Prime cost: what food and labour should run in an independent restaurant
A workable restaurant prime cost benchmark for a full-service room is food plus labour at roughly 65% of sales. Profitable operators in the National Restaurant Association's survey sit near that line. Loss-making ones sit well above it, mostly because of wages.
On this page:
- Prime cost, not turnover, tells you whether the menu works
- Food cost by cuisine
- Labour percentage and the rota
- Reading your P&L against a published survey
- What cost discipline looks like in the entries we read
- Questions
Prime cost, not turnover, tells you whether the menu works
Prime cost is two lines added together. Cost of food and drink sold, plus total labour, including employer taxes and benefits. Divide by net sales and you have the number that decides whether a restaurant survives.
Turnover flatters. A full book on a Saturday says nothing about the margin on each cover. Prime cost tells you whether the menu, the prices and the rota fit together.
It matters to us because it matters to our entrants. Of the 950 winners in our winners directory, 946 are restaurant or food and beverage businesses. The cost pressure in this article is the pressure they describe to us.
Where food cost sits by cuisine, and why a steakhouse differs from a trattoria
The National Restaurant Association's Restaurant Operations Data Abstract gives the clearest published reference. In its latest edition, full-service operators with annual sales of $2 million or more reported a median food and non-alcohol beverage cost of 31.0% of sales. Smaller full-service rooms reported 33.7%.
Volume helps. Bigger rooms buy better and waste less. Cuisine shifts the line too, and the survey does not split it that way, so treat what follows as direction rather than a published range.
A steakhouse carries one of the heaviest plate costs on any menu. Aged beef is expensive and leaves little room to trim. The food percentage runs high, but prep is simpler, so labour can run lower. A trattoria works the other way. Flour, eggs and tinned tomatoes cost little. Fresh pasta, slow sauces and a full brigade cost a lot of hours.
Both models can hit the same prime cost. They get there by balancing opposite lines. That is why food cost alone is a weak measure. Our directory lists 44 Best Italian Restaurant winners and 26 Best Steakhouse winners. Their cost sheets rarely look alike.
Labour percentage: the trade-off between the rota and the service guests remember
Labour is where profit is won or lost. The same Data Abstract shows full-service operators that reported a pre-tax profit ran salaries, wages and benefits at a median of 34.2% of sales. Those that reported a loss ran at 42.9%.
Limited-service figures were lower: 30.0% for profitable operators and 34.1% for those at a loss. Counter service needs fewer hands per cover.
The gap of almost nine points in full service is the whole argument. Food cost moves a point or two with good buying. Labour moves by whole shifts.
Cutting a server saves money and costs the evening. Guests remember a slow table longer than a good plate. The aim is not the smallest rota. It is a rota built on covers per labour hour, adjusted service by service.
Wage rates move the line without anyone changing the rota. In the United States, the Bureau of Labor Statistics publishes average hourly earnings for food services and drinking places every month. In the UK, the Office for National Statistics reports average weekly earnings by sector. Check the reading for your market before you set a labour target.
How do you read your own P&L against a published cost survey?
Start with the same definitions the survey uses. Food cost is opening stock plus purchases minus closing stock. Labour includes employer taxes, pensions and benefits, not just gross pay. Sales are net of tax and service charge.
Here is a worked example with illustrative round numbers for one trading week.
- Opening food stock 4,000, purchases from invoices 6,600, closing stock 4,200. Food cost is 6,400.
- Gross pay 5,800, employer taxes and pension 800. Labour is 6,600.
- Net sales 20,000. Food cost is 32%. Labour is 33%. Prime cost is 13,000, or 65%.
Now set that against the published medians.
| Line | Full service, profitable | Full service, loss-making | Worked example |
|---|---|---|---|
| Food and non-alcohol beverage | 31.0% (higher volume) to 33.7% (lower volume) | Not split by profit | 32% |
| Salaries, wages and benefits | 34.2% | 42.9% | 33% |
| Prime cost (sum) | About 65% to 68% | About 74% to 77% | 65% |
The prime cost rows are our arithmetic, adding the survey's food medians to its labour medians. The survey does not publish a single prime cost figure. If your prime cost sits in the low to mid sixties, the room has margin to absorb a bad month. Above the low seventies, look at the rota before the menu.
Run the calculation weekly, not monthly. A month hides a bad week inside three good ones.
What cost discipline looks like in the entries we read
Our jury scores cuisine, service, consistency and sustainability. It does not score a P&L. Even so, cost discipline shows up in the evidence. You can read the full method on our judging process page.
The strongest entries describe a menu built backwards from plate cost. They name the dish they dropped because it could not hold its margin. They explain how the rota follows the booking sheet. Weak entries describe ambition and leave out how the kitchen pays for it.
One pattern comes up often when the jury discusses consistency. The rooms that score highest on it are rarely the cheapest to run. They are the ones that know their numbers, so a supplier price rise does not become a smaller portion on a Tuesday.
Every award category is judged on merit. No plan, package or payment influences the jury or the result. Entry on the Free plan covers one programme and one category; for paid options, see the fee page.