. It is only the first part of the picture: in the example, 66 cents of each dollar before a single bill for rent or insurance is paid. The [prime cost guide](https://restaurantdoctorai.com/guides/restaurant-prime-cost) gives the ceiling for each kind of restaurant, 65% or less for full service. *Two different numbers* **Prime cost % = (food and drink cost + total labor cost) ÷ net sales × 100:** Profit margin % = (net sales − every cost) ÷ net sales × 100 Both use net sales: what you rang up after discounts and comps, not counting sales tax. Use the same period, a week or a month, for every figure. Prime cost can be checked every Monday, because it needs only invoices, payroll and sales. Margin needs the whole month's bills and often waits for your accountant. That is why prime cost is the better weekly warning light and margin is the month-end score. If prime cost is over its ceiling, less is left to pay for everything else. If it is under, you have not won yet: rent, delivery fees and the rest still come out of what remains. [Free calculatorCheck your prime cost: prime cost calculatorYour sales, food and drink and payroll for a week or a month, against the target for your type of restaurant.](https://restaurantdoctorai.com/calculators/prime-cost) ## What moves margin fastest Each point of a cost, measured against sales, is worth 1% of sales. At $90,000 a month that is $900, or $10,800 a year. At $40,000 a month a point is $400, and at $150,000 it is $1,500. When your margin is 3%, one point is a third of everything you keep. | What changes | Prime cost | Left before tax | Per month | | --- | --- | --- | --- | | Nothing (the example) | 66% | 3% | $2,700 | | Food cost up 1 point | 67% | 2% | $1,800 | | Food cost up 2 points | 68% | 1% | $900 | | Labor up 3 points | 69% | 0% | $0 | | Food and labor each up 2 points | 70% | −1% | −$900 | | Food cost down 2 points | 64% | 5% | $4,500 | | Labor down 2 points | 64% | 5% | $4,500 | *The example restaurant above ($90,000 a month, prime cost 66%, 3% left before tax, $2,700 a month), with one cost changed at a time. Everything not named stays the same. All numbers are examples.* A two-point rise in both costs turns a restaurant that keeps $2,700 a month into one losing $900. The same size of change the other way is worth just as much: two points off either food or labor takes the example from 3% to 5%. Don't cut across the board. Trimming every shift and every dish by the same amount treats a specific problem as a general one. Find which half has been creeping up, then use the [food cost guide](https://restaurantdoctorai.com/guides/food-cost-percentage) or the [labor cost guide](https://restaurantdoctorai.com/guides/restaurant-labor-cost). Rent moves slowly, so check it once with the [rent guide](https://restaurantdoctorai.com/guides/restaurant-rent-percentage), and look at [delivery app commissions](https://restaurantdoctorai.com/guides/delivery-app-commissions) if apps bring in orders. ## Why the same sales can leave very different amounts Two restaurants with $90,000 a month can keep very different amounts, and the survey shows one difference between those that made money and those that didn't. In the Association's [October 8, 2025 article on labor and profitability](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/), the median labor cost (salaries and wages, including benefits) of full-service restaurants that reported a pre-tax profit was 34.2% of sales. For those that reported a loss it was 42.9%. For all full-service respondents it was 36.5%. That is a gap of 8.7 points. On $90,000 a month, 8.7 points is $7,830 a month. These are two groups, not one restaurant, and the survey does not say what caused the gap: it may be wages, hours, sales or a mix. The figure is a reason to look at your schedule against your sales, not a promise that cutting hours will make a profit. The Association's sales-volume article points the same way on food. The larger full-service restaurants had a median food cost of 31.0%, the smaller ones 33.7%, a gap of 2.7 points. Their income before taxes was 4.3% against 1.1%. The article doesn't say why the two groups differ, and neither do we. ## This week 1. **Work out your own margin for last month.** Take net sales, subtract every cost on your P&L, and divide what is left by net sales. Ask your accountant which profit the P&L shows (before or after tax) and whether it counts your own pay, so you compare like with like. 2. **Pick the nearest row in the table.** Full service, full service above or below $2 million a year, or limited service. Put your number next to it, remembering it is a 2024 median, not a goal. 3. **Find what one point is worth to you.** Multiply last month's net sales by 0.01. That is the dollars a month you gain or lose for each point of food cost or labor cost. 4. **Work out your prime cost for last week.** Add last week's food and drink invoices and labor, divide by last week's net sales, and compare it with the ceiling for your type in the [prime cost guide](https://restaurantdoctorai.com/guides/restaurant-prime-cost). Do the same next Monday and the one after. 5. **Pick one half to fix.** Whichever of food or labor has gone up over the last few Mondays is the one to work on first. 45-second profit check Where is your margin going? The free check asks 12 quick questions about your restaurant and shows the cost areas that are most likely costing you profit, in about 45 seconds. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [New Association report helps operators gauge their restaurant performance](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/new-association-report-helps-operators-gauge-their-restaurant-performance), National Restaurant Association, August 20, 2025 - [Higher-volume restaurants reported lower food cost ratios in 2024](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/higher-volume-restaurants-reported-lower-food-cost-ratios-in-2024/), National Restaurant Association, October 16, 2025 - [Variable definitions (Damodaran Online datasets)](https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/variable.htm), NYU Stern School of Business (Aswath Damodaran) - [Margins by Sector (US)](https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/margin.html), NYU Stern School of Business (Aswath Damodaran), January 2026 - [Elevated labor costs had a significant impact on restaurant profitability in 2024](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/), National Restaurant Association, October 8, 2025 --- title: "No tax on tips: what it means for restaurant owners" description: "The federal tips deduction is your servers' income tax break, not a payroll tax cut for you. What the IRS pages say, and what to check with payroll." canonical: "https://restaurantdoctorai.com/blog/no-tax-on-tips-restaurants" published: "2026-10-08" updated: "2026-10-08" --- # No tax on tips: what it means for restaurant owners ## Key takeaways - The tips deduction is the worker's federal income tax break; the final regulations say it does not reduce the wages that Social Security and Medicare tax is figured on. - The IRS list covers servers, bartenders, bussers, cooks, hosts, dishwashers, bakers and pizza delivery drivers, but a mandatory service charge is not a qualified tip. - On the 2026 W-2, the IRS adds a tips amount in Box 12 (code TP) and a tipped-occupation code in Box 14b, so check your payroll setup before year-end. - The FICA tip credit is a separate rule from the deduction. Rules can change and states differ, so confirm your own case with your accountant or payroll provider. “No tax on tips” is a federal income tax deduction that belongs to your tipped workers, not to you. The IRS says a worker can deduct up to $25,000 a year of qualified tips for tax years 2025 through 2028, and the final regulations say it does not change the wages that Social Security and Medicare tax is figured on. So what you pay as the employer on tips stays the same; what changes is paperwork, because the 2026 W-2 has new places for tips and each worker's tipped-occupation code. Read this first This post reports what the IRS pages say as of October 8, 2026, with the date each page gives itself. Tax rules can be updated, and this is not tax advice. It covers federal rules only: state income tax differs by state, and none of the states was checked here. Your accountant or payroll provider knows your case. ## The deduction in one paragraph Tips have always been taxable income. The 2025 law, known as the One, Big, Beautiful Bill Act and called the Working Families Tax Cuts Act on newer IRS pages, added a deduction for [qualified tips](https://www.irs.gov/newsroom/what-the-no-tax-on-tips-deduction-means-for-you): voluntary cash or charged tips from customers, including tips a worker gets through sharing. A deduction lowers the income that federal income tax is figured on. It does not make the tips tax-free in every way, and it does not touch the Social Security and Medicare tax taken from the same tips. These are the limits the IRS lists on that page, and the [IRS tip reporting page](https://www.irs.gov/businesses/small-businesses-self-employed/tip-recordkeeping-and-reporting) gives the same years and cap: - **Up to $25,000 a year** per return, for tax years 2025 through 2028. - **It shrinks at higher incomes:** the deduction phases out when modified adjusted gross income is above $150,000, or $300,000 on a joint return. - **Itemizing is not required.** A worker can take it with the standard deduction. - **A valid Social Security number** is needed, and married workers must file jointly. - **Self-employed people** can't deduct more than the net income from the business where they earned the tips. Workers claim it on their own tax return, on a new form called Schedule 1-A. A server who already filed a 2025 return can [amend it, the IRS says](https://www.irs.gov/forms-pubs/claiming-no-tax-on-tips-deduction-for-occupations-that-customarily-and-regularly-receive-tips-may-require-an-amended-return), to claim or change the deduction. ## Whose tips qualify, by the IRS's list The deduction is tied to the job, not the person. Only tips received in an occupation that customarily and regularly received tips on or before December 31, 2024 count, and the IRS publishes the [list of occupations that receive tips](https://www.irs.gov/forms-pubs/occupations-that-customarily-and-regularly-received-tips-on-or-before-dec-31-2024) (last reviewed June 28, 2026). Each job has a three-digit Treasury Tipped Occupation Code. The [final regulations](https://www.irs.gov/newsroom/treasury-irs-issue-final-regulations-listing-occupations-where-workers-customarily-and-regularly-receive-tips-under-the-one-big-beautiful-bill), announced April 10, 2026, say the list has more than 70 occupations, in eight groups. The restaurant jobs are mostly in the first: | Code | Job on the list | Examples the IRS gives | | --- | --- | --- | | 101 | Bartenders | Barkeep, taproom attendant, sommelier | | 102 | Wait staff | Cocktail waitress, banquet staff | | 104 | Dining room and cafeteria attendants, bartender helpers | Bar back, busser | | 105 | Chefs and cooks | Sous chef, restaurant cook, food truck cook, caterer | | 106 | Food preparation workers | Salad maker, sandwich maker, kitchen steward | | 107 | Fast food and counter workers | Barista, ice cream server | | 108 | Dishwashers | Dish room worker | | 109 | Host staff, restaurant, lounge and coffee shop | Dining room host | | 110 | Bakers | Bread baker, pastry finisher | | 804 | Goods delivery people | Pizza delivery driver, app or platform delivery person | *Jobs on the IRS list that a restaurant, café, bar or pizzeria is likely to have. The code is the Treasury Tipped Occupation Code.* The food and drink group on the page has ten jobs in all (it also lists 103, servers outside a restaurant, such as room service), and the page lists no job called manager or owner. The [final rule](https://www.govinfo.gov/content/pkg/FR-2026-04-13/pdf/2026-07104.pdf) says the list of occupations is closed, while the examples under each code are only examples. Being on the list is only the first test: the worker still has to receive qualified tips. ### What counts as a qualified tip The IRS release on the final regulations says a qualified tip is paid in cash or something like it (check, card, gift card, or a cash-denominated electronic or mobile payment), and comes from a customer or, for an employee, from a tip-sharing arrangement such as a tip pool. It is paid voluntarily and is not subject to negotiation. That is why a card tip on a tip screen counts. Service charges are the part owners should look at. The release says they are generally not qualified tips unless the customer can disregard or modify them, and its own example is an automatic 18% charge for large parties. The W-2 instructions put it plainly: “Mandatory service charges added to the bill are not qualified tips.” The final rule also says a tip screen is voluntary when the customer can bring the tip down to zero, with a slider that goes to zero or an “other” box where zero can be typed. The IRS tip reporting page lists a digital payment prompt that makes the customer pick a tip above zero as an example of a service charge. Ask your accountant how your own screen is set up. Our guide to [restaurant service charges](https://restaurantdoctorai.com/guides/restaurant-service-charge) covers what else changes when you add one. One more exclusion: tips earned in a “specified service trade or business” don't qualify. The IRS has reserved that part of the regulations for later and gave transition relief in the meantime. Ask your accountant whether it could touch your business. ## What changes for you as the employer, and what doesn't The first footnote of the final regulations settles the money question. Tips are wages for Social Security and Medicare (FICA) tax, but the deduction is “not taken into account in determining wages subject to FICA tax.” So the deduction does not reduce your share of that tax, and it does not reduce your workers' share either. I did not find an IRS page that says it lowers an employer's payroll tax. The [IRS tip topic page](https://www.irs.gov/taxtopics/tc761) still says the employer pays its share of Social Security and Medicare tax on tips, with the worker's share collected from wages or from money the worker provides. | Item | Changed? | What the IRS pages say | | --- | --- | --- | | Your share of Social Security and Medicare tax on tips | No | The deduction does not apply for FICA (final regulations). | | Workers reporting tips to you | No | Written report by the 10th of the next month, unless tips from you are under $20 that month (IRS Topic 761). | | Service charges and auto-gratuities | No | Still not tips; the IRS has said so since at least 2012 (final rule). | | Income tax withholding | Not unless a worker gives a new W-4 | A worker who wants paychecks to reflect it gives a new Form W-4 (IRS withholding page, written for 2025). | | The W-2 you give for 2026 | Yes | New Box 12 code TP and new Box 14b for tipped-occupation codes (2026 W-2 instructions). | | FICA tip credit | No change found | The IRS credit page does not mention the deduction. | *What the deduction changes for an employer, and what it leaves alone, as of October 8, 2026.* ### The 2026 W-2 For 2025 the IRS treated reporting as a transition year. Its [release of November 5, 2025](https://www.irs.gov/newsroom/treasury-irs-provide-penalty-relief-for-tax-year-2025-for-information-reporting-on-tips-and-overtime-under-the-one-big-beautiful-bill) says employers and other payors would not face penalties for not separately accounting for cash tips or the worker's occupation, if the return was otherwise complete and correct, and that the 2025 forms would not be updated. That relief was for 2025. I did not find a page giving the same relief for 2026, so ask your accountant whether any applies. The [2026 W-2 instructions](https://www.irs.gov/instructions/iw2w3) (last reviewed April 30, 2026) and the IRS tip reporting page show what the new form holds: - **Box 12, code TP:** the total cash tips the worker reported to you. - **Box 14b (new):** the worker's Treasury Tipped Occupation Code. The old Box 14 “Other” moves to Box 14a. - **Up to two codes** go in Box 14b. A worker with tips from more than two jobs gets codes for any two of them. - **The code 000:** if any of the tips came from a job that doesn't qualify, the instructions say 000 goes in as one of the codes. Ask your payroll provider where it gets these codes, and check them before the year ends. A server who also tends bar can have tips from two jobs, and the form has room for two codes. *Example: one server at an example restaurant, for a year* - Wages: 1,500 hours at $12.00: $18,000 - Large-party service charge paid out as wages: $1,000 - Tips she reported to the restaurant: $20,000 - W-2 Box 12, code TP: $20,000 - W-2 Box 14b (wait staff): 102 - Employer share of FICA on all of it: $39,000 × 7.65%: $2,983.50 **Of that, on her tips: $20,000 × 7.65%:** $1,530 An example, not a real restaurant. The $1,530 is the same with or without the deduction. The $1,000 service charge is wages, so it is not in Box 12 and she can't deduct it. If she qualifies, the deduction can take up to $20,000 off the income her federal income tax is figured on, and that happens on her return, not on your payroll. The employer share is the same 7.65% rate the IRS credit page uses. The employer share of tax on tips is part of what your team costs you, so it belongs in your labor cost. Our [labor cost guide](https://restaurantdoctorai.com/guides/restaurant-labor-cost) counts payroll taxes in the number and leaves the tips themselves out. [Free calculatorCheck your labor cost percentageEnter wages, payroll taxes and sales to see your labor cost against the target for your type of restaurant.](https://restaurantdoctorai.com/calculators/labor-cost) ## The FICA tip credit is a different thing If tipping is customary at your business, the [IRS FICA tip credit](https://www.irs.gov/businesses/small-businesses-self-employed/fica-tip-credit-for-employers) (last reviewed August 16, 2026) gives back some of the employer Social Security and Medicare tax you paid on tips. You figure the credit on Form 8846 and, in most cases, attach it to your tax return. It is a credit against income tax, not a lower payroll deposit. - **Amount:** the 7.65% employer rate times the tips that count, per the IRS's own worked example. - **What doesn't count:** distributed service charges and auto-gratuities, which the page treats as wages, and tips that only bring a worker's pay up to the wage basis the page names. Check that basis on the page or Form 8846's instructions. - **Limits:** it's nonrefundable and counts as a general business credit. Unused credit can be carried back one year or forward up to 20 years. The credit page doesn't mention the new deduction. As the page reads today, the two rules sit side by side: the worker's deduction on income tax, and your credit on the payroll tax you pay. If your accountant has never claimed the credit, ask whether it applies to you. ## What your servers ask, and what to answer Each answer below is what the IRS pages say, in plain words. For anything about one person's tax return, point them to a tax preparer or the IRS. ### “Is my paycheck going up?” Not on its own. The deduction is claimed on the worker's return. The [IRS withholding page](https://www.irs.gov/forms-pubs/how-to-update-withholding-to-account-for-tax-law-changes-for-2025) says a worker who wants the deduction reflected in paychecks gives the employer a new Form W-4, using the IRS worksheet and the deductions line. That page was written for 2025 paychecks, so ask your payroll provider how it handles a new W-4 now. ### “Do I pay less Social Security and Medicare now?” No. The final regulations say the deduction does not apply for that tax. It still comes out of tips as before. ### “Is our auto-gratuity on big parties covered?” The IRS example says an automatic 18% charge for a large party is not a tip, so it isn't a qualified tip. Whether and how to change how you charge it is a question for your accountant. ### “The kitchen shares our tips. Do the cooks get it?” Cooks, food prep and dishwashers are on the IRS list, and the release counts tips received through a tip pool or other sharing arrangement. So a cook's share of customer tips can be a qualified tip, if the other rules are met. Who may be in a pool is a wage-law question, so check it with your state labor agency or accountant. ### “I already filed. Did I miss it?” The IRS page on amended returns says a worker who already filed a 2025 return can file an amended one to claim or change the deduction. Tell them to ask a tax preparer. ## This week 1. **Match your jobs to the IRS list.** Write the three-digit code next to each tipped job on your staff list, using the table above or the IRS page. Note anyone who works two jobs, and any job that isn't on the list. 2. **Ask your payroll provider three questions.** Will the 2026 W-2 show Box 12 code TP and a Box 14b code for each tipped worker? Where do the codes come from? What happens for a worker with two jobs, or one with a job that isn't on the list? 3. **Look at your POS.** Does the tip screen let a guest enter zero? Are auto-gratuities set up as service charges, kept apart from tips in your reports? The IRS treats the two differently. 4. **Check your tip reporting routine.** The IRS page says workers report cash tips in writing by the 10th of the next month. Make sure you can find those reports. If you run a large food or beverage establishment, ask your accountant about Form 8027 and allocated tips, which Topic 761 describes. 5. **Ask your accountant two things.** Are we claiming the FICA tip credit on Form 8846? And how does our state treat tips for state income tax? 6. **Give your team one answer.** It's a federal income tax deduction on the worker's own return, up to the limits, for listed jobs and qualified tips. It does not change Social Security and Medicare tax. Send questions about their own return to a tax preparer. 45-second profit check Is payroll quietly your biggest cost? Tip rules are one corner of labor. The free check asks 12 quick questions, including how you track labor, and grades each cost area so you can see your biggest estimated loss, in about 45 seconds. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [What the “No Tax on Tips” deduction means for you](https://www.irs.gov/newsroom/what-the-no-tax-on-tips-deduction-means-for-you), Internal Revenue Service, March 5, 2026 (last reviewed September 21, 2026) - [Tip recordkeeping and reporting](https://www.irs.gov/businesses/small-businesses-self-employed/tip-recordkeeping-and-reporting), Internal Revenue Service, Last reviewed July 31, 2026 - [Claiming “no tax on tips” deduction for occupations that customarily and regularly receive tips may require an amended return](https://www.irs.gov/forms-pubs/claiming-no-tax-on-tips-deduction-for-occupations-that-customarily-and-regularly-receive-tips-may-require-an-amended-return), Internal Revenue Service, Last reviewed June 4, 2026 - [Occupations that customarily and regularly received tips on or before Dec. 31, 2024](https://www.irs.gov/forms-pubs/occupations-that-customarily-and-regularly-received-tips-on-or-before-dec-31-2024), Internal Revenue Service, Last reviewed June 28, 2026 - [Treasury, IRS issue final regulations listing occupations where workers customarily and regularly receive tips (IR-2026-49)](https://www.irs.gov/newsroom/treasury-irs-issue-final-regulations-listing-occupations-where-workers-customarily-and-regularly-receive-tips-under-the-one-big-beautiful-bill), Internal Revenue Service, April 10, 2026 - [Occupations That Customarily and Regularly Received Tips; Definition of Qualified Tips (T.D. 10044)](https://www.govinfo.gov/content/pkg/FR-2026-04-13/pdf/2026-07104.pdf), Department of the Treasury, Internal Revenue Service, via the Federal Register, April 13, 2026 - [Tax Topic 761, Tips: withholding and reporting](https://www.irs.gov/taxtopics/tc761), Internal Revenue Service, Last reviewed September 24, 2026 - [Treasury, IRS provide penalty relief for tax year 2025 for information reporting on tips and overtime (IR-2025-110)](https://www.irs.gov/newsroom/treasury-irs-provide-penalty-relief-for-tax-year-2025-for-information-reporting-on-tips-and-overtime-under-the-one-big-beautiful-bill), Internal Revenue Service, November 5, 2025 - [General Instructions for Forms W-2 and W-3 (2026)](https://www.irs.gov/instructions/iw2w3), Internal Revenue Service, Last reviewed April 30, 2026 - [FICA Tip Credit for employers](https://www.irs.gov/businesses/small-businesses-self-employed/fica-tip-credit-for-employers), Internal Revenue Service, Last reviewed August 16, 2026 - [How to update withholding to account for tax law changes for 2025](https://www.irs.gov/forms-pubs/how-to-update-withholding-to-account-for-tax-law-changes-for-2025), Internal Revenue Service, Last reviewed September 15, 2026 --- title: "Coffee prices in 2026: what cafés should do about the cup" description: "What USDA, BLS, the World Bank and the ICO show on coffee prices, and how to cost a latte and a drip coffee and decide to hold, raise or change the drink." canonical: "https://restaurantdoctorai.com/blog/coffee-prices-cafes-2026" published: "2026-10-08" updated: "2026-10-08" --- # Coffee prices in 2026: what cafés should do about the cup ## Key takeaways - Shelf coffee cost 4.8% more in August 2026 than a year earlier, though green coffee has eased since late 2025. - Cost a cup as grams of coffee × cost per gram, plus milk and the cup, using your own invoice. - Under 1 point of cost, hold; from 1 to 3, raise by the extra cost; over 3, pair a smaller raise with a recipe change. - Four ways to close the gap: hold the price, raise it by the extra cost, change the recipe or size, or steer guests to drinks that leave more. Cost one cup from your own roaster invoice: grams of coffee in the cup × the price of a gram, plus the milk and the cup. National data shows coffee on the shelf cost 4.8% more in August 2026 than a year earlier and 47% more than two years earlier, even though green coffee has eased since late 2025. Re-cost each drink, see how many points its food and drink cost % has risen, and under 1 point hold the price, from 1 to 3 raise it by the extra cost, and over 3 pair a smaller raise with a recipe change. ## What the primary sources say coffee prices did Coffee has three prices that move at different speeds: the green (unroasted) bean that roasters buy, the roasted coffee on a store shelf, and the bag your roaster sells you. The first two are published. All figures are as of the dates given, and they change every month. ### Green coffee The World Bank's [Pink Sheet](https://thedocs.worldbank.org/en/doc/74e8be41ceb20fa0da750cda2f6b9e4e-0050012026/related/CMO-Pink-Sheet-October-2026.pdf) of October 2, 2026 gives the price of arabica, one of the two main coffee species (robusta is the other), in dollars per kilogram. The per-pound column below is worked out from it (1 kg is 2.2046 lb). | Period | $ per kg | $ per lb | | --- | --- | --- | | 2023 (year) | $4.54 | $2.06 | | 2024 (year) | $5.62 | $2.55 | | 2025 (year) | $8.47 | $3.84 | | Oct to Dec 2025 | $8.78 | $3.98 | | Jan to Mar 2026 | $7.49 | $3.40 | | Apr to Jun 2026 | $7.01 | $3.18 | | Jul to Sep 2026 | $7.71 | $3.50 | | September 2026 (month) | $7.26 | $3.29 | *Arabica green coffee, World Bank Pink Sheet (the International Coffee Organization's indicator price for other mild arabicas). Dollars per kilogram are the Pink Sheet's averages; dollars per pound are worked out from them.* So 2025 averaged 51% above 2024. The highest of the quarters shown is October to December 2025, and the July to September 2026 quarter was 4% below the same quarter a year earlier but still 37% above 2024's average. The International Coffee Organization (ICO) publishes the daily prices behind this. Its [monthly price sheets](https://ico.org/resources/public-market-information/) give the composite indicator in US cents per pound: 330.44 for November 2025, 248.90 for June 2026, 287.29 for August and 262.18 for September. That's a 25% fall from November to June, then a 15% bounce by August and a 9% dip in September. USDA's [Foreign Agricultural Service](https://apps.fas.usda.gov/psdonline/circulars/2026/07/coffee.pdf) said in its July 2026 report that prices had dropped 25% in the last seven months as more supply became available. That same report describes a forecast crop, not a forecast price: a record world harvest of 189.7 million 60-kilogram bags in 2026/27 (July 2026 to June 2027), led by Brazil, Ethiopia, Uganda and Vietnam. Nobody can tell you what your roaster will charge next quarter, and this post doesn't try. ### Coffee on the shelf The Bureau of Labor Statistics' [average price series for ground roast coffee](https://data.bls.gov/timeseries/APU0000717311) is what US shoppers paid per pound, averaged across cities. Its latest month at the time of writing is August 2026. | Month | Price per lb | Change from a year before | | --- | --- | --- | | August 2024 | $6.31 | — | | August 2025 | $8.87 | +40.6% | | December 2025 | $9.05 | — | | April 2026 | $9.72 | series high since 2016 | | August 2026 | $9.30 | +4.8% | *Coffee, 100% ground roast, per pound, U.S. city average, BLS (not seasonally adjusted). The percentage changes are worked out from these prices.* Two things stand out. First, the shelf price kept rising for months after green coffee's October to December 2025 quarter: it hit its high in April 2026, then eased 4.4% by August. USDA's Economic Research Service notes that retail food prices typically lag farm and wholesale prices, which fits. Second, coffee is still far above where it was two years earlier. ERS's [Food Price Outlook](https://www.ers.usda.gov/data-products/food-price-outlook/summary-findings) of September 25, 2026 puts retail coffee prices 6.1% above August 2025. That's a different measure from BLS's average price, so it differs from the 4.8%. It adds that coffee has risen faster than other nonalcoholic drinks, which as a group, coffee included, were up 3.7% over the same year. ### What none of this tells you Your roaster's bag price also pays for roasting, packaging, freight and the roaster's own margin. We found no published source for how much of it is the green bean, so we don't guess. Your roaster may also buy at different times from the dates in the tables, so your price can move earlier or later than they do. Your invoice is the only number that counts for your menu. ## Cost one cup from your bag price You need four numbers: the price of a gram of coffee, the grams in each drink, the milk, and the cup. Weigh the grams on a scale for a week instead of guessing, and write down what you get. *Plate cost of one drink* Cost per gram = bag price ÷ (pounds in the bag × 453.6)Coffee in the cup = grams per drink × cost per gramPlate cost = coffee + milk + cup, lid, sleeve + extrasFood and drink cost % = plate cost ÷ menu price × 100 A pound is 453.6 grams, as BLS lists it. Milk per ounce = price per gallon ÷ 128. Add anything else that goes in the cup: syrup, sugar, a stir stick. Leave out labor, rent and the machine. Example: Corner Café (a made-up café) buys a 5 lb bag of espresso blend for $70.00, which is $14.00 a pound. It pays $4.50 a gallon for milk, and a cup, lid and sleeve cost $0.30. The doses below are examples too: a latte uses 18 g of coffee and 9 oz of milk, and a 12 oz drip coffee uses 22 g. The bag holds 5 × 453.6 = 2,268 g, so a gram costs $70.00 ÷ 2,268 = $0.0309. *Example: Corner Café's two drinks* - Latte: 18 g × $0.0309: $0.56 - + 9 oz milk × $0.035: $0.32 - + Cup, lid, sleeve: $0.30 - = Latte plate cost: $1.18 - Latte food and drink cost: $1.18 ÷ $5.00: 23.6% - Drip: 22 g × $0.0309: $0.68 - + Cup, lid, sleeve, sugar, creamer: $0.35 - = Drip plate cost: $1.03 **Drip cost: $1.03 ÷ $3.00:** 34.3% The latte leaves $3.82 and the drip $1.97. The drip's cost % is higher because its low price has to cover the same cup and 22 g of coffee. Run your own drinks through the plate cost calculator, and keep one number for each: the grams you really use. [Free calculatorCost a drink: plate cost and menu price calculator](https://restaurantdoctorai.com/calculators/recipe-cost) ## What a rise in the bag price does Say Corner Café's roaster raises the 5 lb bag 15%, to $80.50 (an example, not a forecast). A gram now costs $0.0355. | Drink | Coffee in cup | Plate cost | Cost % | | --- | --- | --- | --- | | Latte, bag at $70.00 | $0.56 | $1.18 | 23.6% | | Latte, bag at $80.50 | $0.64 | $1.26 | 25.2% | | Drip, bag at $70.00 | $0.68 | $1.03 | 34.3% | | Drip, bag at $80.50 | $0.78 | $1.13 | 37.7% | *Corner Café's two drinks before and after a 15% rise in the bag price, with menu prices unchanged. All figures are examples.* A 15% rise in the bag adds $0.08 to a latte and $0.10 to a drip. That's small per cup, so look at the whole month. Say the café sells 3,000 lattes at $5.00, 2,500 drip coffees at $3.00, and $7,500 of other items that use no coffee, at a 35% cost. Before the rise it takes in $30,000 and spends $8,740 on food and drink: 29.1%. After the rise, with the same sales and prices, it spends $9,230: 30.8%. That's $490 a month, $5,880 a year, and 1.6 points of food and drink cost. The same arithmetic works downward. If your roaster cuts your price, re-cost and see where each drink sits. You may hold your prices and take the difference as margin rather than cut: that's your decision, not a rule. ## Four ways to close the gap Start with the rise from the table: the latte is $1.18 to $1.26 on a $5.00 menu, the drip $1.03 to $1.13 on a $3.00 menu. ### Hold the price The drip then leaves $1.87 instead of $1.97, so across 2,500 cups the café gives up $250 a month. The latte gives up $240 across 3,000. Holding costs real money, so do it deliberately for a reason, such as a quiet season or a bag price your roaster has told you is coming down, not by default. ### Raise the price by the extra cost To keep dollars per cup, add the extra plate cost: the latte $5.00 + $0.08 = $5.08, so $5.10 on a menu that moves in dimes; the drip $3.00 + $0.10 = $3.10. To get a drink back to its old cost %, divide its new plate cost by the old percentage: the drip is $1.13 ÷ 0.343 = $3.29. If you price in quarters, $5.25 and $3.25 are the next steps up, which are 5% and 8% rises. For comparison, USDA reports that the food-away-from-home index was 3.4% higher in August 2026 than a year before. ### Change the recipe or the size At $0.0355 a gram, dropping the drip from 22 g to 20 g saves $0.07 a cup. Taste it first: change the dose only if the cup still holds up, and tell the person on the bar. Other levers are a smaller default size, or asking your roaster to price a different blend. The Pink Sheet has robusta, the cheaper coffee species, at $3.92 per kg for July to September 2026, about half of arabica's $7.71. What a blend with more of it tastes like is a question for your roaster and your guests, not for a table. ### Steer guests to drinks that leave more Dollars per cup = menu price − plate cost. A latte leaves $3.74 now and a drip $1.87, so each guest who orders a latte instead of a drip adds $1.87. If 100 of the 2,500 monthly drip orders become lattes, that's $187 a month, about three-quarters of the $250 the drip gives up. It helps, but it won't close the gap by itself. ## Which option to pick Decide drink by drink, by how many points its food and drink cost % has risen since you last costed it. These cut-offs are RestaurantDoctorAI's rules of thumb, not an industry standard. - **Under 1 point:** hold, write down the new cost and check again next month. - **1 to 3 points:** raise the price by at least the extra plate cost, rounded up to your usual price ending. The latte is 1.6 points over: $5.10. - **More than 3 points:** split the fix. The drip is 3.3 points over, so raise it about 3%, to $3.10, and take the 20 g dose, which makes the plate $1.06. Here is the whole month if sales hold, which they may not: 3,000 lattes at $5.10 and 2,500 drip coffees at $3.10, plus the same $7,500 of other sales. | Case | Sales | Food and drink cost | Cost % | Left | | --- | --- | --- | --- | --- | | Before the rise ($70.00 bag) | $30,000 | $8,740 | 29.1% | $21,260 | | After the rise, prices held | $30,000 | $9,230 | 30.8% | $20,770 | | After the rise, latte $5.10, drip $3.10 and 20 g | $30,550 | $9,055 | 29.6% | $21,495 | *Corner Café's month (example), at the new bag price of $80.50. 'Left' is sales minus food and drink cost, before labor, rent and everything else.* A raise pays as long as guests keep ordering about as before. Check that against your own counts, as below. The [food cost percentage guide](https://restaurantdoctorai.com/guides/food-cost-percentage) shows how to work out the combined figure for your whole café and the ranges for each type of restaurant. ## What to watch after you change a price - **Cups sold.** Compare each drink's weekly count in your POS item report with the four weeks before the change. In the example, the $3.10 drip with a 20 g dose still beats the old $3.00 drip if it keeps 97% of its sales: $1.97 × 2,500 ÷ $2.04 = 2,414 of its 2,500 cups. - **Your invoice.** If the bag price moves more than 5% from the one you costed at, either way, re-cost. - **Waste.** Dumped batches and failed shots cost coffee too. If you throw out a lot, weigh it for a week and add that share to the grams per drink. ## This week 1. Find your last roaster invoice and work out the cost per gram: bag price ÷ (pounds × 453.6). Do the same for the invoice before it, so you can see how far it has moved. 2. Weigh the coffee in your three best-selling drinks, three times each, and write down the average grams. 3. Cost those three drinks in the plate cost calculator, with milk and the cup, and write down each drink's points over its last costing. 4. Set each change by the cut-offs above, then put a date four weeks out in your calendar to check cups sold and the invoice again. 