Food cost benchmarks by restaurant type
A steakhouse at 35% food cost may be doing everything right. A fast casual at 35% is in trouble. The percentage alone tells you nothing.
Food cost percentage is the cost of the food you sold divided by the food sales that produced it. The industry average sits around 32.4%, and typical ranges run 28% to 35%. That range is wide because it's covering formats with fundamentally different economics.
Targets by concept
| Concept | Food cost |
|---|---|
| Full service | 28–35% of food sales |
| Fast casual | 25–30% |
| Steakhouse, sushi, protein-heavy | 35% and above |
| Cafe and bakery | 28–32% |
The protein-heavy exception is the one people get wrong. A steakhouse running 38% food cost isn't failing at purchasing. It's selling an expensive raw material at a high check average with lower labor intensity per dollar of revenue. The right question isn't whether the percentage is high — it's whether the gross profit dollars per cover justify it.
That distinction matters more than it sounds. A restaurant with 32% food cost on $2M of food sales generates more gross profit than one with 28% on $1.8M. The better-looking percentage produces less money. Percentages are a control tool, not a scoreboard.
Beverage runs on different math
Beverage cost is measured against beverage sales, and every category behaves differently:
| Category | Cost |
|---|---|
| Blended beverage program | 18–24% |
| Spirits and cocktails | 15–22% |
| Draft beer | 20–26% |
| Bottled and canned beer | 24–28% |
| Wine | 30–38% |
| Non-alcoholic | 8–15% |
Set the target by category, not as one blended number. A wine bar blending at 28% is running a good program. A cocktail bar blending at 28% has a problem, and the same number means opposite things.
Note also that pour cost measures margin, not profit. A $200 bottle at a 40% pour cost makes far more money than four $8 beers at 20%. Chasing the percentage can cost you the dollars.
When food cost drifts up
In rough order of how often it's the actual cause:
Menu prices that never followed invoice prices
Wholesale food costs are more than 35% above pre-pandemic levels. Most menus have been repriced once in that span, unevenly. Recost your twenty highest-selling items at this week's prices and you'll usually find the whole gap in three or four dishes.
Portion creep
Nobody decides to over-portion. It happens gradually, on the line, under pressure, and it shows up as a percentage point or two that nobody can explain. Standardized recipes with weights, and a scale on the line, are the only reliable fix.
Waste nobody counts
Prep waste, spoilage, comps, staff meals and remakes are all real food cost and most of them never get recorded. A one-week waste log usually surprises people.
Measuring purchases instead of cost of goods
Dividing invoices by sales ignores inventory movement entirely. Build stock for a holiday and food cost looks terrible; draw it down for a big event and it looks great. The real calculation is opening inventory plus purchases minus closing inventory, divided by sales. Without a physical count, you're looking at noise.
Calculating it properly
The formula that actually works:
Food cost % = (opening inventory + purchases − closing inventory) ÷ food sales
The inventory figures are what make it real. Without them you're measuring purchasing behavior rather than consumption, and the two diverge every week that you build or draw down stock.
A worked example. Opening inventory $22,000, purchases $94,000, closing inventory $19,000, food sales $290,000. Cost of goods is $97,000, so food cost is 33.4%. If you'd used purchases alone you'd have reported 32.4% — a full point better than reality, produced entirely by eating a thousand dollars of stock you'd already paid for.
Count on the same day of the week, at the same time, with the same person where possible. Consistency in the count matters more than precision in it, because you're looking at the trend.
Where the menu itself is the problem
Food cost is an average across everything you sell, which means it moves when your mix moves even if every recipe is unchanged. A dish at 42% food cost that outsells everything else will drag the whole number up while being your best seller.
Sort your items by contribution margin in dollars — menu price minus plate cost — alongside units sold. The four quadrants tell you what to do: high margin and high volume gets protected and promoted, low margin and high volume gets repriced or re-engineered, high margin and low volume gets moved up the menu, and low margin with low volume comes off.
Trimming 15–20% of items is a common outcome, and it usually cuts waste and speeds up service at the same time. Long menus carry inventory that turns slowly, and slow-turning inventory is where spoilage lives.
When food cost is suspiciously low
Worth as much attention as a high number, and far less often examined. Food cost several points under target usually means one of three things: you have a genuine purchasing advantage, portions have shrunk to a point guests have noticed before you have, or your inventory count is wrong.
The middle one is expensive in a way that never appears on the P&L. You see the improved percentage this month and the reduced covers six months from now, and nothing connects them.
The number to watch instead
Food cost is half of prime cost, and it trades against the other half constantly. Buy pre-cut vegetables and food cost rises while prep labor falls. Judge the two together — the industry ceiling for prime cost is 65% of sales, and that's the number that determines whether there's anything left after the kitchen is done.