Plate Costing: What the Data Says Against What the Sector Keeps Repeating

How much does food cost really weigh on a restaurant's cash register?

It weighs a median 32.0% of sales in full service and 32.4% in limited service, per the National Restaurant Association with 2024 close, and that figure works as a ceiling on the business.

What the aggregate buries is internal dispersion. I have sorted sales matrices by descending turnover and the finding repeats venue after venue: six items above 45% that go out every single day, propped up by two appetizers at 18% almost nobody orders. Comfortable arithmetic, business split in two. Under one roof live an operation that earns and another that bleeds, and the second one runs on margin the owner believed was banked. At Masterestaurant that reading opens the first decision the 32% triggers: stop looking at the menu as a block and start looking at each item with its own yield.

The missing productivity: why Latin America's food MSME employs a lot and produces little

Sixty percent of the region's formal employment comes out of MSMEs, and the GDP they return barely reaches 25%, ECLAC figures with a very concrete translator in food service: nobody measures the economic unit being sold. Add the 47.6% labor informality the ILO reports in its Labour Overview 2025 and the frame is complete. An owner blind to the unit cost of the flagship dish cannot project formal payroll, cannot negotiate terms with a supplier, and when credit comes up has only a word to offer as collateral. SATE Institute walked into this analysis through that door: a kitchen's micro-operation ends up surfacing on the macro dashboards development banks watch. Four consequences run in chain from one unfilled recipe card, and the first is called productivity.

The commissions that eat your margin before you ever see it

Two external costs deform any costing done the old way, and the pricier one is not the one you watch. Card commissions average 2.35% per transaction, according to the Texas Restaurant Association in 2025, and swipe fees add up to nearly $187 billion a year in the United States, a National Restaurant Association figure. Delivery bites much harder: Rezku, in its 2026 fee review, puts DoorDash between 15% and 30% per order with a 30% standard marketplace rate, Uber Eats in that same band and Grubhub between 15% and 25%. Add it up. A dish costed at 32% that leaves through a platform charging 30% reaches 62% effective cost before payroll or rent. Do not negotiate the commission: cost a separate menu for the digital channel.

Waste is a costing figure, not an environmental problem

Treat shrinkage as a line in your costing, never as an environmental gesture. Food waste costs the U.S. restaurant industry close to $162 billion a year, per The Restaurant HQ's 2025 count, and in Latin America a loss equivalent to 11.6% of food production moves the question from conscience to solvency. Let me put it without softening: whatever your kitchen throws out, you already paid for it and never collected a cent. Your recipe card records 180 grams of tenderloin, the cook trims by eye and the portion leaves the pass at 210; true food cost lands 17% above the calculated one, and no spreadsheet will warn you. Without input traceability, on top of that, no loss-reduction program under SDG target 12.3 ever gets started.

A smaller market punishes whoever does not know his numbers

Technomic measured with 2024 data that the U.S. full-service segment runs about 18% smaller than in 2019, and the contraction did not spread the blow evenly: it took out operators who entered the pandemic without unit costing and came out to set prices by instinct. Let us run the counterfactual. Two twin locations in 2019, same menu. One raises prices a flat 12% in 2021 because input inflation bites; the other raises 22% on the eight highest-shrinkage items and freezes the rest. Three years on, the first lost its low-ticket crowd without recovering margin where it bled, and the second held traffic with a profitable mix. According to Diego F. Parra, restaurant consultant at Masterestaurant, twenty years of fieldwork across 43 countries settle one conclusion: whoever closes rarely closed for lack of sales.

Opening costs what it costs, and costing decides whether you get it back

Between $75,000 and $150,000 buys you the opening of a small takeout venue in the United States; median opening cost sits at $450 per square foot, with a range of $100 to $800 according to Square in 2024, and the 2025 bottom quartile closed at $175,500, that is $59 per square foot, per Rezku. Now, what those numbers hide: the investment comes back through MARGIN per dish multiplied by turnover, never through total sales. Bill whatever you bill: a venue with a real 41% food cost takes twice as long to return capital as one selling half that at 28%. Watching the daily cash count tells those two scenarios apart only when the flow has already dried up, and by then the conversation with the bank started crooked.

Costing so that a third party can read it

A costing that exists only in the owner's head makes the business illegible to outsiders, and that illegibility carries a market price. No bank running an MSME portfolio, no IDB Lab program, no impact fund rates risk on a hunch; all of them ask for a data series. Masterestaurant S.A.S., technology partner of the SATE model, turns yields, shrinkage, versioned sheets and monthly food cost variance into auditable records, and that is where the jump happens: the establishment stops being an applicant with no history and becomes a subject with structured information, which is the doorway to alternative scoring under SDG 9. There is a tension and I will not dodge it: filling sheets costs hours to a team already stretched. Sequence solves it, twenty items first, 80% of sales, the rest afterwards.

The 3 numbers you should tattoo on yourself

32.0%: full-service median food cost, National Restaurant Association for 2024. Measure each item separately this week and flag in red every one above 35%, however fond of the dish you happen to be. The second number is 30%, the standard marketplace rate Rezku attributes to DoorDash and Uber Eats in its 2026 review; with that on top, publishing your dining-room menu inside the app gives margin away, so build a digital menu with portions and prices calculated to absorb the bite. Left is the 47.6% regional labor informality the ILO reports in its Labour Overview 2025. Put the margin you rescue from costing into formalizing ONE position: whoever controls portions and receives deliveries. That single hire pays for all three numbers by itself.

The differences that change the decision

The formula fits on a single line, and that is not where the disagreement sits: the fight is over the unit of analysis. The myth costs the MENU as a block; reality costs the DISH as an economic unit, with its own yield, trim loss, rotation and price. A menu averaging 31% can hide half a dozen items at 45% that sell every day, subsidised by a couple of appetizers at 18% that move once a week. That arithmetic soothes so well that the operator tells the accountant the business is healthy while the till drains. The second difference stings more, because it asks who gets to read your numbers. As long as the figures travel by memory, the business stays illegible to a bank's credit committee, to IDB Lab, to the agency weighing co-financed equipment. Put them into versioned technical sheets, with supplier, date, gram weight and yield, and that same business starts producing evidence. Diego F.

The differences that change the decision — in practice

Parra frames it this way in his work with Masterestaurant: a restaurant without technical sheets is not informal, it is INVISIBLE, and nobody finances what cannot be seen. Then comes a difference of timing that hardly anyone puts on the table. Costing is an act; recosting is a system. Whoever sets food cost in January and reviews it in December is not costing badly, they cost once and assume eleven times. With protein and oil prices swinging several points inside one quarter, that assumption gets paid in real, measurable margin, and it resurfaces in August as an unexplained cash shortfall. What remains is the angle, which is where consultant and accountant part ways. One cares what it cost; the other cares which decision that cost triggers. A dish at 38% does not get cut by decree: if it drags Tuesday traffic and adds two beverages per table, its contribution margin can beat a 24% dish that only moves on weekends. Menu engineering exists to settle that tension, and its mandatory input is costing dish by dish.