Food cost variance compares theoretical food cost, which is what the sold menu items should have used based on their recipes, with actual food cost, calculated from stock counts and purchases. The difference is the variance.
Actual usage is opening stock plus purchases minus closing stock. Theoretical usage comes from POS sales multiplied by recipe quantities. If the venue sold 200 burgers and each uses a set quantity of beef, the system can say how much beef should be gone, and the count shows how much actually is.
Restaurants, cafés, bars and quick-service chains review variance weekly or monthly to find problems. A positive gap can point to waste, over-portioning, theft, unrecorded comps, receiving mistakes or recipes that no longer match what the kitchen does. Bars use the same approach for liquor pour variance.
Pitfalls include poor counts, unrecorded transfers or wastage and recipes that are not set up properly, all of which produce noise rather than insight. Variance is a prompt to investigate, not proof of theft, and a small percentage is normal.
Tracking variance turns stocktakes into a management tool and helps protect margin in a business where food costs are a large share of expenses. Consistent counting days and methods make week-to-week comparisons far more reliable.