Prime cost: the 6 mistakes draining your EBITDA and the method that stops the leak

Chapter 1 — Why does prime cost decide your EBITDA before any other lever?

Prime cost —food cost plus labor— is the sum that decides your EBITDA before price, marketing or average check, and almost nobody measures it against its theoretical cost.

With the industry net margin pinned at 3%–9% per Statista, there is no cushion: every mismanaged point of prime cost eats the whole profit. I have seen it in dozens of restaurants with healthy sales and a presentable P&L that still leave no cash behind. The problem is rarely selling too little. It is that prime cost gets read on average and at month-end, once the leak has already migrated to the income statement. Labor already weighs more than 25% of expenses in 2024, up from 23% in 2021 (Toast, 2024), and food cost creeps up per portion without anyone noticing. Two levers, one single figure that governs whether the business breathes or drowns. Traditional food cost tells you how much you spent; variance tells you how much you SHOULD have spent and where the difference leaked.

Chapter 2 — Food cost vs. variance: what does each one really tell you?

That gap, with the industry net margin at 3%–9% (Statista), is usually the entire profit of the month. Theoretical cost is what the standardized recipe says each plated dish should have cost;

actual cost is what left the inventory. Subtract one from the other and the leak appears: over-portioning, pilferage, waste, a supplier price that rose unannounced. A restaurant reporting a 30% average food cost can hide a real 38% in its protein category and offset it with beverages at 18%. The average lies by design. The profitable full-service operator closed 2024 with labor at 34.2% of sales versus 36.5% for the average (National Restaurant Association, 2024 data): 2.3 points that are almost all of its margin. Monthly control catches the leak 30 to 45 days late, once it has migrated to the P&L and nothing can be corrected; weekly variance catches it in 5 to 7 days, while you can still adjust portion, supplier or shift.

Chapter 3 — Why does monthly control always arrive 30 days late?

This is the difference between managing and performing an autopsy.

With the sector's pre-tax operating margin at a 10.66% average (NYU Stern / Damodaran, 2024 dataset), a three-point drift sustained for a month and a half before detection wipes out a third of operating profit. Cadence matters as much as the number. At Masterestaurant we measure prime cost by category and by shift, not by location and by month, because the leak is not homogeneous: it happens Friday on the grill or Sunday at the bar. The operator who closes inventory weekly and compares against theoretical fixes the bleeding while it is still a drop, not once it has flooded the basement. Controlling only food cost leaves out half the controllable cost, because labor already exceeds 25% of the restaurant's expenses in 2024 (Toast, 2024) and is as manageable as inventory. Prime cost exists precisely to integrate both legs: if you obsess over the dish spec but schedule shifts out of habit, you fix half the equation.

Chapter 4 — Does controlling only food cost leave out half the controllable cost?

The profitable full-service operator's labor was 34.2% of sales in 2024, against 36.5% for the average (National Restaurant Association, 2024 data);

those 2.3 points come from scheduling against real demand by time band, not from blindly cutting people. Diego F. Parra sums it up in a line I repeat in the boardroom: labor is not controlled by firing, it is controlled by scheduling. An oversized shift Tuesday at 3pm and a short one Saturday at 9pm cost as much as a food cost leak, and almost nobody adds them to prime cost where they belong. Blindly raising prices against 12% input inflation destroys check and table turns; menu engineering by contribution margin raises profit without touching the menu price. The difference is surgical. When the wholesale roaster captures around 67% of the margin per pound of coffee (Bellwether Coffee), passing that hit to the customer via price punishes visit frequency.

Chapter 5 — Does raising prices against input inflation destroy or protect the margin?

The way out is reordering the menu toward the dishes with the highest absolute contribution margin —not the highest percentage margin— and redesigning the affected dish spec.

In a stress simulation with input inflation of 5%, 12% and 20%, the restaurant that adjusts portion, swaps the side and repositions the dish on the menu protects the typical EBITDA of 12%–30% of sales (WhippleWood CPAs, 2026); the one that only raises prices loses traffic and ends up with more margin over fewer tables. Engineering, not reflex. The 90-day prime cost roadmap starts by measuring, not cutting: days 1–30, standardized dish specs and weekly inventory with theoretical cost by category; days 31–60, weekly variance by shift and labor rescheduling against demand; days 61–90, menu engineering and supplier renegotiation with the data in hand. The governing KPI is the theoretical-to-actual gap, targeting to close it below 2 points.

Chapter 6 — What does a 90-day roadmap for the board look like?

The ROI for the board is direct:

with a sector net margin of 3%–9% (Statista), recovering 2.3 points of prime cost —the same gap between profitable and average operator on labor (National Restaurant Association, 2024 data)— doubles or triples net profit without selling one extra dish. And it protects the sale multiple: a single-unit independent trades at 1.5x–3x SDE (Sofer Advisors), and that SDE is exactly what controlled prime cost fattens. The Masterestaurant method turns the P&L into a decision board, not a monthly autopsy. Traditional food cost tells you what you spent; variance tells you what you SHOULD have spent and where the difference leaked. With sector net margin at 3%-9% (Statista), that gap is the entire profit. Monthly control catches the leak 30-45 days late, once it has hit the P&L; weekly variance catches it in 5-7 days, while portion, supplier or shift can still be corrected.

Chapter 7 — The differences that decide your margin

Prime cost integrates food cost AND labor —labor already exceeds 25% of expenses (Toast, 2024)—; controlling food cost alone leaves out half the controllable cost. Menu engineering by contribution margin lifts profit without raising prices; raising prices blindly against 12% input inflation destroys ticket and table turnover.