45-second profit check Coffee is one cost. Check the rest. The free check asks 12 quick questions. In about 45 seconds you get a grade for each cost area and your biggest estimated loss, so you know which cost to fix first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Coffee: World Markets and Trade](https://apps.fas.usda.gov/psdonline/circulars/2026/07/coffee.pdf), USDA Foreign Agricultural Service, July 2026 - [Commodities Price Data (The Pink Sheet)](https://thedocs.worldbank.org/en/doc/74e8be41ceb20fa0da750cda2f6b9e4e-0050012026/related/CMO-Pink-Sheet-October-2026.pdf), World Bank, October 2, 2026 - [Public Market Information: ICO Indicator Prices (I-CIP), monthly](https://ico.org/resources/public-market-information/), International Coffee Organization, September 2026 data - [Average Price Data: Coffee, 100%, ground roast, all sizes, per lb. in U.S. city average (APU0000717311)](https://data.bls.gov/timeseries/APU0000717311), U.S. Bureau of Labor Statistics, August 2026 data - [Food Price Outlook - Summary Findings](https://www.ers.usda.gov/data-products/food-price-outlook/summary-findings), USDA Economic Research Service, September 25, 2026 --- title: "Delivery app fees in 2026: what restaurants really pay" description: "What DoorDash, Uber Eats and Grubhub publish for restaurants, where cities cap fees, what an order leaves you and the app price to charge." canonical: "https://restaurantdoctorai.com/blog/delivery-app-fees-restaurants-2026" published: "2026-10-08" updated: "2026-10-08" --- # Delivery app fees in 2026: what restaurants really pay ## Key takeaways - The apps list plans from 5% to 30%; with Grubhub's delivery fee, a delivered order costs at least 15% on all three. Higher rates buy more reach. - A $30 order with $9.00 of food and $1.50 of packaging leaves $10.50 at a 30% commission and $15.00 at 15%, before labor and rent. - Divide your price by 1 minus the real rate: at 30%, a $30 order needs a $42.86 app price to bring in $30, or $45.00 with packaging. - New York City, Seattle, San Francisco and Portland cap some app fees, per their own pages, so check whether your city does before you judge your plan. The three big apps publish their plans, and they run from 5% to 30% of each order (Grubhub's 5% plan adds a delivery fee from 10%, so a delivered order costs at least 15% on every app), with the higher rates buying more reach and a lower delivery fee for your guest. What you keep is smaller than the plan name suggests. In the example below, a $30 order with $9.00 of food and $1.50 of packaging leaves $10.50 at a 30% commission, and it takes a $42.86 app price to still bring in $30 after the commission ($45.00 if you cover the packaging too). Rates vary by market, plan and contract, and the apps change them. Treat this as a map for reading your own agreement, not a replacement for it. Everything below comes from the platforms' own pricing pages and from city pages, as read on October 8, 2026. ## What the apps publish Each app lists its plans on a pricing page for US merchants. None of the three shows an effective date, so treat the rates as what each page listed on October 8, 2026. All rates are a share of the order's menu price, not of your profit. ### DoorDash The [DoorDash merchant pricing page](https://merchants.doordash.com/en-us/products/pricing) lists three plans. Delivery commission is 15% on Basic, 25% on Plus and 30% on Premier, and pickup is 6% on all three. Basic is a store listing to nearby guests. Plus adds a wider delivery radius, a lower delivery fee for the guest and access to DashPass members at no extra cost. Premier adds the widest range, an in-app Sponsored Listing and a minimum-orders refund: the page says to accept at least 20 orders a month or the commission for that month is refunded, with conditions in its footnotes. The page says the commission covers the listing, Dasher delivery, customer support and credit card processing, and that Promotions and Sponsored Listings are optional extras. It also says you can set different prices for delivery orders. ### Uber Eats The [Uber Eats merchant pricing page](https://merchants.ubereats.com/us/en/pricing/) lists a Marketplace Fee of 20% on Lite, 25% on Plus and 30% on Premium. Plus and Premium open with a 0% intro rate for 30 days. The pickup fee is 7% if you send proof that in-app pickup prices match your in-store prices, and 10% if you don't. Lite gets you found when guests search for your store. Plus adds more visibility, a lower delivery fee for the guest, and ads and offers, and Uber One orders cost an extra 5%. Premium includes Uber One at no extra cost, refunds the fees in any of the first six months you get fewer than 20 orders (conditions apply), and matches ad spending up to $100 a month. There is also a self-delivery plan at 15% (25% when you use Uber's delivery people for some orders) and a Webshop at 2.5% plus $0.29 an order, which the page calls commission-free online ordering. ### Grubhub The [Grubhub pricing and fees page](https://get.grubhub.com/grubhub-pricing-and-fees/) lists a Basic plan at a 5% marketing commission, Plus at 15% and All-access at 20%. Each has the same listing, and the higher plans come with a lower customer delivery fee. Delivery by Grubhub's drivers is a separate fee that starts at 10%, and the page says pricing in select markets may differ. If you deliver yourself, it lists no delivery fee. On the page's starting figures, then, Basic with Grubhub drivers is at least 5% plus 10%, or 15%, which is more than the headline suggests. | App | Lowest plan | Middle plan | Top plan | Pickup | | --- | --- | --- | --- | --- | | DoorDash | 15% | 25% | 30% | 6% | | Uber Eats | 20% | 25% | 30% | 7% (10% without matching prices) | | Grubhub | 5% + delivery from 10% | 15% + delivery from 10% | 20% + delivery from 10% | Not itemized on the page | *Delivery commission on each app's own pricing page, as read on October 8, 2026. Conditions, intro offers and city differences are in each page's footnotes.* Compare plans by what they include, not only by the rate. As the pages describe them, the lowest plans come with the least reach and a higher delivery fee for your guest than the plans above them. ## The other charges owners see The commission isn't always the whole bill, and the three pages are not the same on the extras. - **Card processing.** DoorDash says its commission covers it. Uber lists payment processing among what its Marketplace plans include, though its city footnotes name a separate 3% credit card processing fee for its New York City Lite plan. Grubhub's page says its pricing does not include payment processing fees and gives no rate. DoorDash also says its Online Ordering orders are commission-free but carry a payment processing fee, again with no rate. - **Marketing and promotions.** These are optional on all three, and they are where a bill can grow. DoorDash lists Promotions and Sponsored Listings. Uber lists ads and offers. Grubhub lists a monthly ad credit equal to your commission rate, with conditions in its footnote. A credit you can only use on ads is still money you have to spend to get it. - **Membership orders.** On Uber's Plus plan, Uber One orders cost an extra 5%. - **Delivery by their drivers.** On Grubhub it is a separate fee, as above. Your payout statement shows what really came out, and the page rate is only where to start. Our [delivery commission guide](https://restaurantdoctorai.com/guides/delivery-app-commissions#how-commissions-work) shows how to turn a month of statements into your real commission rate. ## Where cities cap the fees Some cities limit what an app can charge a restaurant. Uber's page says its marketplace fees “may differ depending on applicable regulations in your market”, and Grubhub's says pricing may be subject to regulatory limits depending on your jurisdiction. Below is what four city pages say. This is not legal advice, and the pages are not all current, so ask the city agency named whether your restaurant is covered. | City | What the page says is capped | Page date | | --- | --- | --- | | New York City | 15% delivery, 5% basic service, 3% payment processing, 20% enhanced service (conditions apply) | April 2026 | | Seattle | 15% for delivery services within city limits | December 2022 | | San Francisco | 15% of menu price for covered restaurants (a later exception applies) | November 15, 2022 | | Portland | 15% delivery (with an exception for small orders), 4% takeout, plus the processor's actual fee | Current city code | *What each city page says about fees the apps may charge restaurants. Read the page for the exact wording and who is covered.* ### New York City The city's Department of Consumer and Worker Protection (DCWP) says in its [fee cap flyer](https://www.nyc.gov/assets/dca/downloads/pdf/about/FeeCapFlyer-April2026.pdf) that city law caps these fees at 15% to deliver the order, 5% for a basic service that lets you receive delivery and pickup orders and be listed and searchable, 20% for an enhanced service, and 3% to process an electronic payment. The enhanced fee is allowed only if the app also offers the basic service at or under the 5% cap, and the 3% limit has one exception: an app that pays more than 3% and can show proof. DCWP's [new laws page](https://www.nyc.gov/site/dca/about/new-laws-rules.page) says Local Law 79 of 2025, effective June 30, 2025, added the enhanced service fee and requires monthly itemized fee statements. It adds that the law does not stop a restaurant from charging different prices on an app. The flyer also says DCWP is preparing a report on the caps and asked for comments by May 6, 2026, so the numbers may change. Uber's own page lists its New York City Lite plan at 15% for delivery, 5% for pickup and 3% for card processing. ### Seattle, San Francisco and Portland Seattle's Consumer Protection Division told platforms in a [December 2022 memo](https://seattle.gov/documents/Departments/FAS/RegulatoryServices/food-delivery-platforms/fas-cpd-food-delivery-platforms-cap.pdf) that the city's delivery commission cap, chapter 7.30 of its municipal code, took effect on November 30, 2022. A platform cannot charge more than 15% for delivery services within city limits, but a restaurant can agree to pay more for services beyond delivery. San Francisco's Office of Economic and Workforce Development said in a [November 15, 2022 FAQ](https://www.sf.gov/reports--november-2022--frequently-asked-questions-delivery-service-regulations) that a covered restaurant cannot be charged more than 15% of the menu price of an online order, with taxes and tips left out. It adds that from January 31, 2023 the cap does not apply to an app that offers every covered restaurant a “core delivery service” at 15% or less without requiring other paid services. Core service leaves out advertising and credit card processing. That page is more than three years old, so check what applies now. Portland's [city code, chapter 7.27](https://www.portland.gov/code/7/27) caps delivery at 15% of the purchase price per order and takeout at 4%. An app may pass on a transaction fee equal to what its payment processor charges it, and the code lets an app charge more than the caps only for extra services and only if it also offers the capped options. These are four cities. Others have rules, and most places have none. ## What you keep from one order The commission comes off the menu price, before anything else is paid. Here is one order with example numbers, so you can swap in your own. *Example: Corner Taqueria, one $30 app order at a 30% commission* - Order, at menu price: $30.00 - − App commission, 30%: $9.00 - − Food, 30% of the menu price: $9.00 - − Packaging: $1.50 **Left for labor, rent and profit:** $10.50 Example numbers, not benchmarks. The same $30 of food eaten in the dining room leaves $21.00 after the $9.00 of food, before labor and rent. Now change only the commission. Labor to cook and pack is a cost of its own, so the last column takes off $9.00 for it, the same example figure our delivery guide uses. | Real commission rate | App keeps | Left after food, packaging | Then after $9.00 labor | | --- | --- | --- | --- | | 15% | $4.50 | $15.00 | $6.00 | | 20% | $6.00 | $13.50 | $4.50 | | 25% | $7.50 | $12.00 | $3.00 | | 30% | $9.00 | $10.50 | $1.50 | *The same $30 order, with $9.00 of food and $1.50 of packaging, at different commission rates. Arithmetic on example costs.* Every 5 points of commission costs $1.50 on this order. That is why the plan name matters less than the rate on your statement. Notice, too, that at the top rate almost nothing is left for rent. ## What price on the app restores your margin To keep what the same order leaves you without the commission, divide the price by 1 minus the rate. The app takes its cut from the higher price too, so the markup is bigger than the commission. *Break-even app price* **Break-even app price = in-house price ÷ (1 − real commission rate):** With packaging = (in-house price + packaging) ÷ (1 − real commission rate) Write the rate as a decimal: 30% is 0.30, so you divide by 0.70. To win back the packaging as well, add it to the in-house price before you divide. *Example: the $30 order at a 30% commission* - In-house price plus packaging: $31.50 - ÷ (1 − 0.30): 0.70 - = App price: $45.00 - − App keeps 30%: $13.50 - − Food and packaging: $10.50 **Left, same as the dining room:** $21.00 $31.50 ÷ 0.70 = $45.00, and $45.00 × 30% = $13.50. You receive $31.50, pay $9.00 for food and $1.50 for the box, and keep $21.00, the same as the dining room order. | Real commission rate | Break-even app price | Mark up by at least | Price that also covers the $1.50 packaging | | --- | --- | --- | --- | | 15% | $35.29 | 18% | $37.06 | | 20% | $37.50 | 25% | $39.38 | | 25% | $40.00 | 34% | $42.00 | | 30% | $42.86 | 43% | $45.00 | *App prices that leave the $30 order what it leaves in the dining room. Arithmetic only: price ÷ (1 − rate), markups rounded up.* Know the limits before you raise anything. A higher price can mean fewer app orders, and the apps may set their own rules on pricing: DoorDash says its 6% pickup rate depends on pickup prices matching in-store prices, and Uber says the same for its 7% pickup fee, while DoorDash says you can set different delivery prices. San Francisco and Portland say an app cannot restrict your prices, and DCWP says New York City's law does not stop different app prices. Your own contract decides what applies to you. [Free calculatorYour break-even app price: delivery price calculatorPut in your price and the real commission rate from your payout statement to get the app price that leaves you what a dining room sale does.](https://restaurantdoctorai.com/calculators/delivery-app-price) ## This week 1. **Find your real rate for each app.** Take last month's payout statement, take sales tax and tips out, and divide what the app kept by your menu sales. The [delivery commission guide](https://restaurantdoctorai.com/guides/delivery-app-commissions#how-commissions-work) has the formula. 2. **Open your dashboard and write down your plan.** Note the delivery rate, the pickup rate, and whether card processing, ads, promotions or membership orders are charged on top. 3. **Read your agreement for three things:** any rule on the prices you may charge on the app, what you have agreed to spend on ads or promotions, and how the plan can be changed. 4. **Check whether your city caps fees.** If you are in one of the four cities above, read its page and compare it with your statement. If a charge looks wrong, ask the named agency, not the app. 5. **Re-price with your numbers.** Put your food and packaging costs and your real rate into the calculator, and decide which dishes should leave the app menu. Check the promotions you fund A buy-one-get-one on an app order is paid from your side of the payout. Run it through the same arithmetic before you agree to it. 45-second profit check Are the apps your costliest issue? The free check asks 12 quick questions, including how much of your sales come through delivery apps, and shows your biggest estimated loss in about 45 seconds. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [DoorDash Merchant Plans and Pricing](https://merchants.doordash.com/en-us/products/pricing), DoorDash - [Uber Eats Pricing for Restaurants](https://merchants.ubereats.com/us/en/pricing/), Uber Eats - [Grubhub pricing and fees](https://get.grubhub.com/grubhub-pricing-and-fees/), Grubhub - [We Want Your Feedback on the City's Fee Caps](https://www.nyc.gov/assets/dca/downloads/pdf/about/FeeCapFlyer-April2026.pdf), NYC Department of Consumer and Worker Protection, April 2026 - [New Laws & Rules](https://www.nyc.gov/site/dca/about/new-laws-rules.page), NYC Department of Consumer and Worker Protection - [Delivery Commission Caps and Expanded Delivery Radiuses](https://seattle.gov/documents/Departments/FAS/RegulatoryServices/food-delivery-platforms/fas-cpd-food-delivery-platforms-cap.pdf), City of Seattle, Consumer Protection Division, December 2022 - [Frequently Asked Questions on Delivery Service Regulations](https://www.sf.gov/reports--november-2022--frequently-asked-questions-delivery-service-regulations), City and County of San Francisco, Office of Economic and Workforce Development, November 15, 2022 - [Chapter 7.27 Regulation of Third-Party Food Platforms](https://www.portland.gov/code/7/27), City of Portland, Oregon --- title: "How much should you raise menu prices? Work it out per dish" description: "Set each dish's new price from its plate cost and your target food cost, then check how many guests you can lose before the raise costs you money." canonical: "https://restaurantdoctorai.com/blog/how-much-to-raise-menu-prices" published: "2026-10-03" updated: "2026-10-03" --- # How much should you raise menu prices? Work it out per dish ## Key takeaways - Don't pick one percentage for the whole menu: re-cost each dish on this week's invoices. - Price each dish at plate cost ÷ your target food cost %, so it rises only as far as its own costs need. - In the post's example, a 5% rise in the average check still pays as long as fewer than 7 in 100 covers stay away. Don't pick one percentage for the whole menu. Re-cost each dish on this week's invoices and price it at plate cost ÷ your target food cost %, so each dish rises only as far as its own costs need. Then check the raise against your guest count: in the example below, a 5% rise in the average check still pays as long as fewer than 7 in every 100 covers stay away. ## The national figure isn't your number Prices around you are still rising. The Bureau of Labor Statistics' consumer price index for food away from home, which tracks the prices of meals and snacks bought away from home, mostly at restaurants, rose 3.4% over the 12 months to August 2026, according to its [August 2026 release](https://www.bls.gov/news.release/archives/cpi_09112026.htm) of September 11, 2026. That's an average over every kind of place and dish, not the raise your menu needs. Big jumps also come with a warning sign. The James Beard Foundation's [2026 Independent Restaurant Industry Report](https://www.jamesbeard.org/impact/research-and-reports/2026-independent-restaurant-industry-report) surveyed more than 380 independent owners, chefs and operators from September to November 2025. When it [released the report on February 23, 2026](https://www.prnewswire.com/news-releases/the-james-beard-foundation-releases-2026-independent-restaurant-industry-report-302694350.html), the Foundation said restaurants that had raised menu prices by more than 10% were the most likely to say their profits were down and to expect fewer guests. A survey shows a link, not cause and effect, but it's a reason to go dish by dish. ## Work out each dish's raise from its plate cost A flat raise overcharges for dishes whose costs held still and undercharges for the ones whose costs jumped. Instead, re-cost each plate on this week's invoices, work out its food cost % (plate cost ÷ menu price), and compare it with your target for that dish. *The price that hits your target* **New price = Plate cost ÷ Target food cost %:** Raise % = (New price ÷ Current price − 1) × 100 Plate cost comes from the recipe, priced from this week's invoices; the current price from your menu, before tax. Enter the target as a decimal: 30% is 0.30. *Example: one burger, re-costed* - Menu price: $15.00 - Plate cost when last priced (30%): $4.50 - Plate cost on this week's invoices: $4.80 - Food cost now: $4.80 ÷ $15.00: 32% - Price for 30%: $4.80 ÷ 0.30: $16.00 **Raise: $1.00 on $15.00:** 6.7% The plate cost rose 6.7% ($0.30 on $4.50), so the price rises 6.7% too. At $16.00 the burger leaves $11.20 a plate after food, up from $10.50 before the cost rise. There's also a floor: the dollars the plate cost went up. For the burger that's $0.30, a price of $15.30, which keeps the $10.50 a plate it left before. Below the floor, a dish earns less on every plate than it did before its costs rose. Then decide: - **At or below target:** leave the price alone. Its own costs don't call for a raise. - **10% or less to reach target:** take it in one step, rounded up to a price that fits your menu. - **More than 10% to reach target:** raise at least to the floor now. Close the rest by changing the plate (portion, side, garnish or a cheaper cut), or at your next menu change. If even the floor is more than 10%, change the plate first. The 10% line follows the James Beard finding above. Plate cost only covers food: if wages rose too, check your [prime cost](https://restaurantdoctorai.com/guides/restaurant-prime-cost). For costing a plate step by step, see [plate cost in our food cost guide](https://restaurantdoctorai.com/guides/food-cost-percentage#plate-cost-for-a-single-dish). [Free calculatorWork out a dish's new price: plate cost calculatorAdd the ingredients and your target to get the plate cost, its food cost % and the price that hits the target.](https://restaurantdoctorai.com/calculators/recipe-cost) ## How many guests you can lose before a raise costs you A raise brings in more on every check, but each guest who stops coming takes their whole check with them. The raise pays as long as the share of covers you lose stays under this line: *The most covers you can lose* **Share of covers you can lose = p ÷ (c + p):** c = (Sales − Card fees − Food and drink cost − Packaging) ÷ (Sales − Card fees) p is the rise in your average check, as a decimal (5% is 0.05). Contribution is what's left of sales after card fees, food and drink, and packaging; c is contribution ÷ sales after card fees. Wages and rent stay out, because they don't change when one more guest walks in. Use one month: net sales (before tax and tips) and covers from your POS, card fees from your processing statement, food and drink cost from your P&L, packaging from your invoices. If only some dishes go up, p is smaller than their raises: it's the extra sales the new prices would have brought on last month's product mix, divided by last month's sales. Card fees come off the top because they rise with the check. If you sell through delivery apps, take their commissions off the same way. Here's why it works. Say you serve N covers, the average check brings in P after card fees, and each cover leaves C once its food, drink and packaging are paid, so c = C ÷ P. A raise of p adds p × P to every check, while the food on it costs the same. *Where p ÷ (c + p) comes from* Before: N × CAfter: N × (1 − x) × (C + p × P)Equal: (1 − x) × (C + p × P) = CSo: x = p × P ÷ (C + p × P)Divide by P: x = p ÷ (c + p) N is your covers, and x is the share of them you can lose and still take in the same contribution. *Example: a full-service restaurant, one month* - Covers: 3,000 - Net sales: $120,000 - − Card fees: $3,000 - − Food and drink cost: $38,400 - − Packaging: $600 - = Contribution: $78,000 - c: $78,000 ÷ $117,000: 0.667 - p: new prices add $6,000 ÷ $120,000: 0.05 **Most covers you can lose: 0.05 ÷ 0.717:** 7.0% (209) Check: each cover left $26.00 before the raise ($78,000 ÷ 3,000). After it, each leaves $27.95: $26.00 plus 5% of the $39.00 check after card fees. 2,791 covers × $27.95 = $78,008.45, a little over the $78,000 before. At 2,790 covers it's $77,980.50, a little under. | Rise in average check | c = 0.60 | c = 0.70 | c = 0.80 | | --- | --- | --- | --- | | 3% | 4.8% | 4.1% | 3.6% | | 5% | 7.7% | 6.7% | 5.9% | | 8% | 11.8% | 10.3% | 9.1% | | 10% | 14.3% | 12.5% | 11.1% | *The most covers you can lose before a raise costs you, as a share of covers: p ÷ (c + p), rounded to one decimal. The less each cover leaves you (a lower c), the more guests a raise can afford to lose.* Write your line down Turn the share into covers before the new prices go live: 209 a month in the example. It counts only the costs that come with each cover, so if fewer guests let you cut a shift, you can afford to lose a few more. ## Where to raise first, and what to watch afterwards Start with the dishes where the cost rise costs you the most dollars: the rise per plate times the number sold last month, from your POS product mix report. For example, a $0.30 rise on a burger that sold 600 last month costs $180 a month, while a $1.00 rise on a special that sold 100 costs $100. Fix the burger first. Once the new prices are live, pull covers and average check from your POS every week for at least 4 weeks. Compare each with the same weeks last year, or with the 4 weeks before the change if last year isn't comparable. - **Covers:** measure the change against the gap you were already running. If you were 2% behind last year before the raise and 6% behind after it, the raise cost you about 4% of covers. - **Average check:** it should rise by about p. If it rose less, guests may be trading down or skipping extras, so redo the line with the rise you actually got. - **The call:** if covers fell by less than your line, keep the new prices. If they fell by more for 4 weeks running, roll back the raises on the dishes whose sales fell most, and work on their plate cost instead. ## This week 1. **Re-cost your 10 best sellers by Friday.** Pull last month's product mix report from your POS, price each dish from this week's invoices, and flag every dish above its target food cost. 2. **Work out your line.** From last month's P&L, card processing statement and POS covers, work out c. Then work out p ÷ (c + p) for the raise you're planning, and turn it into covers a week. 3. **Set your baseline before the new menu goes live.** Write down weekly covers and average check from your POS for the last 4 weeks and the same 4 weeks last year. 45-second profit check Is food cost your costliest issue? Menu prices are one lever. The free check asks 12 quick questions, including how often you update recipe costs, and grades each cost area so you can see your biggest estimated loss, in about 45 seconds. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Consumer Price Index – August 2026](https://www.bls.gov/news.release/archives/cpi_09112026.htm), U.S. Bureau of Labor Statistics, September 11, 2026 - [2026 Independent Restaurant Industry Report](https://www.jamesbeard.org/impact/research-and-reports/2026-independent-restaurant-industry-report), James Beard Foundation, February 23, 2026 - [The James Beard Foundation Releases 2026 Independent Restaurant Industry Report](https://www.prnewswire.com/news-releases/the-james-beard-foundation-releases-2026-independent-restaurant-industry-report-302694350.html), James Beard Foundation, via PR Newswire, February 23, 2026 --- title: "Rising beef prices: what to do with your beef dishes" description: "Re-cost each beef dish at your current invoice price, then use how far its food cost rose to decide on a price rise, a smaller portion or a different cut." canonical: "https://restaurantdoctorai.com/blog/rising-beef-prices-restaurants" published: "2026-10-03" updated: "2026-10-03" --- # Rising beef prices: what to do with your beef dishes ## Key takeaways - Re-cost each beef dish at the price per pound on your recent invoices. - Under 1 point of food cost, hold the price; from 1 to 3 points, raise it by at least the extra plate cost. - Over 3 points, pair a smaller rise with a cheaper cut or blend or a smaller portion, and steer guests to dishes that leave more per plate. Re-cost each beef dish at the price per pound on your recent invoices, then see how many points its food cost % has risen since you last costed it. Under 1 point, hold the price. From 1 to 3 points, raise it by at least the extra plate cost. Over 3 points, pair a smaller price rise with a cheaper cut or blend or a smaller portion, and steer guests to dishes that leave more per plate. ## What USDA's numbers say USDA's Economic Research Service updated its [Food Price Outlook](https://www.ers.usda.gov/data-products/food-price-outlook/summary-findings) on September 25, 2026, using prices through August 2026. On beef, it reports: - Wholesale beef prices fell 4.1% from June to July 2026 and 3.0% from July to August, but were still 3.1% higher in August 2026 than in August 2025. - Grocery-store beef and veal prices were 5.9% higher in August 2026 than a year earlier. - Federally inspected beef production was about 2% lower than a year earlier, and USDA expects it to stay below year-earlier levels through the end of 2026. - USDA forecasts 2026 wholesale beef prices 8.5% above 2025 (forecast interval 3.0% to 15.0%), and grocery-store beef and veal 9.4% above (forecast interval 7.4% to 11.6%). Both forecasts compare full-year averages. They don't say what your supplier will charge next month, and neither does this post. The national figures also hide gaps between cuts. Each week, USDA's Agricultural Marketing Service reports what meatpackers get for boxed beef in its [weekly boxed beef report](https://mymarketnews.ams.usda.gov/viewReport/2461), with earlier weeks on the same page: | Week to | Ground beef, 81% lean | Choice brisket | Choice strip loin | | --- | --- | --- | --- | | Oct 2, 2026 | $342.41 | $435.24 | $966.11 | | Oct 3, 2025 | $346.24 | $434.38 | $843.28 | | Oct 4, 2024 | $293.66 | $354.57 | $680.11 | | 1-year change | −1.1% | +0.2% | +14.6% | | 2-year change | +16.6% | +22.8% | +42.1% | *Packer prices in dollars per 100 pounds (weighted averages of negotiated sales), from each date's weekly report. Brisket is IMPS 120 and strip loin IMPS 180, both boneless. Changes are worked out from these prices. Freight and your distributor's costs come on top.* Over one year, ground beef and brisket barely moved at the packer and strip loin rose 14.6%. Over two, all three rose 16.6% or more. Prices also move fast: 81% lean ground beef fell 19.1% in 14 weeks, from $423.12 per 100 pounds in the week to June 26, 2026, to $342.41 in the week to October 2. ## Re-cost each beef dish Use your own invoices, not national numbers. Average the price per pound over your last four weekly invoices, so one high or low delivery doesn't set your menu price. *Plate cost of a beef dish* **Plate cost = price per lb × portion in lb ÷ usable yield + rest of the plate:** Food cost % = plate cost ÷ menu price × 100 Portion in lb = ounces ÷ 16. Usable yield is the share of the beef you buy that ends up in the portion you weigh. Weigh burger patties raw and the yield is 100%. For example, if 10 lb of raw brisket gives 5 lb sliced, the yield is 50%, so each pound served costs twice the invoice price. Take the portion from your recipe card, or weigh a few plates as the line serves them. Take the menu price, before tax, from your POS. Then compare the result with the dish's food cost % when you last costed it. The gap, in points, decides what you do. Never costed it? Use an old invoice from when you last set its price. [Free calculatorRe-cost a beef dish: plate cost and menu price calculator](https://restaurantdoctorai.com/calculators/recipe-cost) Example: a full-service restaurant with $80,000 in monthly sales sells 400 burgers a month at $15.00, each with a 6 oz patty weighed raw and $2.10 of bun, cheese, toppings and fries. Beef was $5.60 a pound when the burger was last costed, two years ago. The last four invoices average $6.72, up 20%. *Example: the burger, re-costed* - Beef: 6 oz (0.375 lb) × $6.72: $2.52 - + Bun, cheese, toppings and fries: $2.10 - = Plate cost now: $4.62 - Food cost now: $4.62 ÷ $15.00: 30.8% - Food cost when last costed: $4.20 ÷ $15.00: 28.0% **Points over:** 2.8 When last costed, the beef was $2.10 (0.375 lb × $5.60) and the plate $4.20. Each burger now leaves $10.38 instead of $10.80: $168 a month across 400 burgers, or $2,016 a year. ## Four ways to close the gap All four start from the example: a $4.62 plate on a $15.00 menu price, a 30.8% food cost. ### Raise the price To keep your dollars per plate, add the extra cost: $15.00 + $0.42 = $15.42, a 2.8% rise. That's always the same number as the points over. At $15.50 on the menu, the burger runs 29.8% and leaves $10.88. To get back to the old 28.0%, divide the plate cost by it: $4.62 ÷ 0.28 = $16.50. That's a 10% rise, the same as the plate cost's own rise. It's also about three times the 3.4% rise in restaurant prices (the food-away-from-home index) from August 2025 to August 2026 that USDA reports. ### Trim the portion Beef at $6.72 a pound is $0.42 an ounce. To bring the plate back to $4.20, the patty gets $2.10 of beef: $2.10 ÷ $0.42 = 5 oz, a sixth less. Guests may notice, regulars first. If the menu says 6 oz, change the menu too. ### Change the cut or blend Ask your supplier to price two or three alternatives: another blend, cut or grade. As the table shows, cuts don't move together. Cost each one in the calculator, then cook and weigh a test batch: a cheaper pound that shrinks more in cooking can cost more per ounce served. Update the menu if it names the cut or grade. ### Push dishes that leave more per plate Dollars per plate = menu price − plate cost. If a chicken sandwich leaves $11.50 (an example figure), each guest who picks it over the burger adds $1.12. Steer 40 of the 400 burger orders a month to it and you add $44.80, about a quarter of the $168 gap. It helps, but it can't close the gap alone. ## Which option to pick Decide by the points the dish's food cost % rose since it was last costed. These cut-offs are RestaurantDoctorAI's rules of thumb, not an industry standard. - **Under 1 point:** hold the price, note the new cost and check again next month. - **1 to 3 points:** raise the price by at least the extra plate cost, rounded up to your usual price ending. A 1% to 3% rise is in line with the 3.4% rise in restaurant prices over the year to August 2026. - **More than 3 points:** split the fix. Raise the price about 3%, close the rest with a cut, blend or portion change, and push dishes that leave more per plate. If nothing closes it, sell the dish as a special at a price that works. If your food and drink cost % for the whole restaurant is already above the top of the target range for your type of restaurant, work toward the price that restores the old %, and let the other options cover the rest. The [food cost percentage guide](https://restaurantdoctorai.com/guides/food-cost-percentage) shows how to work out that number and the targets RestaurantDoctorAI uses. Fix first the dishes with the most at stake: extra cost per plate × plates sold a month, from your POS item sales report. The example burger is 2.8 points over: at least $0.42 more, so $15.50 on the menu. ## What to watch next month - **Price per pound.** Keep a four-week average for each beef item. If it moves more than 5% from the price you costed at, either way, re-cost. - **Each changed dish's food cost %.** Work it out at that average and compare it with the figure you set this month. More than 1 point higher? Run the cut-offs again. - **Sales mix.** Compare each beef dish's weekly sales in your POS item report with the four weeks before the change, and check total covers. In the example, the $15.50 burger beats the $15.00 one if it keeps at least 382 of its 400 monthly sales ($10.38 × 400 ÷ $10.88 = 381.6). Guests who switch dishes aren't lost. ## This week 1. Before your next beef order, pull your last four weekly beef invoices and average the price per pound for each beef item. 2. Pull last month's POS item sales report, list every beef dish and re-cost the five best sellers in the plate cost calculator. Write down each one's points over. 3. For each dish 1 point or more over, set the change by the cut-offs above, and put a date four weeks out in your calendar to re-check its food cost and sales. 45-second profit check Beef is one cost. Check the rest. The free check asks 12 quick questions, including how often you update recipe costs with current supplier prices. In about 45 seconds you get a grade for each cost area and your biggest estimated loss, so you know which cost to fix first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Food Price Outlook - Summary Findings](https://www.ers.usda.gov/data-products/food-price-outlook/summary-findings), USDA Economic Research Service, September 25, 2026 - [National Weekly Boxed Beef Cutout & Boxed Beef Cuts - Negotiated Sales (PDF) (LM_XB459)](https://mymarketnews.ams.usda.gov/viewReport/2461), USDA Agricultural Marketing Service, October 2, 2026 --- title: "Minimum wage increase 2027: what it costs your restaurant" description: "See which states and cities raise the minimum wage on January 1, 2027, and work out what it adds to your restaurant's labor cost and prices." canonical: "https://restaurantdoctorai.com/blog/minimum-wage-increase-2027-restaurants" published: "2026-10-03" updated: "2026-10-12" --- # Minimum wage increase 2027: what it costs your restaurant ## Key takeaways - As of October 12, 2026, at least 13 states had posted January 1, 2027 increases, from 35 cents an hour in Montana to $1.27 in Michigan. - Your extra cost is the hours you pay under the new rate × the raise per hour, plus payroll taxes. - If your labor percentage stays inside your target range, absorb it for now; if not, cut the hours below your sales-per-labor-hour line first. As of October 12, 2026, at least 13 states had posted minimum wage increases for January 1, 2027 on their official sites, from 35 cents an hour in Montana to $1.27 in Michigan, and so had cities such as Denver, Minneapolis and Seattle. Your extra labor cost is the hours you pay under the new rate × the raise per hour, plus payroll taxes. If your labor percentage stays inside your target range, absorb it for now; if not, cut the hours that run below your sales-per-labor-hour line first, and raise prices only for what that doesn't cover. The short version - Extra weekly cost = hours paid under the new rate × raise per hour × (1 + your payroll tax rate). - Price rise to keep your labor percentage = extra cost ÷ labor cost now. To cover just the extra dollars = extra cost ÷ net sales. - Absorb it if labor stays inside your range. If not, trim hours below your cut line first, then reprice for the rest. ## Which increases are confirmed for January 1, 2027 The table lists January 1, 2027 rates posted on official state and city pages as of October 3, 2026, with Arizona and Montana added on October 12. More states and cities announce through the fall, and some change on other dates. If your state adjusts every year and isn't listed, its 2027 rate couldn't be confirmed on its official site by October 12; look again before January. A city or county minimum can be higher than the state's: [California's labor department](https://www.dir.ca.gov/DIRNews/2026/2026-66.html), for one, says employers must pay the higher local rate where it applies. | Place | Now → Jan. 1, 2027 (raise) | | --- | --- | | [Arizona](https://www.azica.gov/labor-department) | $15.15 → $15.65 (+$0.50) | | [California](https://www.dir.ca.gov/DIRNews/2026/2026-66.html) | $16.90 → $17.40 (+$0.50) | | [Colorado](https://cdle.colorado.gov/press-releases/press-release-colorado-minimum-wage-will-increase-to-1571/hour-in-2027) | $15.16 → $15.71 (+$0.55) | | [Denver (city and county)](https://denvergov.org/Government/Agencies-Departments-Offices/Agencies-Departments-Offices-Directory/Auditors-Office/Denver-Labor/Citywide-Minimum-Wage/Minimum-Wage) | $19.29 → $19.84 (+$0.55) | | [Connecticut](https://portal.ct.gov/governor/news/press-releases/2026/08-2026/governor-lamont-announces-minimum-wage-will-increase) | $16.94 → $17.48 (+$0.54) | | [Maine](https://www.maine.gov/labor/news_events/article.shtml?id=13362373) | $15.10 → $15.70 (+$0.60) | | [Michigan](https://www.michigan.gov/leo/bureaus-agencies/ber/wage-and-hour/frequently-asked-questions/wage-and-hour/minimum-wage-and-overtime-faq) | $13.73 → $15.00 (+$1.27) | | [Minnesota](https://dli.mn.gov/news/minimum-wage-rate-adjusted-inflation-jan-1-2027) | $11.41 → $11.87 (+$0.46) | | [Minneapolis (city)](https://minimumwage.minneapolismn.gov/) | $16.37 → $17.02 (+$0.65) | | [Montana](https://erd.dli.mt.gov/labor-standards/wage-and-hour-payment-act/state-minimum-wage) | $10.85 → $11.20 (+$0.35) | | [New Jersey](https://www.nj.gov/labor/lwdhome/press/2026/20260929_min_wage.shtml) | $15.92 → $16.48 (+$0.56) | | [Ohio](https://dam.assets.ohio.gov/image/upload/com.ohio.gov/DICO/WageAndHour/2027_Minumum_Wage_Poster.pdf) | $11.00 → $11.40 (+$0.40) | | [Rhode Island](https://dlt.ri.gov/regulation-and-safety/labor-standards/minimum-wage) | $16.00 → $17.00 (+$1.00) | | [Virginia](https://doli.virginia.gov/programs/labor-law/payment-of-wage/) | $12.77 → $13.75 (+$0.98) | | [Washington](https://www.lni.wa.gov/news-events/article/26-17) | $17.13 → $17.73 (+$0.60) | | [Seattle (city)](https://www.seattle.gov/laborstandards/ordinances/minimum-wage) | $21.30 → $22.14 (+$0.84) | *Hourly rates from each place's official page (linked): the 2026 rate, the January 1, 2027 rate and the raise per hour between them. For California, Minnesota and New Jersey, the 2026 rate is from the U.S. Labor Department's [state minimum wage table](https://www.dol.gov/agencies/whd/minimum-wage/state) (updated July 1, 2026); for Arizona, from the state's [2026 minimum wage notice](https://www.azica.gov/sites/default/files/2025-10/2026%20Minimum%20Wage.pdf).