· Orqestra Team · 7 min read
How to Calculate Food Cost Percentage (With Example)
Food cost percentage is cost of goods sold divided by food sales. Get the formula, plate-level costing, 2026 benchmarks by concept, and two worked examples.
Food cost percentage is the cost of the food you used divided by the food sales you made in the same period, times 100. The formula is (beginning inventory + purchases - ending inventory) ÷ food sales × 100. Most independent restaurants land between 28% and 35%, with pizza and quick service lower and fine dining higher.
That one number tells you whether your menu prices, portions, purchasing, and kitchen discipline are working together. It is the half of prime cost that moves fastest, which is why operators who track it weekly catch problems that monthly trackers miss by a month.
The food cost percentage formula
There are two versions. Use both.
Period food cost (the real one) comes from inventory counts:
- Count your food inventory at the start of the period (beginning inventory).
- Add every food purchase received during the period.
- Count again at the end (ending inventory).
- Food cost = beginning inventory + purchases - ending inventory. This is your cost of goods sold (COGS) for food.
- Food cost percentage = food cost ÷ food sales × 100.
TouchBistro writes it as (total cost of goods sold ÷ total revenue) × 100, with COGS defined as beginning inventory plus purchases minus ending inventory. Katalyst OS uses the same definition and divides by food sales specifically.
Two details matter. First, use food sales, not total sales. If beverages are 20% of your revenue and you divide food cost by total sales, your food cost looks four or five points better than it is. Track beverage cost separately against beverage sales. Second, purchases are not the same as food cost. A big delivery on the last day of the month is purchases, but it sits in ending inventory and does not count as cost until it is used.
Plate food cost (the menu one) is per dish:
Plate cost percentage = cost of ingredients in one serving ÷ menu price × 100
This is what you use to price a dish or decide whether an item belongs on the menu. TouchBistro's version is "(total cost of dish per serving ÷ price of dish to customer) × 100", and it gives the inverse for pricing: ingredient cost ÷ target food cost % = menu price.
Food cost percentage benchmarks by concept (2026)
Published ranges differ by source because some include beverages and some do not, and because chains pull averages down. The table shows three sources side by side so you can see the spread instead of a single number that hides it.
| Concept | TouchBistro | Katalyst OS | Eats365 |
|---|---|---|---|
| Pizza | 20% to 25% | 20% to 25% | 22% to 30% |
| Quick service | 20% to 25% | 25% to 30% | 28% to 32% |
| Fast casual | 20% to 30% | 28% to 32% | 28% to 33% |
| Casual full service | 20% to 30% | 28% to 35% | 30% to 35% |
| Fine dining | 30% to 35% | up to about 40% | 35% to 40% |
| Steakhouse | not listed | 35% to 45% | not listed |
| Coffee shop or cafe | not listed | 25% to 35% | 20% to 25% |
Sources: TouchBistro, Katalyst OS, Eats365.
For an overall target, TouchBistro and Katalyst OS both say most restaurants fall between 28% and 35%. Restaurant365 gives 28% to 32% as the general rule of thumb and notes quick service runs lower and fine dining higher.
Use the table to pick a lane, then set your own target from your own menu. A taqueria at 29% might be fine. A steakhouse at 29% is probably underportioning or overpricing.
Worked example 1: a month of inventory counts
A 70-seat casual full-service restaurant, one location, October 2026. Beverage is tracked separately, so these are food numbers only.
| Input | Amount |
|---|---|
| Food sales for the month | $58,000 |
| Beginning food inventory (Oct 1 count) | $9,400 |
| Food purchases received in October | $19,600 |
| Ending food inventory (Oct 31 count) | $8,800 |
- Food cost = $9,400 + $19,600 - $8,800 = $20,200
- Food cost percentage = $20,200 ÷ $58,000 × 100 = 34.8%
At 34.8%, this restaurant sits at the top of the casual full-service range. If the target is 31%, the gap is 3.8 points, which on $58,000 of food sales is about $2,200 a month. That is the number to chase.
