Prime cost: why 65% is the wrong target for your restaurant

Prime cost is the best single number on a restaurant P&L. The target everyone quotes for it is a blend that fits almost nobody exactly.

A guide from Where the money actually goes

Prime cost is your total cost of sales plus your total labor, divided by total sales. Food, beverage, wages, management salaries, payroll taxes and benefits — everything it costs to put the plate on the table and pour the drink, all in one number.

It matters more than food cost alone because food cost and labor trade against each other constantly. Buy pre-portioned protein and your food cost rises while your prep labor falls. Break down whole animals in house and the reverse happens. Look at either number by itself and you can convince yourself things are improving while the combined figure walks the other way. Prime cost is what closes that loophole, which is why it's the number to run weekly rather than monthly.

Where 65% comes from

The industry rule of thumb puts prime cost between 55% and 65% of sales, with limited-service concepts at the low end and full-service at the high end. Above 65%, the received wisdom goes, there isn't enough left to carry rent, insurance and debt service, and volume stops fixing the problem.

That's a useful ceiling and a bad target. It's an average across concepts whose cost structures have almost nothing in common. A bar selling mostly cocktails at a 20% pour cost and a bakery running 35% labor are being held to the same number, and it can't fit both.

Building your own target

Prime cost has two components and each one is benchmarked against a different base. Cost of sales is measured against the revenue that produced it — food cost against food sales, beverage cost against beverage sales. Labor is measured against total sales. Blend them by your actual mix and you get a target that describes your restaurant instead of the industry.

The published ranges to build from:

Work an example. A full-service room doing 75% food and 25% beverage, at a 30% food target and a 22% beverage target, carries a cost-of-sales target of 27.5% of total sales. Add a 32% labor target and the prime cost target is 59.5% — not 65%. That room is over target at 62% even though it looks fine against the industry rule.

Now the same exercise for a bar at 30% food and 70% beverage. Food at 32% and beverage at 21% gives a cost-of-sales target of 24.3%. Labor at 30% puts prime cost at 54.3%. A bar running 62% prime is eight points over its own target while sitting comfortably inside the industry range — and eight points of a $1.3M business is more than $100,000 a year.

The industry range tells you when you're in trouble. Your mix-adjusted target tells you when you're leaving money on the table.

The wine list problem

Beverage targets deserve their own caution. The 18–24% blended benchmark describes a balanced program. Wine runs structurally higher — often 30% or more — so a wine-focused restaurant can blend at 26–30% and still be run well. Judging that program against 22% would send you hunting for theft that isn't there.

The reverse is also true. If your blended pour cost is 30% and it's driven by spirits or draft rather than wine, that's not a mix story. That's overpouring, pricing that never followed cost increases, or shrinkage, and it's worth a weekly inventory count until you find it.

What to do when you're over

Three levers, in order of how fast they work and how well they hold.

Price

Fastest, and the only one that adds revenue rather than removing cost. A 3% increase drops almost entirely to the bottom line, less card fees. The assumption worth testing is that you hold covers — raise a handful of items first and watch the mix before you touch the whole menu.

Labor

Start with the two slowest hours of each day rather than with headcount. Overstaffing lives at the edges of shifts, and a single scheduled hour removed from the front of a daypart is invisible to guests. One point of labor on a $1.5M room is $15,000 a year, or about twelve scheduled hours a week at a $24 loaded rate.

Food cost

Slowest, and the one that stays fixed once you fix it. Recost your twenty highest-selling items at this week's invoice prices. Wholesale food costs sit more than 35% above pre-pandemic levels, and most menus are still priced against numbers that are two vendor increases old.

Prime cost diagnoses; it doesn't grade. A restaurant can run a textbook 60% prime cost and still lose money if occupancy is 20% of sales, and no amount of food cost work will save it — that's a lease problem wearing a cost-control costume.

Running it weekly

Monthly reporting tells you about a problem four weeks after you could have done something about it. Prime cost is made of the only two costs a manager can move inside a single week, which makes it the one figure worth pulling every Monday.

A weekly prime cost report needs four inputs: sales by category from the POS, purchases from the week's invoices, a physical inventory count, and payroll from the timekeeping system. The count is the part people skip, and skipping it turns the report into noise — purchases divided by sales moves with inventory swings rather than with anything the kitchen did.

Counting a short list beats counting nothing. Ten or fifteen high-value, high-movement items — proteins, cheese, spirits — capture most of the variance and take twenty minutes. A full count monthly and a short count weekly is the practical compromise most operators land on.

Many operators also run thirteen four-week periods instead of twelve calendar months, so every period contains the same number of Fridays and Saturdays. Calendar months don't, which means a month with five weekends looks like an improvement that isn't one.

The number that matters more