* The small print that can change your number: - **New Jersey:** seasonal employers and those with fewer than six employees go from $15.23 to $16.25, a $1.02 raise. - **Ohio:** $11.40 covers businesses with yearly gross receipts of $419,000 or more, per the state's 2027 poster; smaller ones stay at the federal $7.25. - **Colorado:** the state sets a rate for each year, and its [guidance on local minimum wages](https://cdle.colorado.gov/sites/cdle/files/info_%2319_local_min_wages_11.18.24_accessible.pdf) names January 1 as the day it changes. - **Other cities:** Washington's announcement names several more cities above the state rate, and Minnesota's names Minneapolis and St. Paul. Check your city's labor office. **New York: no change on January 1, 2027.** The [state labor department](https://dol.ny.gov/minimum-wage) says a condition in the law that can stop the yearly inflation increase was met, so the minimum stays $17.00 an hour in New York City, Long Island and Westchester and $16.00 elsewhere. The governor has [proposed new rates](https://www.governor.ny.gov/new-york-state-minimum-wage-2027); they need the Legislature's approval and would start March 1, 2027. **Tipped staff.** Where a state raises its tipped cash wage on January 1, 2027 and its page gives the new rate, it's in the table below. California, Minnesota, Montana and Washington allow no tip credit, so tipped staff get the full minimum, according to the Labor Department's [tipped wage table](https://www.dol.gov/agencies/whd/state/minimum-wage/tipped) (July 1, 2026). Elsewhere, tipped rates and tip credit rules vary by state: check that table and your state labor office. | State | Tipped cash wage now → Jan. 1, 2027 (raise) | | --- | --- | | Arizona | $12.15 → $12.65 (+$0.50) | | Colorado | $12.14 → $12.69 (+$0.55) | | Maine | $7.55 → $7.85 (+$0.30) | | Michigan | $5.49 → $6.30 (+$0.81) | | New Jersey | $6.05 → $6.61 (+$0.56) | | Ohio | $5.50 → $5.70 (+$0.20) | *The hourly cash wage you pay a tipped worker, from each state's page above. If tips don't bring the worker up to the full minimum, you pay the difference.* ## What it adds to your labor cost Work it out for one week from your own payroll. Count every hour paid below the new rate, including staff who earn a little more than today's minimum. *Extra labor cost a week* Extra wages = weekly hours × (new rate − current rate), for each person under the new rate Extra labor cost = extra wages × (1 + payroll tax rate) Run tipped hours on their own line, with the tipped cash wage's raise. - **Hours and pay rates:** last pay period's payroll register, or your time clock report. - **Payroll tax rate:** employer taxes ÷ gross wages on a recent payroll report. It runs higher early in the year, when unemployment taxes start over. - **Sales and labor cost now:** net sales from your POS for the same week, and wages, payroll taxes and benefits from payroll or your P&L. *New labor % and the price change* New labor % = (labor cost now + extra cost) ÷ net sales × 100 Price rise to cover the extra cost = extra cost ÷ net sales × 100 Price rise to keep your labor % = extra cost ÷ labor cost now × 100 Both price figures assume the same guests ordering the same dishes. *Example: full-service restaurant in New Jersey, one week* - Net sales (example): $30,000 - Labor cost now, 34.5% of sales (example): $10,350 - Hours at the minimum: 200 regular, 250 tipped (example): 450 - × Raise per hour, both rates: $0.56 - = Extra wages: $252.00 - + Payroll taxes, 10% of that (assumption): $25.20 **Extra labor cost a week:** $277.20 About $14,414 a year (× 52). Labor rises from 34.5% to 35.4% of the same week's sales ($10,627.20 ÷ $30,000). Covering the extra cost takes prices just over 0.9% higher ($277.20 ÷ $30,000); keeping labor at 34.5% takes 2.7% ($277.20 ÷ $10,350). The example restaurant has six or more employees, so the $16.48 rate applies, and its tipped cash wage rises by the same 56 cents. Elsewhere the two raises can differ (in Michigan, $1.27 regular and 81 cents tipped), so run each group of hours at its own raise. Two more costs move with the minimum wage. Budget for them too: - **Staff just above the new rate.** When the minimum catches up with an experienced cook's pay, the gap between new hires and veterans shrinks. That's wage compression. The law doesn't require raises for them, but if you plan any, price them the same way and add them to the total. - **Salaried managers.** In some states the minimum salary for exemption from overtime moves with the minimum wage. From January 1, 2027 it's $72,384 a year in California and $47,101.08 in Maine; in Washington, $82,976.40 for employers with up to 50 workers and $92,196 for larger ones. Below it, a manager no longer meets the salary test for the exemption; check with your state labor office. [Free calculatorRerun your labor cost at the new ratesAdd your extra weekly pay to last week's hourly staff pay, keep the same sales, and see your new labor percentage against the range for your type of restaurant.](https://restaurantdoctorai.com/calculators/labor-cost) ## Reprice, trim hours or absorb it Decide from two of your own numbers: the new labor percentage and your sales per labor hour, hour by hour. Our [restaurant labor cost guide](https://restaurantdoctorai.com/guides/restaurant-labor-cost#labor-cost-targets) explains the ranges and the cut line. 1. **Absorb it for now** if your new labor percentage is at or under the top of the range for your type of restaurant, and recheck it with January's first four weekly payrolls. The example can't: 35.4% is over 35%, the top of the full-service targets RestaurantDoctorAI uses (30–35% of sales). 2. **Trim hours first** if you're over and some hours run below your cut line. Pull sales by hour from your POS and hours worked by hour from your time clock; every hour under the line is a candidate. Hours to cut = extra cost ÷ the hourly cost, with payroll taxes, of the shifts you'd cut. In the example: $277.20 ÷ ($16.48 × 1.10) = 15.3 hours a week of non-tipped time. A tipped hour saves only its cash wage. 3. **Reprice for the rest** if those hours can't cover it. Price rise = the cost trimming doesn't cover ÷ net sales, or up to extra cost ÷ labor cost now to get your old percentage back. In the example, just over 0.9% covers the cost with no trimming, and 2.7% brings labor back to 34.5%. Check the rules before you change pay or hours This is general information, not legal advice. Who's covered, small-employer and tipped rates, and the rules on cutting a shift short differ by state and city. Check with your state labor office (the pages linked above) or an employment lawyer first. ## This week 1. **List everyone paid under your 2027 rate.** From last pay period's payroll register, note each person's rate and weekly hours, then hours × (2027 rate − their rate). Put tipped staff on their own lines, at the tipped cash wage. 2. **Rerun your labor cost for January.** In the [labor cost calculator](https://restaurantdoctorai.com/calculators/labor-cost), enter last week's figures with the extra wages added to hourly staff pay and the extra taxes to payroll taxes, keeping the same net sales. If the new labor percentage is over the top of your range, mark the hours below your cut line before you write January's first schedule. 3. **Set the date for any price change.** If the rules above say reprice, pick the percentage (between your cover-the-cost and keep-your-labor figures) and the day it starts, and book any menu reprint by December 1. 45-second profit check Find out if labor is your costliest issue Answer 12 quick questions, including how your staffing changes when a shift slows down. You'll get a grade for each cost area, your biggest estimated loss and the first fix, free, in about 45 seconds. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [Labor Department (2027 minimum wage poster)](https://www.azica.gov/labor-department), Industrial Commission of Arizona, no date shown; checked October 8, 2026 - [New 2026 Minimum Wage, effective January 1, 2026](https://www.azica.gov/sites/default/files/2025-10/2026%20Minimum%20Wage.pdf), Industrial Commission of Arizona, September 29, 2025 - [ICYMI: California’s minimum wage set to increase to $17.40 per hour on January 1, 2027](https://www.dir.ca.gov/DIRNews/2026/2026-66.html), California Department of Industrial Relations, August 13, 2026 - [Press Release: Colorado Minimum Wage will Increase to $15.71/hour in 2027](https://cdle.colorado.gov/press-releases/press-release-colorado-minimum-wage-will-increase-to-1571/hour-in-2027), Colorado Department of Labor and Employment, August 13, 2026 - [Denver's Minimum Wage](https://denvergov.org/Government/Agencies-Departments-Offices/Agencies-Departments-Offices-Directory/Auditors-Office/Denver-Labor/Citywide-Minimum-Wage/Minimum-Wage), City and County of Denver, Denver Labor, no date shown; checked October 3, 2026 - [Governor Lamont Announces Connecticut’s Minimum Wage Will Increase to $17.48 on January 1, 2027](https://portal.ct.gov/governor/news/press-releases/2026/08-2026/governor-lamont-announces-minimum-wage-will-increase), Office of the Governor of Connecticut, August 5, 2026 - [Maine's Minimum Wage to Increase to $15.70 Per Hour in 2027](https://www.maine.gov/labor/news_events/article.shtml?id=13362373), Maine Department of Labor, September 18, 2026 - [Minimum Wage and Overtime FAQ](https://www.michigan.gov/leo/bureaus-agencies/ber/wage-and-hour/frequently-asked-questions/wage-and-hour/minimum-wage-and-overtime-faq), Michigan Department of Labor and Economic Opportunity, no date shown; checked October 3, 2026 - [Minimum-wage rate adjusted for inflation as of Jan. 1, 2027](https://dli.mn.gov/news/minimum-wage-rate-adjusted-inflation-jan-1-2027), Minnesota Department of Labor and Industry, August 19, 2026 - [Minimum Wage](https://minimumwage.minneapolismn.gov/), City of Minneapolis, no date shown; checked October 3, 2026 - [State Minimum Wage](https://erd.dli.mt.gov/labor-standards/wage-and-hour-payment-act/state-minimum-wage), Montana Department of Labor and Industry, no date shown; checked October 12, 2026 - [New Jersey’s Minimum Wage to Increase to $16.48/Hour for Most Employees on Jan. 1](https://www.nj.gov/labor/lwdhome/press/2026/20260929_min_wage.shtml), New Jersey Department of Labor and Workforce Development, September 29, 2026 - [State of Ohio 2027 Minimum Wage (poster)](https://dam.assets.ohio.gov/image/upload/com.ohio.gov/DICO/WageAndHour/2027_Minumum_Wage_Poster.pdf), Ohio Department of Commerce, revised September 30, 2026 - [Minimum Wage](https://dlt.ri.gov/regulation-and-safety/labor-standards/minimum-wage), Rhode Island Department of Labor and Training, last updated December 31, 2025 - [Payment of Wage](https://doli.virginia.gov/programs/labor-law/payment-of-wage/), Virginia Department of Labor and Industry, last updated August 21, 2026 - [Washington’s minimum wage rises to $17.73 an hour in 2027](https://www.lni.wa.gov/news-events/article/26-17), Washington State Department of Labor & Industries, September 30, 2026 - [Minimum Wage](https://www.seattle.gov/laborstandards/ordinances/minimum-wage), City of Seattle, Office of Labor Standards, no date shown; checked October 3, 2026 - [State Minimum Wage Laws](https://www.dol.gov/agencies/whd/minimum-wage/state), U.S. Department of Labor, Wage and Hour Division, updated July 1, 2026 - [INFO #19: Local Minimum Wages](https://cdle.colorado.gov/sites/cdle/files/info_%2319_local_min_wages_11.18.24_accessible.pdf), Colorado Department of Labor and Employment, Division of Labor Standards and Statistics, last updated November 18, 2024 - [Minimum Wage](https://dol.ny.gov/minimum-wage), New York State Department of Labor, no date shown; checked October 3, 2026 - [New York State Minimum Wage for 2027](https://www.governor.ny.gov/new-york-state-minimum-wage-2027), Office of the Governor of New York, no date shown; checked October 3, 2026 - [Minimum Wages for Tipped Employees](https://www.dol.gov/agencies/whd/state/minimum-wage/tipped), U.S. Department of Labor, Wage and Hour Division, table dated July 1, 2026 --- title: "Restaurant sales down? What to check before you cut" description: "Sales down on last year? Split guests from average check, line up the same weeks, find your break-even, then decide what to cut and what to keep." canonical: "https://restaurantdoctorai.com/blog/restaurant-sales-down" published: "2026-10-03" updated: "2026-10-03" --- # Restaurant sales down? What to check before you cut ## Key takeaways - Find out which number fell first: how many guests came, or what each one spent. - Check it by day, daypart and channel against the same weeks last year. - Then see how far sales sit above break-even: cut the hours guests no longer fill, and keep what brings them in. Before you cut anything, find out which number fell: how many guests came, or what each one spent. Check it by day, daypart and channel against the same weeks last year, then see how far your sales sit above break-even. The first answer tells you where to act and the second how fast: cut the hours guests no longer fill, and keep what brings them in. ## First, find which number fell Sales are your guest count times your average check. So a sales drop means fewer guests, a smaller check, or both, and each needs a different fix. Your POS has both numbers. *Your two numbers* Net sales = guests × average check Guests are covers for dine-in and orders for takeout and delivery. Use net sales, after discounts and comps. | Example, 8 weeks | Last year | This year | Change | | --- | --- | --- | --- | | Lunch guests, dine-in | 1,400 | 1,060 | −340 | | Dinner guests, dine-in | 2,600 | 2,560 | −40 | | Takeout orders | 500 | 420 | −80 | | Delivery app orders | 300 | 280 | −20 | | All guests and orders | 4,800 | 4,320 | −480 (−10%) | | Average check | $40.00 | $42.00 | +$2.00 (+5%) | | Net sales | $192,000 | $181,440 | −$10,560 (−5.5%) | *Example: a full-service restaurant's POS, the last 8 weeks against the same 8 weeks last year.* Over the 8 weeks, sales are down 5.5% but guests are down 10%: the higher check hid a guest problem nearly twice the size. And lunch lost 340 of the 480 missing guests, while dinner barely moved. ### Check 1: guests or average check? Whichever fell more is your main problem. Fewer guests means finding out who stopped coming, and when. A smaller check means guests order less, or discounts and comps grew. If you raised prices in the past year, the check should have risen about as much; if it didn't, look at drinks and desserts per guest in your product mix report. Compare checks within each daypart too: when a cheaper daypart loses guests, the overall average rises with nobody spending more. ### Check 2: which days, dayparts and channels? Split the missing guests by day, daypart and channel. If one accounts for more than half of them, start there, not across the whole week: in the example, lunch. If the loss is spread across every shift, look at what touches all of them, such as prices, ratings or a new competitor. If dine-in fell while app orders grew, some guests may have moved channel rather than left, and [each app order keeps less of the sale](https://restaurantdoctorai.com/guides/delivery-app-commissions). ### Check 3: the same weeks, with the calendar lined up Compare weeks, not months, going back 364 days so every Friday lines up with a Friday. Months don't line up: August 2026 had four Fridays and five Mondays, August 2025 five Fridays and four Mondays. Holidays move too. Labor Day fell on September 1 in 2025 and September 7 in 2026, so on a 364-day comparison each one lines up with an ordinary Monday. Judge the 8 weeks as a whole, or compare the two holiday weeks with each other. The rule: if the drop disappears once the weeks line up, there's nothing to cut. If one or two weeks explain it all, find out what happened in them first. Otherwise, 5 or more of the 8 weeks down is a trend; with 3 or 4, watch the next 2 weeks before you cut. ### Check 4: did costs move at the same time? Put food cost % and labor % for the 8 weeks next to last year's, from your invoices, payroll and POS. Labor % up while guests fell usually means the schedule didn't follow them, unless pay rates rose. A higher food cost % can mean supplier prices rose, or that you're still ordering and prepping for last year's covers: check invoice prices, then cut pars and orders to this year's counts. Neither explains why guests left, but both are fixable this week. ## How deep is the hole? Check your break-even Some costs leave with the sale: food, drink, card fees, packaging. Most stay: rent, salaries, insurance, loans and shifts scheduled the same way every week. Break-even is the sales a month needs to pay them all, and your distance above it is your room to work. *Example: the same restaurant, an average month* - Sales last year: $104,000 - Sales now: $98,280 - Costs that rise with each sale: 40% - Bills that stay the same: $54,000 - Break-even: $54,000 ÷ 0.60: $90,000 - Left over last year: $8,400 **Left over now:** $4,968 Monthly sales are weekly sales × 52 ÷ 12 ($24,000 and $22,680). Food and drink are 33% of sales and other per-sale costs 7%, leaving 60 cents of each dollar. Bills: rent and utilities $10,000, salaries $11,000, fixed hourly shifts $22,000, everything else $11,000, payroll taxes included. Left over = sales × 0.60 − $54,000. Sales fell 5.5%; what's left over after every bill fell about 41%. Each $100 of lost sales took $60 with it, because only $40 of costs left with the sale. Now compare your cushion, sales minus break-even, with the monthly drop: - Below break-even: every week costs cash. Match hours to your guests this week. - Above it by less than the drop: one more drop like it puts you under. Act on costs this week. - Above it by more than the drop: you have time to find the cause. Don't wait for next month's P&L. The example's cushion is $8,280 against a $5,720 drop: room for one more drop like it, not two. [Free calculatorWork out yours: break-even calculatorOne typical month of sales and costs gives the sales you need and your cushion.](https://restaurantdoctorai.com/calculators/break-even) ## What to cut, and what to keep ### Flex labor to the guests you actually have Shifts written for last year's guest counts don't shrink on their own. In the example, $33,000 a month of salaries and hourly pay went from 31.7% of sales to 33.6%, with no raises. The rule: if labor % is up on the same weeks last year while guests are down, rewrite next week's schedule from forecast covers, starting with the daypart that lost the most guests, and use a [sales-per-labor-hour cut line](https://restaurantdoctorai.com/guides/restaurant-labor-cost#sales-per-labor-hour) on the shift. In the example, each $1,000 a month of hours it doesn't need lowers break-even by about $1,667 in monthly sales ($1,000 ÷ 0.60). Rules on schedule changes and on pay for shifts cut short differ by state and city. If you're not sure which apply to you, ask [your state labor office](https://www.dol.gov/agencies/whd/state/contacts), or your city's labor office where your city has its own scheduling rules, before you change shifts. ### Don't cut what brings guests in Guests notice portions, quality, speed at the rush and a clean room. Cut those on the shifts that held up and a lunch problem can become a dinner problem. The rule: if a regular would see or taste the cut on a busy shift, don't make it. Cut idle hours, prep for covers that don't come and spending no guest sees. If your star rating slipped while guests fell, fix what the reviews name first. ### Fix the costs that lose money at any level of sales Some of these costs have nothing to do with the drop, such as dishes priced off old invoices or app orders at menu prices. Fixing them raises the share of each sale you keep (60 cents in the example), which lowers break-even too. Start with [the six usual places the money goes](https://restaurantdoctorai.com/guides/restaurant-busy-but-not-making-money#where-the-money-goes): they cost money when you're slow, not just when you're busy. ## Is it just you? What the national numbers say In the [National Restaurant Association's monthly survey of operators](https://restaurant.org/research-and-media/research/restaurant-economic-insights/economic-indicators/same-store-sales-and-customer-traffic/), published September 30, 2026, 51% said same-store sales in August 2026 were higher than in August 2025, and 31% said lower. Traffic was weaker: 43% said customer traffic rose over the same 12 months and 46% said it fell. August was the 18th month in the last 19 in which more operators reported lower traffic than higher. More operators saw fewer guests (46%) than saw lower sales (31%): in some restaurants, as in the example, rising checks are covering for fewer guests. Some guests are trimming extras too. In the [Association's third-quarter consumer survey](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/affordability-pressures-mount,-but-consumers-continue-to-prioritize-restaurants/), published September 17, 2026, 34% of consumers said they were ordering fewer add-ons, such as desserts and drinks. That's the backdrop, not your diagnosis. A national number can't tell you why your guests left; your own reports can. ## This week 1. Pull the last 8 weeks and the same 8 weeks last year (364 days earlier) from your POS: net sales, guests and average check, by daypart and channel. Do it before you write next week's schedule. 2. Enter last month's sales and costs in the [break-even calculator](https://restaurantdoctorai.com/calculators/break-even). If your cushion is smaller than the monthly drop, act on costs this week. 3. Write next week's schedule to forecast covers, starting with the daypart that lost the most guests. Next Monday, compare labor % with the same week last year. 45-second profit check See where the money goes Answer 12 quick questions, including how full your dining room is through the week, how you bring guests back, your Google and Yelp rating and what happens to staffing on a slow shift. You'll get a free grade and your biggest estimated loss in about 45 seconds, so you know what to fix first. [Start the free check](https://restaurantdoctorai.com/#diagnostic) ## Sources - [State Labor Offices](https://www.dol.gov/agencies/whd/state/contacts), U.S. Department of Labor, Wage and Hour Division, January 1, 2026 - [Same-store sales and customer traffic](https://restaurant.org/research-and-media/research/restaurant-economic-insights/economic-indicators/same-store-sales-and-customer-traffic/), National Restaurant Association, September 30, 2026 - [Affordability Pressures Mount, but Consumers Continue to Prioritize Restaurants](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/affordability-pressures-mount,-but-consumers-continue-to-prioritize-restaurants/), National Restaurant Association, September 17, 2026 --- title: "How AI sees your website" description: "Restaurant website check: what a basic AI reader gets from your home page beside what guests see, with your grade and first three fixes free. Full report $4.99." canonical: "https://restaurantdoctorai.com/online-check" --- # How AI sees your website Type your restaurant's name. In about 30 seconds, see what a basic AI reader gets from your home page next to what a guest sees, with your grade and the first three things to fix, free. The full report, with every check and how you compare nearby, is $4.99. Want to see what's in it first? [See an example report](https://restaurantdoctorai.com/online-check/example) - What can AI read on my site? - Does my website lose orders? - How do I compare nearby? No sign-up. It reads your public Google listing and your public website; the free part saves nothing. Uses Google Maps: by using it you agree to the [Google Maps Additional Terms](https://maps.google.com/help/terms_maps/) and [Google's Privacy Policy](https://policies.google.com/privacy). Curious what we look at? [See every check we make](https://restaurantdoctorai.com/online-check/checks). --- title: "Every check in How AI sees your website" description: "Every check in our restaurant website check, How AI sees your website: your Google listing, website and phone speed, with points and sources." canonical: "https://restaurantdoctorai.com/online-check/checks" --- # Every check in How AI sees your website Last updated November 3, 2026 Before a guest ever walks in, they've looked you up: your Google listing, your website, how fast it opens on their phone. Our check, [How AI sees your website](https://restaurantdoctorai.com/online-check), looks at what they see, in up to 49 checks. Here is every one, with the points it counts for and where its facts come from, so you can see exactly how your score is made. To see them in a finished report, [see an example report](https://restaurantdoctorai.com/online-check/example), for a made-up restaurant. ## How the score is worked out - A check that passes earns all its points, “Could be better” earns half, “Needs fixing” earns none. - Your score is the points earned out of the points of the checks we could make, out of 100. All 49 checks together count for 186 points. - A check we couldn't make shows “Couldn't check” and is left out of the score. It never counts as a fail: if we can't read something, we don't mark you down for it. - The letter: A from 90, B from 80, C from 70, D from 60, F below 60. - Some checks only apply where they make sense: a way to book a table for sit-down restaurants, a reason to order direct for sites that take orders, picture descriptions for pages with pictures, the description's length for pages that have one, and your town and food words when Google gives us your town and kind of restaurant. ## Your Google listing: 11 checks Most guests meet you here first: on Google Maps, deciding between you and the place down the street. Where the facts come from: Your Google Business Profile as Google's Places API returns it; the places nearby that Google lists as the same kind of restaurant; and where you come up in the first 20 places Google returns for two searches in your town. “Couldn't check” when: Google returns no places like yours nearby to compare with, or the Google Maps searches don't answer. | Check | Heading | Points | | --- | --- | --- | | Opening hours on Google | Profile | 8 | | Phone number on Google | Profile | 5 | | Website link on Google | Profile | 8 | | Photos on Google | Profile | 8 | | Star rating against places nearby | Reviews | 8 | | Number of reviews against places nearby | Reviews | 8 | | A description on Google | Profile | 3 | | Price range on Google | Profile | 2 | | Dine-in, takeout and delivery on Google | Profile | 3 | | “{your kind of food} in your town” on Google Maps | Google Maps ranking | 6 | | “Best {your kind of food} in your town” on Google Maps | Google Maps ranking | 6 | ## Your website: 31 checks The guests who tap through want your menu, your hours and a way to order, without hunting for them. Where the facts come from: The home page your Google listing links to, read the way a browser receives it (and the menu page it links to, when the menu isn't on the home page). We read only what's in the page itself: words, links, and the title and description written for Google. “Couldn't check” when: Your listing has no website; the site doesn't answer or answers with an error; it shows a security check (“Just a moment…”, “Access denied”, a CAPTCHA) instead of its page; or we can read only a little text on it, most often because it's drawn by scripts. Then a check of its words is “Couldn't check”, never a fail. | Check | Heading | Points | | --- | --- | --- | | Secure website (https) | Address | 4 | | Made for phones | Phones | 6 | | Menu a phone and Google can read | Menu and ordering | 10 | | Online orders on your own terms | Menu and ordering | 12 | | Tap to call | Finding and reaching you | 4 | | Title and description for Google | What Google reads | 4 | | Your own web address | Address | 3 | | Your Google link lands on your site | Address | 2 | | Google is allowed to list your page | What Google reads | 6 | | A main headline on the page | What Google reads | 2 | | Your town in your headline, title and description | What Google reads | 3 | | What you serve in your headline, title and description | What Google reads | 3 | | A description the right length | What Google reads | 1 | | Restaurant details Google can read | What Google reads | 3 | | A preview when your link is shared | When your link is shared | 2 | | A preview card for X (Twitter) | When your link is shared | 1 | | A small icon for browser tabs | When your link is shared | 1 | | Enough words about your restaurant | What your page says | 2 | | Your story | What your page says | 1 | | Answers to common questions | What your page says | 1 | | Guests' reviews on your site | What your page says | 1 | | Pictures with descriptions | What your page says | 1 | | Links to your social pages | What your page says | 1 | | Your address written on your site | Finding and reaching you | 3 | | Your hours written on your site | Finding and reaching you | 3 | | Your phone number written on your site | Finding and reaching you | 3 | | A way to reach you besides calling | Finding and reaching you | 2 | | A way to sign up for your news | Finding and reaching you | 2 | | Gift cards | Finding and reaching you | 1 | | A reason to order from you, not an app | Menu and ordering | 2 | | A way to book a table | Menu and ordering | 2 | ## Phone speed: 7 checks On a phone, a page that takes too long to show is a page many people leave before it loads. Where the facts come from: Google's own speed test (PageSpeed Insights, mobile), with Google's goal for each measure, and real Chrome visitors' speed over the last 28 days (Chrome UX Report) when Google has enough of them. “Couldn't check” when: Google's test can't load your page in time, or Google doesn't have enough real visitors' data for your site. | Check | Heading | Points | | --- | --- | --- | | Google's phone speed test | Google's speed test | 15 | | Time until the main content shows | Google's speed test | 3 | | Time until anything shows | Google's speed test | 2 | | The page holding still while it loads | Google's speed test | 2 | | Time the page is frozen to taps | Google's speed test | 2 | | How fast the page fills in | Google's speed test | 1 | | Speed for your real visitors | Real visitors | 4 | ## AI assistants: outside your grade Your result also has its own section on AI, under the same name. It starts free, with a picture of your home page on a phone beside the plain text a basic AI reader gets from it, and seven plain answers, whether your site matches your Google listing among them. Your full report adds why each answer came out as it did, one or two next steps, and two rows on ordering through assistants. There's no score for any of it, and none of it is part of your grade: it earns no points and takes none away. - Whether AI can get in - What you sell - Your prices - Your hours - Your phone and address - How guests can order or book - Does your site match your Google listing? - Ordering through an assistant - Your menu on Google We read only your home page and at most one menu page, and assistants differ, so this is what a basic reader gets. To say whether AI can get in, we read your site's robots.txt file, the rules crawlers read, and its tags, and we can't see a firewall. Your website is one source: assistants often get restaurant facts elsewhere too. Google doesn't show your menu on Google from outside, so we say what to look at. Checked November 1, 2026. Our [guide to ordering through assistants](https://restaurantdoctorai.com/blog/order-through-chatgpt-claude-google-maps) has the rest. ## What the check doesn't do - It doesn't turn your score into dollars. A score says what your listing and website show, not what you'd earn by changing them. - It doesn't sell you a website or an ordering system, so no check leans toward one. - It reads only what's public: your Google listing, your website's page and robots.txt file, and Google's speed test. It doesn't sign in to anything. - Google Maps' own list can differ a little with who is searching and from where, so a ranking is where you came up for us, at that moment. ## See your own Type your restaurant's name. In about a minute you'll see your grade and the first three things to fix, free. The full report, with each check for your place and how to fix it, is $4.99. [Run the check](https://restaurantdoctorai.com/online-check) --- title: "An AI that answers the calls you're missing right now." description: "An AI answers your restaurant's missed calls, takes pickup orders from your menu and sends them to a kitchen screen. $149 a month after 3 free days." canonical: "https://restaurantdoctorai.com/phone-ordering" --- # An AI that answers the calls you're missing right now. It picks up when your team can't, takes the order from your menu, reads it back and sends it to your kitchen, so you can stop losing orders to a ringing phone. A person is always one word away. - Keep your phone number - Guests pay at pickup - 3 days free, cancel anytime [See your dashboard before you sign up Click through it with an example restaurant: setup, calls, voice, kitchen tablet.](https://restaurantdoctorai.com/phone-ordering/setup#example) Already using it? [Sign in to your dashboard](https://restaurantdoctorai.com/phone/login) ## Phone orders that land in your kitchen, not in voicemail - **The calls you miss, answered** At rush hour, when the line's busy, after close: the AI picks up, so a hungry guest isn't sent to voicemail. - **The whole order, read back** Sizes, add-ons and notes from your menu, priced by our code, read back, and placed only after a yes. - **Straight to your kitchen** It rings on a tablet by the pass with a copy by email. Staff tap Accept, Ready, Picked up. AI assistant · Example Café On a call - Guest: Actually, can you make the latte a large? - AI: Done, a large latte. That's $1.00 more. Anything else? Order updated: Latte, Small → Large Natural conversation ## Talks like your friendliest person at the counter. Picks up day and night. Guests can change their minds mid-order; it keeps up, re-prices from your menu and reads the order back again. In your kitchen Every order rings on a tablet by the pass, with a copy by email. Safety first It never tells a guest a dish is allergy-safe. It never tells a guest a dish is safe. It connects them to your team, or flags the ticket in red. Is the banana nut muffin okay for a nut allergy? For allergies I'll get someone from the team who can check for you. One moment, I'll connect you. Connected to your staff ## It sells only what's on your menu Upload a photo or PDF and we read in every item, its price, its sizes and add-ons. You check it, mark what's sold out, and save. Every order is priced by our code from your saved menu, never by the AI, so the total the guest hears is the total on the ticket. ## What it does - Answers calls your team misses, says it's an AI, and gives your hours when you're closed - Takes pickup orders from your menu, with sizes and add-ons, and reads them back - Gives your hours and pickup wait, and says when you're closed - Hands the call to your staff when a guest says "staff" - Shows you every call's transcript and order on your dashboard - Answers your questions about your calls and orders on the dashboard: type "what sold best last week?" and Ask looks it up ## What it won't do - Tell a guest a dish is safe for an allergy: it hands the call to a person - Take a card number over the phone: guests pay when they pick up - Invent an item, a price or a deal that isn't on your menu - Call anyone, or text a guest who didn't ask - Keep recordings of voices: transcripts only, deleted after 90 days ## Is it working right now? Every 10 minutes we test the whole phone line for real: that the AI is switched on, that phone numbers reach it, that each call's conversation is sent to us, that our database answers and that it prices an order correctly. A step passes only when the answer is right. No call is placed. ## How it starts 1. Upload your menu A photo or PDF. We read it in; you check each item, its price and its choices. 2. Forward missed calls Keep your number. Calls your team can't pick up, or that come when the line's busy, go to the AI. 3. The AI takes the order From your menu only. It reads it back, gets a yes and a name, and gives the pickup time. 4. It rings in the kitchen On a tablet by the pass, with a copy by email. Staff tap Accept, Ready, Picked up. [See every setup step with an example restaurant](https://restaurantdoctorai.com/phone-ordering/setup) ## The same AI number books tables too Turn on table bookings and the AI that takes your orders also takes reservations, by your rules, so you can fill tables on the calls you miss. One account, one bill. $79 a month on top of the phone line, with 500 reminder texts a month. Optional: turn it on or off anytime. [Try the example table book](https://restaurantdoctorai.com/bookings/demo) [More about AI table bookings](https://restaurantdoctorai.com/table-bookings) - Books tables on the calls it answers: name, day, time and party size, at a table or the bar, with wheelchair access when asked - Keeps to your hours and your tables, and offers the nearest free times when the one asked for is full - Texts a reminder an hour before to guests who said yes to texts; a reply of 2 frees the table on your book - Gives your host stand a table book on any phone or tablet: tonight's bookings, walk-ins and a floor view of every table ## Questions Do I keep my phone number? Yes. You turn on call forwarding with your phone company, so missed and busy calls go to your AI number. Your staff still answer when they can, and you can turn forwarding off anytime. What if the AI gets an order wrong? It reads every order back and places it only after the guest says yes. Prices come from your saved menu, never from the AI. Every call's transcript is on your dashboard, so you can see exactly what was said. What happens with allergy questions? The AI never says a dish is safe. It offers to connect the guest to your staff. If they'd rather just order, the allergy goes on the ticket, which the kitchen screen flags in red. Can it take table bookings too? Yes, with the table bookings add-on: $79 a month on top of the phone line, on the same AI number. It books by your hours and tables, offers other times when you're full, texts guests a reminder (500 texts a month included, 500 more for $10) and gives your host stand a table book. Turn it on or off from your dashboard. You can click through an example book at /bookings/demo. Does it work with my POS? Not yet. Orders arrive on our kitchen screen (any tablet) and by email, and your staff ring them into your POS. We're working on Square first. How do guests pay? When they pick up, the way they do now. The AI never takes card numbers. What does setup involve? Six steps on your dashboard: upload a photo of your menu and check it, set your hours, add a staff number and your address, get your AI number, turn on call forwarding, then pair the kitchen tablet and place a test order. Each step checks itself, and you can stop and come back anytime. You can click through every step with an example restaurant on our setup page (/phone-ordering/setup). Can I choose the voice, or have it speak Spanish? Yes. On your dashboard you pick from six voices (press play to hear each one) and the language: English, Spanish, or both, where the AI answers in whichever language the caller speaks. The voice doesn't change the price. Who can see my orders? Only the tablets you pair. A tablet shows your orders after someone types a 6-digit code from your dashboard, which lasts 10 minutes and works once. Lost a tablet or someone left? Sign out all tablets from the dashboard and pair the ones you keep again. How do I get back to my dashboard? Sign in with the email you signed up with (restaurantdoctorai.com/phone/login). We email you a link that works once, for 30 minutes, and that device stays signed in. No password to remember. What does it cost? $149 a month after 3 free days, which start with your AI's first call, your own test call included, so set up first and call when you're ready (the trial ends at most 7 days after you sign up, with up to 30 calls; past that, calls go to your staff until it ends). One free trial per email. It covers about 300 calls a month. No contract, no setup fee; cancel on your dashboard and it stops at the end of the period you've paid for. ## Stop missing the calls at rush hour $149 a month after 3 free days from your first call (up to 30 calls), about 300 calls a month included. No contract, no setup fee, cancel anytime. Table bookings on the same number: $79 a month more, optional. --- title: "Your AI phone line books tables too" description: "The AI that answers your missed calls can book tables too, from your own table book, and texts a reminder to guests who say yes. $79 a month add-on." canonical: "https://restaurantdoctorai.com/table-bookings" --- # Your AI phone line books tables too The AI that answers your missed calls also books tables: from your own table book, by your hours and your rules, so you can fill tables on the calls nobody picks up. Your host stand sees every booking on a phone or tablet. $79 a month on top of the AI phone line, with 500 reminder texts a month. [Try it with your own tables](https://restaurantdoctorai.com/bookings/demo?setup=1) [See the AI phone line](https://restaurantdoctorai.com/phone-ordering) Already have the AI phone line? [Add it from your bookings app](https://restaurantdoctorai.com/bookings?tab=settings). Put in your hours and tables and see your own floor and tonight's book, with made-up guests. It stays in your browser, and if you sign up they carry over, for you to check before they're saved. - Same AI number as your orders - A flat monthly price, no fee per guest - Your guest book stays yours 1. AI assistant Hi, you've reached Casa Verde. I'm their automated AI assistant, and this call is recorded and transcribed. Say "staff" anytime to reach a person. What can I get for you? 2. Caller Do you have a table for four on Friday at 7:30? 