Restaurant Food Cost Percentage: Calculate It and Fix What Is Driving It

August 24, 2026

Your Monday inventory count lands at 8:12 a.m. Sales were strong. The walk-in looks normal. Yet your food cost percentage jumped again, and nobody can tell you why.

Your chef blames vendor prices. Your bookkeeper questions the count. Your manager remembers three large comps and a broken reach-in. Everyone could be right.

That's the problem.

The percentage tells you what happened. It doesn't tell you where it happened. And it definitely doesn't tell you what to cut first.

I use three views to make the number useful: actual cost, theoretical cost, and contribution dollars. You need all three. Otherwise, you can improve one percentage while quietly making your business worse.

What food cost percentage tells you

Your food cost percentage shows how much of your food sales went toward the food you used during the same period. The basic relationship is simple:

Food cost percentage = cost of goods sold ÷ food sales × 100

Your cost of goods sold, or COGS, comes from inventory:

COGS = beginning inventory + purchases − ending inventory

This food cost formula is an operating measure. It helps you compare weeks, locations, categories, and your actual result against your plan.

But it's not profit.

Your restaurant still has labor, rent, card fees, repairs, utilities, and every other cost after food. A low number can sit beside a weak profit. A higher number can work when an item produces strong contribution dollars and needs little labor.

The National Restaurant Association Educational Foundation menu-management chapter makes the same distinction. Food cost differs by menu category, while contribution margin measures the dollars left after item food cost.

So one percentage is not the target. Your real question is whether your menu leaves enough dollars to run the restaurant.

Use a benchmark as a screen, not a target

You'll see broad restaurant benchmarks online. They can help you spot an unusual result. They can't set your target.

Your concept changes the answer. So do your prices, sales mix, labor model, service style, and ingredient quality. A steakhouse and a bakery can run different food economics and both make money.

And your categories differ too. Your bottled beer, burger, steak, and dessert shouldn't carry identical targets. Forcing them there can damage the menu.

I'd compare your result in four ways: against your budget, against the same location last period, against similar locations, and against theoretical usage. That gives you context. One generic average doesn't.

How to calculate food cost percentage each week

A clean period comes first. A week often works because it catches problems fast without making your team count the whole kitchen every night.

Your team should use the same count time each week. Count with the same units. Include food purchases and credits from that period. Then match the result with food sales from those exact dates.

Here's a labeled arithmetic example:

  • Beginning inventory: $18,400
  • Purchases: $7,250
  • Ending inventory: $16,900
  • Food sales: $29,750

First, calculate COGS:

$18,400 + $7,250 − $16,900 = $8,750

Then divide by food sales:

$8,750 ÷ $29,750 × 100 = 29.4%

The arithmetic is easy. But a clean result is not a correct result. Your inputs are harder.

A case entered as one pound can wreck the result. So can a late invoice, an unrecorded transfer, or a Monday count taken before one location receives a delivery. Your spreadsheet will still produce a clean answer. It'll just be wrong.

And compare food cost with food sales. Don't divide food COGS by total revenue that includes alcohol, service charges, or retail items. Match the numerator and denominator.

Make every count comparable

Good counting feels boring. That's a compliment.

A shelf-to-sheet order keeps your team on the same path. Lock each item to one count unit. Record open cases with a clear fraction or weight. Separate food, beverage, and packaging when your reporting needs separate decisions.

Then investigate the strange result before you act. A sudden low percentage can signal a missed invoice or inflated ending inventory. Low isn't always good.

For one location, a disciplined inventory spreadsheet may be enough. But the owner must stay clear. Someone needs to control items, units, count times, and corrections.

Calculate item food cost separately

Your weekly result shows total use. An item calculation shows recipe economics:

Item food cost percentage = recipe cost per serving ÷ selling price × 100

Suppose your current recipe costs $4.65 and sells for $16.50. The item result is 28.2%. That's another arithmetic example, not a market benchmark.

