Food cost: traditional method vs the Masterestaurant method

Best for the independent under fifteen tables: contribution margin, not percentage

If you run a single location with fewer than fifteen tables and nobody on your payroll costs recipes full time, the Masterestaurant CONTRIBUTION MARGIN method beats percentage food cost, and the arithmetic settles it: at 900 covers a month with an average margin of 9 USD per dish you gather 8,100 USD toward rent, payroll and utilities, while a mix averaging 12 USD leaves you 10,800 USD, and those 2,700 USD of difference come from selling something else, never from buying cheaper. The percentage would tell you both scenarios look fine as long as each one hovers near the industry median, which in 2024 stood at 32.0% of sales in full service according to the National Restaurant Association. Two operations identical on the dashboard, and one pays the rent while the other does not. Because one number averages the whole menu and erases the sales mix, which is exactly where a small operation's cash gets decided.

Why does the percentage mislead precisely when the place is small

The National Restaurant Association measured that full service restaurants under two million dollars in annual sales closed 2024 with food cost at 33.7%, against 31.0% for those at two million or more, and the easy reading —«small guys buy worse»— is barely half true: the small operator also has less volume to dilute one badly costed dish. A thirteen table place turning twice sells roughly 780 covers a month; if the most heavily pushed plate yields 4,100 pesos and the runner up yields 12,700, shifting a hundred sales from the first to the second adds 860,000 pesos of margin without touching a single supplier. No percentage will ever show you that lever. If your menu runs past forty references and your purchase prices move every two weeks, run a WEEKLY close by product family —protein, dairy, dry goods, beverage— instead of the full monthly inventory.

Best for long menus with unstable suppliers: weekly close by product family

The monthly one arrives late by design: a two point variance on 60,000 USD of sales has already eaten 1,200 USD by the time you see it, and that money does not get recovered, it gets recorded. With a weekly cut the window drops to seven days and the maximum damage falls to about 280 USD, a figure you can actually fix with a portion change or one call to the supplier. Context pushes the same way: the United States Department of Agriculture projected a 2.9% rise in food away from home prices for 2025, and nine out of ten operators named food cost as their main pressure, according to the National Restaurant Association. Contribution margin is the wrong tool in three concrete scenarios, and it is worth saying so before selling it. First, if you run low ticket fast casual with an eight item menu, the percentage is enough: with limited service food cost at a 32.4% median in 2024 (National Restaurant Association) and a short menu, the mix barely moves and margin tells you nothing new.

When NOT to choose the popular option?

Second, if your problem is waste rather than menu: United States foodservice generated 12.5 million tons of surplus food in 2024 according to ReFED, and the lever there is portion control and purchasing, not menu reordering.

Third, if poor cash management already has you cornered —82% of small business closures are associated with it, per the U.S. Bank study cited by Inc.—, cash first, menu engineering later. Four signals tell you the method being sold to you will not survive a real service, and all four surface in ten minutes. One: the system pushes kitchen payroll into the plate cost «to see the real cost»; that payroll is semi fixed and belongs in the break even calculation, and folding it into the plate produces a price that neither negotiates nor sells. Two: the dashboard shows a global percentage and no ranking by unit margin, meaning it shows you the average and hides the decision.

Red flags when comparing costing methods

Three: the recipe card carries no process waste or butchering yield, so your protein cost sits underestimated by somewhere between 8% and 15% from day one. Four: the report lands on the 8th of the following month. Data arriving eight days late on 2,000 USD of daily sales is history, not management. If you have already decided to move prices within the next ninety days, cost dish by dish using the REAL purchase price of the last four weeks, not the standard on the recipe card. The reason is cash: if your protein climbed 9% and you apply a flat 5% increase across the menu, you overcharge the vegetable plates —which did not climb— and keep losing on the meat ones, punishing demand where you had margin and repairing nothing where you bleed. On 60,000 USD of monthly sales with 40% concentrated in three meat dishes, that miscalibration costs close to 1,100 USD a month.

Best for anyone raising prices this quarter: plate costing with real purchase data

The reference ceiling is hard: 32% food cost per dish is the MAXIMUM tolerable, not the target, and the healthy industry band runs from 28% to 35% according to the National Restaurant Association. If nobody on your team is dedicated to costing —the case for most independents— the minimum that holds over time is three things and not one more: recipe cards with real waste for the ten dishes that make 70% of sales, a weekly cut of purchases against sales by family, and a monthly ranking of unit margin per dish sorted from highest to lowest. Diego F. Parra, twenty years between kitchens and boardrooms across 43 countries, repeats it in every Masterestaurant audit: the problem is almost never that food cost runs high, it is that nobody knows WHICH DISH runs high or what that ignorance costs per service. We call it CAPITAL LEAKAGE, and its signature is awkward: it never shows up as a line on the income statement, it shows up diluted in a margin that drops a point per quarter.

What happens if you change nothing this year?

Leave the dashboard as it is, percentage spotless and mix untouched, and the result is not a visible fall but an erosion dressed up as normality.

Assume 60,000 USD of monthly sales, food cost at 32% and purchase prices rising the 2.9% projected by the United States Department of Agriculture: without adjusting menu or portions, your monthly cost moves from 19,200 to 19,757 USD, some 557 USD taken straight out of margin. Twelve months later that is 6,684 USD, nearly three months of the typical American restaurant electricity bill, which Toast measured at roughly 2,300 USD per month. And you will not see it coming, because the percentage keeps reading 32 and change while every point costs more money. Start with the cheap move: pull today's unit margin ranking for your ten best sellers and compare it against what the floor team pushes.

Four differences that move the cash

Percentage measures purchasing efficiency; contribution margin measures your ability to pay the structure. At 900 covers a month with a 9 USD average margin, you have 8,100 USD for rent, payroll and utilities. At 12 USD, you have 10,800. Buying cheaper does not produce that gap; selling a different mix does. The traditional method arrives late by design. It closes with monthly inventory, and by then a 2-point variance on 60,000 USD of sales has already eaten 1,200 USD nobody will recover. A weekly close by product family cuts that window to seven days and turns a loss into a correction. Percentage pushes you to promote the wrong dish. A plate at 24% leaving 5 USD looks like a win on the dashboard; one at 30% leaving 13 looks like a problem. If servers push the first, every table that turns leaves 8 USD behind, literally.

Four differences that move the cash — in practice

The traditional approach blurs CapEx and OpEx and ends up burying depreciation inside plate cost. That dish comes out overcosted, the price goes up, turnover falls, and the fixed cost you were trying to cover now spreads across fewer covers, so the problem worsens on the next lap.