3. AI assistant 7:30 on Friday is full. I can do 7 o'clock or 8:15. Which works for you? 4. Caller 8:15, please. It's Maria. 5. AI assistant Can I text you a confirmation now and a reminder an hour before, at the number you're calling from? You can reply STOP anytime. 6. Caller Yes, please. 7. AI assistant You're booked, Maria: Friday at 8:15 PM, a table for four. Your confirmation is on its way. *An example call. The AI's words vary; its rules don't.* ## How a booking call goes 1. It checks your real book Every time is checked against your tables, hours, turn times and the bookings already in, during the call. It never offers a table that isn't free. 2. Full? It offers the nearest times The nearest free times that day or another, and the bar for a party of two when you've turned that on. 3. Big parties go to your team Above the party size you set, and for private events, set menus and deposits, it connects the caller to your staff. 4. Read back, then texted It reads the booking back. With the guest's yes, it texts a confirmation and a reminder an hour before; a reply of 2 frees the table. ## Your host stand sees it all The same table book on any phone or tablet, installed from the browser: no app store, no new system to learn. - Tonight's book Who's coming, who's late, who replied to their reminder, on any phone or tablet. - Every table at a glance Free, in use, held for a booking or over its usual time, with the guest's name on it. - Walk-ins and big parties Seat a walk-in in two taps, or put a big party on tables pushed together. - A gentle table check A table seated past its usual time gets one quiet question: still eating, or free? - Ask about your book Type a question in plain words. Ask reads your book; it can't change it, so check anything important. - Every change has a name Each booking keeps its history: what changed, when, and who did it, the AI, the guest by text, or your team. [Click through the example table book](https://restaurantdoctorai.com/bookings/demo) ## What it does - Says it's an AI and that the call is recorded, before anything else - Books only tables your book shows free, checked during the call - Reads every booking back before it's final - Hands the call to your staff when a guest says "staff" - Keeps to your hours, closed days, turn times and how many guests you seat at once ## What it won't do - Invent a free table, or book a time you're closed - Book a party above your phone limit, a private event or anything needing a deposit: that's your team's call - Text a guest who didn't say yes - Send a promo without a texted YES, or before 9 AM or after 8 PM - Take a card number or a deposit by phone ## Your guests' details, handled with care - Your guest book is yours: export it as a spreadsheet anytime. - Delete a guest when they ask. Their "stop texts" is kept, so they're never texted again. - Guest cards go 18 months after the last visit; notes about allergies or access needs go 90 days after the booking. - Every text names your restaurant and how to stop. A guest who replies STOP gets no more texts. - Questions typed into Ask go to Google's Gemini with names and notes, never full phone numbers, and aren't used to train it. - We never sell guests' details. ## Questions Do guests know they're talking to an AI? Yes. Every call starts with it: "Hi, you've reached your restaurant. I'm their automated AI assistant, and this call is recorded and transcribed. Say "staff" anytime to reach a person. What can I get for you?" A guest who says "staff" is connected to your team. Does it work with OpenTable or Resy? It doesn't connect to them. It keeps its own table book, the one your host stand uses, so a booking by phone and a walk-in can never take the same table. If you also take bookings somewhere else, the settings ask about it and warn you about double bookings. What about big parties and special requests? You set the biggest party the AI books. Bigger parties, private events, set menus, deposits and complaints go to your staff, with the caller's number on your screen if nobody picks up. Can my team still take bookings by hand? Yes. Add a booking from a phone call, seat a walk-in, block a time for a private event, all in the same book the AI uses. What does it cost? $79 a month on top of the AI phone line ($149 a month), with 500 reminder texts a month included; 500 more for $10 if you need them. One account, one bill. Turn it on or off anytime. How do I turn it on? On your phone line's dashboard, open Bookings and switch it on. Then set your tables and booking hours, and the AI starts booking on the calls it answers. If you set up your tables in the example on the same device, your table book offers to add them for you to check. ## Book tables on the calls you miss $79 a month on top of the AI phone line, with 500 reminder texts. One account, one bill, one AI number. [Try the example table book](https://restaurantdoctorai.com/bookings/demo) [Start the AI phone line](https://restaurantdoctorai.com/phone-ordering#price) --- title: "See what your guests really say, and what to fix first" description: "We read your restaurant's newest Google reviews: the stars, what guests praise and complain about, and what to do about each complaint. $6.99, paid once." canonical: "https://restaurantdoctorai.com/reviews" --- # See what your guests really say, and what to fix first We read your restaurant's newest Google reviews and show the stars, what guests praise and complain about in their own words, and what to do about each complaint, from official rules and our own tools, so you can fix what guests complain about most. 1. Pick your restaurant Find it by name, as on Google, and see its rating beside the places nearby. Then approve that we read its newest Google reviews, up to 100. 2. Pay $6.99 once Through Stripe. No account, nothing renews, and a 14-day refund for any reason. 3. See what to do In about two minutes: the stars, the chart, the places nearby side by side, replies ready to post, and what to do about each complaint, step by step. Kept 90 days under a private link we email you. ## What Review Watch sends you Every week we look at your 20 newest Google reviews. When a 1- or 2-star one comes in and you haven't replied, you get an email like this one, with a reply drafted, so you can answer it the same day. Every month you also get a link to a new reading that opens on what changed since the last one. Example email · from a made-up café A new unhappy review came in “Ordered on DoorDash and the drinks showed up half spilled in the bag.” A reply you could post: Thank you for letting us know about your order. We are sorry your drinks arrived spilled and your muffin was missing. [what you've changed] Please reach out to us directly at [your phone or email]. ## Where the reviews come from Before you pay, the rating, the places nearby and Google's review summary come straight from Google Maps while the page is open, and we don't keep them. After you pay, we fetch the restaurant's newest public Google reviews through Outscraper, a review data company, and send them to Google's Gemini AI to be read and to TypeSafe for a second opinion on the writing; we don't store the reviews. We keep the reading (the counts, the summary, the rules and a few short quotes) and your email for 90 days, then delete them. See the [privacy policy](https://restaurantdoctorai.com/privacy). Review Watch, $9.99 a month: a full reading now and every month, and an email with a reply drafted for each new unhappy Google review. Cancel anytime from its page, and it stops at the end of the month you've paid for. --- title: "Why I build with AI, and what it should never do" description: "The founder on why RestaurantDoctorAI uses AI, the jobs it does well, the one it should never do, and how it already helps restaurant owners today." canonical: "https://restaurantdoctorai.com/ai" --- # Why I build with AI, and what it should never do AI takes the jobs nobody has a free hand for, so your team has time for what only people can do. Someone asked me what AI should never do in a restaurant. My answer was quick: cook the food. That one's yours, and your cooks'. On this site, AI picks up the phone when nobody can, reads what your guests write, and helps explain your numbers in plain words. It isn't here to replace your cooks or your servers. The decisions stay with you. [Take the free check](https://restaurantdoctorai.com/#diagnostic) 12 questions, about 45 seconds ## The jobs AI does well, and the one it should never do AI isn't better than people. It's better at a few jobs, the ones that wear people down: - **It's always awake.** It can answer the phone during the rush, when nobody has a free hand. After close, ours tells callers your hours. - **It never gets tired.** The hundredth call of the day gets the same patience as the first. - **It reads everything.** A hundred reviews, start to finish, without skimming the long ones. - **It checks every number, every time.** Not just when someone has time. - **It puts notes where your kitchen will see them.** On our phone line, a caller's allergy goes on the ticket and shows in red on the kitchen screen. Your kitchen still decides what's safe. And then there's everything only people can do: making a guest feel welcome, tasting a sauce and knowing it needs salt, reading a room, deciding what your restaurant stands for. Cooking the food. The Bureau of Labor Statistics projects restaurant cook jobs will grow 12% from 2025 to 2035[[1]](https://restaurantdoctorai.com/ai#source-1). In 2024, 69% of operators said technology will help their staff rather than replace them[[2]](https://restaurantdoctorai.com/ai#source-2). I agree with them. ## Why I went this route I don't think of AI as just a tool. I think of it as an extension of us, of each person who uses it. People close to me run restaurants, and I've seen how hard it is to tell where the money goes when you're busy every night. I wanted to build help for them, and I came to AI out of frustration. When I look at something, I tend to see what it needs in order to work. A rocket engine needs fuel to burn. A rock breaks glass because of what it's made of. Most days I can picture how a thing should be built before I start. That's a blessing and a weight, because seeing it isn't the same as building it. For years I looked for help. I hired freelancers and I worked with other people. It was hard to find someone who saw the idea the way I did. AI changed that. For the first time, the only limit on what I can build is my imagination and how hard I'm willing to push. For the first time in my life, I feel free when it comes to making things. That's what made me sure. ## What I believe I believe AI can do for a small restaurant what a bookkeeper and a front desk would do, if you could afford both: watch the numbers, answer the phone, read what guests say, and never get tired of it. That matters, because running a restaurant is hard right now. In 2025, 42% of restaurant operators said their restaurant wasn't profitable[[3]](https://restaurantdoctorai.com/ai#source-3). In August 2026, people who work in restaurants and hotels quit at nearly twice the rate of workers overall[[4]](https://restaurantdoctorai.com/ai#source-4). And 89% of consumers expect business owners to respond to reviews[[5]](https://restaurantdoctorai.com/ai#source-5), on top of everything else you already do in a day. Here is what I believe the next 5 to 10 years can bring: owners who can see where their money goes, and why. Money lost to avoidable costs, spotted in the same week instead of at the end of the month. Help a single restaurant can afford, not only the big chains. Less of your team's day spent on the jobs nobody wants, and more of it with guests. These are my beliefs, not promises about your restaurant. They're the reason I built this site. ## Where AI gets it wrong, and how I keep it in check AI also makes mistakes, and I'd rather say so than have you find out. Google itself warns that AI like its Gemini apps can “hallucinate” and present inaccurate information as fact[[6]](https://restaurantdoctorai.com/ai#source-6). When reporters tested New York City's own business chatbot, it said employers could take a cut of their workers' tips[[7]](https://restaurantdoctorai.com/ai#source-7). New York law forbids that[[8]](https://restaurantdoctorai.com/ai#source-8). So I keep AI on a short leash here. It never sets a dollar figure in your report or a price on a phone order. On the phone line, the person a caller reaches is one of your own staff, on the number you choose. Nothing is posted in your name: the review reader drafts replies, and you post them yourself. The decisions stay with you and your team. ## How AI already helps on this site - **The free check uses no AI at all, on purpose.** Your grade and every dollar figure come from fixed formulas, so the same answers always give the same numbers. Our [Methodology page](https://restaurantdoctorai.com/methodology) shows how each one is worked out. - **In the full report, AI helps write the words around your numbers.** Google's Gemini drafts the plain-English explanations, and our software checks that every figure matches our own math. Then a second AI, from a separate company called TypeSafe, reads the text before it's sent. The full report is paid, and you see its price before you pay anything. - **The AI phone line answers the calls your team misses.** It says it's an AI, takes pickup orders from your menu, reads each one back and places it only after the guest says yes. Prices come from your menu, never from the AI. A guest who says “staff” goes to your team, and it never tells anyone a dish is safe for an allergy. In English, Spanish or both. It keeps written transcripts only, never a recording of a voice, and deletes them after 90 days. So you can stay with the guests in front of you when the phone rings at the rush. [See the AI phone line](https://restaurantdoctorai.com/phone-ordering) - **The review reader reads your newest Google reviews, up to 100,** and shows what guests praise and complain about. It drafts a reply to each, and you read it, change it and post it yourself. So you can decide which complaint to fix first, and answer without giving up your night. [See the review reader](https://restaurantdoctorai.com/reviews) - **Our AI restaurant expert answers your questions** about food cost, labor, tips and state rules, from sourced guides and our own calculators. So you can get a plain-English answer when the question comes up, even after close. [Ask a question](https://restaurantdoctorai.com/ask) - **The plate cost calculator can read a recipe from a photo or PDF** and fill in the ingredients for you to check, so you can price a dish without typing it all in. - **AI table bookings for the phone line.** The same AI books tables from your own table book, by your hours and rules. [See AI table bookings](https://restaurantdoctorai.com/table-bookings) - **Our free calculators work inside AI assistants, too.** They're listed in the official MCP Registry, a public list of tools for AI assistants. Once you add our [free connector](https://restaurantdoctorai.com/ai-assistants), the assistant you already use can run our numbers for you. ## Why I believe this is just the beginning Three things tell me we're at the very start. AI has become much cheaper, fast. Running AI at the level of the first ChatGPT got more than 280 times cheaper between November 2022 and October 2024[[9]](https://restaurantdoctorai.com/ai#source-9). To me, that means help that once only big chains could pay for can now reach a single restaurant. It's spreading faster than anything I've seen. Stanford's AI Index found that generative AI reached 53% adoption within three years, faster than the personal computer or the internet. The US is lower, at about 28%[[10]](https://restaurantdoctorai.com/ai#source-10). And restaurants have barely started. About a quarter of operators (26%) say they use AI tools, and only 6% use it to take orders[[11]](https://restaurantdoctorai.com/ai#source-11). So if you haven't tried any of it yet, you're in good company. History says the big gains come later. Electric power arrived in the early 1880s. American factories didn't get much more productive from it until the 1920s, once owners rebuilt how the work was done around it[[12]](https://restaurantdoctorai.com/ai#source-12). I believe AI will follow the same path: a little help at first, then whole new ways of running a restaurant. I believe this is a once-in-a-lifetime moment. I've never been this sure of anything I've built with. ## Thank you I'm grateful for this technology. It let me build something I could only see in my head before. And it lets me put help in front of owners who never had a finance team, or anyone to pick up the calls they miss. If I'm honest, one thing about AI makes me sad. I won't be here in 50 or 100 years to see where it goes. I'd love to see what restaurants look like then. I'll settle for helping build the start of it. ## Everything we do, we do with you in mind. If you'd like to see what this looks like for your restaurant, start with the free check. It takes 12 questions and about 45 seconds, with no logins and no AI touching your numbers. [Take the free check](https://restaurantdoctorai.com/#diagnostic) [Try the AI phone line demo](https://restaurantdoctorai.com/phone-ordering) The phone line's AI never calls anyone back, and no recording of your voice is kept. Felix S. Pavel Founder, RestaurantDoctorAI ## Sources 1. [U.S. Bureau of Labor Statistics, Occupational Outlook Handbook: Cooks (restaurant cooks, 2025 to 2035)](https://www.bls.gov/ooh/food-preparation-and-serving/cooks.htm) 2. [National Restaurant Association, "Make room for restaurant automation", October 10, 2024](https://restaurant.org/education-and-resources/resource-library/make-room-for-restaurant-automation/) 3. [National Restaurant Association, State of the Restaurant Industry 2026, press release, February 12, 2026](https://www.restaurant.org/research-and-media/media/press-releases/persistent-cost-increases-and-enduring-demand-will-shape-the-restaurant-industry-in-2026) 4. [U.S. Bureau of Labor Statistics, "Quits rate 3.5 percent in accommodation and food services, August 2026", October 13, 2026](https://www.bls.gov/opub/ted/2026/quits-rate-3-5-percent-in-accommodation-and-food-services-august-2026.htm) 5. [BrightLocal, Local Consumer Review Survey 2026, February 11, 2026](https://www.brightlocal.com/research/local-consumer-review-survey/) 6. [Google, Gemini Apps Privacy Hub](https://support.google.com/gemini/answer/13594961) 7. [The Markup, "NYC's AI Chatbot Tells Businesses to Break the Law", March 29, 2024](https://themarkup.org/news/2024/03/29/nycs-ai-chatbot-tells-businesses-to-break-the-law) 8. [New York Labor Law §196-d](https://www.nysenate.gov/legislation/laws/LAB/196-D) 9. [Stanford HAI, AI Index Report 2025](https://hai.stanford.edu/ai-index/2025-ai-index-report) 10. [Stanford HAI, "Inside the AI Index: 12 Takeaways from the 2026 Report", April 13, 2026](https://hai.stanford.edu/news/inside-the-ai-index-12-takeaways-from-the-2026-report) 11. [Restaurant Dive, "NRA: Over 25% of restaurant operators use AI", February 18, 2026](https://www.restaurantdive.com/news/national-restaurant-assocation-operator-artificial-intelligence-adoption/812418/) 12. [Paul A. David, "The Dynamo and the Computer", American Economic Review, May 1990; dates from Tim Harford, Slate, June 9, 2007](https://www.slate.com/articles/arts/the_undercover_economist/2007/06/the_shock_of_the_new.html) --- title: "Use RestaurantDoctorAI in your AI assistant" description: "Add the free profit check, recipe costing and calculators to Claude or another AI assistant: one address, no sign-in, and what you send isn't stored." canonical: "https://restaurantdoctorai.com/ai-assistants" --- # Use RestaurantDoctorAI in your AI assistant Ask Claude, or another AI assistant that connects to MCP servers, to run our free restaurant profit check. It asks you the same 12 multiple-choice questions as the check on this website and gives you the same free result. It can also cost one of your recipes as a draft, which you then check on our plate cost calculator, run our ten free calculators with your own figures (among them an estimate of your FICA tip credit, a check of your card fees and what your booking app costs per guest), check your website, list your tax, payroll and permit deadlines, and look things up in our guides and blog posts, with the source of each figure. Server address `https://restaurantdoctorai.com/api/mcp` No account, sign-in or key needed. ## Is it working right now? Every 10 minutes our own server connects to this one over the internet, the way your assistant does, and runs seven real steps: it says hello, lists the tools, gets the 12 questions, runs a free check, runs a calculator on fixed figures, searches the library and runs every other tool. A step passes only when the answer is right, not just when one comes back. These tests aren't counted as use. ## What it does The server gives your assistant twenty-two tools. All of them only read. The answers aren't stored; only an anonymous daily count of checks is kept. The same goes for recipes: only a daily count of drafts is kept. And for the figures you give a calculator: only a daily count of calculator runs is kept. And for a website you have checked: only a daily count of website checks is kept. And for your deadline calendar: only a daily count of calendars is kept. And for what your assistant looks up in our library: only a daily count of searches (a list counts as one) and reads is kept. To see exactly what your assistant reads and gets back, word for word, see [how our AI tools work](https://restaurantdoctorai.com/ai-assistants/how-it-works). - `restaurantdoctor_get_questions` lists the 12 questions and the answers each one accepts. Money ranges are in euros, pounds sterling, Swiss francs or Canadian, Australian or New Zealand dollars where that's your currency, and in US dollars anywhere else. - `restaurantdoctor_run_free_check` scores your answers against RestaurantDoctorAI's targets for your type of restaurant and returns your free result. - `restaurantdoctor_cost_recipe` costs a recipe's ingredient lines as a draft, in US dollars, as described below. - `restaurantdoctor_check_website` checks your restaurant's own website as our check, [How AI sees your website](https://restaurantdoctorai.com/online-check), does: secure, made for phones, a menu phones and Google can read, online orders on your own terms or only through the delivery apps, tap to call, a title and description for Google, and Google's phone speed test. It gives your score and grade, the three things to fix first and what a basic AI reader can tell about you, and links to the full check, which adds your Google listing, every check with what to fix, the reasons and the places near you, on this website. - `restaurantdoctor_deadlines` lists your restaurant's coming tax, payroll, license and permit deadlines from your state and a few yes/no answers, as our [deadline calendar](https://restaurantdoctorai.com/deadline-calendar) does: each date with its rule, what a miss costs and the official source, what to check with your state where we haven't confirmed a date, and a link to the same calendar on this website, where you can add your own license and permit dates and get free email reminders. Reminders, not legal or tax advice. - `restaurantdoctor_search_knowledge` searches our sourced library: our [guides](https://restaurantdoctorai.com/guides) and [blog posts](https://restaurantdoctorai.com/blog), the step-by-step fixes from our paid report, RestaurantDoctorAI's targets by type of restaurant, sourced industry facts and our expert knowledge base of sourced how-tos, rules and checklists. It returns the passages that match, each with its link and source, so your assistant can answer from them and cite them. `restaurantdoctor_list_knowledge` lists what's in it, by kind, with the date of each guide and blog post, and `restaurantdoctor_read_knowledge` reads one of them in full, a long one in parts. - `restaurantdoctor_prime_cost_consult` runs our guided consult, Fix my prime cost, one step at a time from the answers your assistant has so far: your prime cost, a few more questions, then the biggest problem and the step-by-step fix to start with, with its sources. Set steps and our calculators do this, with no AI touching a figure, and it takes no dish names. A guide, not a promise of results, and not legal, tax or financial advice. - The other thirteen run our free calculators and printable sheets, as described below. Your free result shows: - Your grade, from A to F, and how many of your 12 answers need attention. - A grade for each part of the business. - Your estimated money lost to avoidable costs, every problem together, a month and a year, or a note when your answers show no major problem. - Your costliest issue and its estimated cost a month, or your biggest opportunity when nothing is costing you. - What one point of prime cost (1% of your sales) is worth a month at your sales range. - The first fix in your plan. - A link that opens the same result on this website, where you can change your answers. Estimates from 12 multiple-choice answers and RestaurantDoctorAI's targets; not financial, tax or legal advice. Your recipe draft shows: - Each ingredient line as your assistant sent it, matched to a USDA food and weighed, after trimming. - What each line costs, at the prices you give your assistant or, where you haven't, at US public prices (store averages, wholesale prices and weekly store sale ads, which run low), with their source and date. None of them is what you pay. - What a line still needs before it can be costed, such as your price for it. Nothing is costed at 0. - The cost per plate as a range, and the lines whose prices move it most. - A link that opens the draft on our plate cost calculator. There you check each line against your recipe and add your prices; once your prices cover most of the cost, it shows one plate cost and the menu price for your target. A draft from the recipe as sent, not checked against the original. It doesn't check allergens or work out nutrition. The calculator tools are our [free calculators](https://restaurantdoctorai.com/calculators), run with the figures you give your assistant, in any currency (the FICA tip credit, card fees and booking fee checks are for US restaurants, in US dollars). Each one gives the same result as the calculator on this website, says how it was worked out, and links to that calculator with your figures filled in. A figure you don't give is never counted as 0: your assistant asks you for it. - `restaurantdoctor_prime_cost` works out your prime cost (food, drink and labor as a share of your sales) for a week or a month, against the target for your type of restaurant. - `restaurantdoctor_labor_cost` works out your labor cost as a share of your sales, the staff hours next week's sales can pay for, or the sales per labor hour below which the first person goes home. - `restaurantdoctor_break_even` works out the sales you need each month, week and day open to pay every bill. - `restaurantdoctor_restaurant_budget` splits the sales you expect in a month into food and drink, labor, rent and other costs by RestaurantDoctorAI's targets for your type of restaurant, with what's left after every bill and your break-even. - `restaurantdoctor_plate_cost` works out what one plate costs from your own pack prices, or takes a plate cost you know, and the menu price for your target food cost. - `restaurantdoctor_delivery_price` works out a dish's break-even price on a delivery app, from its price on your menu and the app's real commission rate. - `restaurantdoctor_menu_calories` works out a dish's calories from its recipe, rounded the way menus in the US, Ontario, or the UK and EU show them. - `restaurantdoctor_fica_tip_credit` estimates the FICA tip credit (IRS Form 8846) your business may have generated in a year, from your state, the tips reported through your payroll and your tipped staff's hours: first the reasons the real answer may be nothing, then the estimate, what it's worth after tax and the years that may still be open, to take to your accountant. - `restaurantdoctor_card_fees` works out what a month's card fees cost you for every 100 dollars of card sales, against a fair range for your type of restaurant, with what's realistic for your processor, what to ask it and the fees to question on your statement. - `restaurantdoctor_booking_fees` works out what your booking app, such as OpenTable, Resy, Tock, Yelp Guest Manager or Toast Tables, costs you a month, a year and for each guest it books, from your plan, your invoice and your guests last month, and what share of a guest's spend that is, with the platform's published prices and the date they were read. The last three are our free printable sheets, run with the figures you give your assistant, in any currency. Each one links to its sheet on this website, to print or download. A figure you don't give is never counted as 0 here either. - `restaurantdoctor_cash_out` works out one drawer's cash close for a night: the cash it should hold, how far your count is over or short, what to deposit, the difference between your card sales and the processor's batch, and your comps, voids and discounts as a share of net sales. It sets no normal for over or short: you set your own. - `restaurantdoctor_tip_out` works out a server's tip-out to the roles that support them as a percentage of their sales or tips, or shares a tip pool by hours and points, with what the Department of Labor publishes about tips and its sources. It works out the split from your numbers; it doesn't say whether a tip-out or a tip pool is allowed. - `restaurantdoctor_par_levels` works out par levels and what to order for up to 50 items, from how much you use, the days each order has to cover and what's on hand, in whole packs where you give a pack size. Estimates from the figures you give; not financial, tax or legal advice. The calorie count is not a laboratory analysis. We don't file anything and we aren't accountants. ## Add it to Claude 1. In Claude on the web or the desktop app, open [Customize › Connectors](https://claude.ai/customize/connectors). 2. Click **+**, then **Add custom connector**. 3. Name it RestaurantDoctorAI and paste the server address above. 4. If Claude asks how people sign in, choose **No sign-in**. Then click **Add**. 5. In a chat, click **+**, open **Connectors** and turn RestaurantDoctorAI on. Or [open Claude's form with the address filled in](https://claude.ai/customize/connectors?modal=add-custom-connector&connectorName=RestaurantDoctorAI&connectorUrl=https%3A%2F%2Frestaurantdoctorai.com%2Fapi%2Fmcp), check it and click **Add**. - It works on every Claude plan. The Free plan allows one custom connector. - On a Team or Enterprise plan, an Owner adds it under **Organization settings › Connectors**, then each member clicks **Connect**. - Adding connectors in Claude's phone apps is still in beta, so add it on the web or in the desktop app. It then works in the phone apps too. ## Other AI assistants Any assistant or app that can connect to a remote MCP server can use the same address. Look for a setting to add an MCP server or connector by its address, and choose no sign-in if it asks. - Address: https://restaurantdoctorai.com/api/mcp - Transport: MCP over Streamable HTTP - Sign-in: None - Tools: restaurantdoctor_get_questions restaurantdoctor_run_free_check restaurantdoctor_cost_recipe restaurantdoctor_check_website restaurantdoctor_deadlines restaurantdoctor_search_knowledge restaurantdoctor_list_knowledge restaurantdoctor_read_knowledge restaurantdoctor_prime_cost restaurantdoctor_labor_cost restaurantdoctor_break_even restaurantdoctor_restaurant_budget restaurantdoctor_plate_cost restaurantdoctor_delivery_price restaurantdoctor_menu_calories restaurantdoctor_fica_tip_credit restaurantdoctor_card_fees restaurantdoctor_booking_fees restaurantdoctor_cash_out restaurantdoctor_tip_out restaurantdoctor_par_levels restaurantdoctor_prime_cost_consult ## Without MCP: the web API The same tools, except the website check, are also a plain web API for any app or AI tool that can send a web request. No key and no sign-up. The list gives each tool's inputs and an example; send the inputs as JSON to the tool's own address, and the answer comes back as JSON with the same text an assistant gets. - The list: GET https://restaurantdoctorai.com/api/tools - Run a tool: POST https://restaurantdoctorai.com/api/tools/{tool name} - Limit: 60 calls a minute from one address - Discovery: [Our files for AI tools, listed below](https://restaurantdoctorai.com/ai-assistants#files) Our main pages, restaurant type pages, guides, blog posts and calculator pages, their lists and our questions page also come as plain text (Markdown) to a request that asks for it with Accept: text/markdown. The main pages are home, pricing, how it works, How AI sees your website and its list of checks, the AI phone line, table bookings, the review reader, about, our methodology, templates and the three AI pages. What's sent through the API is treated as below: used only to work out the answer, never stored. The address a call comes from counts toward the limit on calls a minute, and that count is cleared out soon after. ## Files for AI tools AI tools read these files to find out what's here and how to connect. You don't need them to add us to Claude or another assistant: the server address at the top of this page is enough. - [https://restaurantdoctorai.com/llms.txt](https://restaurantdoctorai.com/llms.txt): A short guide to this site for AI tools, in plain text: what we offer, our tools and our main pages. - [https://restaurantdoctorai.com/llms-full.txt](https://restaurantdoctorai.com/llms-full.txt): That guide, then our questions page, calculators, guides, blog posts, main pages and restaurant type pages in full, as one plain text file. The home, pricing and how it works pages aren't in it: they're built around a live price, and that guide describes what they sell. - [https://restaurantdoctorai.com/openapi.json](https://restaurantdoctorai.com/openapi.json): The free web API above, in OpenAPI 3.1: every tool, its inputs and the form its answer comes in. - [https://restaurantdoctorai.com/.well-known/api-catalog](https://restaurantdoctorai.com/.well-known/api-catalog): A list of our APIs (the web API, the MCP server, the A2A connection and our targets data) and where each one is described. - [https://restaurantdoctorai.com/.well-known/mcp/server-card.json](https://restaurantdoctorai.com/.well-known/mcp/server-card.json): The free MCP server's details: its address, how it connects and that it needs no sign-in. - [https://restaurantdoctorai.com/.well-known/agent-card.json](https://restaurantdoctorai.com/.well-known/agent-card.json): How agents that use the A2A protocol (Agent2Agent) find our A2A connection: its address and what it does. - [https://restaurantdoctorai.com/api/a2a](https://restaurantdoctorai.com/api/a2a): Our A2A connection. It does one thing: search our library. Agents send it POST requests. Opening it in a browser shows a short note for people. - [https://restaurantdoctorai.com/.well-known/ucp](https://restaurantdoctorai.com/.well-known/ucp): Our profile for UCP, the Universal Commerce Protocol. It's empty on purpose: none of our tools use UCP, so there's nothing to call through it. ## Where it's listed The free server is listed in these directories, so you can find it where you already look. The date is the day we last opened each listing and saw the server on it. - [Official MCP Registry](https://registry.modelcontextprotocol.io/v0.1/servers?search=com.restaurantdoctorai%2Fprofit-check). The MCP project's own list of servers. Ours is com.restaurantdoctorai/profit-check, and other directories copy from it. Checked October 14, 2026. - [Glama](https://glama.ai/mcp/connectors/com.restaurantdoctorai/profit-check). A directory of MCP servers that checks each one is online and lists its tools. Checked October 14, 2026. - [mcpservers.org](https://mcpservers.org/servers/restaurantdoctorai-com-ai-assistants). A hand-kept directory of MCP servers. Checked October 14, 2026. - [GitHub MCP Registry](https://github.com/mcp). GitHub's own list of MCP servers, which VS Code searches too. Checked October 21, 2026. - [mcptry](https://mcptry.com/servers/com.restaurantdoctorai/profit-check). A directory that checks MCP servers really work and lets you try their tools in its playground. Checked October 28, 2026. Know a directory we should be in? Email [support@restaurantdoctorai.com](mailto:support@restaurantdoctorai.com). ## Things to ask - “Run the RestaurantDoctorAI free profit check for my restaurant.” - “I run a full-service restaurant in Canada. Ask me the RestaurantDoctorAI questions and tell me my grade.” - “Where is my bar losing money? Use the RestaurantDoctorAI profit check.” - “Cost this recipe with RestaurantDoctorAI. It makes 8 plates, and I pay 28 dollars for 5 lb of ground beef.” Then paste the recipe, or add a photo of it. - “Last week we sold 24,000, bought 7,400 of food and drink and paid 6,900 in wages. What's my prime cost? Use RestaurantDoctorAI.” - “Our burger is 14 dollars in the restaurant and the delivery app takes 30%. What should it cost on the app? Use RestaurantDoctorAI.” - “We're a restaurant in Texas and reported 600,000 dollars in tips through payroll last year. Is there a FICA tip credit we're missing? Use RestaurantDoctorAI.” - “Last month we took 80,000 dollars on cards and our bank's processor charged 2,810 in fees. Are our card fees too high? Use RestaurantDoctorAI.” - “We're on OpenTable's Core plan. Last month 900 guests booked through OpenTable and 400 on our website, and a guest spends about 45 dollars. What does a guest from OpenTable cost us? Use RestaurantDoctorAI.” ## What's sent, and what isn't - For the free check, only your 12 multiple-choice answers and your country are sent to our server. The tool doesn't ask for your name, your restaurant's name, your email or any figures. - We use the answers only to work out your free result and don't store them. We keep only an anonymous count of how many checks are run each day. - For a recipe draft, only the recipe your assistant sends and any prices you give it reach our server. We use them only to work out the draft and don't store them; we count only how many drafts are made each day. - For a calculator, only the figures your assistant sends reach our server, and for a dish, its name and ingredients; for the FICA tip credit, card fees and booking fee checks, also your state and your answers, such as how the business files its taxes, who processes your cards or which booking app you use. We use them only to work out the answer and don't store them; we count only how many calculator runs are made each day. - For a website check, only your website's address reaches our server. We read that public page (and its menu page, if it links one) and send the address to Google PageSpeed Insights for its phone speed test. We don't store the address or what the page says; we count only how many website checks are run each day. - For your deadline calendar, only your state and your answers (such as whether you have staff or serve alcohol, and how your business is set up) reach our server. We use them only to work out your dates and don't store them; we count only how many calendars are made each day. - For a library search, only the words your assistant searches for reach our server; for a list, the kind of entry and which page; and for a read, the id of the entry it reads. We use them only to find the entries and don't store them; we count only how many searches (a list counts as one) and reads are made each day. - The links back to this website carry your answers, your recipe or your figures after the “#” in the address, a part your browser doesn't send to us. The page removes them from the address bar when it opens. - Your conversation is handled by the company that provides your assistant (for example Anthropic, for Claude), under its own privacy policy. We receive only the answers and country it sends to our free check, the recipe and prices it sends to our recipe tool, the figures and answers it sends to a calculator, the website address it sends to the website check, the state and answers it sends to the deadline calendar, or the words, kind or id it sends to our library, plus the usual technical details of the request (such as the IP address it came from), which our hosting provider keeps briefly in its logs. Our [Privacy Policy](https://restaurantdoctorai.com/privacy) has the details. ## Buying through your agent The tools above never take payment, send email or create a report. A second server, at `https://restaurantdoctorai.com/api/mcp/agent`, lets an agent buy for you: - `restaurantdoctor_buy_report` prepares the full report from your 12 answers, your restaurant's name and your email, and returns a payment link. Your agent can ask for the report with your real numbers and a check-up instead (we email you the report page's link, and you add last month's numbers there within 24 hours), and can add the Restaurant Fix Kit: food cost, labor, delivery and waste workbooks for Excel, with printable PDFs. - `restaurantdoctor_buy_online_check` finds your restaurant on Google Maps and returns a payment link for a restaurant's full check (How AI sees your website). - `restaurantdoctor_buy_report_codes` returns a payment link for a pack of 10 prepaid report codes, for advisors who run the check for the restaurants they advise. Once it's paid, the codes go back to the agent and are emailed to the address it gives. A pack of 50 is bought on our [advisors page](https://restaurantdoctorai.com/advisors). - `restaurantdoctor_use_code` spends a prepaid report code on one restaurant's report. Nothing is charged until a link is paid: by an agent that pays with Stripe's machine payments, from a wallet you set up and within the limits you set, or by you, when you open the link in a browser and pay on this website's checkout. Your agent should ask you first. Our [pricing page](https://restaurantdoctorai.com/pricing#ai-agents) lists what each costs, and our [Terms](https://restaurantdoctorai.com/terms#payment) say who sells what. Questions or a tool that isn't working? Email [support@restaurantdoctorai.com](mailto:support@restaurantdoctorai.com). --- title: "How our AI tools work" description: "What an AI assistant sees when it uses RestaurantDoctorAI: how it finds our tools, the words it reads, one tool's answer start to finish, and what we keep." canonical: "https://restaurantdoctorai.com/ai-assistants/how-it-works" --- # How our AI tools work When you ask Claude or another AI assistant to use RestaurantDoctorAI, it never sees our code. It reads a few short pieces of plain text, sends the figures you gave it, and gets an answer back that shows its working. This page shows those pieces, word for word, taken from the live tools themselves. 