Now look at contribution dollars:

$16.50 − $4.65 = $11.85

Those dollars help cover everything after ingredients. This is why a lower percentage isn't always the better item. Your $9 item at 22% leaves $7.02. Your $16.50 item leaves $11.85. Volume, labor, and waste still matter, but contribution changes the decision.

Our menu engineering guide shows how to combine popularity and contribution margin. That view should guide you before you cut a high-cost bestseller.

Why actual and theoretical costs disagree

Your actual food cost comes from inventory. It includes what guests ate, what spoiled, what staff over-portioned, what got remade, and what disappeared.

Your theoretical food cost comes from current recipes multiplied by actual menu item sales. It estimates what you should have used if every recipe, yield, and portion stayed on standard. Some systems call this ideal food cost.

They won't match perfectly.

But the gap matters. Subtract theoretical percentage from actual percentage to get your variance:

Actual cost percentage − theoretical cost percentage = variance

If your actual result is 31.8% and your theoretical result is 28.9%, your food cost variance is 2.9 percentage points. That gap gives you a place to investigate. It doesn't prove theft.

Bad mapping looks like kitchen waste. The problem is not the math.

Run the four-part variance test

Your team should check four causes in order: price, portion, waste, and data.

  • Price. Compare invoice prices with the prices inside your recipes. Your invoice review should catch substitutions, pack-size changes, freight, and missed credits. The USDA Food Price Outlook can provide broad price context, but your invoices show your actual vendor pricing.
  • Portion. Weigh expensive proteins and batch recipes. Your portion review should cover trim yield, ladles, scoops, and prep sheets. Portion control fails when your written six-ounce portion plates seven.
  • Waste. Separate spoilage, prep trim, remakes, overproduction, and returned dishes. The FDA explains the scale and resource cost of food loss and waste, though its national estimate isn't a restaurant waste rate. Your own waste log has to guide waste reduction.
  • Data. Audit count units, recipe units, invoice items, POS mappings, transfers, comps, and deleted buttons. Fix this before blaming the kitchen.

And don't change all four at once. You'll lose the signal.

Two examples that show why context matters

A single percentage can hide two different operating problems. Your response should follow the cause.

One location, two leaks

Assume your restaurant reports 31.8% actual cost and 28.9% theoretical cost. The 2.9-point variance says actual usage exceeded recipe-based usage.

Now your invoice review finds that chicken rose from the recipe file's old price. That makes the theoretical number too low. Update the price first.

After the update, theoretical cost moves to 29.8%. Your true use variance is now 2.0 points, not 2.9. The remaining gap may come from portions, waste, comps, or counts.

See the difference? This is not just bookkeeping.

A vendor-price response wouldn't fix over-portioning. A portion crackdown wouldn't fix an old recipe price. You need to separate purchase variance from use variance before you act.

That's why we connect invoice, recipe, inventory, and POS data for our clients. I want the team solving the real problem, not arguing over four reports.

A group average can hide a local leak

Now take a multi-location group with three restaurants. This is another arithmetic scenario.

  • Location A: $100,000 food sales at 29% cost
  • Location B: $100,000 food sales at 29% cost
  • Location C: $50,000 food sales at 37% cost

The group spends $76,500 on $250,000 of food sales. That produces a 30.6% group result.

It looks manageable.

But Location C is far from its peers. The group average hides it because Location C has lower sales. If you only review the rollup, you can miss a local receiving, portion, waste, or mapping problem.

So review each location first. Then review the group. Standard item names and count units make that comparison possible.

Ten ways to improve food cost without hurting guests

The fastest fix isn't always a price increase. Start with the controls that improve your information and protect the plate.

