Actual food cost for a week or a month, calculated properly from inventory rather than from invoices — including what comps and staff meals really cost you.
—
| Cost of goods sold | — |
| Inventory movement | — |
| COGS at your target % | — |
| Difference vs target | — |
COGS = beginning inventory + purchases − ending inventory, and food cost % is COGS ÷ food sales. Purchases alone are not your food cost: a big delivery on the last day of the period sits in ending inventory and belongs to next period, which is exactly why "invoices ÷ sales" swings wildly week to week and tells you nothing.
Comps, staff meals and transfers to the bar are subtracted here because the product left the kitchen without generating food sales. Leaving them in quietly inflates your food cost and sends you hunting for a theft problem that is really a free-staff-lunch policy.
Run this on the same weekday every period, counting at the same time of day, before any delivery. Consistency in when you count matters more than precision in what you count — a number that moves for real reasons is worth far more than an exact number you take twice a year.
COGS ÷ food sales, where COGS = beginning inventory + purchases − ending inventory. Subtract comps, staff meals and transfers out, because that product left the kitchen without producing food sales.
Because a large delivery near the end of a period sits in ending inventory and belongs to the next period. Invoices ÷ sales swings wildly week to week and cannot tell you whether you have a real cost problem.
Weekly if you can, monthly at minimum, always on the same weekday at the same time of day and before any delivery. Consistency in when you count matters more than precision in what you count.