22 free tools Each one only reads No sign-in One address, no account or key Nothing stored Only an anonymous daily count ## 1. How an assistant finds us 1. **A public directory.** Our tools are listed in the official MCP Registry, the directory AI assistants use, with one address: `https://restaurantdoctorai.com/api/mcp`. 2. **A short greeting.** When an assistant connects, our server introduces itself with a paragraph saying what each tool is for and what we keep. 3. **The tool list.** The assistant asks what tools there are and gets 22, each with a name, a description and the inputs it accepts. That's everything it knows before it acts. Here is the greeting, exactly as it reads it: The server's greeting RestaurantDoctorAI's free restaurant profit check, recipe costing, calculators, website check, deadline calendar, sourced library and guided prime cost consult. restaurantdoctor_get_questions lists the 12 questions and the values they accept, worded for the restaurant's country and currency; restaurantdoctor_run_free_check scores a full set of answers and returns the free result with a link to the same result on the website. The check takes no names, emails or figures. The answers aren't stored; only an anonymous daily count of checks is kept. restaurantdoctor_prime_cost_consult runs the guided consult "Fix my prime cost" one step at a time from the answers so far, with set questions and the calculators and no AI in between, and ends in the biggest problem at those figures, a fix from RestaurantDoctorAI's playbook and its sources; it keeps nothing between calls, and each step counts as a calculator run. restaurantdoctor_cost_recipe costs a recipe's ingredient lines as a draft, with a link where the owner checks them on the website. The recipe isn't stored; only an anonymous daily count of drafts is kept. restaurantdoctor_prime_cost, restaurantdoctor_labor_cost, restaurantdoctor_break_even, restaurantdoctor_restaurant_budget, restaurantdoctor_plate_cost, restaurantdoctor_delivery_price, restaurantdoctor_menu_calories, restaurantdoctor_fica_tip_credit, restaurantdoctor_card_fees and restaurantdoctor_booking_fees work out the website's free calculators from the owner's own figures, in any currency (the FICA tip credit, card fees and booking fee checks: a US restaurant's, in US dollars), each with a link that opens the same calculator with the figures filled in. restaurantdoctor_restaurant_budget splits a month's expected sales by RestaurantDoctorAI's targets for the type of restaurant, with what's left after every bill and break-even. restaurantdoctor_fica_tip_credit estimates a US restaurant's FICA tip credit (Form 8846) from a year's tips, with the reasons the real answer may be zero first; it files nothing and isn't tax advice. restaurantdoctor_card_fees compares a month's card fees with a dated fair range for the type of restaurant. restaurantdoctor_booking_fees works out what a reservation app costs a month, a year and per guest booked, and lists other apps' prices as information, not a recommendation. restaurantdoctor_cash_out works out a night's cash close for one drawer, with no normal of its own for over or short. restaurantdoctor_tip_out works out a server's tip-out or a tip pool; it doesn't say whether a split is allowed. restaurantdoctor_par_levels works out par levels and what to order for up to 50 items from their use, the days an order covers and what's on hand. These three link to the website's free printable sheets, not to a calculator. The figures aren't stored; only an anonymous daily count of calculator runs is kept. restaurantdoctor_check_website scores a restaurant's own website (and Google's phone speed test), says the three things to fix first, and links to the full online check on the website. The website isn't stored; only an anonymous daily count of checks is kept. restaurantdoctor_deadlines lists a US restaurant's coming tax, payroll, license and permit deadlines from its state and yes/no answers, each with its rule and official source, and links to the same calendar on the website for reminders. The answers aren't stored; only an anonymous daily count of calendars is kept. restaurantdoctor_search_knowledge searches RestaurantDoctorAI's sourced library (its guides, blog posts, playbook fixes, targets, sourced facts and expert knowledge) and returns passages with their links and sources, restaurantdoctor_list_knowledge lists its entries by kind with the guides' and posts' dates, and restaurantdoctor_read_knowledge reads one entry in full, in parts for a long one; each figure comes with its source, and RestaurantDoctorAI's targets are its own, not industry benchmarks. The query isn't stored; only an anonymous daily count of searches and reads is kept. A list counts as a search. Tools that look for files instead can start from our [list of files for AI tools](https://restaurantdoctorai.com/ai-assistants#files). ## 2. One tool, start to finish: labor cost Take the labor cost calculator. Before the assistant uses it, this is the description it reads: Works out a restaurant's labor cost from the owner's own payroll figures for one week or month, as the free labor cost calculator on restaurantdoctorai.com does, in one of its three modes: share (labor as a % of net sales, against the range for its type of restaurant, with what the points over are worth), hours (next period's labor budget and the hourly staff hours it pays for once salaries are paid) or line (the cut line: the sales per labor hour below which the first person goes home). Takes net sales, hourly pay and hours worked, plus optional salaries, payroll taxes and benefits, expected sales, a target and more, in any currency; a figure that isn't sent is never counted as 0. Returns the calculator's result, how it was worked out, and a link that opens the calculator with the same figures filled in. The figures aren't stored; only an anonymous daily count of calculator runs is kept. It then sees what it may send. Three inputs are needed: `mode`, `format` and `period`. The rest are optional, and each says where to find the number. For net sales it reads: Last period's net sales, in the owner's own currency like every amount here, from the POS: after discounts and comps, before tax and tips, plus any service charge paid out to staff. Needed for share, and for line without forecast. Say a full-service restaurant's week had $22,500 in net sales, $5,175 of hourly pay over 345 hours, $1,800 of salaries and $900 of payroll taxes and benefits (example figures). This is the answer the assistant gets back: *What the assistant receives (the link is shortened here)* ``` RestaurantDoctorAI labor cost calculator, labor cost % (full service, a week's figures, in the currency they were sent in) Labor cost is 35% of net sales. Pay, taxes and benefits: 7,875 of 22,500 in net sales. That's inside the 30–35% range for full service, right at the top. A slow week or a pay rise would push it over. Each point of labor cost (1% of your sales) is worth about 225 a week, or 11,700 a year, at these sales: about 13 hours of hourly staff. Wages alone are 31% of sales; payroll taxes and benefits add 4 points. With taxes and benefits at 12.9% of pay, an hour of hourly staff costs you 16.94, not the 15.00 of pay alone. Count pay by the days worked, not the day it was paid: a payroll run this week can pay for last week's shifts. How it was worked out: - Labor: 5,175 + 1,800 + 900 = 7,875 - Share of sales: 7,875 ÷ 22,500 × 100 = 35% - Range: 30–35% for full service - One point: 1% of 22,500 = 225 - A year: 225 × 52 = 11,700 - Wages alone: 6,975 ÷ 22,500 × 100 = 31% - Taxes and benefits on pay: 900 ÷ 6,975 × 100 = 12.9% - An hour of hourly staff: 5,175 × 7,875 ÷ 6,975 ÷ 345, rounded up = 16.94 - One point in hours: 225 ÷ 16.94 = 13.28, about 13 Open it in the calculator, with these figures filled in: https://restaurantdoctorai.com/calculators/labor-cost?utm_source=ai-assistant&utm_campaign=calculator#calc=… For the whole restaurant, the free profit check (restaurantdoctor_run_free_check) gives a grade and the costliest issue from 12 multiple-choice questions. ``` The same answer also comes as tidy data the assistant can work with. It shows how each figure was worked out, so it can be checked, and it ends with a link that opens the [labor cost calculator](https://restaurantdoctorai.com/calculators/labor-cost) on this website with the same figures filled in. ## 3. When something is missing or wrong The tool never answers with a bare “error”. It says plainly what's missing or what doesn't fit, so the assistant can ask you rather than guess a number. These are its real replies: - Figures left outMissing: the net sales expected next week and the hours hourly staff worked last week. They're the owner's own figures, with last week's figures all for the same week. A figure that isn't sent is never counted as 0. - Sales below zeroCan't be below 0: it takes a plain number from 0 up. - A type of restaurant it doesn't knowInvalid option: expected one of "quick_service"|"fast_casual"|"full_service"|"fine_dining"|"bar_lounge"|"cafe" ## 4. What we keep Nothing you send is stored: not the answers to the check, not your figures, not a recipe, not a website you check. For each kind of use we keep one anonymous count a day, such as how many calculator runs there were, with no figures and nothing that says who ran them. The server's greeting says the same. The details are in our [privacy policy](https://restaurantdoctorai.com/privacy). All 22 tools only read. None of them takes payment, sends an email or creates an account. One limit to know: the website check reads a real website and runs Google's phone speed test, so it has a daily limit shared by all assistants. When it's reached, the tool says so and points to our check, [How AI sees your website](https://restaurantdoctorai.com/online-check), on this website instead. A request is also capped in size, so a tool can't be sent more than it needs. ## 5. How we keep the answers honest - Each calculator tool runs the same calculation as its page on this website, so an assistant and the page can't give you different numbers. Our tests check the two against each other. - Every 10 minutes our own server connects to the tools the way an assistant does and checks the answers are right. You can see the result on the [AI assistants page](https://restaurantdoctorai.com/ai-assistants#status). - The descriptions only describe. None of them tells an assistant to use another tool or what to say to you; pointers to the free check come only inside an answer, after the tool has run. ## 6. All 22 tools, as an assistant sees them ### The free profit check The same 12 questions and the same free result as the check on this website. Profit check questions `restaurantdoctor_get_questions` What the assistant reads Lists the 12 multiple-choice questions of the free restaurant profit check, with the answer values each one accepts. Money ranges, delivery apps and review sites follow the restaurant's country (default US). What it can send (0 needed, 1 optional) - `country`Optional Where the restaurant is: a two-letter ISO 3166-1 code such as US, CA, AU or DE (GB for the United Kingdom). Sets the currency and local wording. Defaults to US. Run the free profit check `restaurantdoctor_run_free_check` What the assistant reads Scores a restaurant's 12 answers against RestaurantDoctorAI's targets for its type of restaurant and returns the free result the website shows: an overall grade, grades by area, the estimated money lost to avoidable costs, all of them together, a month and a year, the costliest issue or biggest opportunity with its estimated money a month, what one point of prime cost is worth, and the first fix. Includes a link that opens the same result on restaurantdoctorai.com. Takes only the 12 answers and the country, each answer one of the values its question accepts. The answers aren't stored; only an anonymous daily count of checks is kept. What it can send (12 needed, 1 optional) - `serviceModel`Needed What kind of restaurant do you run? quick_service = Quick service (Counter or kiosk, fast turnover); fast_casual = Fast casual (Order at the counter, served at the table); full_service = Full service / casual dining (Servers, moderate check); fine_dining = Fine dining (Chef-driven, higher check); bar_lounge = Bar, taproom or lounge (Drinks are about half of sales or more); cafe = Café or coffee shop (Drinks lead, with pastries and light food). - `monthlyRevenue`Needed About how much do you sell in a typical month? Amounts are in US dollars, or in the local currency where that's the euro, pound sterling, Swiss franc, or Canadian, Australian or New Zealand dollar. under_25k = Under $25,000 (Under $300K a year); 25k_60k = $25,000 – $60,000 ($300K – $720K a year); 60k_120k = $60,000 – $120,000 ($720K – $1.44M a year); 120k_200k = $120,000 – $200,000 ($1.44M – $2.4M a year); over_200k = Over $200,000 (Over $2.4M a year). - `primeCostAwareness`Needed What's your prime cost: food, drink and labor costs as a share of sales? under_58 = Under 58%; 58_to_64 = 58% – 64%; over_65 = 65% or more; not_tracked_weekly = I don't track it regularly. - `plateCostingFrequency`Needed How often do you update recipe costs with current supplier prices? live_monthly_software = Monthly or live, with software (Synced to supplier invoices); 6_to_12m_manual = Once or twice a year (In a spreadsheet); new_menu_only = Only when the menu changes; rarely_gut_feeling = Rarely (Prices are set by feel or by competitors). - `wasteTrackingMethod`Needed How do you track kitchen waste and spoilage? daily_signed_log = A daily log the chef signs off; weekly_visual_estimate = A rough estimate during weekly inventory; bulk_loss_only = Only when a big batch or case goes bad; no_formal_logging = We don't track it. - `beverageMonitoring`Needed How do you control drink pours? no_alcohol = We don't serve alcohol; measured_pour_keg_audits = Measured pours, plus weekly pour-cost checks; occasional_spot_checks = Spot checks when bottles run out early; free_pour_no_audit = Free pours, no pour-cost checks. - `laborFlexing`Needed When a shift gets slow, what happens to staffing? realtime_cuts_splh = We cut staff based on sales per labor hour; fixed_shifts_regardless = Everyone works their scheduled shift anyway; overtime_above_8_percent = Overtime is common (More than 8% of hours); chronic_understaffing_owner_covers = We're short-staffed (The owner or GM works the line). - `occupancyCostRatio`Needed What share of sales goes to rent, property costs and utilities? under_8_percent = Under 8%; 8_to_12_percent = 8% – 12%; over_12_percent = More than 12%; unsure_ratio = I'm not sure. - `deliveryDependency`Needed How much of your sales come through DoorDash, Uber Eats or similar apps? zero_direct_only = None (Dine-in and our own ordering only); under_15_with_markup = Under 15% (With higher app prices to cover fees); 15_to_35_standard_cut = 15% – 35% (At regular menu prices); over_35_heavy_drag = More than 35%. - `seatingDynamics`Needed How full is your dining room across the week? consistent_weekround = Steady all week; packed_weekends_dead_midweek = Packed on weekends, quiet midweek; strong_lunch_weak_dinner = Busy lunches, slow dinners; erratic_unpredictable = Hard to predict week to week. - `contactCapture`Needed How do you bring guests back? automated_pos_loyalty_50plus = Loyalty or email list with automatic campaigns (50+ new contacts a week); occasional_checkout_emails = We collect some emails but rarely use them; no_database_social_only = Social media and walk-ins only; paid_ads_no_list = Paid ads, but no guest list of our own. - `reputationHealth`Needed What's your average rating on Google and Yelp? 4_5_to_5_0 = 4.5 – 5.0 stars; 4_0_to_4_4 = 4.0 – 4.4 stars; 3_5_to_3_9 = 3.5 – 3.9 stars; under_3_5 = Below 3.5 stars. - `country`Optional Where the restaurant is: a two-letter ISO 3166-1 code such as US, CA, AU or DE (GB for the United Kingdom). Sets the currency and local wording. Defaults to US. ### Calculators The website's free calculators, and a draft plate cost from a recipe's ingredient lines. Cost a recipe (draft) `restaurantdoctor_cost_recipe` What the assistant reads Reads a recipe's ingredient lines as written, weighs each one with USDA FoodData Central and USDA yields, and prices it with the owner's own prices when given, else dated US public prices (BLS and USDA ERS store averages, USDA wholesale prices, and USDA weekly store-ad sale prices, which run low). Returns a draft in US dollars: each line's grams, cost and price source, what each line still needs, the cost per plate as a range, and a link that opens the draft on restaurantdoctorai.com, where the owner checks each line and adds prices for one plate cost and a menu price. The draft never gives one plate cost or a menu price, and says nothing about allergens or nutrition. The recipe isn't stored; only an anonymous daily count of drafts is kept. What it can send (1 needed, 3 optional) - `recipe`Needed The recipe as written: its name, its servings line ("Serves 6") and its ingredient lines, one a line, each amount and unit exactly as the recipe or the photo gives it. Method steps may follow under a heading such as "Instructions"; only ingredient lines are costed. The lines are read as sent, not converted, rounded or corrected: the owner checks them on the website. At most 8000 characters and 60 ingredient lines. - `servings`Optional How many plates the recipe makes, when its text doesn't say. - `source`Optional Where the recipe came from: text the owner typed or pasted (text), a photo or scan of a printed or typed page (typed-photo), a handwritten card (handwriting), or a PDF (pdf). It sets how many lines the owner confirms on the website: handwriting, every line. - `prices`Optional The owner's own prices, when they give them. Lines without one use US public prices: store averages (BLS, USDA ERS), wholesale prices (USDA) and, for items those don't cover, USDA weekly store-ad sale prices, which run low. None is the owner's cost. Tap water is costed at 0. Prime cost calculator `restaurantdoctor_prime_cost` What the assistant reads Works out a restaurant's prime cost from the owner's own figures for one week or month, as the free prime cost calculator on restaurantdoctorai.com does: food and drink used plus labor, as a share of net sales, against the target for its type of restaurant, with what the points over or under are worth, food and labor each against their own range, and which to look at first. Takes net sales, food and drink purchases and wages, plus optional payroll taxes and benefits, stock counts and the owner's unpaid hours, in any currency; a figure that isn't sent is never counted as 0. Returns the calculator's result, how it was worked out, and a link that opens the calculator with the same figures filled in. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (2 needed, 7 optional) - `format`Needed The type of restaurant, which sets the targets: prime cost, and the ranges for food and drink and for labor: quick_service = Quick service (Counter or kiosk, fast turnover); fast_casual = Fast casual (Order at the counter, served at the table); full_service = Full service / casual dining (Servers, moderate check); fine_dining = Fine dining (Chef-driven, higher check); bar_lounge = Bar, taproom or lounge (Drinks are about half of sales or more); cafe = Café or coffee shop (Drinks lead, with pastries and light food). - `period`Needed Whether every figure covers one week or one month, the same dates for all of them. Sets how a year is worked out (52 weeks or 12 months). - `sales`Optional Net sales for the period, in the owner's own currency like every figure here, from the POS sales summary: after discounts and comps, before tax and tips, with any service charge paid out to staff added back. Needed for a result, and more than 0; if missing, the reply asks for it. - `purchases`Optional Food and drink supplier invoices for the same dates, added up by delivery date, minus credits. No packaging. Needed for a result; if missing, the reply asks for it. - `wages`Optional Every wage and salary for the same dates, from payroll, managers included (a salary counts as a 52nd of a year for a week, a 12th for a month). No tips, but a service charge paid to staff stays in. Needed for a result; if missing, the reply asks for it. - `payrollCosts`Optional Optional: the employer's payroll taxes and benefits for the period, such as health insurance or a pension, when the wages figure leaves them out. - `openingStock`Optional Optional: stock at cost at the start of the period. Used only with closingStock: food and drink then becomes what was used, opening + purchases - closing. - `closingStock`Optional Optional: stock at cost at the end of the period, counted on the same items the invoices cover. Used only with openingStock. - `ownerPay`Optional Optional: when the owner works in the restaurant without a salary, what they'd pay someone to do their hours, plus the payroll taxes on it. It's a second figure; the main one stays as the books have it. Labor cost calculator `restaurantdoctor_labor_cost` What the assistant reads Works out a restaurant's labor cost from the owner's own payroll figures for one week or month, as the free labor cost calculator on restaurantdoctorai.com does, in one of its three modes: share (labor as a % of net sales, against the range for its type of restaurant, with what the points over are worth), hours (next period's labor budget and the hourly staff hours it pays for once salaries are paid) or line (the cut line: the sales per labor hour below which the first person goes home). Takes net sales, hourly pay and hours worked, plus optional salaries, payroll taxes and benefits, expected sales, a target and more, in any currency; a figure that isn't sent is never counted as 0. Returns the calculator's result, how it was worked out, and a link that opens the calculator with the same figures filled in. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (3 needed, 10 optional) - `mode`Needed What to work out, as the page's tabs. share: last period's labor cost, % of net sales, against the range (needs sales, hourlyWages). hours: the labor budget for next period's expected sales and the hourly staff hours it pays for (needs forecast, hourlyWages, hours). line: the cut line, the sales per labor hour below which the first person goes home (needs hourlyWages, hours, and sales or forecast). - `format`Needed The type of restaurant, which sets the labor range and the starting target: quick_service = Quick service (Counter or kiosk, fast turnover); fast_casual = Fast casual (Order at the counter, served at the table); full_service = Full service / casual dining (Servers, moderate check); fine_dining = Fine dining (Chef-driven, higher check); bar_lounge = Bar, taproom or lounge (Drinks are about half of sales or more); cafe = Café or coffee shop (Drinks lead, with pastries and light food). - `period`Needed Whether the figures are a week's or a month's: last period's, all for the same dates, and forecast for the next. Sets how a year is worked out (52 weeks or 12 months). - `sales`Optional Last period's net sales, in the owner's own currency like every amount here, from the POS: after discounts and comps, before tax and tips, plus any service charge paid out to staff. Needed for share, and for line without forecast. - `hourlyWages`Optional Last period's gross pay for everyone paid by the hour, from payroll: by the days worked, overtime in, tips out. 0 if all are salaried (share only). Needed in every mode. - `hours`Optional Hours those staff worked last period, from the time clock: overtime in, salaried managers out. Needed for hours and line. - `salaries`Optional Optional: salaried pay for the period (a yearly salary ÷ 52, or ÷ 12 for a month). - `payrollCosts`Optional Optional: the employer's payroll taxes and benefits on that pay, such as health insurance. Left out, labor reads lower, unless the pay already includes them. - `overtimeHours`Optional Optional, for share: how many of the hours worked were overtime. - `ownerPay`Optional Optional, for share: if the owner works without a salary, what someone doing their hours would cost, payroll taxes in. Shown as a second figure. - `forecast`Optional Net sales expected next period, more than 0: the last 4 weeks' (or 3 months') average, adjusted for events and holidays. Needed for hours; for line, the line is then for these sales. - `target`Optional Optional, for hours and line: the owner's labor target, % of net sales (1 to 99). Left out: the top of the range, where the page starts. - `hourSales`Optional Optional, for line: net sales in the last hour of a shift, for how many hourly staff it pays for. Break-even calculator `restaurantdoctor_break_even` What the assistant reads Works out a restaurant's break-even sales from the owner's figures for one typical month, as the free break-even calculator on restaurantdoctorai.com does: the sales a month needs to pay every bill, also a week, a day open and guests a day; how far the month was above or short of it; what was left over after every bill; what a point of cost moves it by; with and without the owner's own pay; and if prices were 10% lower or higher (an illustration). Hourly pay can rise with sales or stay fixed, or both ways as a range. Takes the month's figures in any currency; a figure not sent is never counted as 0. Returns the result, the working and a link that opens the calculator with the figures in. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (0 needed, 12 optional) - `sales`Optional Net sales for one typical month, in the owner's own currency: the POS sales summary, before tax and tips, app orders at menu price. Not a holiday rush: every figure is for that month. Needed, more than 0; if missing, the reply asks for it. - `purchases`Optional That month's food and drink supplier invoices, by delivery date. Needed; if missing, the reply asks for it. - `hourlyPay`Optional That month's hourly wages from payroll, plus the employer's payroll taxes and benefits; no tips. Paid every two weeks: one pay run × 26 ÷ 12. Needed (0 with no hourly staff); if missing, the reply asks for it. - `otherVariable`Optional That month's other costs that rise with every sale: card fees, app commissions (app sales minus payouts, without tax and tips), packaging, percentage rent, royalties, ad fund fees; nothing in the food and drink invoices. Needed (0 if none); if missing, the reply asks for it. - `hourly`Optional Whether hourly staff time goes up and down with sales: yes, shifts are added and cut as it gets busy or quiet; no, about the same team every shift (a bill that stays the same); unsure (the default): both ways, as a range. - `fixedBills`Optional A month of the bills that stay the same: rent and utilities, salaries with their payroll taxes and benefits, insurance, software, marketing, loan and lease payments in full. No depreciation. Yearly bills ÷ 12. Not the owner's pay. Needed; if missing, the reply asks for it. - `ownerPay`Optional Optional: the owner's own pay a month, if not on payroll. Counted as a bill; break-even is also given without it. - `averageSpend`Optional Optional: average spend per guest, the POS average check or net sales ÷ guests. Turns sales into guests. - `daysPerWeek`Optional Optional: days open a week, a whole number from 1 to 7. Turns the month into a day open (and guests a day, with averageSpend). - `openingStock`Optional Optional: stock at cost at the start of the month. Used only with closingStock: food and drink is then opening + purchases - closing. Skip both if stock stayed steady. - `closingStock`Optional Optional: stock at cost at the end of the month. Used only with openingStock. - `goal`Optional Optional: what the owner would like left over a month after every bill, before income tax. Adds the sales for it. Restaurant budget calculator `restaurantdoctor_restaurant_budget` What the assistant reads Works out a restaurant's monthly budget from the owner's expected sales and type of restaurant, as the free restaurant budget calculator on restaurantdoctorai.com does: food and drink, labor, prime cost and rent with utilities as money ranges by RestaurantDoctorAI's targets for that type, other operating costs at the owner's share or a 20% working assumption, what's left after every bill, and break-even sales with labor counted both ways. Takes expected net sales and the type, plus optional rent, the owner's own pay and the other costs share, in any currency; a figure that isn't sent is never counted as 0. Returns the calculator's result, how it was worked out, and a link that opens the calculator with the same figures filled in. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (1 needed, 4 optional) - `format`Needed The type of restaurant, which sets RestaurantDoctorAI's targets the budget is split by: quick_service = Quick service (Counter or kiosk, fast turnover); fast_casual = Fast casual (Order at the counter, served at the table); full_service = Full service / casual dining (Servers, moderate check); fine_dining = Fine dining (Chef-driven, higher check); bar_lounge = Bar, taproom or lounge (Drinks are about half of sales or more); cafe = Café or coffee shop (Drinks lead, with pastries and light food). - `sales`Optional Expected net sales for one month, in the owner's own currency like every figure here: before tax and tips. Needed for a result, and more than 0; if missing, the reply asks for it. - `rent`Optional Optional: the owner's own rent, property charges and utilities for a month, together, as the occupancy targets count them. Replaces the occupancy range. - `ownerPay`Optional Optional: the owner's own pay for a month, when it isn't on payroll. Comes off what's left, and counts as a bill that stays the same. - `otherPct`Optional Optional: other operating costs as a % of sales (card fees, packaging, cleaning, repairs, marketing, insurance, software, accounting, loan payments). Left out, 20% is used: a working assumption, not a target or a published figure. Plate cost and menu price calculator `restaurantdoctor_plate_cost` What the assistant reads Works out what one plate of a dish costs and what to charge for it, as the free plate cost calculator on restaurantdoctorai.com does: from each ingredient's pack price and the amount used (with a usable % for trim), or from a plate cost the owner already knows; then the food cost at the menu price, what's left per plate, the price for their target food cost and the prices inside the range for their type of restaurant, with the tax taken out of a tax-inclusive menu. It works from the owner's own prices and amounts, in any currency; a figure that isn't sent is never counted as 0. Returns the calculator's result, how it was worked out, and a link that opens the calculator with the same figures filled in. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (2 needed, 9 optional) - `mode`Needed ingredients: from each ingredient's pack price and the amount used. known: from a plate cost the owner already has. - `format`Needed The type of restaurant, which sets the food cost range and the starting target: quick_service = Quick service (Counter or kiosk, fast turnover); fast_casual = Fast casual (Order at the counter, served at the table); full_service = Full service / casual dining (Servers, moderate check); fine_dining = Fine dining (Chef-driven, higher check); bar_lounge = Bar, taproom or lounge (Drinks are about half of sales or more); cafe = Café or coffee shop (Drinks lead, with pastries and light food). - `ingredients`Optional With mode ingredients: each ingredient. A sauce or side already costed is one: its cost for a pack of 1 each, 1 each used. - `portions`Optional With mode ingredients, needed: how many plates the amounts make, 1 for a single plate. - `extras`Optional Optional, mode ingredients: small extras without a line of their own (salt, oil, garnish), % of the ingredients; often 5 to 10. - `plateCost`Optional With mode known, needed: what the ingredients on one plate cost, from a recipe card or costing sheet, in the owner's own currency like every money figure here. - `menuPrice`Optional Optional: the dish's price on the menu. Left out, the result gives the prices to charge, as for a new dish. - `taxIncluded`Optional true when menu prices include VAT or GST (UK, EU, Australia, New Zealand); left out when tax is added at the till (US, Canada). - `taxRate`Optional With taxIncluded, needed: the tax rate in this dish's price, %. Food and drinks can have different rates. - `target`Optional Optional: the owner's target food cost, %. Left out, the middle of the format's range (32 for full service); none for a bar. - `sold`Optional Optional: how many of this dish sold last month (POS item sales), for the figures a month. Delivery app price calculator `restaurantdoctor_delivery_price` What the assistant reads Works out a dish's break-even price on a delivery app, the app price that leaves the owner the same money as selling it in their dining room, as the free delivery app price calculator on restaurantdoctorai.com does: the in-house price ÷ (1 − the app's real commission rate), the markup on the in-house price that takes (rounded up to a whole percent), and what the app keeps. Takes the dish's in-house price, in any currency, and one app's real commission rate from its payout statements; a figure that isn't sent is never counted as 0. Returns the calculator's result, how it was worked out, and a link that opens the calculator with the same figures filled in. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (0 needed, 2 optional) - `price`Optional The dish's price on the owner's own menu, as a dining room guest pays it, in the owner's own currency (14 or 18.5). Needed for a result, and more than 0; if missing, the reply asks for it. - `rate`Optional One delivery app's real commission rate, as a percent: 30 for 30%, not 0.30. From a month of that app's payout statements: (menu sales before tax and tips − the app's payout without tax and tips) ÷ those menu sales × 100, so marketing fees, promotions the owner funds and refunds are in it; the rate in the app's plan leaves them out. Each app is worked out on its own. Needed for a result: 0, or from 1 to 89; if missing, the reply asks for it. Menu calorie counter `restaurantdoctor_menu_calories` What the assistant reads Works out a dish's calories from its recipe, as the free menu calorie counter on restaurantdoctorai.com does: each ingredient's calories from USDA FoodData Central (a list of 118 common foods) or its pack label, added up, shared by the plates it makes or by portion weight, and rounded once the way US, Ontario or UK and EU menus show calories. Ingredients are matched to that list, never guessed: words that fit several foods come back with the foods to pick from, and words that fit none with the nearest foods, or as not on the list. A pack label's kcal per 100 g comes from the owner's own label, never an estimate: without one, that ingredient can't be counted yet. Returns the menu figure, the exact one, how it was worked out with each food's FDC ID, what to check before printing, and a link that opens the counter with the same recipe filled in. It's an estimate, not a lab test. