Restaurant food cost percentage: how to calculate and lower it

Food cost is the line on the P&L that owners worry about most and understand least. Most know their percentage was "around 32" last month. Far fewer can tell you why, or what it should have been. The number by itself doesn't tell you much. What matters is the gap between what you spent and what you should have spent, and where that gap comes from.

Here's how I calculate food cost percentage, how I find the leaks, and the order I'd work in to bring it down.

How to calculate food cost percentage

Food cost % = food cost of goods sold ÷ food sales

And food cost of goods sold is:

COGS = beginning inventory + purchases − ending inventory

Two details matter here:

• Divide by food sales, not total sales. If you divide by total sales, a strong bar month will make your kitchen look better than it is.
• Use COGS, not purchases. Purchases tell you what you bought. COGS tells you what you used. If you ordered heavy before a holiday, purchases will scare you for no reason.

Hypothetical example: you start the month with $10,000 of food inventory, buy $30,000, and end with $9,000 on the shelf. Your food COGS is $31,000. If food sales were $100,000, your food cost is 31%.

What's a good food cost percentage?

The NRA put the 2024 median at about 32% of food sales, with strong operators running in the 28–30% range.

Your target depends on your concept. A menu built around premium proteins will run higher than a pasta-and-salad menu, and that can be fine if labor is lean. That's why I always look at food cost next to labor, as part of prime cost, and never alone.

Actual vs. theoretical food cost

This is the part most restaurants skip, and it's where the money is.

• Actual food cost is what the inventory math above says you used.
• Theoretical food cost is what you should have used, based on what you sold and what each recipe costs.

To get theoretical, you need two things: a costed recipe for every menu item, and an item-level sales report from your POS. Multiply each item's plate cost by the number sold, add it up, and divide by food sales.

The gap between actual and theoretical is your variance. That's the number that tells you something is wrong: over-portioning, waste, spoilage, receiving errors, theft, or recipe costs that are out of date.

Hypothetical example: theoretical food cost is 28% and actual is 32%. On $100,000 a month in food sales, that four-point gap is $4,000 a month you paid for and didn't sell. That's the size of the prize, and it tells you to look on the floor, not just at your menu prices.

If actual and theoretical are close but both are high, the problem is your menu pricing or your recipes. If theoretical looks good and actual is high, the problem is in the building.

Yield: the invoice price isn't your real cost

The price on the invoice isn't what you're actually paying for an ingredient. Yield decides that.

Hypothetical example: you buy a product at $4.00/lb. After trimming, cooking and portioning, 55% of that weight is sellable. Your real cost is $4.00 ÷ 0.55 = $7.27 per usable pound. A 6 oz portion costs about $2.73, not the $1.50 you'd get from the invoice price.

Now say another supplier offers a similar product at $3.60/lb, but it yields 48%. That's $7.50 per usable pound. The "cheaper" product costs you more.

What I'd do:

• Run yield tests on your top proteins and highest-spend produce.
• Redo them when you change suppliers, specs or cooking methods.
• Cost every recipe from yielded cost, not invoice cost.
• Compare suppliers on cost per usable pound.

Portioning

Portioning is where good recipes go to die. A cook adding an extra ounce "to make it look right" on a busy night doesn't feel like much. Across a few hundred plates a week, it is.

• Give the line the tools. Scales at every protein station, the right scoops and ladles, portion cups for sauces and dressings.
• Post the spec. Recipe card and a plating photo at the station. If it isn't written down, every cook has their own version.
• Pre-portion where you can. Bagging or panning proteins during prep is easier to control than portioning on the line during a rush.
• Spot-check. Pull a plate during service and weigh it. Do it often enough that it's normal, not a punishment.

Order guides, pars and waste

Over-ordering doesn't just tie up cash. It turns into spoilage, and spoilage is food cost.

• Build an order guide with pars by day of week. Your Friday par isn't your Tuesday par.
• Drive prep from a forecast, not habit. If the prep list says "make two pans" every day regardless of sales, you'll throw out the difference.
• Keep a waste log and actually read it. Weekly, with the chef. Patterns show up fast: the same item over-prepped every Monday, the same product spoiling in the same corner of the walk-in.
• Practice FIFO. Label and date everything. Old product in the back means it gets thrown out.
• Cross-utilize. Trim and leftovers that can go into a special, a family meal or a soup aren't waste.

Check every invoice

Receiving is one of the easiest places to lose money and one of the easiest to fix.

• Check deliveries against the invoice before the driver leaves: counts, weights and condition.
• Check prices against what you agreed to. Prices drift when nobody's checking.
• Collect credits for shorts, substitutions and damaged product. Track them until they show up on a statement.
• Watch for substitutions. A different brand or pack size can change both price and yield.
• Track price increases on your top items. When a big-spend item moves, your recipe costs and menu prices may need to move too.

If you haven't rebid your main supplier in a year or more, that's worth doing too. Pull 60–90 days of invoices, list your top 25–40 items by spend, write exact specs, and get two or three quotes on the same list. More on that on the supplier negotiation page.

Menu pricing and menu mix

Once the kitchen is running tight, look at the menu itself.

• Recost the menu when costs move, not once a year. Your top sellers deserve attention every time a major input changes.
• Look at contribution margin, not just percentage. A higher-cost item can still make you more dollars per plate. You bank dollars, not percentages.
• Find popular low-margin items. Your product mix report shows what sells. Put it next to recipe costs and you'll see which favorites are quietly dragging food cost up. Those are candidates for a small price change, a spec change or a better position on the menu.
• Price modifiers. Add-ons and substitutions that ring in free are a food cost leak. Make sure your POS charges for them.
• Price delivery menus to cover commissions. If third-party pricing matches the dine-in menu, you may be losing money on those orders.

Where I'd start

If your food cost is higher than you want and you're not sure why, here's the order I'd go in:

1. Get an accurate inventory count and calculate real COGS.
2. Cost your top 15–20 sellers from yielded cost.
3. Compare actual to theoretical. The size of the gap tells you where to look.
4. Fix receiving and portioning first. They're fast and they don't touch the guest experience.
5. Then tackle pricing and suppliers.

Run your own numbers through the Prime Cost Health Check to see how food cost fits with labor. If you want someone to build the yield tests, recipe costing and weekly tracking with your team, that's the core of the Prime Cost Sprint.