BenchmarksUpdated

Prime Cost: The Formula, a Good Percentage by Venue Type, and Which Half to Fix (2026)

Prime cost is cost of goods sold plus total labour, divided by net sales. Most operators target 55% to 65%, but the median US limited-service restaurant ran at 65% in 2024. Here are the bands by venue type, a weekly calculation, and how to tell whether food or labour is the problem.

Alex Riesenkampff

7 August 2026 · Updated 6 October 2026 · 13 min read · Markdown

Prime cost is your cost of goods sold plus your total labour cost, divided by net sales. A good prime cost for most independent venues is 55% to 65% of net sales: about 55% to 60% for a counter or quick-service operation, and 60% to 65% for table service. Those are the targets operator guides publish. What restaurants run is higher. The National Restaurant Association's 2025 operations survey found a median prime cost of 65 cents in every sales dollar among US limited-service restaurants in 2024, the top of the target range.

So 65% is a ceiling to stay under, and the useful question is which half of the number is pushing you towards it. Prime cost joins the goods side, which our menu pricing calculator covers dish by dish, to the wage side, which our labour cost benchmarks cover on their own. If you run a coffee shop that is busy but still not making money, where a UK cafe's takings go runs the same sum in pounds, with VAT and the owner's own hours included.

What is the prime cost formula?

Prime cost = cost of goods sold + total labour cost, and prime cost percentage = prime cost ÷ net sales × 100. Restaurant365 states it the same way: "Total Cost of Goods Sold + Total Labor Costs = Prime Cost". Everything else on the P&L, from rent to card fees, sits outside it.

Two definitions decide whether your number means anything. Cost of goods sold is what you used: opening stock plus purchases minus closing stock. Total labour is everything a person costs you, on-costs included.

The owner's hours are where small venues fool themselves. If you work 40 hours a week on the floor, a prime cost without them describes a business you could not hand to a manager. Service charge that goes to staff belongs on neither side; our tronc guide covers how that money should flow.

Here is one week at a neighbourhood cafe, worked through:

Leave out the owner's hours and the same week reads 58.8%, which looks inside the quick-service band. With them it reads 62.5%, above the 55% to 60% band for a counter cafe, which is the honest answer for a cafe that relies on its owner.

What is a good prime cost percentage by venue type?

Operator guides converge on 55% to 60% for quick service, 60% to 65% for full service and up to 68% for fine dining, while the measured US median for limited-service restaurants was 65% in 2024. The bands come from restaurant software and accounting firms. The National Restaurant Association's survey of more than 900 restaurants is the closest thing to a measured figure, and it lands at the top.

The two full-service medians together point to the high 60s, above every published target. Labour is the reason. The NRA puts full-service labour at a median of 36.5% of sales in 2024, against an average of roughly 33% across its 2010, 2013 and 2016 editions, while food cost fell to 32.0% from an average of 34%.

The same survey shows what the band is for. Full-service operators that made a pre-tax profit in 2024 ran labour at a median of 34.2% of sales; those that made a loss ran it at 42.9%. In limited service the split was 30.0% against 34.1%. And even the median full-service restaurant kept only 2.8% of sales as pre-tax income.

Both ends of the range carry a warning. TouchBistro notes that "a prime cost of 50% or less could indicate that your food quality is low, you’re overcharging your customers, or you’re overworking your staff." Above 70%, it says, "it becomes very difficult to be profitable."

The default, then: take the band for your format as the target, aim for the lower half, and treat 65% as the line where you act. A bar can sit lower because drinks cost less to buy than plates do to make. useforcs.com puts beverage cost of goods at roughly 18% to 24%, against 28% to 35% for food.

How do you calculate prime cost every week?

A weekly prime cost catches a cost problem while you can still change next week's order and rota, where a monthly one arrives after up to four weeks of it have been paid for. 7shifts puts it plainly: "The good operators know their prime cost weekly, not monthly." MarketMan names the reason most don't: pulling the numbers by hand from a POS, payroll and a stack of invoices "eats an afternoon nobody has."

Tim Skinner, licensee at the Devonshire Arms in Marylebone, told The Morning Advertiser in June 2026: "Labour has always been the biggest cost in hospitality, between that and cost of sales, and it’s just got bigger." He called labour "the biggest controllable cost on a weekly basis." A weekly number is how you control it.