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (2 needed, 5 optional) - `country`Needed Where the menu is: sets the energy figure and its rounding. us = US; ontario = Canada (Ontario); ukeu = UK, Ireland or EU (kcal, with the EU's energy factors). - `dish`Optional Optional: the dish's name, for the record. - `ingredients`Optional What goes into the dish or batch, one entry each. Needed for a result; if missing, the reply asks for it. - `portioning`Needed plates: one plate, or a batch shared into plates (12 pancakes). weight: portions weighed from a cooked batch (a ladle of soup). - `plates`Optional With plates: how many plates it makes, 1 for a single plate. Needed then; if missing, the reply asks for it. - `cookedWeight`Optional With weight: the whole batch, weighed after cooking. Needed then; if missing, the reply asks for it. - `portionWeight`Optional With weight: one portion. Needed then; if missing, the reply asks for it. FICA tip credit check `restaurantdoctor_fica_tip_credit` What the assistant reads Estimates a US restaurant's FICA tip credit (IRS Form 8846, the 45B credit) for a year like the one in its payroll figures, as the free FICA tip credit check on restaurantdoctorai.com does: the employer's 7.65% Social Security and Medicare tax on reported tips, less the tips needed to bring tipped staff's cash wage up to the 5.15 dollars an hour the law fixes for restaurants. Returns the reasons the real answer may be zero first (already claimed, no federal income tax owed, service charges aren't tips, only the business can claim it), then the credit as a range, its worth after the deduction it replaces, the years that may still be open, and sources. It answers questions about taxes on tips, Form 8846, the 45B credit or money the business may be missing on tips. It needs the state, a year's tips reported through payroll (never service charges) and, where the cash wage is under 5.15 dollars, tipped staff's hours that year; whether the return already includes Form 8846 comes first, if known. A figure that isn't sent is never counted as 0. It files nothing, isn't tax advice and takes no payment: the result is an estimate to take to the accountant, with a link that opens the same check with the figures filled in. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (0 needed, 7 optional) - `alreadyClaimed`Optional Optional, asked first: whether the business's return already includes Form 8846 (Form 3800, Part III, line 4f, or the credit on a K-1 from the business): yes, no or unsure (left out: unsure). Yes means there may be nothing left to find. - `state`Optional The US state the restaurant is in: a two-letter code (TX) or its name. It sets tipped staff's cash wage from the US Department of Labor's table. Needed for a result unless cashWage is sent; if missing, the reply asks for it. - `cashWage`Optional Optional: the hourly cash wage tipped staff were paid before tips that year, in US dollars (2.13), if it wasn't the state's tipped wage; the lowest if it changed. Used instead of the state's. - `reportedTips`Optional A year's tips reported through payroll for one business, in US dollars: line 5b of the year's four Forms 941 added up, W-2 box 7 totals, or the year's total from the payroll tips report (Toast, Gusto). Service charges, automatic gratuities and delivery fees don't count: they're wages. Needed for a result; if missing, the reply asks for it. - `tippedHours`Optional The hours tipped staff worked that year, all together, from the payroll hours report. Needed for a result only when the cash wage is under $5.15 an hour (Texas: $2.13); then, if missing, the reply asks for it. - `businessType`Optional Optional: how the business files its taxes, which sets what the credit is worth after tax: c_corp = C corporation; s_corp = S corporation (or an LLC taxed as one); partnership = Partnership (or an LLC taxed as one); sole_owner = Sole owner (or a one-owner LLC); unsure = Not sure (left out: unsure). - `owesIncomeTax`Optional Optional: whether the business (for a pass-through, its owners) usually owes federal income tax: yes, no or unsure (left out: unsure). Card fees check `restaurantdoctor_card_fees` What the assistant reads Checks a US restaurant's card processing fees for one month, as the free card fees check on restaurantdoctorai.com does: the cost per 100 dollars of card sales (the effective rate), against a dated fair range for that type of restaurant (an estimate of a competitive interchange-plus account from Visa's published restaurant interchange, network fees and a fair markup); how far above it the fees are, a month and a year; how they compare with Square's, Toast's, Clover's or Stripe's published rate; what's realistic for that processor, words to read out to it, the fees to question on the statement by name, and the caveats. It answers whether card fees are too high, what the effective rate is, and what to ask the processor. It needs the month's total card sales and total card fees from the processor's statement (every fee, monthly and annual ones included), who processes the cards and what kind of restaurant it is; the number of card sales or the average sale narrows it. A figure that isn't sent is never counted as 0. It doesn't read statements, contact a processor or take payment. Returns the result, how it was worked out, and a link that opens the same check with the figures filled in. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (2 needed, 5 optional) - `cardSales`Optional A month's card sales in US dollars, from the processor's statement or the POS card report: everything charged to cards, tips included, before fees, without delivery app orders. Needed for a result; if missing, the reply asks for it. - `totalFees`Optional The same month's total card fees in US dollars, from the same statement: every fee the processor took (percentage, per-sale, monthly, annual, PCI and statement fees), not one line. Needed for a result; if missing, the reply asks for it. - `processor`Needed Who processes the cards (the name at the top of the statement): bank = a bank or independent processor; other = another processor, or not sure. - `format`Needed The type of restaurant, which sets the likely mix of debit and credit cards: quick_service = Quick service (Counter or kiosk, fast turnover); fast_casual = Fast casual (Order at the counter, served at the table); full_service = Full service / casual dining (Servers, moderate check); fine_dining = Fine dining (Chef-driven, higher check); bar_lounge = Bar, taproom or lounge (Drinks are about half of sales or more); cafe = Café or coffee shop (Drinks lead, with pastries and light food). - `cardCount`Optional Optional: the month's number of card sales (transactions), from the same statement. It narrows the fair range to the owner's own sale size. - `averageTicket`Optional Optional: the average card sale in US dollars, when cardCount isn't known. It narrows the fair range too. - `pricingModel`Optional Optional, if the owner knows it: flat = Flat rate: one rate for every card; interchange_plus = Interchange-plus: interchange at cost plus a stated markup; tiered = Tiered: qualified, mid-qualified and non-qualified rates; not_sure = Not sure. It changes the advice, not the fair range. Booking fee check `restaurantdoctor_booking_fees` What the assistant reads Works out what a US restaurant's reservation app costs it, as the free booking fee check on restaurantdoctorai.com does: a month, a year, per guest who booked through it and as a share of the average spend, from the plan's published price (dated, sourced) or the owner's invoice, plus fees per guest and the percentage on deposits. Of the apps checked, only OpenTable charges per guest: for guests from its own app and website, possibly from Google and partners, and on Basic from the restaurant's own website. Returns a one-sentence headline first, the month line by line, what isn't counted, what the owner can do (book direct, mark no-shows, the contract's notice date), other apps' published prices as information, not a recommendation, the caveats and sources. It answers what OpenTable, Resy, Tock, Yelp, Toast Tables, SpotOn or SevenRooms costs a restaurant per reservation or per guest. It needs the app and plan (on the invoice) and, for OpenTable, last month's guests from its covers report by where they booked; the rest is optional. A figure that isn't sent is never counted as 0. Not affiliated with any app; it reads no account. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (1 needed, 12 optional) - `platform`Needed The app and plan, on the invoice. tock_ = Tock, now sold by Resy; other = any app not listed, with monthlyFee. - `networkCovers`Optional Guests seated last month who booked through the app itself (its own app or site). OpenTable: the Network line of its covers report; needed for OpenTable. Other apps: every guest booked through it, wherever they booked. - `partnerCovers`Optional OpenTable only: guests who booked through Google and other partners, the report's Partner or Affiliate line. Left out: not counted. - `websiteCovers`Optional OpenTable only: guests who booked through the restaurant's own website (a fee each on Basic). - `websiteFlatAddOn`Optional OpenTable Basic only: true if the invoice shows the monthly flat fee for website bookings. - `monthlyFee`Optional The invoice's monthly fee: plan plus add-ons, before fees per guest. Left out: the published price. Needed for sevenrooms and other. - `networkCoverFee`Optional OpenTable or other only: the fee per guest on the invoice, only if it isn't the published one. - `partnerCoverFee`Optional OpenTable only: the fee per partner guest, from its Billing History tab. Left out: a range. - `prepaidAmount`Optional Money the app collected last month, from its statement: deposits, prepayments, no-show charges. tock_: prepayments only. - `prepaidFeePercent`Optional The app's percentage of that money, as a percent (2 for 2%, not 0.02), only if it isn't the published one. - `averageSpend`Optional The average spend per guest: net sales ÷ guests, from the POS. - `firstMonth`Optional OpenTable Basic only: true if last month was the first 30 days (network guests free). - `renewalDate`Optional Not on opentable_basic or spoton_reserve: the contract's renewal date, YYYY-MM-DD, from the contract; month-to-month plans have none. ### A guided consult "Fix my prime cost", one question at a time: set steps and the calculators find the biggest problem, with no AI touching a figure. Prime cost consult `restaurantdoctor_prime_cost_consult` What the assistant reads Runs the guided consult "Fix my prime cost" one step at a time from the answers so far, as RestaurantDoctorAI's AI restaurant expert does on restaurantdoctorai.com: the 6 set questions (type of restaurant, a week or a month, net sales, food and drink invoices, wages, payroll taxes and benefits), a check on any figure that looks like a slip, the free prime cost calculator's result, then up to 5 more questions on the half that comes first, ending in one finding: the biggest problem at these figures, the step-by-step fix to start with, its sources and at most two other things worth a look. Fixed rules and the calculators produce every figure and sentence, with no AI in between. Nothing is kept between calls: the answers so far come in on every call, each one optional, and the reply is the first step still open, with the fields and set answers that answer it. Any number of answers can come in one call. A figure that isn't sent is never counted as 0. A dish is only its price, plate cost and number sold, with no name. A guide, not a promise of results, and not legal, tax or financial advice. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (0 needed, 23 optional) - `format`Optional The type of restaurant, which sets the targets: prime cost, and the ranges for food and drink and for labor: quick_service = Quick service (Counter or kiosk, fast turnover); fast_casual = Fast casual (Order at the counter, served at the table); full_service = Full service / casual dining (Servers, moderate check); fine_dining = Fine dining (Chef-driven, higher check); bar_lounge = Bar, taproom or lounge (Drinks are about half of sales or more); cafe = Café or coffee shop (Drinks lead, with pastries and light food). - `period`Optional Whether every figure covers one week or one month, the same dates for all of them. Sets how a year is worked out (52 weeks or 12 months). - `payrollIn`Optional True when the wages figure already has payroll taxes and benefits in it. A payrollCosts figure then isn't counted. - `estimated`Optional Which figures are the owner's own rough estimate: each reads "(your estimate)" where the figures are shown. - `skipped`Optional The steps the owner passed on: payrollCosts, stock, hourlyWages, hours, overtimeHours and owner. The others stay open. - `confirmed`Optional The checks the owner says are right, by the id each check step gives. - `firstHalf`Optional Which half comes first when food and drink and labor tie: the answer to the tie step. - `resultSeen`Optional True once the result step has been seen. The questions on why come after it, and the finding doesn't repeat the result's lines. - `costsKnown`Optional Whether the owner knows what their best sellers cost to plate. - `dishes`Optional Up to 3 best sellers, drinks included, for the same dates as the other figures. All 3 can come in one call, and a call's list is the whole list. - `dishesDone`Optional True when the list of dishes is finished at fewer than 3, or empty. - `hourlyWages`Optional How much of the wages figure was hourly pay. The rest counts as salaries. - `hours`Optional The hours hourly staff worked, overtime included, from the time clock. - `overtimeHours`Optional How many of those hours were overtime. - `unpaid`Optional False when the owner doesn't work in the restaurant without a salary. - `alsoOther`Optional True for the other half's questions as well, when it's over its range too. They come after the finding and aren't part of it. - `sales`Optional Net sales for the period, in the owner's own currency like every figure here, from the POS sales summary: after discounts and comps, before tax and tips, with any service charge paid out to staff added back. Needed for a result, and more than 0; if missing, the reply asks for it. - `purchases`Optional Food and drink supplier invoices for the same dates, added up by delivery date, minus credits. No packaging. Needed for a result; if missing, the reply asks for it. - `wages`Optional Every wage and salary for the same dates, from payroll, managers included (a salary counts as a 52nd of a year for a week, a 12th for a month). No tips, but a service charge paid to staff stays in. Needed for a result; if missing, the reply asks for it. - `payrollCosts`Optional Optional: the employer's payroll taxes and benefits for the period, such as health insurance or a pension, when the wages figure leaves them out. - `openingStock`Optional Optional: stock at cost at the start of the period. Used only with closingStock: food and drink then becomes what was used, opening + purchases - closing. - `closingStock`Optional Optional: stock at cost at the end of the period, counted on the same items the invoices cover. Used only with openingStock. - `ownerPay`Optional Optional: when the owner works in the restaurant without a salary, what they'd pay someone to do their hours, plus the payroll taxes on it. It's a second figure; the main one stays as the books have it. ### Printable sheets The figures for the free templates: a night's cash close, a tip-out, and par levels to order to. Nightly cash-out `restaurantdoctor_cash_out` What the assistant reads Works out a restaurant's nightly cash close for one drawer and one day or shift, as the free cash-out sheet on restaurantdoctorai.com does: expected cash (starting bank + cash sales - cash tips paid out - paid outs), over or short against the cash counted, the deposit (counted minus the starting bank), the card difference (processor batch minus POS card sales) when both are sent, and comps, voids and discounts as a share of net sales. Takes the starting bank, cash sales and cash counted, plus optional paid outs, card figures and sales totals, in any currency; a figure that isn't sent is never counted as 0. It sets no normal for over or short. Returns the result, how it was worked out, and a link to the free cash-out sheet. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (0 needed, 11 optional) - `startingBank`Optional The cash in the drawer at open: one drawer, one day or shift, in the owner's own currency like every figure here. Needed; if missing, the reply asks for it. - `cashSales`Optional That day's cash payments, from the POS end-of-day (Z) report. Needed; if missing, the reply asks for it. - `cashTipsPaidOut`Optional Optional: card tips paid to staff in cash from this drawer. Taken off expected cash. - `paidOuts`Optional Optional: cash taken out of the drawer for supplies and the like. Taken off expected cash. - `cashCounted`Optional The cash counted at close, bank included. Needed; if missing, the reply asks for it. - `cardSales`Optional Optional: that day's card sales from the POS. With cardBatch, gives the card difference. - `cardBatch`Optional Optional: what the card processor settled for the same day. With cardSales. - `netSales`Optional Optional: that day's net sales from the POS, before tax and tips. Turns comps, voids and discounts into shares. - `comps`Optional Optional: the day's comps total from the POS. Needs netSales for its share. - `voids`Optional Optional: the day's voids total from the POS. Needs netSales for its share. - `discounts`Optional Optional: the day's discounts total from the POS. Needs netSales for its share. Tip-out sheet `restaurantdoctor_tip_out` What the assistant reads Works out how a restaurant shares tips, two ways, as the free tip-out sheet on restaurantdoctorai.com does. method percent: a server's tip-out to support roles, each a percentage of the server's net sales or tips, with the total tip-out, what the server keeps from their tips, and an error-level check when the tip-out is more than the tips. method pool: a tip pool shared by hours, with optional points for each role, rounded to the cent with the rounding remainder placed on the largest share. Takes the figures in any currency; a figure that isn't sent is never counted as 0. Returns the split, how it was worked out, the Department of Labor's published points on tips with their sources, and a link to the free tip-out sheet. It only works out a split from the numbers sent; it doesn't say whether a tip-out or a tip pool is allowed. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (1 needed, 6 optional) - `method`Needed percent: one server's tip-out to support roles, each a % of the base. pool: a tip pool shared by hours, with optional points. - `baseKind`Optional With percent, needed: netSales or tips, what the roles' percentages are taken of. - `base`Optional With percent, needed: the server's net sales for the shift (before tax and tips), or their tips, as baseKind says. In the owner's own currency. - `tips`Optional With percent: the server's tips for the shift, to work out what they keep. Not used when baseKind is tips. - `roles`Optional With percent, needed: the support roles tipped out. - `pool`Optional With pool, needed: the tips in the pool for the period. - `members`Optional With pool, needed: the people or roles sharing the pool. Prep list and par sheet `restaurantdoctor_par_levels` What the assistant reads Works out par levels and what to order for up to 50 kitchen or bar items, as the free prep and par sheet on restaurantdoctorai.com does. Par = use per day × days covered + safety stock (days covered: from the day of the count and the order until the next delivery after this one arrives); order = par minus what's on hand, never below 0, rounded up to whole packs when a pack size is sent. Takes each item's use per day (or per week and the days open a week), days covered, optional safety stock and pack size, and what's on hand, in the owner's own units; a figure that isn't sent is never counted as 0, and an item missing one is returned as incomplete. Returns each item's par and order, the items that need an order, and a link to the free prep and par sheet. The figures aren't stored; only an anonymous daily count of calculator runs is kept. What it can send (0 needed, 3 optional) - `items`Optional Needed: the items, up to 50. - `daysOpenPerWeek`Optional Days open a week, 1 to 7. Needed only for items sent with usePerWeek. - `daysCovered`Optional Days each order covers, for items without their own: from the day of the count and the order to the day the next delivery after this one arrives. ### Checks and dates A check of a restaurant's own website, and its coming tax, payroll and permit deadlines. Check a restaurant's website `restaurantdoctor_check_website` What the assistant reads Checks a restaurant's own website as the online check on restaurantdoctorai.com does: secure (https) and on its own address, made for phones, a menu written as text that phones and Google can read, online orders on its own terms or only through delivery apps, the address, hours and phone written on the page, tap to call, a way to get in touch, email sign-up and gift cards, what Google reads (a headline, title and description, not hidden from search, restaurant structured data), the preview a shared link shows, the page's text, story, questions, reviews, photo descriptions and social links, and Google's phone speed test with its measures one by one and real visitors' speed when Google has it (15 to 40 seconds; phoneSpeed: false skips it). Returns the score and grade, each area's score, the three checks to fix first with what to do about each, what a basic AI reader can tell about each question in a word, and a link to the full online check, which adds the Google listing, every check with what to fix, the reasons and places nearby on the website. A check it couldn't make (a page drawn by scripts, a security screen instead of the page) is left out of the score, never failed; every check, its points and where its facts come from are listed at https://restaurantdoctorai.com/online-check/checks. Takes only the website address. The website isn't stored; only an anonymous daily count of checks is kept. What it can send (1 needed, 1 optional) - `website`Needed The restaurant's own website address, such as joesdiner.com or https://www.joesdiner.com. Not a DoorDash, Uber Eats or Google Maps page. - `phoneSpeed`Optional Whether to run Google's phone speed test too (the default). It takes 15 to 40 seconds; false skips it. List a restaurant's deadlines `restaurantdoctor_deadlines` What the assistant reads Lists a US restaurant's coming tax, payroll, license and permit deadlines, as the deadline calendar on restaurantdoctorai.com does: federal dates (payroll forms, tip reports, W-2s, 1099s, returns, estimated tax) and the state's (sales tax, unemployment reports, yearly reports, franchise taxes, minimum wage changes, liquor and health permit dates where one date applies to everyone), each moved for weekends and holidays by the agency's rule, with the rule, what a miss costs and the official source. A state date is given only where the state's own page confirms it; the rest come as what to check with the agency. Also the rules for each hire, and a link that opens the same calendar on the website, where the owner adds their own license and permit dates and gets free email reminders. Takes the state and yes/no answers, no names or emails; scope (federal or state) and withinDays (the next so many days) cut the list. Reminders, not legal or tax advice. The answers aren't stored; only an anonymous daily count of calendars is kept. What it can send (3 needed, 14 optional) - `state`Needed The restaurant's US state, two letters (DC for Washington DC). - `staff`Needed Has anyone on payroll. - `tips`Optional Customers usually tip the staff. - `alcohol`Needed Sells alcohol. - `business`Optional How the business is set up: sole (sole owner, no LLC), llc_single, llc_multi, partnership (no LLC), s_corp or c_corp. Decides the tax return and the state's yearly report. Left out, both return dates come with an "If". - `formed`Optional The month the LLC or corporation was formed with the state, YYYY-MM. Some states set the yearly report by it. - `salesTax`Optional How often the state has the restaurant file sales tax (on its permit letter or state account). - `deposits`Optional Who pays the payroll taxes to the IRS: provider (a payroll company), monthly (the owner, by the 15th), semiweekly (the owner, after each payday). - `contractors`Optional Pays a contractor, or rent to a landlord that isn't a corporation, $2,000 or more a year (1099 forms). - `estimated`Optional The owner pays quarterly estimated tax on the business's profit. - `bigTipped`Optional Tipped, and staff worked more than 80 hours in total on a typical day last year, all locations together (Form 8027). - `fullTime50`Optional 50 or more full-time staff on average, counting part-timers' hours, related businesses together (ACA forms). - `caterer`Optional Mainly catering or a food truck, not a restaurant or bar (OSHA's injury log). - `city`Optional A city with its own minimum wage, as an id like los-angeles or chicago; left out, the state's rate applies. - `months`Optional How many months ahead to list, 1 to 15 (default 3). The calendar runs to the end of 2027. withinDays, when sent, sets the window instead. - `scope`Optional Optional, to cut the list: federal (the IRS's and other federal dates only), state (the state's and the city's only) or all (the default). The rules for each hire are listed either way. - `withinDays`Optional Optional: only the dates in the next so many days, 1 to 460, from today in the state (7 for the coming week). Sent, it replaces months. ### The library Our guides, blog posts and sourced facts, so an assistant can quote them with their source. Search RestaurantDoctorAI's library `restaurantdoctor_search_knowledge` What the assistant reads Searches RestaurantDoctorAI's sourced library: the website's guides and blog posts, RestaurantDoctorAI's step-by-step fixes, RestaurantDoctorAI's targets by type of restaurant, sourced industry facts, and an expert knowledge base of sourced how-tos, rules and checklists (costs, labor, money, growth, compliance, operations). Keyword search: "break-even", "breakeven" and "break even" are one term, and plurals match. Returns the best passages, each with its id, kind, title, link and source; kind narrows the search to one kind of entry. RestaurantDoctorAI's targets are its own, not industry benchmarks. The query isn't stored; only an anonymous daily count of searches and reads is kept. What it can send (1 needed, 2 optional) - `query`Needed What to look up, in plain words, e.g. 'how to lower food cost', 'tip pool rules' or 'delivery app commission'. - `limit`Optional The most passages to return, 1 to 8 (default 5). Up to 5, a passage far weaker than the best match is left out. - `kind`Optional Only this kind of entry: guide (the website's guides), post (blog posts), play (RestaurantDoctorAI's step-by-step fixes), targets (RestaurantDoctorAI's targets by type of restaurant), fact (sourced industry facts) or knowledge (sourced how-tos, rules and checklists). Left out, every kind. List RestaurantDoctorAI's library `restaurantdoctor_list_knowledge` What the assistant reads Lists the entries of RestaurantDoctorAI's sourced library, all or one kind (guide, post, play, targets, fact or knowledge), up to 50 at a time: each with its id, kind, title, link and source, and for guides and blog posts the day it was published and last updated. It shows what the library has, such as every blog post and its date. Pages with offset and limit; nextOffset gives the next page. Nothing about the request is stored; only an anonymous daily count of searches (a list counts as one) is kept. What it can send (0 needed, 3 optional) - `kind`Optional Which kind of entry to list: guide (the website's guides), post (blog posts), play (RestaurantDoctorAI's step-by-step fixes), targets (RestaurantDoctorAI's targets by type of restaurant), fact (sourced industry facts) or knowledge (sourced how-tos, rules and checklists); or all (the default). - `offset`Optional How many entries to skip (default 0). The next page starts at the nextOffset the previous list returned. - `limit`Optional How many entries to return, 1 to 50 (default 20). Read a library entry `restaurantdoctor_read_knowledge` What the assistant reads Reads one entry of RestaurantDoctorAI's library in full (a guide, a blog post, a playbook fix, the targets, a fact or a knowledge entry), by its id (such as 'guide:restaurant-prime-cost'), with its title, link and source. A long entry comes 8,000 characters at a time, with its total length: the next part starts at the nextOffset it returns. The query isn't stored; only an anonymous daily count of searches and reads is kept. What it can send (1 needed, 1 optional) - `id`Needed An entry id, as the library's search and list give it, e.g. 'guide:restaurant-prime-cost' or 'post:restaurant-break-even-point'. - `offset`Optional Where to start, in characters (default 0, the start). A long entry comes 8,000 characters at a time: the next part starts at the nextOffset the previous read returned. ## The paid tools Buying is kept apart, on a second server with 4 tools of its own: the full report, a restaurant's full check (How AI sees your website), a pack of report codes, and using a code. Payment goes through Stripe. How it works is on the [AI assistants page](https://restaurantdoctorai.com/ai-assistants#payment). Try it in your own assistant One address, no account, and what you send isn't stored. [How to connect](https://restaurantdoctorai.com/ai-assistants) --- title: "Who builds RestaurantDoctorAI, and why" description: "Who builds RestaurantDoctorAI and why, what it believes, how it makes money and who runs it: Gomorra Labs LLC. No accounts, and reports paid once." canonical: "https://restaurantdoctorai.com/about" --- # Who builds RestaurantDoctorAI, and why Felix S. Pavel Founder, RestaurantDoctorAI People close to me run restaurants, and I've seen how hard it is to tell where the money goes when you're busy every night. I built RestaurantDoctorAI to give independent owners a fast, plain-English first look, for free. ## Why I built it Good advice pays. In a [study of small businesses](https://www.kellogg.northwestern.edu/academics-research/research/detail/2018/the-impact-of-consulting-services-on-small-and-medium), the ones that got a year of consulting had about 50% more employees five years later than the ones that didn't. The catch is the price: restaurant consultants who [publish their rates](https://restaurantdoctorai.com/blog/restaurant-consultant-vs-profit-report) charge around $995 for a day, or $90 to $150 an hour. Most independent owners never get that help. AI changes that. Today we can put a first set of numbers in front of any owner in seconds: RestaurantDoctorAI's targets, checked against published figures where they exist, fixed formulas, and a playbook of common fixes. I built RestaurantDoctorAI so every independent owner can take advantage of it: a free first look at where the money goes, in plain English, and a plan you can buy once if you want it. It doesn't replace a good consultant. It tells you where to look first, so if you do hire one, you spend their hours on the fix, not the search. ## Three things I believe They decide what the site shows you, what it charges and what it recommends. 1. **You should see the problem before you pay for anything.** A lot of restaurant software wants a demo call, your POS login or a monthly plan before it shows you a single number. We won't. Answer 12 questions, and in about 45 seconds you see your grade and your costliest issue, free, with no account and no logins. What you do after that is up to you. [Where your answers go](https://restaurantdoctorai.com/privacy) 2. **An independent restaurant deserves the numbers a chain gets.** Big chains have finance teams watching prime cost every week. Most independent owners have no one. So the report puts our own targets for your type of restaurant, which follow the National Restaurant Association's figures where it publishes them, and fixed formulas to work on your answers, and explains them in plain English with a plan for this week. AI helps write the words; it never sets a number. It's an estimate from your answers, not an audit of your books, and it says so. [How we grade](https://restaurantdoctorai.com/methodology) 3. **You should pay once, not every month.** A lot of restaurant tech bills you every month, whether it helps or not. Here the check, the calculators and the guides are free, and the full report is one payment that never renews. When a company shows up under your costliest issue, it's because its product is made for that problem, never because of what it pays, and the link says it's paid. [How paid links work](https://restaurantdoctorai.com/partner-links) Felix S. Pavel, founder of RestaurantDoctorAI ## How the site makes money The profit check, the calculators and the guides are free. These are the ways the business is paid, and you always know which one you're looking at. There are no ads, and your details go to a partner company only if you press a button that says so. We don't sell your information to anyone else. RestaurantDoctorAI is run by Gomorra Labs LLC, an independent company. It isn't owned by a delivery app, a POS company or a booking platform, and a company that pays us never changes your grade or what your report says. - **One-time reports** The profit check is free. The full report is one payment: no subscription, nothing renews and there's no account to cancel. [Every price, on one page](https://restaurantdoctorai.com/pricing) - **How AI sees your website** Your grade and first three fixes are free. The full report is one payment that never renews. [Check my restaurant](https://restaurantdoctorai.com/online-check) - **The Restaurant Fix Kit** Food cost, labor, delivery and waste workbooks for Excel, with printable PDFs. It's an extra you choose: the box at checkout is never ticked for you. - **Report codes for advisors** Consultants, accountants and associations can prepay for reports in packs of 10 or 50 codes, which never expire. [For advisors](https://restaurantdoctorai.com/advisors) - **The AI restaurant expert's pass** The first 5 questions to our AI restaurant expert are free. After that, a 30-day pass is one payment that never renews. [Ask the expert](https://restaurantdoctorai.com/ask) - **Purchases by AI agents** An AI agent with a wallet can buy a report, a restaurant's full check (How AI sees your website) or a pack of codes for you through our agent server, paying with Stripe. For those purchases we are the seller. [What an agent can buy](https://restaurantdoctorai.com/pricing#ai-agents) - **Paid links** Under your costliest issue we may show a company whose product is made for it, marked “Paid link”. If you sign up through it, the company pays us a fee, which adds nothing to what you pay. None are showing right now. [What a paid link is](https://restaurantdoctorai.com/partner-links) ## Who runs it, and how to reach us RestaurantDoctorAI is a product of Gomorra Labs LLC, a Wyoming limited liability company. Questions, corrections or ideas? Write to us. How your information is handled is in our [Privacy Policy](https://restaurantdoctorai.com/privacy), and the companies that help run the site are on our [Service providers](https://restaurantdoctorai.com/subprocessors) page. - Email: [support@restaurantdoctorai.com](mailto:support@restaurantdoctorai.com) - Company: Gomorra Labs LLC - Registered in: Wyoming, United States [How every number is worked out](https://restaurantdoctorai.com/methodology) Page last updated November 4, 2026 ## See where your restaurant's money goes Your grade and your costliest issue, free. [Get my free grade](https://restaurantdoctorai.com/#diagnostic) 12 questions · about 45 seconds · no POS or bank logins --- title: "How we grade your restaurant" description: "How the free check grades a restaurant: the 12 questions, the targets for each type, how the dollars are estimated, what the AI does and our sources." canonical: "https://restaurantdoctorai.com/methodology" --- # How we grade your restaurant Your grade and every dollar figure come from fixed formulas, held against RestaurantDoctorAI's targets for your type of restaurant. Here's how they work, where the targets come from and what they can't tell you. AI helps write the paid report's words. It never sets a number. Last reviewed: October 14, 2026. Spotted a mistake? Email [support@restaurantdoctorai.com](mailto:support@restaurantdoctorai.com) At a glance - Questions: 12 - Types of restaurant, each with its own targets: 6 - Cap on all avoidable costs together: 5.5–8% - Cap on all sales you could add: 4–6% - Numbers the AI sets: None Caps are shares of your sales a month, low end to high end. The free grade uses no AI at all. ## 12 questions about how you run things Your type of restaurant picks the targets, and your sales range sets the scale. Every other answer is held against what a well-run restaurant of your type does, and becomes one of these findings: - Problem: Costing you money now, with a dollar range. - Opportunity: Sales you could add, with a dollar range. - Watch: A blind spot to keep an eye on. Some come with sales you could add. - Healthy: Nothing to fix. - **Your baseline** 1. What kind of restaurant do you run? 2. About how much do you sell in a typical month? 3. What's your prime cost: food, drink and labor costs as a share of sales? - **Kitchen & food cost** 1. How often do you update recipe costs with current supplier prices? 2. How do you track kitchen waste and spoilage? 3. How do you control drink pours? - **Labor & overhead** 1. When a shift gets slow, what happens to staffing? 2. What share of sales goes to rent, property costs and utilities? - **Delivery** 1. How much of your sales come through DoorDash, Uber Eats or similar apps? - **Guests & revenue** 1. How full is your dining room across the week? 2. How do you bring guests back? 3. What's your average rating on Google and Yelp? - No AI in the free check. Your free grade is worked out in your browser, by the same code that scores the paid report. ## The targets for your type of restaurant Your first answer picks one row. Every figure is a share of your sales, and the free calculators use the same targets. They're also published as data, at [/targets.json](https://restaurantdoctorai.com/targets.json). Targets for each type of restaurant, as a share of sales | Type of restaurant | Prime cost, at most | Food and drink cost | Labor | Rent and utilities | Drinks' share of sales | | --- | --- | --- | --- | --- | --- | | Quick service | 60% | 26–32% | 26–31% | 6–10% | 5% | | Fast casual | 60% | 28–33% | 27–32% | 6–10% | 8% | | Full service / casual dining | 65% | 30–34% | 30–35% | 7–12% | 20% | | Fine dining | 70% | 32–38% | 34–38% | 7–12% | 30% | | Bar, taproom or lounge | 57% | 20–26% | 25–32% | 7–12% | 55% | | Café or coffee shop | 60% | 24–32% | 28–34% | 6–10% | 65% | - **Quick service** - Prime cost, at most: 60% - Food and drink cost: 26–32% - Labor: 26–31% - Rent and utilities: 6–10% - Drinks' share of sales: 5% - **Fast casual** - Prime cost, at most: 60% - Food and drink cost: 28–33% - Labor: 27–32% - Rent and utilities: 6–10% - Drinks' share of sales: 8% - **Full service / casual dining** - Prime cost, at most: 65% - Food and drink cost: 30–34% - Labor: 30–35% - Rent and utilities: 7–12% - Drinks' share of sales: 20% - **Fine dining** - Prime cost, at most: 70% - Food and drink cost: 32–38% - Labor: 34–38% - Rent and utilities: 7–12% - Drinks' share of sales: 30% - **Bar, taproom or lounge** - Prime cost, at most: 57% - Food and drink cost: 20–26% - Labor: 25–32% - Rent and utilities: 7–12% - Drinks' share of sales: 55% - **Café or coffee shop** - Prime cost, at most: 60% - Food and drink cost: 24–32% - Labor: 28–34% - Rent and utilities: 6–10% - Drinks' share of sales: 65% - Prime cost: Food and drink cost plus labor. It's a ceiling: at or under it is healthy. - Food and drink cost: What you buy from suppliers for the food and drink you sell. - Labor: Your total payroll: wages, salaries and your costs on them, such as payroll taxes. - Rent and utilities: Rent, property charges and utilities together. - Drinks' share of sales: Not a target: what we assume for your type, to size the bar's pour estimate. A café's drinks are mostly coffee, so its pour estimate counts only the alcohol it serves. Where the targets come from They're the ranges we use for a healthy restaurant of each type. Where an independent source publishes a line, as for the prime cost ceilings at quick service and table service, ours match it; the fine dining and bar ceilings are our own, and so are the café ceiling, set at the quick-service line, and the café ranges: no survey covers cafés on their own. We check them against the National Restaurant Association's latest medians. Many restaurants run over them; what matters is where yours stands. Our rent ranges include utilities, so they run higher than rent-only figures. [See the sources](https://restaurantdoctorai.com/methodology#sources-targets-title). ## How the grade is worked out The grade measures how much money looks to be lost, against your sales. First the money lost becomes risk points, then the points become a letter. *Risk points* the middle of your avoidable-loss estimate, as % of sales + 0.4 for each point your own prime cost runs over its target If you don't track prime cost, the grade counts it as the highest answer, 65% or more: not knowing it never grades better than knowing it's high. The estimate in your report stays an estimate. 