How to Find Restaurant Food-Cost Variance

Build one reliable recipe before scaling up

Choose a frequently sold dish with ingredients that are practical to count. Record the standard quantity of each ingredient, the unit, expected yield and allowed modifiers. Distinguish raw weight from usable or cooked weight. If the recipe uses grams but purchasing records kilograms, define the conversion once and test it rather than relying on staff to calculate it mentally during every receipt.

Keep the recipe version and effective date so a changed portion size does not silently rewrite the interpretation of earlier sales. Odoo’s bill-of-materials documentation explains component structures in the documented product version; restaurant recipe depletion still needs verification in your actual POS and inventory setup. Ask to see the complete flow, not merely a recipe screen.

Calculate theoretical consumption from sales

For one ingredient, theoretical usage equals the number of portions sold multiplied by the standard quantity per portion, adjusted for recorded modifiers and other legitimate recipe variations. Use completed sales on a defined date range and handle voids or returns consistently. If several dishes use the ingredient, add their expected usage rather than judging the ingredient against one dish alone.

Illustrative example: 200 sold portions at 0.20 kg each imply 40 kg of theoretical usage. That is an expectation, not a physical measurement. It is only credible if the recipe, sales count and unit conversion are correct. Keep the calculation visible so a chef or manager can challenge an input instead of dismissing an unexplained variance percentage.

Calculate actual usage on the same basis

For the same period and location, a simple quantity model is opening stock plus receipts plus transfers in, minus transfers out and closing stock. Make any supplier-return treatment explicit so it is not counted twice. Ensure counts use comparable timing and include the relevant storage and preparation locations. Missing a freezer from one count can create an apparent consumption spike.

Continuing the illustrative example, opening stock of 20 kg plus receipts of 50 kg minus closing stock of 25 kg gives 45 kg of actual consumption, assuming no transfers or supplier returns. Against 40 kg theoretical usage, the difference is 5 kg. Do not immediately label it theft or waste: first establish what legitimate movements and measurement issues explain it.

Separate explained usage from unexplained variance

Record staff meals, spoilage, preparation loss and complimentary portions using clear categories. In the example, 3 kg of documented waste and staff consumption could explain part of the 5 kg difference, leaving 2 kg to investigate. Check whether any of those quantities are already included in the recipe yield before subtracting them again. A well-labelled log prevents double counting.

At an illustrative ingredient cost of AED 30 per kg, 2 kg represents AED 60 at that chosen cost basis. It is not automatically a realised cash saving or the full margin effect. Keep quantity investigation separate from supplier-price changes and accounting valuation. Improve inventory accuracy first if receiving and stock counts cannot be trusted.

Turn the weekly review into a kitchen action

Review a small set of high-impact ingredients with the chef and stock owner. Ask whether the recipe changed, portions drifted, a purchase unit was wrong or waste went unrecorded. Assign one action and owner for each material exception. Examples include recalibrating a portion measure, correcting a conversion or changing how damaged deliveries are recorded.

Check the same ingredient in the following period using the same definitions. A reduction in unexplained variance is useful evidence; a lower number produced by changing the recipe to match whatever was consumed is not. Avoid using the report purely to blame a shift. The aim is a process people trust enough to record problems honestly.

Use reliable costs for menu decisions

Once quantities are credible, combine recipe cost with selling price and volume to review contribution by dish. State whether figures include or exclude tax and which other costs are outside the calculation. Ingredient margin is not the same as restaurant profit. Packaging, delivery charges, labour and other overheads can change the commercial decision even when food cost looks attractive.

When evaluating restaurant systems, request this worked flow using your own dish and a controlled sample of transactions. Our POS buying guide shows how to test the surrounding sales and kitchen processes. Expand recipe coverage only after the first calculations reconcile and somebody owns ongoing updates.

A checklist to use in your next review