Is your problem food cost or labour cost?

A prime cost over target does not tell you which half is wrong, so split it before you cut anything: for casual full service, food above about 34% of sales or labour above about 35% marks the problem side. Those thresholds come from useforcs.com's diagnostic for casual full-service restaurants. If both are high, it advises fixing labour first, "since labor changes typically show up in the P&L within 2 to 4 weeks while food changes take 6 to 8 weeks."

The labour fix usually lives in one specific daypart. One cocktail bar in Germany, an anonymised Super44 merchant, found its last opening hour from Tuesday to Thursday was not paying for itself. It closed earlier on those nights after a four-week trial with a roll-back plan. Skinner did the same at the Palmerston, a food-led pub in East Dulwich: "It was clear that nothing was really happening on Monday and Tuesday food wise, and it was costing us a lot of money," he said. The kitchen now opens Wednesday to Sunday.

On the goods side, our guide to supplier negotiation covers what moves a distributor's price, and the food GP calculator shows where a single dish is leaking.

Does a good prime cost mean you are making money?

No: a venue inside its prime cost band can still run out of cash if occupancy costs take more than about 10% of revenue. Armitage Accounting, a US firm that works with restaurants, says venues "with occupancy costs over 10% of gross revenue consistently experience cash flow challenges." Its target splits that 10% into 6% base rent, 2% common-area maintenance and property taxes, and 2% utilities. For a UK venue, read service charge and business rates into the middle slice.

Rent alone varies by format. Paytronix gives 5% to 8% of sales for fast casual, 7% to 8% for casual dining and 8% to 12% for fine dining. Put both numbers on one line and the room left over shrinks fast:

A cheap, long lease buys room to run prime cost a few points hotter. A new lease at 13% of sales means your prime cost has to sit near the bottom of its band. Check occupancy before you decide your prime cost target. If the gap is structural, our guide to cash flow in the first six months covers what to do before it becomes a crisis.

When is a higher prime cost the right call?

Making food in-house moves cost from goods to labour, and buying it prepared moves it back: two cafes can land on the same 62% prime cost with opposite splits. Prime cost cannot tell you which is better. It can tell you whether the choice is paying for itself.

The figures are illustrative; the point is the trade. Compare the landed cost of the bought-in version with the full cost of making it, including the hours. Then ask whether "made here" earns a higher price or more repeat visits. A fine-dining room sits at the top of its band on purpose: fewer covers at a higher spend mean each plate and each skilled hour earn more. A higher prime cost is a problem only when it rises without a decision behind it.

What happens when you cut prime cost too hard?

Cutting portions or hours to hit a ratio can cost more in lost sales than it saves: one Super44 cafe that opened two hours later on Tuesdays saw Tuesday takings fall from €299 to €191. Super44 measured the change in the cafe's own till data, and the lost takings outweighed the wages saved. The rota looked leaner and the cafe was poorer for it.

Chains learn the same lesson in public. In mid-2024, Wells Fargo analysts weighed 75 identical Chipotle burrito bowls from eight New York locations and found the heaviest bowls weighed as much as 87% more than the smallest, CBS News reported. Chipotle's food, beverage and packaging costs rose to 30.6% of revenue in the third quarter of 2024, from 29.7% a year earlier. The company put that down to avocado and dairy inflation, a protein mix shift from its Smoked Brisket offer, and higher ingredient usage as it focused on "ensuring consistent and generous portions." Part of the rise was a portion decision, made on purpose.

Tim Bird, co-founder of Cheshire Cat Pubs & Bars, has not changed rotas, roles or staffing levels in response to rising costs. "You could come along and make draconian measures and they just don’t work," he told The Morning Advertiser.

Track both halves weekly and you can tell a deliberate choice from a slow drift. Super44 does this from your connected POS and rota: it keeps your prime cost current, split into goods and labour by daypart, with the specific action behind whichever half has moved. If the labour half is the one moving, our guide to sales per labour hour shows how to read it one daypart at a time, and our overtime guide covers the causes that inflate it. If the goods half is moving, menu engineering beyond the matrix covers which menu changes have held up in real sales data.