How food and labor costs shape your margin.

A cost-control textbook calls food and labor the two largest expenses in most foodservice operations^[1]. Before the pandemic, the National Restaurant Association reports, each took about 33 cents of a typical independent restaurant's sales dollar^[2]. In the association's figures for 2024, the median food and non-alcohol beverage cost was 32.0% of sales at full-service restaurants and 32.4% at limited-service ones^[3]. Median salaries, wages and benefits were 36.5% of sales at full-service and 31.7% at limited-service, up from about 33% and 28% in the association's 2010, 2013 and 2016 reports^[4].

The association puts median pre-tax income for full-service restaurants at 2.8% of sales in 2024^[5]. In the same data, the median labor cost was 34.2% of sales for full-service operators that made a profit and 42.9% for those that did not^[6]. On a 2.8% pre-tax margin, one point of food or labor cost is more than a third of pre-tax income.

The USDA reports that food-away-from-home prices, which include restaurant meals, rose 4.1% in 2024 and 3.8% in 2025, above their historical average of 3.5% a year^[7]. The Bureau of Labor Statistics puts average hourly earnings at US food services and drinking places at $22.17 in August 2026 (preliminary), against $21.25 a year earlier^[8]. Whether your prices kept up with your costs shows in prime cost as a share of sales.

Prime cost is the cost of food and beverage sold plus total labor cost. The same textbook puts it on the P&L as a measure of how well management controls those two costs^[1].

One can be traded for the other. Buying croissants par-baked moves cost from labor to food; laminating the dough in-house moves it back. Either one read alone can improve while the sum gets worse.

Labor means more than wages. For cost control it includes payroll taxes, workers' compensation, health insurance, retirement contributions and staff meals^[9]. Advisers writing for Baker Tilly recommend that full figure for prime cost^[10].

Published prime cost bands are rules of thumb, not measurements. One CPA firm gives 60% to 65% of sales for full-service and 55% to 60% for quick-service^[11]. The Baker Tilly advisers give 58% to 62%, and say a profit is hard to make above 65%^[10]. A third firm writes that there is no standard figure for labor on its own^[12]. None of these is a target for your shops.

Actual food cost for a period is opening inventory plus purchases minus closing inventory; divided by food sales, it gives the food cost percentage^[13]. One period's closing inventory is the next one's opening inventory^[13], so a wrong count misstates two periods, in opposite directions.

The textbook has a name for what the food sold should have cost with every standard met and nothing avoidable lost: attainable, standard or ideal product cost^[14]. Recipe cost is one way to estimate it: each item's recipe, priced ingredient by ingredient. The table's food cost is recipe cost.

The textbook treats actual cost of 100% to 110% of attainable as generally achievable; above 110%, the gap can point to waste, theft, spoilage, overportioning or inaccurate recipe costing^[14].

Labor has the same pair: scheduled hours and clocked hours, compared shift by shift^[15]. Clocked hours are in payroll.

When product moves between shops or out of the commissary, its cost has to leave the sending unit's cost of sales and enter the receiving unit's, so that each unit's food cost percentage is right^[14]. A tray of morning buns sent across town and never recorded makes the sending shop look expensive and the receiving shop cheap. The group total hides it.

The textbook notes that kitchens commonly make 5% to 10% more than they expect to sell, so they do not run out^[16]. Unsold carryover has to be logged and reused, or it is lost in storage and becomes waste that raises food cost^[16].

Donated food leaves without a sale, so it belongs on the count; unrecorded, it reads as waste. How a donation is treated for tax is a question for your accountant.

A missing price counted as zero makes recipe cost too low. The gap to actual then looks like waste in the kitchen, when the cause is the recipe costing^[14]. A recipe missing one price should have no cost, not a lower one.

Damia Labs builds back-office software for restaurant groups, one group at a time. Each shop counts its items once a day against its PAR, with deliveries received, transfers and donations on the same sheet. That is a stock count, not the inventory count behind actual food cost. The controller reads recipe food cost, rostered labor and prime cost by shop and by period. Transfers and donations become journal entries at recipe cost, prepared for Restaurant365 in its format.

None of these figures is an actual. Recipe cost is not invoice cost. Rostered labor is scheduled hours times a rate; it includes taxes and benefits only if the rate does, and it is not payroll. Purchases and payroll stay in your books, where your controller sets them against these figures. There is no target, only last period beside this one.