6.8 points of EBITDA leaking away: how we set the target food cost by restaurant type and fixed a trattoria with the Restaurant Model Canvas

What is the correct target food cost by restaurant type?

The target food cost is NOT a universal 30%: it is set by type of operation. A high-margin pizzeria lives at 22-26%, a steakhouse tolerates 34-38% because of the protein load, and a fresh-product trattoria anchors near 28%.

Copying the number next door means setting a goal that does not match your cash register. In this case —an Italian trattoria with 14 tables, a 41 USD average ticket, dining room dominant at 72% of sales— the owner arrived running against an imported 30% that never added up. Full service carries high costs: according to ReFED (2024), full-service restaurants generate more than 43% of the total foodservice surplus. With perishable fresh product, that waste floor demands its own target. We set 28% theoretical, and that is when the profit the sales promised —but the bank account denied month after month— finally started to fit. The line the owner arrived with sums up the drama of thousands of operations: he was selling well and the money evaporated in production.

«We're billing better than ever and the bank is still empty»

The keyword of his anguish —why my restaurant loses money with good sales— is not solved by looking at the P&L at month's end. In six years of operation he had never separated theoretical cost from real cost. His declared food cost ran 35-37% against a mental target of 30%, and every point above 28% in a trattoria with a 41 USD ticket and 14 tables translates into thousands of dollars a year leaving inventory without passing through the register. The mistake I see over and over: confusing revenue with margin. The National Restaurant Association reported restaurant price inflation of 8.8% in March 2023, the highest in more than two decades; without daily control, that cost pressure eats profit before the owner sees it in the month-end income statement. The number that matters is not the real food cost by itself, but the DEVIATION between theoretical cost —what the dishes sold should cost per standard recipe— and real cost, what actually left inventory.

The gap between theoretical and real cost: that's where the leak lives

That gap is the leak: waste, over-portioning, unrecorded comps, petty theft. In this trattoria we measured a 7-point deviation: theoretical 28%, real 35%. According to The Restaurant HQ (2025), an average restaurant wastes between 4% and 10% of the inventory it buys; with Italian fresh product —tomato, basil, cheeses, fish of the day— the high end of that range is the norm, not the exception. Seven points on annual product sales running into the hundreds of thousands of dollars is not an accounting detail: it is the difference between positive EBITDA and a slow decapitalization the owner only notices once the bank is already empty. The tool we applied was the food cost variance dashboard of the Masterestaurant method: every dish with a costed standard recipe, theoretical cost calculated on real daily sales, and the deviation measured against the 28% target every morning, not at month's end.

The Masterestaurant method: standard recipe and daily deviation

Diego F. Parra puts it this way: the month-end P&L is a death certificate, not a dashboard; by the time you see food cost inflated in the income statement, the capital is already gone. We standardized 22 recipes covering 80% of sales, weighed real portions against the technical sheet, and logged every comp and waste event. At Masterestaurant we have seen this pattern in hundreds of full-service operations: control is not expensive software, it is the discipline of comparing theoretical against real every day and acting on the deviation before a full month of leakage piles up. The measurable result of the case: the trattoria cut its real food cost from 35% to 29% in fourteen weeks, closing the 7-point deviation to just 1 point over the 28% theoretical target. On a sales structure of 72% dining room and a 41 USD ticket, those 6 recovered points translated into operating profit that previously escaped in waste and over-portioning.

The result: from 35% real to 29% in fourteen weeks

The most expensive input was not technology: it was 22 standardized recipes and fifteen minutes of counting each morning. Price pressure was real —ground beef went from 4.56 to 5.63 USD per pound per USDA (2026) and eggs rose 21.9% in 2025 per USDA— but by measuring deviation daily the operation could adjust portions and purchasing before those increases eroded the margin. The register stopped lying: for the first time in six years, the P&L profit matched the bank balance. The transferable lesson is to set a target food cost by type of operation and measure the deviation daily, but the first step changes by your size. Small independent (1 location, owner on the floor): this week hand-cost the 10 recipes that are 70% of your sales and weigh real portions across one full service; that is where you will see your first gap. Mid-size (2-4 locations, with a chef): standardize technical sheets on the top 20-25 recipes and set up a weekly inventory count per location with a named owner.

Transferable lessons: your first step this week by size

Multi-site group (5+, with administration): implement the centralized daily theoretical-vs-real deviation dashboard and compare food cost across same-format locations to hunt the outlier. The principle is identical in all three —target by type, daily measurement of the gap—; only the system level changes. None needs costly software to start: it needs the discipline of comparing theoretical against real and acting on the deviation. This case is not a universal recipe, and it is worth marking its limits to avoid survivorship bias. First: in a limited-service format or high-margin pizzeria I would not expect the same jump, because its target food cost already lives at 22-26% and its fresh-product waste is much lower —in Canadá limited service is 46.4% of foodservice sales per Statistics Canadá (2024) precisely because of its lighter cost structure—; the lever there is different. Second: if the real leak were not waste but a structural pricing or menu-mix problem, standardizing recipes would not solve it.

Limits of this case: where I would NOT expect the same result

Third: in markets with extreme input inflation or protein volatility —recall the 8.8% peak from the National Restaurant Association (2023)—, a fixed 28% target can become unrealistic and menu prices should be reviewed before squeezing cost alone. The method diagnoses; the target number is always calibrated to context. Target food cost is NOT a universal number: a pizzeria lives at 22-26%, a steakhouse tolerates 34-38% for its protein cost, a fresh-product trattoria anchors near 28%. Copying the neighbor's 30% means setting a goal that doesn't match your operation type. The number that matters is not the actual food cost by itself, but the VARIANCE between theoretical cost (what the dishes sold should cost per standard recipe) and actual cost (what really left inventory). That gap is the leak: waste, over-portioning, unrecorded comps, pilferage. The month-end P&L is a death certificate, not a dashboard.

The difference almost nobody measures: theoretical vs actual cost

By the time you see inflated food cost on the income statement, the capital has already leaked. Control must be daily and anchored to a target by restaurant type, not monthly and anchored to an industry average.