Food cost percentage, calculated honestly.

Food cost percentage is cost of goods sold divided by food revenue for the same period. Cost of goods sold is opening stock, plus purchases, minus closing stock. Invoices alone are not cost of goods sold, they are purchases, and using them instead is the single most common reason a restaurant's food cost looks fine on paper while the bank balance disagrees. Without a count at both ends you are measuring buying, not using.

It varies enormously by format, so a single industry benchmark is close to useless. What matters more is your own trend over several months and whether it moves when prices move. As a rough orientation: dough and rice-led formats sit lower because the base ingredient is cheap, while protein-led menus sit higher and always will. A number is only good or bad relative to the menu that produced it.

When food cost drifts up and the menu has not changed, it is nearly always one of four things, in roughly this order of frequency. None of them show up on an invoice.

A full stock count every week is not realistic for most kitchens and does not need to be. Count the lines that carry the money, which for most restaurants is meat, fish, cheese and spirits, and count them weekly. Everything else can go monthly. This gets you most of the accuracy for a fraction of the effort, and it catches a portioning problem in days rather than in a quarter.

When the percentage moves the instinct is to reprice, and it is usually the wrong first move because it exports your problem to the customer. Work through portion control, waste and supplier terms first. Reprice when input costs have genuinely moved and you have exhausted the rest, then do it visibly and on fewer items than you think, targeting the dishes where the customer is least price-sensitive.