1. A Strong Under 1.5 points 2. B Healthy 1.5 to under 3 points 3. C Needs attention 3 to under 4.5 points 4. D At risk 4.5 to under 6 points 5. F Critical 6 points or more - When your own numbers put prime cost over its target, or your answer puts it at or over it, the grade is never an A. Not tracking it counts as that answer too. - The grade covers costs. Your rating changes the sales you could add, not the letter, and a restaurant rated below 3.5 stars is never called Strong. ### Each area's letter The free result also gives Prime cost, Kitchen & food cost, Labor & overhead and Delivery a letter of their own. An area is an A when nothing in it is losing money. Otherwise its letter comes from the middle of its own estimate, as a share of sales: B under 0.7%, C under 1.6%, D under 3%, F at 3% or more. That's stricter than the whole restaurant's scale, since each area is one part of it. Prime cost gets its letter from its points over target, and is at least a C whenever it is or may be over. Delivery gets no letter if you don't use the apps, and Guests & revenue gets none at all: it's sales you could add, not a loss. *Worked example: the restaurant in our sample report* - Type: Full service / casual dining - Sales ($60,000 – $120,000): $90,000 a month - Costly problems found: 4 - Together, as % of sales: 1.9–4.2% - Risk points (the middle): 3.05 **Grade:** C · Needs attention Its prime cost is estimated from its answers, so it adds no points. Its report shows $1,730–$3,820 a month lost to avoidable costs, and $1,530–$3,690 a month in sales it could add. ## How the dollar figures are estimated Every figure is a range, low to high, because it comes from multiple-choice answers rather than your books. 1. **Start from one sales figure** Each sales range uses one working figure a month: The sales figure used for each sales range | Your answer | Sales we use, a month | | --- | --- | | Under $25,000 | $15,000 | | $25,000 – $60,000 | $42,500 | | $60,000 – $120,000 | $90,000 | | $120,000 – $200,000 | $160,000 | | Over $200,000 | $250,000 | “Under $25,000” has no middle, and a restaurant that small is rarely near the top of it, so it uses $15,000. “Over $200,000” has no top, so it uses $250,000. Outside the US, the same figures are in your currency. 2. **Give each answer its range** Each answer that points to money lost, or to sales you could add, counts for a range, as a share of your sales. For a full service / casual dining restaurant, for example: - “Only when the menu changes”: Problem0.6–1.3% of sales - “Everyone works their scheduled shift anyway”: Problem0.8–1.7% of sales - “Packed on weekends, quiet midweek”: Opportunity1–2.5% of sales you could add The bar's ranges grow with how much of your sales is drinks, the last column of the targets. Every answer and its range, for a full service / casual dining restaurant #### What's your prime cost: food, drink and labor costs as a share of sales? - Under 58%: HealthyNo dollar figure - 58% – 64%: HealthyNo dollar figure - 65% or more: WatchNo dollar figure - I don't track it regularly: WatchNo dollar figure #### How often do you update recipe costs with current supplier prices? - Monthly or live, with software: HealthyNo dollar figure - Once or twice a year: Problem0.3–0.7% of sales - Only when the menu changes: Problem0.6–1.3% of sales - Rarely: Problem0.9–1.6% of sales #### How do you track kitchen waste and spoilage? - A daily log the chef signs off: HealthyNo dollar figure - A rough estimate during weekly inventory: Problem0.15–0.4% of sales - Only when a big batch or case goes bad: Problem0.3–0.8% of sales - We don't track it: Problem0.5–1.2% of sales #### How do you control drink pours? - We don't serve alcohol: HealthyNo dollar figure - Measured pours, plus weekly pour-cost checks: HealthyNo dollar figure - Spot checks when bottles run out early: Problem0.22–0.44% of sales - Free pours, no pour-cost checks: Problem0.44–0.88% of sales #### When a shift gets slow, what happens to staffing? - We cut staff based on sales per labor hour: HealthyNo dollar figure - Everyone works their scheduled shift anyway: Problem0.8–1.7% of sales - Overtime is common: Problem0.5–1.2% of sales - We're short-staffed: WatchNo dollar figure #### What share of sales goes to rent, property costs and utilities? - Under 8%: HealthyNo dollar figure - 8% – 12%: HealthyNo dollar figure - More than 12%: WatchNo dollar figure - I'm not sure: WatchNo dollar figure #### How much of your sales come through DoorDash, Uber Eats or similar apps? - None: HealthyNo dollar figure - Under 15%: HealthyNo dollar figure - 15% – 35%: Problem1–2% of sales - More than 35%: Problem1.6–3.2% of sales #### How full is your dining room across the week? - Steady all week: HealthyNo dollar figure - Packed on weekends, quiet midweek: Opportunity1–2.5% of sales you could add - Busy lunches, slow dinners: Opportunity1–2% of sales you could add - Hard to predict week to week: Watch0.5–1% of sales you could add #### How do you bring guests back? - Loyalty or email list with automatic campaigns: HealthyNo dollar figure - We collect some emails but rarely use them: Watch0.4–0.8% of sales you could add - Social media and walk-ins only: Opportunity0.8–1.6% of sales you could add - Paid ads, but no guest list of our own: Opportunity0.8–1.8% of sales you could add #### What's your average rating on Google and Yelp? - 4.5 – 5.0 stars: HealthyNo dollar figure - 4.0 – 4.4 stars: Watch0.3–0.8% of sales you could add - 3.5 – 3.9 stars: Opportunity1–2.5% of sales you could add - Below 3.5 stars: Opportunity2.5–5% of sales you could add 3. **Cap the totals** Problems overlap: fixing one often shrinks another, so they can't simply be added up. All avoidable costs together are capped at 5.5% of sales at the low end and 8% at the high end, and all sales you could add at 4% and 6%. When a total reaches its cap, every figure under it is scaled down to fit. 4. **Count the prime cost you tell us** If your own prime cost answer is certainly over the ceiling for your type (“65% or more” at a quick service, say), that gap is money too. 25–50% of the points over are counted as recoverable, from the least your answer allows to its working figure: the same planning assumption as with your real numbers. The food, drink and labor problems are where that gap comes from, so this line counts only the part of it they don't already count, and the total stays under the cap. An answer right at the ceiling, or one that straddles it, stays something to watch, with no money. 5. **Count habits at half when prime cost is on target** If your own prime cost answer is at or under the ceiling for your type, the food, drink and labor habit problems count at half, and the report says so: “Your prime cost is within target, so this is counted at half.” 6. **Turn the shares into dollars** Each range is applied to the sales figure and rounded, and the figures under each total are fitted so they add up to it exactly. 7. **With your real numbers** In the report with your real numbers, the sales you type in replace the working figure, and the costs you type in decide those findings instead of your answers. When a cost runs over its range, 25–50% of the points over are counted as recoverable: a planning assumption, since no restaurant wins back every point. Delivery is never sized at more than half of what the apps actually kept. What the apps really keep is your app sales minus what the apps paid you, divided by those sales, with both figures before sales tax and tips. A payout that still holds sales tax the app passed on to you makes the share look too low: $1,000 of food plus $80 of tax at a 25% commission pays out $830, which would read as 17% instead of 25%. ## What the AI does, and what it never does The free check uses no AI. In the paid report, Google's Gemini writes the words around the figures. The figures come from the formulas on this page. ### The AI writes - The one-line verdict at the top - A short summary - The paragraph on your prime cost - A title and a short explanation for each of your costliest issues - Why each fix in your plan suits your restaurant, and what it goes after ### The AI never - Sets your grade, a dollar figure, a target or a percentage - Picks your fixes or writes their steps: those come word for word from a playbook of standard restaurant fixes, chosen by code for your answers - Sees your email address or payment details - Touches the free check ### How we check what the AI writes The AI only helps with the wording. Before any of it reaches your report, our software checks it against three fixed rules: - Every number must come from our own calculations, copied exactly. The AI can't add a number or change one. - Short, plain sentences: none longer than 22 words, and none of the 24 buzzwords on our banned list. - Nothing that contradicts your answers. For example, it can't call your staff idle when your labor cost is within the normal range. If the writing breaks a rule, the AI tries again. If a part still breaks one, we replace that part with simple wording built from your own numbers, so no report goes out with a broken rule. The AI is Google's Gemini; what it receives, and what it never sees, is on our [service providers](https://restaurantdoctorai.com/subprocessors) page. ## How the formulas are tested - **The grade and the estimates** Automated tests score thousands of random sets of answers. They check that every total equals the figures under it, no total passes its cap, and a worse answer never gives a better grade or a smaller estimate. - **The calculators** Each one is checked against examples worked out by hand, and most also on thousands of generated numbers. Where a calculator and the paid report work out the same figure, they must give the same answer. - **The check on the AI** Its tests feed it drafts that break each rule, to make sure every one is caught, and check that the plain wording that replaces a broken part keeps every rule. ## What the check can't know - Your books: It works from multiple-choice answers, so every figure is an estimate: a range, not a measurement. - Your restaurant up close: It can't see your menu, your prices, your lease, your team or your local market. - Norms outside the US: The targets are set for US restaurants. They run close to the norms in Canada, the UK, Ireland, Australia and New Zealand; anywhere else, treat them as a rough guide. - What you'll win back: Recoverable money is a planning assumption. No restaurant recovers every point. *What every report says* This diagnostic is an operational benchmarking assessment for managerial decision support. It is not an audit opinion, tax advice or legal counsel. ## Sources Named in plain text, without logos: listing an organization doesn't mean it endorses us. ### Behind the targets and the estimates Used to set and check the targets for each type of restaurant, and the delivery figures. - National Restaurant Association [Release on its 2025 Restaurant Operations Data Abstract](https://restaurant.org/research-and-media/media/press-releases/new-resource-from-national-restaurant-association-provides-insights-into-operational-realities/) Used for: Checking the prime cost ceilings against what the median counter-service restaurant actually runs. Published August 20, 2025 - National Restaurant Association [Food cost in 2024, from the same data abstract](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/) Used for: Checking the food and drink cost ranges against the median full-service and counter-service restaurant. Published September 10, 2025 - National Restaurant Association [Labor costs and profitability in 2024, from the same data abstract](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/) Used for: Checking the labor ranges against the median restaurant that made a profit. Published October 8, 2025 - National Restaurant Association [Occupancy costs in 2024, from the same data abstract](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-occupancy-costs-were-more-than-5-of-sales-in-2024/) Used for: Checking the rent ranges. The Association counts utilities on a separate line, so its figures run lower than ours. Published September 3, 2025 - ENERGY STAR (US Environmental Protection Agency) [Ask the experts: cutting utility costs in a commercial kitchen](https://www.energystar.gov/products/ask-the-experts/how-cut-utility-costs-your-commercial-kitchen) Used for: Why our rent ranges include utilities: energy alone is a real share of a restaurant's sales. - DoorDash and Uber Eats Their US merchant plan pages Used for: The 15–30% commission that delivery findings quote for US restaurants. Checked September 2026 ### Behind the calculators The plate cost calculator and the menu calorie counter copy their food figures from these public datasets, and never retype them. - US Department of Agriculture [FoodData Central, SR Legacy](https://fdc.nal.usda.gov/download-datasets) Used for: Calories and nutrients in the calorie counter; portion weights and densities in the plate cost calculator. April 2018 release, the last - US Department of Agriculture [Table of Cooking Yields for Meat and Poultry, Release 2(Download)](https://ndownloader.figshare.com/files/43365402) Used for: How much meat and poultry weigh after cooking. 2014 - US Department of Agriculture [Food Buying Guide for Child Nutrition Programs](https://foodbuyingguide.fns.usda.gov/) Used for: Cooked yields for rice, pasta, beans and lentils, and the sizes of common cans and jars. - US Bureau of Labor Statistics [Average Price Data, US city average](https://download.bls.gov/pub/time.series/ap/) Used for: Store-shelf prices for common ingredients, shown with their month, until you type your own. Through August 2026 - US Department of Agriculture, Economic Research Service [Fruit and Vegetable Prices](https://www.ers.usda.gov/data-products/fruit-and-vegetable-prices) Used for: Store prices per pound for fruit, vegetables and beans, shown with their year. 2023 prices - US Department of Agriculture, Agricultural Marketing Service [Market News: boxed beef, chicken and weekly grocery-store ad reports](https://www.ams.usda.gov/market-news) Used for: Wholesale and store-ad prices for ingredients the other price lists don't cover, labeled as such. Reports of October 2, 2026 ### In the guides and the blog Each [guide](https://restaurantdoctorai.com/guides) and [blog post](https://restaurantdoctorai.com/blog) links the sources behind its figures, right where it uses them. ## Changes 1. October 14, 2026 Cafés and coffee shops are now a type of their own, with targets of their own (see the table above): they were graded as quick service. The page and the guides also now call the targets what they are, RestaurantDoctorAI's targets, some of which match published figures and the rest are our own. No other type's grade or number changed. 2. October 11, 2026 A prime cost answer that is certainly over the ceiling for your type now counts as money too, only the part the food, drink and labor findings don't already count. The top sales range is split in two, $120,000 to $200,000 and over $200,000, so a bigger restaurant's figures aren't read too low. 3. October 8, 2026 The report with your real numbers asks for delivery app payouts without the sales tax or tips the apps passed on to you, so what the apps kept isn't read too low. 4. October 6, 2026 This page: how the check grades, where the targets come from, what the AI does and what the check can't know. 5. October 1, 2026 The first free calculators, built on the same targets as the check. 6. September 30, 2026 Every rate checked against published sources, including the National Restaurant Association's latest figures. Waste, recipe costing, overtime and fixed shifts now count for less, and counter-service targets and rent ranges match today's medians. Every draft the AI writes is checked before it goes in a report, and the costs you type in decide those findings instead of your answers. 7. September 29, 2026 The report with your real numbers: type in last month's sales and costs, and the report uses them. 8. September 26, 2026 Owners outside the US can pick their country, with figures in their own currency. 9. September 24, 2026 First version of the check, with totals capped so the estimates stay believable. ## See where your restaurant stands 12 questions, free, with no POS or bank logins. [Get my free grade](https://restaurantdoctorai.com/#diagnostic) --- title: "Free restaurant templates" description: "Free printable and spreadsheet templates for a restaurant's cash-out, tip-out, prep list and par sheet. Print the PDF or open the .xlsx. No email asked." canonical: "https://restaurantdoctorai.com/templates" --- # Free restaurant templates Sheets for the jobs a restaurant does every night and every week: closing the drawer, sharing tips, and setting what to prep and order. Print the PDF or open the spreadsheet, with no email asked, and check the math with your own numbers before you rely on it. - **Restaurant cash-out sheet** Close each cash drawer at the end of the day: expected cash from the Z report, over or short, and the deposit, so you can spot a gap the same night. [Open the template: Restaurant cash-out sheet](https://restaurantdoctorai.com/templates/restaurant-cash-out-sheet) [PDF to print: Restaurant cash-out sheet](https://restaurantdoctorai.com/downloads/restaurant-cash-out-sheet.pdf) [.xlsx for Excel or Google Sheets: Restaurant cash-out sheet](https://restaurantdoctorai.com/downloads/restaurant-cash-out-sheet.xlsx) - **Tip-out sheet and calculator** Share a server's tips by percentage, or split a tip pool by hours, to the cent, so you can show each person how their share was worked out. [Open the template: Tip-out sheet and calculator](https://restaurantdoctorai.com/templates/tip-out-sheet) [PDF to print: Tip-out sheet and calculator](https://restaurantdoctorai.com/downloads/tip-out-sheet.pdf) [.xlsx for Excel or Google Sheets: Tip-out sheet and calculator](https://restaurantdoctorai.com/downloads/tip-out-sheet.xlsx) - **Prep list and par sheet** Set a par for each item from what you use, then see what to prep and order, so you can order to a number you worked out rather than by eye. [Open the template: Prep list and par sheet](https://restaurantdoctorai.com/templates/prep-list-and-par-sheet) [PDF to print: Prep list and par sheet](https://restaurantdoctorai.com/downloads/prep-list-and-par-sheet.pdf) [.xlsx for Excel or Google Sheets: Prep list and par sheet](https://restaurantdoctorai.com/downloads/prep-list-and-par-sheet.xlsx) ## What a restaurant has to post and keep Some paper isn't a choice. The federal labor posters, and the pay, tip and food safety records a restaurant keeps, are in one guide, each with its free official source: [What a restaurant has to post and keep](https://restaurantdoctorai.com/guides/restaurant-required-posters-and-records). The prime cost, labor cost, break-even, budget and plate cost calculators also download their results as a spreadsheet. See all the [free restaurant calculators](https://restaurantdoctorai.com/calculators). 45-second profit check See where your restaurant stands Answer 12 quick questions and get your restaurant's grade and its biggest estimated loss, free, in about 45 seconds, so you know where to start. [Start the free check](https://restaurantdoctorai.com/#diagnostic) --- title: "Restaurant cost targets by type of restaurant" description: "Prime cost, food, labor and rent targets for quick service, fast casual, full service, fine dining, bars and cafés, with the NRA's 2024 medians." canonical: "https://restaurantdoctorai.com/restaurant-types" --- # Restaurant cost targets by type of restaurant Updated October 14, 2026 What's healthy depends on how you serve. A counter needs fewer paid hours per dollar of sales than a dining room with servers, and a bar's drinks call for a different cost target than a kitchen's food. Here are the targets RestaurantDoctorAI uses for each type, the published medians beside them, and how to tell which type you are. ## Targets by type of restaurant | Type | Prime cost | Food and drink | Labor | Rent and utilities | | --- | --- | --- | --- | --- | | [Quick service](https://restaurantdoctorai.com/restaurant-types/quick-service) | 60% or less | 26–32% | 26–31% | 10% or less | | [Fast casual](https://restaurantdoctorai.com/restaurant-types/quick-service#fast-casual) | 60% or less | 28–33% | 27–32% | 10% or less | | [Full service / casual dining](https://restaurantdoctorai.com/restaurant-types/full-service) | 65% or less | 30–34% | 30–35% | 12% or less | | Fine dining | 70% or less | 32–38% | 34–38% | 12% or less | | [Bar, taproom or lounge](https://restaurantdoctorai.com/restaurant-types/bars) | 57% or less | 20–26% | 25–32% | 12% or less | | [Café or coffee shop](https://restaurantdoctorai.com/restaurant-types/cafes) | 60% or less | 24–32% | 28–34% | 10% or less | *RestaurantDoctorAI's targets, as a % of net sales: the ones its free profit check grades against. The quick service and full service prime cost ceilings match the lines [RestaurantOwner.com](https://www.restaurantowner.com/public/Why-Prime-Cost-Is-the-Most-Important-Number-That-Should-Be-on-Your-PL.cfm) gives. No independent source publishes ceilings for fine dining, bars or cafés: those three are our own, and the café one sits at the quick-service line. Rent and utilities are shown as a ceiling: under it is fine, however low.* ## What the typical restaurant reported for 2024 The National Restaurant Association (NRA) surveyed more than 900 restaurants for its 2025 Restaurant Operations Data Abstract. It splits them two ways: full service, and limited service, which counts quick service and fast casual together. It publishes no free figures for fine dining or bars. | NRA 2024 median | Full service | Limited service | | --- | --- | --- | | Pretax income | 2.8% | 4.0% | | Prime cost | Not published free | 65% | | Food and soft drinks | 32.0% | 32.4% | | Labor, benefits in | 36.5% | 31.7% | | Occupancy | 5.7% | 5.2% | *NRA medians for 2024, % of sales: [pretax income and prime cost](https://restaurant.org/research-and-media/media/press-releases/new-resource-from-national-restaurant-association-provides-insights-into-operational-realities/), [food](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/), [labor](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/) and [occupancy](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-occupancy-costs-were-more-than-5-of-sales-in-2024/), checked on October 10, 2026. Its food figure leaves alcohol out.* The typical restaurant misses some of our targets, and the type pages say where. The ceilings are healthy lines, not typical ones: the median full-service restaurant kept 2.8% of sales before taxes. ## Which type are you? - [Quick service See its targets](https://restaurantdoctorai.com/restaurant-types/quick-service) - [Fast casual See its targets](https://restaurantdoctorai.com/restaurant-types/quick-service#fast-casual) - [Full service / casual dining See its targets](https://restaurantdoctorai.com/restaurant-types/full-service) - [Fine dining Grade yours free](https://restaurantdoctorai.com/?type=fine_dining#diagnostic) - [Bar, taproom or lounge See its targets](https://restaurantdoctorai.com/restaurant-types/bars) - [Café or coffee shop See its targets](https://restaurantdoctorai.com/restaurant-types/cafes) - [Not sure which? The free check describes each](https://restaurantdoctorai.com/#diagnostic) Go by how guests order and are served, then check the drinks. The free check describes each type like this, and assumes the share of sales drinks make up shown here. | Type | How it works | Drinks | | --- | --- | --- | | [Quick service](https://restaurantdoctorai.com/restaurant-types/quick-service) | Counter or kiosk, fast turnover | 5% | | [Fast casual](https://restaurantdoctorai.com/restaurant-types/quick-service#fast-casual) | Order at the counter, served at the table | 8% | | [Full service / casual dining](https://restaurantdoctorai.com/restaurant-types/full-service) | Servers, moderate check | 20% | | Fine dining | Chef-driven, higher check | 30% | | [Bar, taproom or lounge](https://restaurantdoctorai.com/restaurant-types/bars) | Drinks are about half of sales or more | 55% | | [Café or coffee shop](https://restaurantdoctorai.com/restaurant-types/cafes) | Drinks lead, with pastries and light food | 65% | *How the free check describes each type, and the share of sales it assumes drinks make up.* - Guests order and pay at a counter or kiosk and take the food themselves: quick service. - Guests order at a counter and staff bring the food to the table: fast casual. - Servers take the order at the table: full service. If the menu is chef-driven and the check is high, fine dining. - Espresso and other drinks lead, with pastries and light food: a café or coffee shop. - Drinks, mostly beer, wine and spirits, bring in about half of sales or more: a bar, taproom or lounge, whatever the kitchen serves. ## Pizzerias, diners and food trucks These have no separate targets yet. Here's the closest type, and what to watch. | Type | Closest type | | --- | --- | | Takeout and delivery pizzeria | Quick service | | Pizzeria with servers | Full service | | Diner | Full service | | Food truck | None: the check isn't built for trucks | *RestaurantDoctorAI has no separate targets for these yet.* - **Pizzerias.** Go by service: counter and delivery is quick service, servers at the table is full service. If apps bring in a real share of orders, see [what delivery commissions cost you](https://restaurantdoctorai.com/guides/delivery-app-commissions). More in [pizzerias on the quick-service page](https://restaurantdoctorai.com/restaurant-types/quick-service#pizzerias). - **Diners.** Servers at the table and a moderate check: full service, whatever the hours. See [diners on the full-service page](https://restaurantdoctorai.com/restaurant-types/full-service#diners). - **Food trucks.** The free check asks about a dining room and rent, so it isn't built for trucks. The [break-even](https://restaurantdoctorai.com/calculators/break-even) and [prime cost](https://restaurantdoctorai.com/calculators/prime-cost) calculators still work with a truck's numbers. ## Pages by type Each has its targets beside what's published, an example month, where the money goes, and the free check set to that type. - [Full service / casual dining Full-service restaurant costs: targets and 2024 medians Targets for full service and casual dining beside the NRA's 2024 medians, an example month, where the money goes, and tipped wages.](https://restaurantdoctorai.com/restaurant-types/full-service) - [Quick service · Fast casual Quick-service restaurant costs: targets and 2024 medians Targets for quick service and fast casual beside the NRA's 2024 medians, an example month, where the money goes, and California's chain wage.](https://restaurantdoctorai.com/restaurant-types/quick-service) - [Bar, taproom or lounge Bar, pub and taproom costs: targets and where money goes Our own targets for bars, pubs and taprooms, an example month, and the problems behind the bar: pours, comps and shrink.](https://restaurantdoctorai.com/restaurant-types/bars) - [Café or coffee shop Café and coffee shop costs: targets and where money goes Our own targets for cafés and coffee shops, an example month, and four places to check: cups and milk, card fees on small sales, pastry waste and idle hours.](https://restaurantdoctorai.com/restaurant-types/cafes) Fine dining has no page yet. Its targets are in the table above. For a bar's drinks on their own, see [bar pour cost](https://restaurantdoctorai.com/guides/bar-pour-cost). 45-second profit check Grade your restaurant against its type Answer 12 quick questions, starting with your type of restaurant, and get your grade and your costliest issue, free, in about 45 seconds, so you know where to start. [Start the free check](https://restaurantdoctorai.com/#diagnostic) Start with your type already picked: [Quick service](https://restaurantdoctorai.com/?type=quick_service#diagnostic) · [Fast casual](https://restaurantdoctorai.com/?type=fast_casual#diagnostic) · [Full service / casual dining](https://restaurantdoctorai.com/?type=full_service#diagnostic) · [Fine dining](https://restaurantdoctorai.com/?type=fine_dining#diagnostic) · [Bar, taproom or lounge](https://restaurantdoctorai.com/?type=bar_lounge#diagnostic) · [Café or coffee shop](https://restaurantdoctorai.com/?type=cafe#diagnostic) --- title: "Full-service restaurant costs: targets and 2024 medians" description: "Prime cost, food, labor and rent targets for full service and casual dining, the NRA's 2024 medians beside them, an example month and tipped wages." canonical: "https://restaurantdoctorai.com/restaurant-types/full-service" published: "2026-10-10" updated: "2026-10-14" --- # Full-service restaurant costs: targets and 2024 medians By [Felix S. Pavel](https://restaurantdoctorai.com/about) · Updated October 14, 2026 · 7 min read In a full-service restaurant, servers take the order at the table and bring it out. That service is paid hours, so labor takes a bigger share of each dollar than at a counter, and every other target moves with it. Here are the targets RestaurantDoctorAI uses for full service and casual dining, what full-service restaurants reported for 2024, and where the money usually goes. The quick answer Healthy targets for a full-service restaurant: prime cost 65% or less, food and drink 30–34%, labor 30–35%, and rent and utilities 12% or less of sales. Prime cost % = (food and drink cost + labor cost) ÷ net sales × 100 | Full service | RestaurantDoctorAI's target | | --- | --- | | Prime cost | 65% or less | | Food and drink | 30–34% | | Labor, payroll taxes and benefits in | 30–35% | | Rent and utilities | 12% or less | The typical full-service restaurant kept 2.8 cents of each sales dollar before taxes in 2024, and 1.1 cents if it sold under $2 million a year ([National Restaurant Association](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/higher-volume-restaurants-reported-lower-food-cost-ratios-in-2024/)). The targets are the ones the free profit check grades against. ## Our targets beside the 2024 medians The National Restaurant Association (NRA) surveyed more than 900 restaurants for its 2025 Restaurant Operations Data Abstract, on 2024. Its free pages give these medians for full service. The second column is ours. | Line | RestaurantDoctorAI's target | NRA 2024 median | Split by the NRA | | --- | --- | --- | --- | | Prime cost | 65% or less | Not published free | — | | Food and drink | 30–34% | 32.0% (food and soft drinks) | 31.0% at $2M+, 33.7% under | | Labor | 30–35% | 36.5% | 34.2% if profitable, 42.9% at a loss | | Rent and utilities | 12% or less | 5.7% | 6.0% in cities, 5.5% in suburbs | | Pretax income | No target | 2.8% | 4.3% at $2M+, 1.1% under | *% of net sales. RestaurantDoctorAI's targets are its own. The medians are the NRA's for 2024: [pretax income](https://restaurant.org/research-and-media/media/press-releases/new-resource-from-national-restaurant-association-provides-insights-into-operational-realities/), [food](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/), [by sales volume](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/higher-volume-restaurants-reported-lower-food-cost-ratios-in-2024/), [labor](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/) and [occupancy](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-occupancy-costs-were-more-than-5-of-sales-in-2024/), each checked on October 10, 2026.* **Prime cost.** The NRA doesn't publish a free full-service prime cost median. Our 65% ceiling is the line [RestaurantOwner.com](https://www.restaurantowner.com/public/Why-Prime-Cost-Is-the-Most-Important-Number-That-Should-Be-on-Your-PL.cfm) gives for table service: prime cost “no more than 65 percent of total sales.” **Labor.** The top of our range, 35%, sits just above the 34.2% median of full-service restaurants that made a profit. The median across all of them, 36.5%, is over it: the typical full-service restaurant misses this target, and the ones that lost money ran 42.9%. Labor is where profit and loss split. **Rent and utilities.** We print a ceiling, 12% or less, not a range. The NRA's median was 5.7%, but its page doesn't say what the figure counts, and ours counts rent, charges on top of rent and utilities together. Under the ceiling is fine, however low. Above it, rent squeezes profit. ## Why full-service numbers online disagree Before you compare your numbers with one you read, check what it counts. Four things move it: - **Alcohol.** The NRA's food figure leaves alcohol out. Our food and drink target counts it in. Where a full-service restaurant serves alcohol, it averaged 21% of total sales, the NRA told a federal regulator in 2023 ([its comments](https://restaurant.org/issues-and-advocacy/comments-updates-to-trade-practice-regulations)). So the two figures measure different things. - **What counts as labor.** The NRA's labor medians include benefits. Our targets include payroll taxes and benefits. A figure for wages alone reads lower than either. - **Sales volume.** Full-service restaurants selling $2 million or more a year ran food at 31.0% and kept 4.3% before taxes. Under $2 million: 33.7% and 1.1%. A national figure blends both. - **Vendor ranges.** Software and payment companies publish their own ranges. [Restaurant365](https://www.restaurant365.com/blog/what-is-the-average-profit-margin-for-a-restaurant/), a software vendor, says full-service margins “can fall into the 3-5% profit margin range” and names no survey. Treat a range with no source as a guess. ## An example month, worked through This is an example, not an average: one full-service restaurant with $90,000 in sales a month, food and drink at 32%, inside the 30–34% range, and labor at the NRA's 36.5% median. *Example full-service restaurant, one month* - Net sales: $90,000 - Food and drink, 32%: $28,800 - + Labor, 36.5%: $32,850 - = Prime cost, 68.5%: $61,650 - Ceiling at 65% of sales: $58,500 **Over the ceiling, a month:** $3,150 That's $37,800 a year, and all of it is labor: food and drink is inside its range. With labor at 33%, inside its range too, prime cost would be 65%, right at the ceiling. Each point of sales here is $900 a month. [Free calculator Work out yours: prime cost calculator Your food and drink, labor and net sales for a week or a month, against the full-service ceiling.](https://restaurantdoctorai.com/calculators/prime-cost) ## Where full-service restaurants lose money Each of these is common at full service. Each has one first fix and one tool or guide for it. ### Labor that doesn't flex with the room Servers, runners, bussers and hosts are scheduled for the covers you expect. A slow night still pays for all of them. Profitable full-service restaurants ran labor at 34.2% of sales; loss-making ones, 42.9%. **The fix:** set a cut line, the sales per labor hour below which the first person goes home, and check it on the hour. [How to set your cut line](https://restaurantdoctorai.com/guides/sales-per-labor-hour). ### Tipped wages that rise by law Tipped staff are a full-service cost a counter barely has, and state and city minimums change on set dates. **The fix:** work out what the next raise costs a year before it lands, and price for it. [Tipped minimum wage by state, with a raise calculator](https://restaurantdoctorai.com/guides/tipped-minimum-wage). ### Sales too low for the fixed costs Full-service restaurants under $2 million in sales kept a median 1.1% before taxes, against 4.3% for bigger ones. A dining room's rent and salaried staff cost the same on a quiet week. **The fix:** know the sales a month needs to pay every bill, and the covers a day that takes. [Break-even calculator](https://restaurantdoctorai.com/calculators/break-even). ### Plate costs that drift up between menu changes Full-service menu prices rose 3.5% in the year to August 2026, against 2.2% for groceries ([Bureau of Labor Statistics](https://www.bls.gov/news.release/cpi.t02.htm)). With prices already up, a price rise is a poor fix for a recipe nobody has re-costed. **The fix:** re-cost your five best sellers with this month's supplier prices before you change a price. [Plate cost calculator](https://restaurantdoctorai.com/calculators/recipe-cost). ### Drinks poured by eye Alcohol averaged 21% of sales where full-service restaurants serve it. Free pours and unrung drinks come straight off that line. **The fix:** measure every pour with a jigger, and count the bar each week. [How to work out and fix pour cost](https://restaurantdoctorai.com/guides/bar-pour-cost). ### Rent set for a busier year A lease doesn't move when sales do. Above 12% of sales, rent and utilities squeeze profit. **The fix:** work out your share on last month's sales, and the sales your rent needs, before the lease comes up. [How much rent a restaurant can afford](https://restaurantdoctorai.com/guides/restaurant-rent-percentage). ## Tipped wages: the full-service line item Under the federal rule, an employer may pay tipped staff a cash wage of $2.13 an hour and count up to $5.12 of their tips toward the $7.25 minimum, if tips make up the difference. Staff count as tipped when they get more than $30 a month in tips. Many states set more. These states require the full minimum before tips, so there's no tip credit at all: - Alaska: $14.00 an hour before tips. - California: $16.90 an hour before tips. A higher city or county minimum comes first. - Minnesota: $11.41 an hour before tips. - Montana: $10.85 an hour before tips. For businesses with yearly sales over $110,000. Smaller businesses the federal law doesn't cover: $4.00. - Nevada: $12.00 an hour before tips. - Oregon: $15.55 an hour before tips. Standard rate. Portland metro: $16.80. Nonurban counties: $14.55. - Washington: $17.13 an hour before tips. From the US Department of Labor's [table of tipped minimum wages](https://www.dol.gov/agencies/whd/state/minimum-wage/tipped) (revised July 1, 2026), checked on October 10, 2026. A city can set more than its state. Check your state's labor office before you change pay. Where there's no tip credit RestaurantOwner.com notes that states with no tip credit make a 65% prime cost harder to reach, and still treats it as the line. Where servers earn the full minimum, the schedule and the menu prices carry the difference. [Every state's tipped wage, and the 2027 changes](https://restaurantdoctorai.com/guides/tipped-minimum-wage). ## Diners: full service, long hours A diner with servers at the booths and the counter is full service in the free check: servers and a moderate check, whatever the hours. If guests order and pay at a counter and carry their own food, pick [quick service](https://restaurantdoctorai.com/restaurant-types/quick-service) instead. There are no separate diner targets. Three problems matter most: - **A full crew through the slow hours.