To know your restaurant's profit daily you need three numbers: net sales, labour cost and food cost. Labour plus food is your prime cost, the two lines that decide whether a restaurant makes money. Everything else is comparatively fixed. Tracked daily, prime cost tells you today what a month-end P&L tells you six weeks too late.
The one formula that matters
Prime cost = total labour cost + total cost of goods sold. As a percentage: prime cost % = (labour cost + food cost) ÷ net sales × 100. This single number is the fastest honest read on whether a service, a day or a branch made money.
Where each number comes from
- Net sales: from your POS, after discounts, comps and voids, and excluding the 15% VAT you collect on behalf of ZATCA
- Labour cost: from scheduling and payroll: wages plus employer costs for hours actually worked, not hours scheduled
- Food cost, opening inventory + purchases − closing inventory for the period
The third one is where most operators stop, because counting inventory daily is impractical. The workaround is theoretical food cost: what your recipes say you should have consumed given what the POS sold. It is available instantly, and the gap between it and your periodic actual count is itself the number worth watching.
Why the month-end P&L is too late
A P&L closed on the 10th of the following month describes decisions you made up to 40 days earlier. If portion sizes drifted in week one, you served four more weeks at the wrong margin before the report told you. The cost is not the reporting delay itself, it is every shift you ran between the problem starting and you finding out.
Getting to daily without a spreadsheet
Daily prime cost requires POS, scheduling and purchasing to be readable in one place. That is an integration problem, not an accounting one. With pre-built connectors to the systems you already run, the three numbers assemble themselves and the calculation happens continuously, which is why operators on DataGrid report 75% less time spent building reports.
What to do with the number
Set a prime cost target per branch, then read the daily figure against it rather than against last month. A single day above target is noise. Three consecutive days above target is a structural problem in either the rota or the recipes, and it is cheap to fix in week one and expensive to fix in week six.