Restaurant prime cost: what good looks like

Prime cost is cost of goods sold plus total labor, expressed as a percentage of sales. It is the single most useful number in a restaurant because it covers the two costs an operator actually controls week to week. Rent is fixed, insurance is fixed, the utility bill is what it is. Food, beverage and labor are decisions, and they are made hundreds of times a day.

The benchmarks differ by format, and comparing across formats is how operators talk themselves into false comfort. Full-service independents should target 60 to 65 percent of sales. Fast-casual and counter-service runs 55 to 60, because the labor model is lighter. Bars with a strong beverage mix should land 50 to 58, since pour margins carry more of the sales line. Coffee shops sit around 55 to 62, with a heavier labor share against low ticket values. Anything above 70 in any format means the establishment is not earning its rent, and the only question is how fast that shows up in the bank balance.

Calculate it monthly at minimum, weekly if you can. Cost of goods is opening inventory plus purchases minus closing inventory — not simply what you bought that month. Total labor is every dollar the establishment pays for people: wages, salaried managers, payroll taxes, benefits and any agency cover. Divide the sum by net sales for the same period.

Three errors make most operators' number wrong. The first is using purchases instead of true cost of goods, which makes a heavy buying week look like a cost problem and a light one look like a win. The second is excluding management salaries and payroll taxes from labor, which understates the number by three to six points and hides the real cost of an overweight salaried layer. The third is comparing a month against a month with a different day count or holiday pattern rather than against the same period last year.

When prime cost drifts, split it before reacting. If cost of goods moved and labor did not, look at supplier pricing, portioning, waste and theft — in that order of likelihood for most operations. If labor moved and cost of goods did not, look at scheduled hours versus forecast covers, overtime, and whether the schedule was built from a forecast at all. Cutting hours to fix a food cost problem is one of the most common and most damaging mistakes an operator can make.

The operational target worth adopting is a weekly prime cost number posted where managers see it, alongside sales per labor hour. Monthly reporting tells you what happened; weekly reporting lets you fix it while the month is still in play. Operators who move to a weekly rhythm typically find one to two points within a quarter simply because the feedback loop closed.

If you want a fast read on where your number sits against your format, our margin health check gives you a benchmarked answer in a few minutes.