** Breakfast and lunch can be busy and the afternoon quiet. If, when a shift gets slow, “everyone works their scheduled shift anyway”, the free check estimates $720 to $1,530 a month on the example's $90,000 in sales. [Set a cut line](https://restaurantdoctorai.com/guides/sales-per-labor-hour). - **A long menu costed once.** Egg prices in stores fell 20.6% in the year to September 2026 ([Bureau of Labor Statistics](https://data.bls.gov/timeseries/CUUR0000SEFH)). Prices that move that far, either way, leave last year's plate costs wrong. Re-cost your breakfast best sellers with this month's supplier prices. [Plate cost calculator](https://restaurantdoctorai.com/calculators/recipe-cost). - **Waste nobody counts.** A long menu means many open containers. If you answer “we don't track it”, the check estimates $450 to $1,080 a month. A daily log the chef signs off shows over-prep while you can still change the order. ## Questions full-service owners ask ### What is a good profit margin for a full-service restaurant? The median full-service restaurant kept 2.8% of sales before taxes in 2024: 4.3% at $2 million or more in sales, 1.1% under it, by the NRA's figures. Prime cost under 65% is what leaves room for one. ### Is 36% labor too high for a full-service restaurant? It's over our 30–35% range and close to the NRA's 36.5% median for all full-service restaurants. Profitable ones ran 34.2%. If food and drink is low, prime cost can still come in under 65%, so check the two together. ### What is the 30/30/30 rule for restaurants? A rule of thumb with no published source: 30% of sales to food and drink, 30% to labor, 30% to everything else. At full service it rarely holds. The NRA's labor median alone was 36.5%. [How the rule compares with real restaurants](https://restaurantdoctorai.com/guides/restaurant-prime-cost#30-30-30-rule). ### Is a 32% food cost good for a full-service restaurant? Yes, it's inside our 30–34% range for food and drink, and the NRA's median for food and soft drinks was 32.0%. A good food cost still needs labor inside its range to keep prime cost under 65%. ### Is casual dining the same as full service? In the free check, yes: “Full service / casual dining” means servers and a moderate check. Fine dining, chef-driven with a higher check, has its own targets, which are [on the restaurant types page](https://restaurantdoctorai.com/restaurant-types). ## Check your own numbers Two free calculators for the numbers on this page. No sign-up. [Free calculator Labor cost calculator Labor as a share of sales, payroll taxes and benefits in, against the 30–35% range, plus your cut line.](https://restaurantdoctorai.com/calculators/labor-cost) [Free calculator Break-even calculator The sales a month needs to pay every bill, and the covers a day that takes.](https://restaurantdoctorai.com/calculators/break-even) 45-second profit check Grade your full-service restaurant Answer 12 quick questions, with full service already picked, and get your grade and your biggest estimated loss, free, in about 45 seconds, so you know where to start. [Start the free check](https://restaurantdoctorai.com/?type=full_service#diagnostic) Want every issue with its dollar estimate and a plan? [See what the full report adds](https://restaurantdoctorai.com/pricing). RestaurantDoctorAI's targets are its own; every outside figure is linked where it's quoted. General information, not financial, tax or legal advice. --- title: "Quick-service restaurant costs: targets and 2024 medians" description: "Prime cost, food, labor and rent targets for quick service and fast casual, the NRA's 2024 medians beside them, an example month and what to fix first." canonical: "https://restaurantdoctorai.com/restaurant-types/quick-service" published: "2026-10-10" updated: "2026-10-14" --- # Quick-service restaurant costs: targets and 2024 medians By [Felix S. Pavel](https://restaurantdoctorai.com/about) · Updated October 14, 2026 · 8 min read At a quick-service restaurant, guests order at a counter or kiosk and the food comes fast. Fewer paid hours stand behind each sale than in a dining room, so the targets sit lower than full service's. Here are the targets RestaurantDoctorAI uses for quick service, what counter-service restaurants reported for 2024, and where the money goes. Fast casual has its own section below. The quick answer Healthy targets for a quick-service restaurant: prime cost 60% or less, food and drink 26–32%, labor 26–31%, and rent and utilities 10% or less of sales. Prime cost % = (food and drink cost + labor cost) ÷ net sales × 100 | Quick service | RestaurantDoctorAI's target | | --- | --- | | Prime cost | 60% or less | | Food and drink | 26–32% | | Labor, payroll taxes and benefits in | 26–31% | | Rent and utilities | 10% or less | The typical limited-service restaurant, counters of every kind, spent 65 cents of each sales dollar on food, drink and labor in 2024, and kept 4.0% before taxes ([National Restaurant Association](https://restaurant.org/research-and-media/media/press-releases/new-resource-from-national-restaurant-association-provides-insights-into-operational-realities/)). The targets are the ones the free profit check grades against. ## Our targets beside the 2024 medians The National Restaurant Association (NRA) surveyed more than 900 restaurants for its 2025 Restaurant Operations Data Abstract, on 2024. It reports counter service as one group, limited service, so its medians cover quick service and fast casual together. No neutral source we found splits the two. | Line | RestaurantDoctorAI's target: quick service | RestaurantDoctorAI's target: fast casual | NRA 2024 median | | --- | --- | --- | --- | | Prime cost | 60% or less | 60% or less | 65% | | Food and drink | 26–32% | 28–33% | 32.4% (food and soft drinks) | | Labor | 26–31% | 27–32% | 31.7% | | Rent and utilities | 10% or less | 10% or less | 5.2% | | Pretax income | No target | No target | 4.0% | *% of net sales. RestaurantDoctorAI's targets are its own. The medians are the NRA's for limited service in 2024: [prime cost and pretax income](https://restaurant.org/research-and-media/media/press-releases/new-resource-from-national-restaurant-association-provides-insights-into-operational-realities/), [food](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/), [labor](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/) and [occupancy](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-occupancy-costs-were-more-than-5-of-sales-in-2024/), each checked on October 10, 2026.* **The typical counter restaurant misses our targets, and we say so.** Its 65% prime cost is 5 points over our 60% ceiling, and its 31.7% labor is 0.7 points over the top of our quick-service range. The ceiling is a healthy line, not a typical one: the median operator kept only 4.0% before taxes. Our 60% matches the line [RestaurantOwner.com](https://www.restaurantowner.com/public/Why-Prime-Cost-Is-the-Most-Important-Number-That-Should-Be-on-Your-PL.cfm) gives: keep quick-service prime cost “at 60 percent of total sales or less.” **Labor is where profit and loss split.** Limited-service restaurants that made a profit ran labor at 30.0% of sales. The ones that lost money ran 34.1%. **Rent and utilities.** We print a ceiling, 10% or less, not a range. The NRA's median was 5.2%: 6.0% in cities, 5.0% in suburbs and 3.2% in small towns. Its page doesn't say what the figure counts, and ours counts rent, charges on top of rent and utilities together. Under the ceiling is fine, however low. ## Why quick-service numbers online disagree - **Margins with no source.** [Restaurant365](https://www.restaurant365.com/blog/what-is-the-average-profit-margin-for-a-restaurant/), a software vendor, says “fast casual and quick-service restaurants average 6–9%” and names no survey. The NRA's median for limited service was 4.0% before taxes. - **Payroll taxes and benefits.** Ours count both; the NRA's medians count benefits. A “labor %” built from hourly wages alone, common in scheduling apps, comes out lower. - **Two types counted as one.** A burger counter and a made-to-order bowl shop land in the same limited-service group, so its medians mix the two. - **Drinks.** The NRA's food figure leaves alcohol out; our food and drink target counts it in. At counters it matters less: where limited-service restaurants serve alcohol, it averaged 6% of sales, the NRA told a federal regulator in 2023 ([its comments](https://restaurant.org/issues-and-advocacy/comments-updates-to-trade-practice-regulations)). ## An example month, worked through This is an example, not an average: one quick-service restaurant with $50,000 in sales a month, food and drink at 30%, inside the 26–32% range, and labor at the NRA's 31.7% limited-service median. *Example quick-service restaurant, one month* - Net sales: $50,000 - Food and drink, 30%: $15,000 - + Labor, 31.7%: $15,850 - = Prime cost, 61.7%: $30,850 - Ceiling at 60% of sales: $30,000 **Over the ceiling, a month:** $850 That's $10,200 a year. Food is fine; labor is 0.7 points over the 31% top of its range. At this volume each point of sales is $500 a month. [Free calculator Work out yours: prime cost calculator Your food and drink, labor and net sales for a week or a month, against the quick-service ceiling.](https://restaurantdoctorai.com/calculators/prime-cost) ## Where quick-service restaurants lose money Six problems that show up at counters, each with the first thing to do and where to work it out. ### Staff on the clock between rushes A counter's sales come in peaks. Staff who stay on through the afternoon lull are paid for idle hours. **The fix:** stagger start times so extra staff arrive with the rush, and send the first person home when sales per labor hour drop under your cut line. [How to set your cut line](https://restaurantdoctorai.com/guides/sales-per-labor-hour). ### Staff who leave In accommodation and food services, an average 4.2% of workers quit each month in 2025, against 2.0% across all jobs ([Bureau of Labor Statistics](https://data.bls.gov/timeseries/JTU720000000000000QUR)). Each new hire means paid hours of training. **The fix:** train a second person on each station so a call-out doesn't become a double. [Seven steps to lower labor cost](https://restaurantdoctorai.com/guides/restaurant-labor-cost#how-to-reduce-labor-cost). ### Menu prices already up Limited-service menu prices rose 3.2% in the year to August 2026, against 2.2% for groceries ([Bureau of Labor Statistics](https://www.bls.gov/news.release/cpi.t02.htm)). A price rise is a poor fix for a recipe nobody has re-costed. **The fix:** re-cost your best sellers with this month's supplier prices, and fix portions before prices. [Plate cost calculator](https://restaurantdoctorai.com/calculators/recipe-cost). ### Delivery orders priced like counter orders The delivery apps take a commission on every order, before any promotions. At counter prices, it comes out of a margin that was already thin. **The fix:** work out the app price that leaves each dish what it earns at the counter. [Delivery app price calculator](https://restaurantdoctorai.com/calculators/delivery-app-price). ### Prime cost nobody checks weekly Food and labor can each look fine on their own and still add up past 60%: the tops of our quick-service ranges add to 63%. **The fix:** add the two every week, against sales for the same days. [How to track prime cost](https://restaurantdoctorai.com/guides/restaurant-prime-cost). ### Rent that outgrew the sales Above 10% of sales, rent and utilities squeeze a counter's thin margin. **The fix:** work out the sales your rent needs before the lease comes up. [How much rent a restaurant can afford](https://restaurantdoctorai.com/guides/restaurant-rent-percentage). ## California's $20 fast food wage covers chains, not independents California's fast food minimum wage, $20.00 an hour from April 1, 2024, applies to a limited-service restaurant that's part of a chain of at least 60 establishments nationwide, whether its owner runs the brand or is a franchisee ([California Department of Industrial Relations](https://www.dir.ca.gov/dlse/Fast-Food-Minimum-Wage-FAQ.htm), checked October 10, 2026). An independent counter with fewer locations isn't covered by it. The state minimum still applies: $16.90 an hour in the US Department of Labor's table, or more where a city or county sets more. Hiring next to a chain You may still be hiring against chain jobs that pay the fast food rate. Work out what a raise would cost a year before you match one, with the [labor cost calculator](https://restaurantdoctorai.com/calculators/labor-cost). ## Fast casual: the same group, its own targets In the free check, fast casual means order at the counter, served at the table; quick service means counter or kiosk, fast turnover. Food made to order on a line takes more prep per order, so our fast-casual ranges sit a little higher: food and drink 28–33% against 26–32%, labor 27–32% against 26–31%. The prime cost ceiling is the same, 60% or less, so fast casual has less room for both to run high at once. No neutral source splits fast casual from quick service, so these ranges are RestaurantDoctorAI's own. The NRA's limited-service medians, 32.4% food and 31.7% labor, sit inside both fast-casual ranges and just over the top of quick service's. - **Plate cost on build-your-own orders.** Every extra scoop changes the plate. Cost your most popular build, not the menu photo's. [Plate cost calculator](https://restaurantdoctorai.com/calculators/recipe-cost). - **Labor on the line.** A made-to-order line needs its stations staffed at the rush. Set the cut line from the top of the fast-casual range, 32%. [Work out your cut line](https://restaurantdoctorai.com/calculators/labor-cost#cut-line). [Run the free check as a fast-casual restaurant](https://restaurantdoctorai.com/?type=fast_casual#diagnostic), with fast casual already picked. ## Pizzerias: counter, takeout and delivery A pizzeria where guests order at the counter, by phone or online, then pick up or get it delivered, is quick service in the free check: counter or kiosk, fast turnover. A sit-down pizzeria with servers at the tables is [full service](https://restaurantdoctorai.com/restaurant-types/full-service). There are no separate pizzeria targets. Three problems matter most: - **Delivery apps at menu prices.** The apps' US plans take up to 30% of each order, by [DoorDash's](https://merchants.doordash.com/en-us/products/pricing) and [Uber Eats'](https://merchants.ubereats.com/us/en/pricing/) pricing pages, as listed on October 8, 2026. If 15% to 35% of your sales come through them at regular menu prices, the free check estimates $500 to $1,000 a month on the example's $50,000 in sales. [Delivery app price calculator](https://restaurantdoctorai.com/calculators/delivery-app-price). - **Your own drivers' miles.** The IRS business mileage rate is 76 cents a mile from July 1, 2026 ([IRS](https://www.irs.gov/tax-professionals/standard-mileage-rates)). Use it as a yardstick: price the delivery fee against the miles a run takes. - **Pies costed once.** Cheese in stores cost 3.0% less in September 2026 than a year before ([Bureau of Labor Statistics](https://data.bls.gov/timeseries/CUUR0000SEFJ02)). Prices move both ways, so re-cost your best-selling pies with this month's supplier prices. [Plate cost calculator](https://restaurantdoctorai.com/calculators/recipe-cost). ## Questions quick-service owners ask ### What is a good profit margin for a quick-service restaurant? The median limited-service restaurant kept 4.0% of sales before taxes in 2024, by the NRA's figures. Higher figures you'll see online usually come from vendors with no survey behind them. Prime cost under 60% is what leaves room for more. ### Is 25% labor cost high for a fast food restaurant? No, it's under our 26–31% quick-service range and under the 30.0% median of profitable limited-service restaurants. Check that it counts payroll taxes, benefits, managers' salaries and any hours you work yourself unpaid. ### Is a 32.8% food cost acceptable? At quick service it's just over our 26–32% range; at fast casual it's inside 28–33%. The NRA's limited-service median for food and soft drinks was 32.4%. Either way, check it with labor: prime cost under 60% is the line that matters. ### What is the 30/30/30 rule for restaurants? A rule of thumb with no published source: 30% of sales to food and drink, 30% to labor, 30% to everything else. Food and labor at 30% each make 60%, our quick-service ceiling. The typical limited-service restaurant spent 65%. [How the rule compares with real restaurants](https://restaurantdoctorai.com/guides/restaurant-prime-cost#30-30-30-rule). ### Is fast casual the same as quick service? The NRA counts them together, as limited service. The free check treats them as two types with their own targets, a point or two apart. If guests order at a counter and staff bring the food to the table, you're fast casual. ## Check your own numbers Free calculators for the numbers on this page. No sign-up. [Free calculator Labor cost calculator Labor as a share of sales, payroll taxes and benefits in, against your range, plus your cut line.](https://restaurantdoctorai.com/calculators/labor-cost) [Free calculator Delivery app price calculator The app price that covers the commission on each dish.](https://restaurantdoctorai.com/calculators/delivery-app-price) 45-second profit check Grade your quick-service restaurant Answer 12 quick questions, with quick service already picked, and get your grade and your biggest estimated loss, free, in about 45 seconds, so you know where to start. [Start the free check](https://restaurantdoctorai.com/?type=quick_service#diagnostic) Want every issue with its dollar estimate and a plan? [See what the full report adds](https://restaurantdoctorai.com/pricing). RestaurantDoctorAI's targets are its own; every outside figure is linked where it's quoted. General information, not financial, tax or legal advice. --- title: "Bar, pub and taproom costs: targets and where money goes" description: "Prime cost, drink, labor and rent targets for bars, pubs and taprooms, why they're our own, an example month, and the problems: pours, comps and shrink." canonical: "https://restaurantdoctorai.com/restaurant-types/bars" published: "2026-10-14" updated: "2026-10-14" --- # Bar, pub and taproom costs: targets and where money goes By [Felix S. Pavel](https://restaurantdoctorai.com/about) · Updated October 14, 2026 · 7 min read A bar, pub or taproom earns most of its money from drinks. In the free check, a bar is a place where drinks are about half of sales or more, whatever the kitchen serves, and the check assumes about 55%. With sales that lean on drinks, a bar gets its own targets, lower than a restaurant's. Here are the targets RestaurantDoctorAI uses for bars, why they're our own, and the three problems the check looks for behind the bar: pours, comps and shrink. The quick answer Healthy targets for a bar, pub or taproom: prime cost 57% or less, food and drink 20–26%, labor 25–32%, and rent and utilities 12% or less of sales. Prime cost % = (food and drink cost + labor cost) ÷ net sales × 100 | Bar, pub or taproom | RestaurantDoctorAI's target | | --- | --- | | Prime cost | 57% or less | | Food and drink | 20–26% | | Labor, payroll taxes and benefits in | 25–32% | | Rent and utilities | 12% or less | No independent source publishes targets for bars. These are RestaurantDoctorAI's own, the ones the free profit check grades a bar against. ## Our targets, and why no median sits beside them The full-service and quick-service pages put our targets beside the National Restaurant Association's 2024 medians. For bars there is nothing to put there: the NRA publishes no free figures for bars, and we found no government or university source that does. So the second column says what is published, and where. | Line | RestaurantDoctorAI's target | What's published | | --- | --- | --- | | Prime cost | 57% or less | Nothing for bars. RestaurantOwner.com gives 65% for table service and 60% for quick service. | | Food and drink | 20–26% | Nothing for bars. The NRA's food figures leave alcohol out. | | Labor | 25–32% | Nothing for bars | | Rent and utilities | 12% or less | Nothing for bars | | Drinks' share of sales | About 55% | 21% at full-service restaurants that serve alcohol | *% of net sales. RestaurantDoctorAI's targets are its own. Sources: [RestaurantOwner.com](https://www.restaurantowner.com/public/Why-Prime-Cost-Is-the-Most-Important-Number-That-Should-Be-on-Your-PL.cfm) and the NRA's [2023 comments](https://restaurant.org/issues-and-advocacy/comments-updates-to-trade-practice-regulations), checked on October 14, 2026.* **Prime cost: the 57% ceiling is ours.** No independent source publishes one for bars. It sits under our 65% full-service line because our food and drink range for bars, 20–26%, sits under full service's, 30–34%: a bar's sales lean on drinks. The labor ranges are close, 25–32% against 30–35%. **The two ranges can't both run at the top.** The tops of our bar ranges add to 58%, a point over the ceiling. If labor runs high because the bar is open late, food and drink has to sit lower, and the other way round. **Rent and utilities.** We print a ceiling, 12% or less, not a range. Under it is fine, however low. Above it, rent squeezes profit. Liquor liability insurance isn't part of this line or of any of our targets: keep it as its own line, and know what it costs a year before the policy renews. ## Why bar numbers online disagree - **Pour cost is not food and drink cost.** Pour cost divides what the bar poured by bar sales alone. Our 20–26% target counts the kitchen too, against all sales. A bar with a kitchen reads higher on ours than on its pour cost. [Where pour cost fits in your numbers](https://restaurantdoctorai.com/guides/bar-pour-cost#where-pour-cost-fits-in-your-numbers). - **Vendor figures.** The pour cost and bar margin figures you'll find online mostly come from point-of-sale and inventory software companies, worked out from their own customers. Our pour cost guide's target comes from one of them, and it says so. No independent survey publishes one. - **The drink mix.** Beer, wine and spirits each cost something different per drink. When guests order more of one and less of another, a blended pour cost moves with nothing lost at all. Work it out by category as well as in total. - **Drinks with or without alcohol.** Some figures count only alcohol. Ours count every drink and every plate. ## An example month, worked through This is an example, not an average: one bar with $50,000 in sales a month, $27,500 of it from drinks, food and drink cost at 28% and labor at 30%, inside its 25–32% range. *Example bar, one month* - Net sales: $50,000 - Food and drink, 28%: $14,000 - + Labor, 30%: $15,000 - = Prime cost, 58%: $29,000 - Ceiling at 57% of sales: $28,500 **Over the ceiling, a month:** $500 That's $6,000 a year, and all of it is food and drink: 2 points over the 26% top of its range. Each point of sales here is $500 a month. Behind the bar is the first place to look: for a bar this size that free-pours, the free check's own estimate is $605 to $1,210 a month. [Free calculator Work out yours: prime cost calculator Your food and drink, labor and net sales for a week or a month, against the bar ceiling.](https://restaurantdoctorai.com/calculators/prime-cost) ## Where bars lose money The free check asks one question about the bar: “How do you control drink pours?” The first three problems below are what it looks for. The dollar figures are its own estimates for the $50,000 example bar, before it caps the total of all of them. ### Pours nobody measures A free pour runs a little long, and on more than half of a bar's sales a little adds up. If your answer is “Free pours, no pour-cost checks”, the check counts $605 to $1,210 a month at these sales. **The fix:** every spirit pour through a jigger, and the standard pour for each drink written on a card behind the bar. [How to work out and fix pour cost](https://restaurantdoctorai.com/guides/bar-pour-cost). ### Comps nobody writes down A round for a regular, a remade drink, a staff drink after close. Each is fine on its own. Unrecorded, they look exactly like theft or over-pouring in the count. **The fix:** ring every free drink into the POS as a comp, or log it on a sheet with a manager's initials, so the count can tell them apart. [Why pour cost drifts upward](https://restaurantdoctorai.com/guides/bar-pour-cost#why-pour-cost-drifts). ### Shrink found only when a bottle runs dry If your answer is “Spot checks when bottles run out early”, the missing product is found only once it's gone. The check counts $303 to $605 a month at these sales. **The fix:** count the bar every week, on the same day, by the same person. Compare your top 20 bottles used with what the POS says was sold, and kegs used with pints rung in. [Fixing a high pour cost](https://restaurantdoctorai.com/guides/bar-pour-cost#fixing-a-high-pour-cost). ### A full crew on a quiet night Bars run late, and a Tuesday at 11 p.m. costs the same per hour as a Saturday. If everyone works their scheduled shift anyway, the check counts $400 to $850 a month at these sales. **The fix:** set a cut line, the sales per labor hour below which the first person goes home, and check it on the hour. [How to set your cut line](https://restaurantdoctorai.com/guides/sales-per-labor-hour). ### Prices raised to cover a gap Drink prices at bars and restaurants rose 2.6% in the year to September 2026, while alcohol in stores rose 0.2% (Bureau of Labor Statistics: [away from home](https://data.bls.gov/timeseries/CUUR0000SEFX), [at home](https://data.bls.gov/timeseries/CUUR0000SEFW)). Store prices aren't what a bar pays its distributor, but guests see both. **The fix:** close the pour and comp gaps before you raise a price. ### Fewer guests who drink 54% of US adults said they drink alcohol in 2025, down from 62% in 2023 ([Gallup](https://news.gallup.com/poll/693362/drinking-rate-new-low-alcohol-concerns-surge.aspx)). **The fix:** cost your alcohol-free drinks and your food the way you cost a cocktail, so the guest who doesn't drink still earns the seat. [Plate cost calculator](https://restaurantdoctorai.com/calculators/recipe-cost). ## Bartender pay, tips included The median bartender earned $16.51 an hour in May 2025, and $16.01 in drinking places, against $17.07 in full-service restaurants. These figures include tips ([Bureau of Labor Statistics](https://www.bls.gov/ooh/food-preparation-and-serving/bartenders.htm)). What you pay before tips depends on your state's tipped wage rules, and our labor target counts what you pay, payroll taxes and benefits in, never the tips guests leave. Before a raise lands State and city tipped minimums change on set dates. Work out what the next one costs a year before it does. [Tipped minimum wage by state, with a raise calculator](https://restaurantdoctorai.com/guides/tipped-minimum-wage). ## Taprooms and brewpubs The Brewers Association calls a brewery that sells 25% or more of its beer on site a taproom brewery when it runs no significant food service, and a brewpub when it does ([its definitions](https://www.brewersassociation.org/statistics-and-data/craft-beer-industry-market-segments/)). In the free check, go by sales, not the license: - A taproom where beer and other drinks are about half of sales or more: bar, taproom or lounge. - A brewpub whose kitchen sells more than its taps, with servers at the tables: [full service](https://restaurantdoctorai.com/restaurant-types/full-service). - Either way, compare kegs used with pints rung in each week. A keg that pours fewer pints than it should is shrink. ## Questions bar owners ask ### What is a good prime cost for a bar? RestaurantDoctorAI uses 57% or less. It's our own line: no independent source publishes one for bars. Prime cost under it leaves room for rent, insurance and profit. ### What pour cost should a bar aim for? The pour cost guide gives the target we use and where it comes from, with the formula for a week and for one drink. [Bar pour cost: how to calculate and fix it](https://restaurantdoctorai.com/guides/bar-pour-cost). ### Is a pub with a kitchen a bar or a restaurant? Go by sales. If drinks bring in about half or more, pick bar, taproom or lounge in the free check. If food brings in more and servers take orders at the table, pick full service. ### Is a 30% labor cost high for a bar? No, it's inside our 25–32% range. Check it with food and drink: at 30% labor, food and drink has to stay at 27% or under to keep prime cost at 57% or less. ## Check your own numbers Free calculators for the numbers on this page. No sign-up. [Free calculator Labor cost calculator Labor as a share of sales, payroll taxes and benefits in, against the 25–32% range, plus your cut line.](https://restaurantdoctorai.com/calculators/labor-cost) [Free calculator Break-even calculator The sales a month needs to pay every bill, insurance included.](https://restaurantdoctorai.com/calculators/break-even) 45-second profit check Grade your bar Answer 12 quick questions, with bar, taproom or lounge already picked, and get your grade and your biggest estimated loss, free, in about 45 seconds, so you know where to start. [Start the free check](https://restaurantdoctorai.com/?type=bar_lounge#diagnostic) Want every issue with its dollar estimate and a plan? [See what the full report adds](https://restaurantdoctorai.com/pricing). RestaurantDoctorAI's targets are its own; every outside figure is linked where it's quoted. General information, not financial, tax or legal advice. --- title: "Café and coffee shop costs: targets and where money goes" description: "Prime cost, food and drink, labor and rent targets for cafés and coffee shops, why they're our own, an example month and four places the money goes." canonical: "https://restaurantdoctorai.com/restaurant-types/cafes" published: "2026-10-14" updated: "2026-10-14" --- # Café and coffee shop costs: targets and where money goes By [Felix S. Pavel](https://restaurantdoctorai.com/about) · Updated October 14, 2026 · 7 min read A café or coffee shop earns most of its money from drinks: about 65% of sales in the free check, with pastries and light food beside them. Small tickets and a cup under every drink set its costs apart from a restaurant's, so a café gets its own targets. Here are the targets RestaurantDoctorAI uses for cafés, why they're our own, and four places to check for lost money. The quick answer Healthy targets for a café or coffee shop: prime cost 60% or less, food and drink 24–32%, labor 28–34%, and rent and utilities 10% or less of sales. Prime cost % = (food and drink cost + labor cost) ÷ net sales × 100 | Café or coffee shop | RestaurantDoctorAI's target | | --- | --- | | Prime cost | 60% or less | | Food and drink | 24–32% | | Labor, payroll taxes and benefits in | 28–34% | | Rent and utilities | 10% or less | We found no café-only cost survey. These are RestaurantDoctorAI's own, the ones the free profit check grades a café against. ## Our targets, and what's published beside them The National Restaurant Association (NRA) reports counter service as one group, limited service, so its medians mix cafés with burger counters and bowl shops. We found no café-only cost survey from the NRA, a government agency or a point-of-sale company that shows its sample. So the table puts two published sets beside our targets: the NRA's medians for 2024, and one coffee chain's results. | Line | RestaurantDoctorAI's target | NRA median, limited service, 2024 | Dutch Bros shops, 2025 | | --- | --- | --- | --- | | Prime cost | 60% or less | 65% | 52.8% | | Food and drink | 24–32% | 32.4% (food and soft drinks) | 25.9% (drinks, food and packaging) | | Labor | 28–34% | 31.7% | 26.9% | | Rent and utilities | 10% or less | 5.2% | Not comparable | *% of net sales. RestaurantDoctorAI's targets are its own. NRA medians for limited service in 2024: [prime cost](https://restaurant.org/research-and-media/media/press-releases/new-resource-from-national-restaurant-association-provides-insights-into-operational-realities/), [food](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-operators-kept-food-cost-ratios-in-check-in-2024/), [labor](https://restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/elevated-labor-costs-had-a-significant-impact-on-restaurant-profitability-in-2024/) and [occupancy](https://www.restaurant.org/research-and-media/research/restaurant-economic-insights/analysis-commentary/restaurant-occupancy-costs-were-more-than-5-of-sales-in-2024/). Dutch Bros' company-operated shops in fiscal 2025, from its [results release](https://www.sec.gov/Archives/edgar/data/1866581/000186658126000003/a2025-q4_ex991.htm) of February 12, 2026. Checked on October 14, 2026.* **Prime cost: the 60% ceiling is ours.** It is the line [RestaurantOwner.com](https://www.restaurantowner.com/public/Why-Prime-Cost-Is-the-Most-Important-Number-That-Should-Be-on-Your-PL.cfm) gives for quick service, and the ceiling we use there. The NRA's median counter restaurant ran 65%, 5 points over it. A café at the median is over our line, and we say so. The ceiling is a healthy line, not a typical one. [How to track prime cost every week](https://restaurantdoctorai.com/guides/restaurant-prime-cost). **Dutch Bros is a floor, not a target.** It is a drive-thru chain with high sales in each shop, and its 52.8% prime cost shows what that model reached in 2025. An independent café, with fewer cups an hour and no drive-thru, shouldn't read it as its own target. **Food and drink: 24–32%.** The low end sits near Dutch Bros' 25.9%. The top sits just under the NRA's limited-service median, 32.4%. Coffee drinks and pastries cost different shares of their prices, so work out each group too. [Café food cost for resold pastries](https://restaurantdoctorai.com/blog/cafe-food-cost-resold-items). **Labor: 28–34%.** It runs from below the NRA's median for limited-service restaurants that made a profit, 30.0%, up to just under the median for those that lost money, 34.1%. Small tickets keep it a little above our quick-service range. **The two ranges can't both run at the top.** The tops of our café ranges add to 66%, 6 points over the ceiling. When labor runs high, food and drink has to sit lower, and the other way round. **Rent and utilities.** We print a ceiling, 10% or less, not a range. The NRA's median was 5.2%. Under the ceiling is fine, however low. **Drinks.** We assume drinks are about 65% of a café's sales. Starbucks' company-operated stores sold 73% beverages and 23% food in fiscal 2025, by its [annual report](https://www.sec.gov/Archives/edgar/data/829224/000082922425000114/sbux-20250928.htm). We found no measured figure for independent cafés. ## Why café numbers online disagree - **Chains and independents.** Public filings come from chains with hundreds of shops and drive-thrus. A single café buys and sells at a different scale. - **What the margin counts.** Many café margin figures don't say whether the owner's own pay was taken out first. A margin that leaves it out reads higher than one that counts it. - **Coffee cost against coffee sales.** Guides often quote coffee cost against coffee sales alone, and may leave cups out. Our food and drink target counts everything on the menu, against all sales. ## An example month, worked through This is an example, not an average: one café with $40,000 in sales a month, $26,000 of it from drinks. Food and drink cost is 30% and labor 33%: each inside its range. *Example café, one month* - Net sales: $40,000 - Food and drink, 30%: $12,000 - + Labor, 33%: $13,200 - = Prime cost, 63%: $25,200 - Ceiling at 60% of sales: $24,000 **Over the ceiling, a month:** $1,200 That's $14,400 a year, and neither line is outside its own range: only the two together are over. At this volume each point of sales is $400 a month. [Free calculator Work out yours: prime cost calculator Your food and drink, labor and net sales for a week or a month, against the café ceiling.](https://restaurantdoctorai.com/calculators/prime-cost) ## Where cafés lose money Four places to check. The dollar figures are worked examples from round numbers, not measured costs: use your own invoices and statements. ### Cups, lids, sleeves and milk on every drink Every drink carries a cup, a lid, a sleeve and milk, whatever the coffee costs. On a low price they are a large share. Say cup, lid and sleeve cost 30 cents together and the café sells 4,000 hot drinks a month: $1,200, or 3 points of the example's $40,000 in sales. Milk adds to it: steaming 12 ounces for an 8-ounce drink pays for 4 ounces nobody drinks, every time. **What to check:** read your cup and lid invoices against the drinks you sold, and weigh the milk poured and the milk left in the pitcher for your three best sellers. [Plate cost calculator](https://restaurantdoctorai.com/calculators/recipe-cost). ### Card fees on small sales Most card processors charge a percentage plus a fixed amount on each sale. The fixed part is the same on a coffee as on a large order, so it takes more of a small sale. Say it is 10 cents: that is 2.5 points of a $4 sale and under 0.1 point of a $150 one. These are example amounts, not any processor's rate. **What to check:** your statement for the percentage and the per-sale fee, and what the month's fees come to against its card sales. [Credit card fees calculator](https://restaurantdoctorai.com/calculators/credit-card-fees). ### Pastries and food that don't sell by close A pastry, a sandwich or a pot of soup that goes in the bin is food and drink cost with no sale against it. A bought-in pastry costs the same whether it sells or not. **What to check:** count what is left at close, by item, for a few weeks, and compare it with what you baked or ordered. Set the next order from those counts. [Costing resold pastries](https://restaurantdoctorai.com/blog/cafe-food-cost-resold-items). ### Staff on the clock between rushes A café's sales come in waves: the morning rush, then a quiet afternoon. If the same crew works every hour anyway, the free check counts $320 to $680 a month at $40,000 in sales. **What to check:** sales per labor hour for each hour of the day, against your cut line. [How to set your cut line](https://restaurantdoctorai.com/guides/sales-per-labor-hour). ## Questions café owners ask ### What is a good profit margin for a coffee shop? We found no café-only survey that shows how it was measured. The NRA's median limited-service restaurant kept 4.0% of sales before taxes in 2024. Higher figures online often don't say what they count. Prime cost under 60% is what leaves room for rent and profit. ### Is a 30% food and drink cost good for a café? It is inside our 24–32% range. Check it with labor: at 33% labor, food and drink at 30% puts prime cost over the ceiling, so it has to sit at 27% or under. ### Am I a café in the free check? Pick café or coffee shop if drinks lead, with pastries and light food. If guests order at a counter and the food is mostly meals, pick [quick service](https://restaurantdoctorai.com/restaurant-types/quick-service) or fast casual. If servers take orders at the table, pick [full service](https://restaurantdoctorai.com/restaurant-types/full-service). ## Check your own numbers Free calculators for the numbers on this page. No sign-up. [Free calculator Plate cost calculator What one drink or pastry costs with cup, lid and milk in, against its price.](https://restaurantdoctorai.com/calculators/recipe-cost) [Free calculator Restaurant budget calculator A month of sales, costs and rent laid out line by line.](https://restaurantdoctorai.com/calculators/restaurant-budget) [Free calculator Credit card fees calculator What card fees cost on your sales, with small tickets counted.](https://restaurantdoctorai.com/calculators/credit-card-fees) 45-second profit check Grade your café Answer 12 quick questions, with café or coffee shop already picked, and get your grade and your biggest estimated loss, free, in about 45 seconds, so you know where to start. [Start the free check](https://restaurantdoctorai.com/?type=cafe#diagnostic) Want every issue with its dollar estimate and a plan? [See what the full report adds](https://restaurantdoctorai.com/pricing). RestaurantDoctorAI's targets are its own; every outside figure is linked where it's quoted. General information, not financial, tax or legal advice.