What prime cost is and why operators watch it weekly
Prime cost is the sum of two lines: cost of sales and total labour. The Uniform System of Accounts for Restaurants, the industry's standard chart of accounts, defines it that way, and the reason it gets its own name is that those two lines are the only large costs a manager can move inside a week. Rent is fixed by a lease. Insurance is fixed by a policy. Food cost and labour respond to decisions made this morning.
Watching them together rather than separately is the whole point. Food cost and labour trade against each other constantly: buying pre-cut vegetables raises cost of sales and lowers prep hours; making pasta in house does the opposite. Judge either one alone and you will congratulate a kitchen for a 27% food cost that it bought with eleven extra labour hours. Prime cost closes that loophole because both sides of the trade land in the same number.
The weekly cadence matters as much as the formula. A monthly P&L arrives two to three weeks after the month it describes, by which point the schedule that caused the problem is four weeks old. A prime cost run every Tuesday on last week's numbers is imprecise — you are estimating inventory rather than counting it fully — but it is early enough to change next week's schedule and next week's order.
The formula, and the two places it goes wrong
Prime cost = cost of sales + fully loaded labour, and prime cost % = prime cost ÷ total sales × 100. The arithmetic is trivial; the definitions are where the errors live.
Cost of sales is not purchases. It is opening inventory + purchases − closing inventory. A week in which you bought three cases of steak you have not yet sold shows a high purchase figure and a normal cost of sales, because those steaks are still on the shelf. Substituting purchases for cost of sales makes prime cost swing wildly with delivery timing and teaches you nothing. If you cannot count inventory weekly, count the high-value categories — proteins and liquor — and carry the rest at a steady estimate.
Labour is not gross wages. Employing someone costs their wage plus the employer's share of payroll taxes, workers' compensation premiums, insurance and any benefits. That burden is real money leaving the business, so it belongs inside prime cost. Take the burden percentage from your own payroll reports rather than assuming one: divide total employer payroll costs by gross wages for a recent quarter and use that figure. The default here is 20%, which is a placeholder, not a benchmark.
Two conventions to keep straight on the supporting ratios. Food cost percentage uses food sales as its denominator, and beverage cost percentage uses beverage sales, because that is how menu and pour costing work. Labour percentage and prime cost percentage use total sales, because labour serves both revenue streams and cannot be split cleanly. This calculator follows both conventions, which is why the outputs and the build-up table use different denominators — the table says so in its note.
Worked example: a full-service week on $100,000 of sales
These are the values the calculator loads with, and they describe a busy independent restaurant's week.
- Total sales. $80,000 food + $20,000 beverage = $100,000.
- Cost of sales. $24,000 food + $4,000 beverage = $28,000.
- Fully loaded labour. $22,000 hourly + $8,000 salaried = $30,000 of gross wages. With a 20% burden: 30,000 × 1.20 = $36,000.
- Prime cost. 28,000 + 36,000 = $64,000.
- Prime cost percentage. 64,000 ÷ 100,000 = 64.00%.
- The ratios underneath. Food cost 24,000 ÷ 80,000 = 30.0%. Beverage cost 4,000 ÷ 20,000 = 20.0%. Labour 36,000 ÷ 100,000 = 36.0%.
- Gap to a 60% target. 64,000 − (100,000 × 0.60) = $4,000 above target for the week.
That $4,000 is the actionable number, and it is worth translating before you act. Against the $28,000 cost of sales it is a 14.3% cut in what you buy — 4,000 ÷ 28,000 — which no purchasing negotiation delivers. Against the $36,000 labour bill it is an 11.1% cut, and at a fully loaded rate of $24 an hour that is 4,000 ÷ 24 = about 167 hours, or roughly four full-time positions. Framing the same gap both ways turns it into two very different management problems, and the labour side is usually the tractable one.
Annualised, $4,000 a week is $208,000 — which is why operators chase points, not pennies. One percentage point of prime cost on this restaurant is $1,000 a week.
What prime cost percentage is normal, and what to do about yours
The widely used industry rule of thumb puts full-service restaurants near 60-65% and quick service somewhat lower, with 60% often quoted as the line to stay under. Treat those as orientation, not as law: a high-rent urban site with counter service and a rural full-service dining room can both be healthy at very different prime costs, because what matters is what is left after prime cost against what the rest of the business costs.