Notice what happened with purchases. The restaurant bought $19,600 but used $20,200, because inventory dropped by $600. If the manager had used purchases alone, food cost would have shown 33.8%, one full point too low. The counts matter. A consistent count sheet (see the restaurant inventory count sheet template) is what makes the number trustworthy.
Worked example 2: one dish on the menu
Same restaurant, costing its best-selling burger. Ingredient costs come from the latest invoices, not from memory.
| Ingredient | Cost per serving |
|---|---|
| 6 oz ground beef patty | $2.10 |
| Brioche bun | $0.45 |
| Cheddar slice | $0.35 |
| Lettuce, tomato, onion | $0.30 |
| House sauce | $0.15 |
| Fries (4 oz) | $0.60 |
| Pickle and garnish | $0.10 |
| Total plate cost | $4.05 |
Menu price is $14.00.
- Plate cost percentage = $4.05 ÷ $14.00 × 100 = 28.9%
That is a healthy number for a burger. If beef goes up and the patty hits $2.60, plate cost becomes $4.55 and the percentage becomes 32.5%. To hold a 30% target, the price would need to be $4.55 ÷ 0.30 = $15.17, so $15.25 on the menu. Running this math on your top 20 items each quarter is the fastest way to keep period food cost in line. The food cost calculator does the plate math and the period math in one place.
A dish-level food cost that looks fine while the period food cost runs high is a sign the gap is in the kitchen, not the menu. That comparison is the whole idea behind theoretical vs actual food cost.
How often to calculate it
Weekly. Katalyst OS calls weekly "the operating standard" and points out that a monthly count gives you twelve chances a year to notice a problem that compounds daily. TouchBistro recommends calculating both weekly and monthly. Restaurant365 says monthly is the minimum.
A practical schedule for an independent restaurant:
- Weekly: a short count of the 30 to 50 items that make up most of your spend (proteins, cheese, oil, produce you buy in volume). Compute food cost from that count plus the week's invoices and the week's food sales. Accept that it is approximate.
- Monthly: a full count of everything, which gives you the number for the P&L and corrects any drift in the weekly estimate.
- Quarterly: re-cost your top 20 dishes from current invoices.
If your inventory lives in a system that tracks purchases and counts for you, the weekly number takes a few minutes. Orqestra's inventory module keeps counts, purchases, and par levels together for this reason, but a spreadsheet and a clipboard produce the same math.
What to do when food cost is high
Work through these in order. Each one is cheap to check.
- Confirm the count. Recount the five most expensive items. A miscounted case of beef swings the number more than anything else on this list.
- Check invoices against quoted prices. Vendor price creep on a dozen items adds a point without anyone noticing.
- Compare theoretical to actual. Multiply recipe costs by items sold and compare to your inventory-based cost. If actual is more than two or three points above theoretical, the gap is waste, portioning, or theft, not pricing.
- Watch portions on your top five sellers. A 7 oz patty served as a 6 oz patty costs you 17% more beef on every burger.
- Log waste for two weeks. Write down what gets thrown out and why. Overprepping and spoilage usually show up fast.
- Fix comps and voids. Food that left the kitchen but never rang up is food cost with no sales against it.
- Reprice or re-engineer. Only after the steps above. Raise prices on high-volume, low-margin items first, and remove or rework items with plate costs above 40% that do not sell.
Common mistakes
- Dividing by total sales instead of food sales. Alcohol and soft drinks flatter the number.
- Using purchases instead of inventory-based COGS. Delivery timing swings it by a point or more either way.
- Counting inconsistently. A count at 9 a.m. one week and 11 p.m. the next, or by two different people with different units, produces noise that looks like a trend.
- Costing dishes from old prices. Recipe costs built in 2024 do not reflect 2026 invoices.
- Forgetting yield. A $9 per pound whole chicken does not give you a pound of usable meat. Cost the usable portion.
- Benchmarking against the wrong concept. A pizzeria comparing itself to a fine dining range will relax when it should not.
- Reacting to one bad week. Look for three weeks in the same direction before changing prices or portions.