Why your food cost looks fine on paper and terrible in the bank

A restaurant that sells 100 plates of biryani should have consumed a predictable quantity of rice, chicken, oil and spice. That is theoretical usage. What actually left the store is actual usage. The gap between the two is where your margin goes.

Where the gap comes from

• Over-portioning — the most common and least visible cause.
• Wastage and spoilage that never gets logged.
• Staff meals taken out of production stock.
• Voids and remakes that consume ingredients twice.
• Receiving short against a supplier invoice that was paid in full.
• Outright theft, which is usually the smallest of these.

Costing the dish, not the invoice

Most operators track food cost at the invoice level: total purchases divided by total sales. That number tells you something is wrong but never what. Recipe-level costing tells you which dish is bleeding.

Once each menu item carries a recipe, every sale deducts its component ingredients from stock automatically. Compare that theoretical deduction to a physical count and the variance is itemised — by ingredient, by outlet, by day.

You cannot fix a 4% food-cost problem. You can fix a specific dish that is over-portioning paneer by 15 grams.

Menu engineering follows naturally

When you know the true cost of each dish and you can see its sell-through, the menu splits into four quadrants: high-margin bestsellers to protect, high-margin slow movers to promote, low-margin bestsellers to re-cost or re-portion, and low-margin slow movers to cut.

That analysis is impossible without recipe data and it is nearly automatic once you have it.

Start small

You do not need to cost the entire menu on day one. Start with your top twenty items by volume — they usually account for the bulk of consumption. Get those recipes accurate, run a weekly variance report, and extend from there.