What should restaurant prime cost be?

What prime cost is. Food and beverage cost plus labor, as a share of sales. Labor means wages, payroll taxes and benefits, not just the hourly rate. Rent, insurance and the rest hardly move from week to week. These two lines move every week, and they are the two you can still do something about before Friday.

The number everybody quotes. Search the question and you get a range: 60 to 65 percent for full service, 55 to 60 for quick service. One of the pages that ranks for it says a “successful, sustainable” restaurant runs about 60 percent, and does not say where the number came from. The calculators that print the same range do not say either. The figure may be roughly right. It is still a figure with no survey behind it that anybody will show you.

What is actually published. The National Restaurant Association’s Restaurant Operations Data Abstract, 2025 edition, reporting 2024 figures from more than nine hundred restaurants. For full service, three medians. Food and non-alcohol beverage cost: 32.0% of sales. Salaries and wages including benefits: 36.5%. Income before income taxes: 2.8%. And one split that is worth more than any median: operators who made a pre-tax profit ran wages at 34.2% of sales, and operators who lost money ran 42.9%.

For limited service the Association does publish a prime cost, sixty-five cents of every sales dollar. For full service it does not.

Why you cannot add 32 and 36.5. It looks like a prime cost. 68.5 percent. It is two medians from two separate lines, each worked out on its own, and the restaurant in the middle of the food line is not the restaurant in the middle of the wage line. Medians do not add. There is a second problem: the food figure is “of sales” and the report does not say whether that means total sales or food sales, so you cannot even be sure what you would be adding it to. If somebody hands you 68.5 with the Association’s name on it, they did the arithmetic. The Association did not.

So what should yours be? Three honest answers.

• Lower than it was, or the same, and when it is not, you know why by Monday. The only prime cost that means anything is yours, this week, against your own normal.
• On the labor side, use the one published split that is real. If wages sit nearer 43 percent than 34, that is the line to work on, whatever your food cost is doing. That is where the restaurants that lost money were.
• Measured the same week, from the invoices, not the month after from the bookkeeper. A prime cost at month end is history. A prime cost on Monday is a decision you can still make.

How to see it every week, by hand. Three numbers and a piece of paper.

• Sales: net sales for the week from your point of sale, before tax and tips.
• Labor: every hour worked that week at what it costs you, plus payroll taxes and benefits. The Association’s number includes benefits, so yours has to, or you are comparing against nothing.
• Food: the invoices that arrived that week. That is purchases, not usage, and the two differ when the walk-in fills up or empties out. Count the walk-in once a month and correct it. Purchases every week, a count every month, is how a small restaurant can actually do this.

Divide labor plus food by sales. Write it under last week’s. After two months you have the only comparison that was ever going to help you: your own restaurant, a week ago.

The part the spreadsheet cannot do. The invoices. Sales and hours are in a system already. The food number lives on paper in a clip by the back door, and it is the line that moves the most. Most operators who give up on weekly prime cost give up there.

That is the number on the first line of a Syphor morning page: what you kept, after every bill, last night, with the invoices already in it. Against your own normal. Never against a median.