Food Cost Percentage Formula (With a Real Worked Weekly Example)

Every operator knows the food cost formula. Most run it wrong, and not because they can’t do the division. The math is simple. The inputs are where it falls apart, and that’s where I’ve watched good operators fool themselves for years.

The formula #

Food Cost % = COGS ÷ Food Sales × 100

COGS = Beginning Inventory + Purchases − Ending Inventory

That is the standard. Toast, Lightspeed, and Restaurant365 all teach it this way, and the math is correct as far as it goes. It’s just incomplete.

Here’s the version that actually predicts your P&L:

True COGS = Beginning Inventory + Purchases + Transfers In − Transfers Out − Employee Meals − Comps − Ending Inventory

Those four extra adjustments separate “COGS as reported” from “COGS as actually consumed in service.” Skip them and your food cost number is off by 1-4 points on a typical week (Restaurant365, MarginEdge). That’s the difference between a good month and a bad one, hiding in plain sight.

What each adjustment does #

Transfers in / out. If your kitchen pulls $200 of produce from the bar’s garnish station for a special, that’s a transfer in to the kitchen. The receiving department got cost it didn’t sell. The giving department lost cost without revenue. Both sides need adjusting or both sides report wrong, and the bar takes the blame for a hit the kitchen ate.

Employee meals. Staff meals cost real food and generate zero revenue. Pull them out of COGS and you get the cost-of-sales-only number. Industry-average employee meal cost runs 1-3% of food sales (Toast).

Comps and voids. A comp is food you served and didn’t get paid for. It belongs in marketing or guest-recovery cost, not food cost. Pull comps out of food COGS and the number turns into a cleaner operational read.

Skip these adjustments and your reported 32% food cost is really 29.5%. Sounds like good news. Your real cost is lower than you thought. The problem is the trend. Say employee meals jumped from 1.5% to 3% because the new sous chef is letting the line cooks eat whatever they want off the line. Your reported food cost sits flat at 32% while actual operational food cost climbs, and you won’t catch it unless you adjust. By the time it shows up somewhere you do look, it’s already a habit on the line.

A worked example #

Casual full-service restaurant, one-week period.

• Beginning food inventory: $12,400
• Food purchases (invoices booked this week): $24,800
• Ending food inventory: $11,900
• Transfers in (from bar to kitchen): $150
• Transfers out (kitchen to bar for syrup, garnish, citrus): $480
• Employee meals: $620
• Comps and voids: $310
• Food sales for the week: $86,400

Reported COGS (basic formula): $12,400 + $24,800 − $11,900 = $25,300

Reported food cost: $25,300 ÷ $86,400 = 29.3%

True COGS (full formula): $25,300 + $150 − $480 − $620 − $310 = $24,040

True food cost: $24,040 ÷ $86,400 = 27.8%

That 1.5-point difference isn’t noise. On annualized $4.5M revenue at 60% food sales share, it’s $40,500/year of cost living in a category nobody is watching. Run the same gap across every operation that doesn’t adjust and you’ve got the reason so much of the industry sits at the top end of the healthy food cost band. They’re reading the wrong number.

The 28-35% target and what changes it #

Industry-standard food cost benchmark for casual full-service: 28-35% (Toast, Lightspeed, NRA Industry Factbook).

Concept shifts the band:

• Pizza: 20-28% (cheese and dough run cheap; sauce is essentially free)
• Burgers / QSR: 25-32%
• Casual full-service: 28-35%
• Steakhouse: 32-40%
• Sushi / seafood: 30-38% (perishability + high-cost protein)
• Italian (non-pizza): 28-34%
• Mexican: 28-32%
• Fine dining: 32-40%
• Bakery / coffee shop: 28-35% (depends heavily on espresso vs pastry mix)

The right number for your concept is whatever leaves enough margin after labor, rent, and overhead to hit your net profit goal. Don’t chase the median, the median is full of operations that are barely making it. For most independents you want to land 2-4 points below the segment median. If casual full-service runs 32% median, your operating target is 28-30%.

Three reasons your number is wrong even when the math is right #

1. Bad inventory counting. The biggest single error source, and it’s almost never the math. Counting frequency matters less than counting consistently. Weekly counts done Sunday at 9pm by the same person on the same form are reliable. Monthly counts done at random times by whoever’s around are garbage, and you’ll trust the garbage because it’s written down.

The other counting failure is open product. A half-empty case of romaine is not zero. A quarter-bottle of olive oil is not zero. A bin of dry pasta with a guess at the level is a guess, not a count. Put weight scales on the items that move real money, eyeball the rest, and stop pretending the eyeballed stuff is precise.

2. Invoice timing. Book the $4,200 protein order on Thursday but the meat actually rolls in Friday at midnight, already sitting in next week’s inventory, and you’ve inflated this week’s purchases by $4,200. Your COGS reads $4,200 high for no reason. Book by receipt date, not invoice date, and it goes away.

3. Recipe non-compliance. Your recipe calls for 4 oz of cheese on a burger. Your line is pulling 5. The spec sheet is honest, but the operational reality is 25% over and nobody weighed it because everybody eyeballs cheese. Theoretical food cost stays at spec while actual food cost runs hot, and the difference shows up as variance. That’s the whole reason a food cost problem always ends in the same place: run a variance check.

What this looks like in the calculator #

The food cost calculator on this site runs the full formula, transfers and employee meals and comps included. Plug in your weekly inventory and purchases and the adjustments, and it spits back true food cost percentage against your concept benchmark. When the number runs hot, pair it with the inventory variance calculator to see whether the gap is on the operational side or the pricing side.

What to do today #

If you’ve never run the adjusted formula, pull last week’s numbers and run it both ways. The gap between the simple formula and the adjusted one is the cost category you’ve been ignoring. For most operations that gap is 1-3 points of food cost, which on a $1.5M food revenue base is $15-45K a year you’re leaving on the table.

The fix isn’t complicated. It’s a checklist. Count inventory consistently, book purchases by receipt date, track employee meals and comps separately, and run the full formula every week. None of that is hard. Doing it every single week when the place is slammed and you’re short two cooks, that’s the hard part. The discipline beats the math every time.

Sources: Toast, Lightspeed, Restaurant365, MarginEdge, NRA 2024 Industry Factbook.