Prime cost is food cost plus labour cost, together, as a percentage of sales. It is the single most useful number in a restaurant, and the reason is that the two halves trade against each other — so watching either one alone will mislead you.
Why the combined number matters
A kitchen can cut food cost by buying worse ingredients or by prepping less — and prepping less means more labour, or a more skilled and expensive cook. Buying pre-portioned costs more per pound and less per hour.
So a restaurant congratulating itself on a falling food cost while labour quietly rises has not improved anything. Prime cost catches that; the individual numbers do not.
Prime cost = (cost of goods sold + total labour) ÷ sales.
Total labour means everything: wages, employer payroll taxes, benefits, and management salaries. Leaving management out is the most common way restaurants flatter the figure.
What a normal number looks like
The widely used benchmark is that full-service restaurants aim to keep prime cost around the low-to-mid sixties per cent of sales, with quick service running lower because labour is lighter.
Treat that as a starting reference rather than a target handed down from anywhere authoritative — the right number depends on your rent, your model and your market. A restaurant with unusually low rent can carry a higher prime cost; one paying city-centre rent cannot.
The more useful comparison is your own number over time. A prime cost that has moved three points in a quarter is telling you something specific, whatever the benchmark says.
Calculating it properly
Cost of goods sold is not what you spent on food this month. It is opening inventory plus purchases minus closing inventory. If you are not counting inventory, you are tracking purchases, which moves with delivery timing rather than with what you used.
That distinction matters. A month with two large deliveries at the end looks terrible and is not.
Labour is wages plus payroll taxes plus benefits plus management. Use the same definition every period or the trend is meaningless. The labour cost formula and the three places it goes wrong goes through what gets left out.
Sales is net of comps and discounts.
Run it weekly if you can. Monthly is too slow to act on — by the time a month closes, the problem has had four weeks to run.
What moves it, in order
Portioning. The single largest controllable factor in food cost, and it drifts constantly without anyone deciding to change it. A dish consistently plated ten per cent over spec is ten per cent of that dish's food cost, every time it sells.
Waste and spoilage. Over-ordering, poor rotation, prep that does not get used.
Scheduling against actual covers. Not against a fixed pattern. Most restaurants know their busy hours precisely and still staff to a template.
Menu mix. Which dishes sell, not just what they cost. A high-margin dish that nobody orders does nothing.
Theft and comps. Uncomfortable and real, and the reason to look at voids and comps by staff member rather than in aggregate.
The two halves pull against each other
Worth stating explicitly because it is where decisions actually get made:
Buying prepped raises food cost and lowers labour. For a restaurant short of skilled kitchen staff this is frequently the right trade even though the food cost percentage gets worse.
Simplifying the menu lowers both — less inventory, less waste, less prep, faster service. It is the most under-used lever in restaurants and the most resisted.
Raising prices improves both percentages immediately and risks volume. Small and frequent beats large and rare.
What prime cost does not tell you
It says nothing about whether you have enough customers.
A restaurant with excellent prime cost and empty tables is still losing money, because rent, utilities and insurance do not care how efficiently you cooked. Prime cost is about how well you convert sales; it is silent on whether the sales arrive.
Which is why it is worth looking at alongside covers, and alongside the things that bring covers in — including the reservation calls that ring out during service, which appear in no report anywhere. Restaurant profit margins: what is normal covers what reaches the bottom line after all of it.
The habit worth building
Count inventory weekly on the same day. Pull labour for the same period with the same definition. Divide by net sales. Write the number down.
Four weeks of that and you will see a trend. Twelve and you will know your own business better than any benchmark can tell you.