The honest test is arithmetic on your own P&L rather than a benchmark. Add your occupancy costs, utilities, marketing, repairs, insurance, administrative expenses and any debt service as a percentage of sales, add the profit margin you need, and subtract the total from 100%. What remains is your prime cost target. If occupancy and other operating costs run 28% of sales and you need 7% to be worth doing, your target is 65% — and 62% would be a good week rather than a mediocre one.
Read the split, not just the total. Two restaurants at 64% prime are in different trouble: one at 34% food and 30% labour has a purchasing and portioning problem; one at 26% food and 38% labour has a scheduling problem. The recipe cost per serving calculator attacks the first by giving you a theoretical plate cost to compare your actual food cost against. The gap between theoretical and actual is where waste, over-portioning and theft live.
Watch the direction of travel above all. A single week is noisy — one large event, one delivery counted on the wrong side of the inventory date, one holiday — so plot four weeks and act on the trend.
What each prime cost percentage leaves behind
| Prime cost | Prime cost $ | Left after prime cost | Profit if other costs run 28% of sales | Annualised profit (52 weeks) |
|---|---|---|---|---|
| 50% | $50,000 | $50,000 | 22.0% ($22,000) | $1,144,000 |
| 55% | $55,000 | $45,000 | 17.0% ($17,000) | $884,000 |
| 60% | $60,000 | $40,000 | 12.0% ($12,000) | $624,000 |
| 64% | $64,000 | $36,000 | 8.0% ($8,000) | $416,000 |
| 65% | $65,000 | $35,000 | 7.0% ($7,000) | $364,000 |
| 70% | $70,000 | $30,000 | 2.0% ($2,000) | $104,000 |
| 72% | $72,000 | $28,000 | 0.0% ($0) | $0 |
| 75% | $75,000 | $25,000 | −3.0% (−$3,000) | −$156,000 |
Every figure is computed from the two stated assumptions: $100,000 of sales and other operating costs at 28% of sales. Substitute your own other-cost percentage and the break-even prime cost moves to 100% minus that figure.
Follow one chart of accounts, consistently
Prime cost only trends usefully if the same expenses land in the same buckets every week. The Uniform System of Accounts for Restaurants exists to settle those questions — whether employee meals belong in cost of sales or in labour, whether delivery fees are cost of sales or an operating expense — and following any consistent scheme beats arguing them afresh each period. The most common inconsistency is management salary: include it, and include it every week, because a restaurant whose manager is unpaid is not cheaper to run, only cheaper to report.
Errors that make prime cost look better than it is
- Using purchases instead of cost of sales. Without the inventory adjustment, the number tracks delivery schedules rather than consumption.
- Leaving out payroll burden. Gross wages understate the cost of employing people by whatever your taxes, insurance and benefits actually run.
- Leaving out owner or manager labour. Unpaid owner hours are a real cost that reappears the day you have to hire a replacement.
- Using gross sales including sales tax. Tax collected is not revenue; including it inflates the denominator and flatters every percentage.
- Mismatched periods. A payroll week ending Sunday against a sales week ending Saturday introduces a full day of error every period.
- Ignoring comps and voids. Food given away costs the same to produce as food sold, so it raises cost of sales without adding sales.
- Counting catering revenue but not its labour. Off-site events carry travel and setup hours that never touch the restaurant schedule.
How prime cost connects to the rest of your cost control
Prime cost is the scoreboard; the other calculators in this category are the plays. Your food cost percentage is the aggregate result of every plate you sell, and the only way to know what it should be is to cost the plates: the recipe cost per serving calculator gives you a theoretical cost per dish, which you weight by sales mix to get a theoretical food cost for the whole menu. When actual runs above theoretical, the difference is waste, over-portioning, comps or shrinkage, and none of those are fixed by raising prices.
Underneath the plate cost sits purchasing. The food yield percentage calculator and the edible portion cost calculator tell you what your ingredients truly cost after trim, which is where a supposedly cheap supplier turns out to be expensive. And if you run events, the catering food quantity per person calculator keeps the ordering side honest, since over-ordering for a 200-cover event shows up in cost of sales three weeks later as a mystery.
What prime cost cannot tell you is whether your prices are right, whether the menu mix is working, or whether you are busy enough. It is a cost ratio, not a demand signal. A restaurant with an excellent 55% prime cost and half-empty dining rooms is still failing; a restaurant at 66% with a queue may simply need to raise prices by three percent. Read prime cost alongside sales trend and cover count, never alone.