Why monthly financials are the wrong tool here
Restaurant margins are thin enough that a four-point swing in food cost is the difference between a good month and a loss. Both of the costs that produce that swing — food and labor — move every single day, driven by portioning, waste, theft, over-ordering, scheduling and a hundred small decisions made on the floor.
A monthly P&L that lands on the 20th tells you that something went wrong across a period that ended three weeks ago, which is far too late to identify the shift it happened on, let alone correct it. By the time you read it, you have already repeated the mistake for three more weeks.
This is why well-run restaurants do not manage on monthly financials. They manage on a weekly prime cost report, and use the monthly close for accuracy, taxes and trend rather than for control.
What prime cost is, and the targets
Prime cost is cost of goods sold plus total labor, expressed as a percentage of sales. COGS means food and beverage. Total labor means all of it — hourly wages, management salaries, payroll taxes, workers compensation and benefits, not just the wage line.
It is the right number to manage because it captures the two largest controllable costs in one figure, and because the tradeoff between them is real: cutting prep labor tends to raise food cost through worse yield, and cheaper ingredients tend to raise labor through more handling. Managing them separately produces exactly that whack-a-mole. Managing the combined number does not.
- Full-service, table service: 60-65% of sales
- Quick service and fast casual: 55-60%, on lower labor
- Bar-led concepts: often below 55%, on much stronger beverage margin
- Fine dining: frequently 65-70%, on high food cost and high service labor
Above 70% the model does not work
Occupancy, utilities, insurance, marketing, repairs and administration typically consume another 25-30% of sales. At 70% prime cost there is nothing left for any of it, and no amount of volume fixes a structural margin problem — it just makes the losses bigger.
Running the weekly number
The weekly prime cost is not a miniature month-end close. It is a fast, deliberately approximate calculation that a manager can complete in under an hour, and its value is entirely in the speed.
- Pull net sales for the week from the point of sale, excluding tax and comps
- Count inventory at the same time every week — a Sunday night or Monday morning count, held consistent
- Calculate usage: beginning inventory plus purchases less ending inventory
- Pull total labor from the payroll or scheduling system, and gross it up for taxes and benefits at your known burden rate
- Divide the sum by net sales, and compare against both your target and the prior four weeks
Theoretical versus actual: where the money is going
Actual food cost tells you what you spent. It does not tell you why. The diagnostic that does is the gap between theoretical and actual.
Theoretical food cost is what your ingredients should have cost given exactly what you sold: every item's recipe cost multiplied by the units sold that week, per the point of sale. Actual food cost is what you really consumed, from the inventory count. The difference is the leak.
A gap of one to two points is normal — trim waste, small portioning variance, measurement error. Three points or more is a problem with a specific, findable cause, and the list of causes is short enough to work through systematically.
- Portioning drift. The most common cause by a wide margin, and the one scales fix immediately.
- Waste and spoilage from over-prepping or over-ordering perishables against a forecast nobody updated.
- Unrecorded comps and staff meals — food that left the kitchen with no corresponding sale.
- Receiving errors. Invoiced for 40 pounds, delivered 36, nobody weighed it.
- Price increases absorbed silently because recipe costs were last updated eighteen months ago.
- Theft, which is real but is far less common than the five above and should be the last conclusion rather than the first.
Recipe costs go stale fast
Theoretical food cost is only as good as the recipe costs behind it. Re-cost your top 20 menu items quarterly at minimum. A restaurant running on two-year-old recipe costs is comparing actual spending against a number that stopped being true a long time ago.
Scheduling labor to forecast, not to habit
Labor is the half of prime cost you can adjust with the most notice, and most restaurants schedule from last week's schedule rather than from a sales forecast.
The discipline is to forecast sales by day and by daypart, convert that into required labor hours using your known sales-per-labor-hour, and schedule against the result. Then compare scheduled hours to actual hours every week and find out where the overage happened — usually early clock-ins, late clock-outs and shifts nobody cut when the dining room emptied.
Sales per labor hour is the metric that makes this concrete. Track it by daypart rather than in aggregate, because a weak Tuesday lunch and a strong Friday dinner average into a number that hides both.
What the monthly close still owes you
Weekly prime cost is a control tool built for speed, and it is approximate by design. It does not replace real accounting, and running one without the other produces a restaurant that is well controlled and badly reported.
- Accruals for invoices received after the period and for payroll spanning the period end
- A full inventory count reconciled to the general ledger, not just the weekly operational count
- Occupancy, insurance, utilities, marketing, repairs and administrative costs, none of which appear in prime cost
- Sales tax and tip reporting, which carry their own filing deadlines and penalties
- Depreciation on leasehold improvements and equipment, and any deferred rent or lease accounting
- Period-over-period comparison on a consistent calendar — most multi-unit operators use 4-4-5 or thirteen four-week periods precisely so that each period contains the same number of Fridays and Saturdays