  1. Fix count rules. Use one schedule, one shelf order, and one unit for each item.
  2. Update invoice prices. Refresh recipe costs when vendors change prices or pack sizes.
  3. Flash-count expensive items. Check proteins, oils, seafood, and other high-value products between full counts.
  4. Standardize portions. Give your team scales, scoops, ladles, photos, and current recipes.
  5. Log waste by reason. Separate spoilage, prep loss, overproduction, remakes, and returns.
  6. Reconcile receiving. Match purchase orders, delivered quantities, invoices, credits, and substitutions.
  7. Forecast prep from sales mix. Use comparable days, dayparts, channels, and local events.
  8. Review modifiers and comps. Unpriced add-ons, packaging, discounts, and voids can distort item economics.
  9. Engineer the menu. Protect contribution dollars before removing a popular item or cutting quality.
  10. Change prices or vendors after diagnosis. Make the commercial move when the operating data supports it.

Waste work deserves care. The National Restaurant Association's menu guidance connects portion choices with waste prevention. A smaller default with an easy add-on may serve guests better than quietly shrinking every plate.

And talk with your team. Your line cooks often know which garnish gets dumped, which prep batch runs long, and which modifier creates remakes. Your report should test their observations, not replace them.

What your restaurant systems must agree on

Your POS knows what sold. Your inventory system knows what you counted. Your invoice tool knows what you bought. Your recipe file knows what each item should use. Your accounting system knows what posted.

Those systems need shared definitions.

At minimum, align these fields:

  • item and modifier names
  • purchase and count units
  • pack sizes and conversion factors
  • recipe yields and portions
  • location and transfer records
  • sales and accounting periods
  • food, beverage, and packaging categories

This isn't glamorous work. It's the work.

Your POS integrations should move dependable data, not duplicate bad mappings faster. Test one item from invoice through recipe, sale, inventory use, and accounting. Then repeat with a modifier and a transfer.

Still, software isn't always the answer. A clean spreadsheet can serve one stable location. Restaurant inventory software starts earning its cost when locations, vendors, recipes, and manual reconciliation multiply.

I also look for ownership. When the POS vendor blames inventory and inventory blames accounting, your operator becomes the unpaid systems integrator. That's a bad support model.

Build a number your team can act on

Your food cost percentage should lead to a question, a test, and an owner. It shouldn't lead to a random menu cut.

Calculate actual cost. Build a current theoretical cost. Review contribution dollars. Then test price, portion, waste, and data in that order.

That's where restaurant technology helps. Flyght connects and manages the POS, network, inventory, integrations, and support around your operation. And when reports disagree, our team owns the path to an answer.

If your food-cost meeting keeps ending in arguments about whose number is right, talk with Flyght. We'll help you build a reporting system your operators can trust.

Frequently asked questions

What is a good food cost percentage for a restaurant?

A good food cost percentage is one that supports your concept's labor, overhead, guest value, and profit goals. Use broad industry ranges only as a screen. Compare your result with your budget, similar periods, similar locations, theoretical usage, and contribution dollars.

How do I calculate food cost percentage weekly?

Add beginning inventory to purchases, then subtract ending inventory. Divide that COGS result by food sales for the same week and multiply by 100. Count at the same time, use fixed units, and include transfers, credits, and late invoices correctly.

Why is actual food cost higher than theoretical food cost?

Actual cost can exceed theoretical cost because of outdated ingredient prices, over-portioning, low yields, spoilage, overproduction, remakes, theft, or data errors. Update recipe prices first. Then test portion, waste, and system mappings before assigning blame.

What is the fastest way to lower food cost?

Start with count accuracy, current invoice prices, high-cost-item counts, and portion checks. These controls can expose the cause without changing the guest experience. Raise prices, switch vendors, or change recipes only after the data shows which action fits.

Should I raise menu prices when food cost rises?

Maybe, but don't make price your automatic response. Check vendor prices, recipe costs, portions, waste, sales mix, and contribution dollars first. If the item still can't support your target economics, test a price change against guest demand and local value.

Put this guide to work

Choose one reporting period and lock its cutoff. Reconcile purchases, credits, transfers, and beginning and ending inventory before comparing actual cost with recipe-based usage. Investigate unit conversions and stale recipe prices before interpreting variance as waste or loss.

Abstract hero background with green gradient stripes and glowing accents.