Restaurant food cost 2026: how to calculate and steer your margin

Food cost — the food-to-revenue ratio — is the restaurant operator's survival metric. One extra food-cost point on a venue turning over €600,000 means €6,000 of margin vanishing over the year, with nobody noticing immediately. This article explains how to calculate it correctly, distinguish theoretical from actual, and above all how to steer it.

Food cost is the cost of raw materials consumed divided by the revenue they generated, expressed as a percentage. A 30% food cost means that for every €100 of dishes sold (net of VAT), €30 went on raw materials. The target depends on the concept:

These ranges are indicative: a steakhouse will structurally run a higher food cost than a pasta concept, and that is fine as long as price and rotation follow.

Actual food cost is not computed on purchases alone. The correct formula goes through consumption, which accounts for the change in stock:

Food consumption = Opening stock + Purchases − Closing stock

Food cost % = Food consumption / Net revenue

The most common mistake is dividing the month's purchases by the month's revenue. That is wrong: a month with stock build-up would show an artificially high food cost, and vice versa. Without a valued stocktake, food cost is meaningless — which is why a monthly stocktake is non-negotiable.

Theoretical food cost is what your recipes should produce. It is built from recipe cards: for each dish, you add the cost of every ingredient (to the gram), then multiply by quantities sold. It is the "ideal" food cost — what you would get if nothing was lost.

Actual food cost is measured by stocktake. It captures everything the theoretical ignores: waste, over-portioning, comps, breakage, theft, till errors.

The theoretical/actual gap is the most useful indicator in the restaurant. A 1–2 point gap is normal. Beyond 3 points, look for the leak: non-standardised portions, waste and expired stock, untracked comps, till entry errors, or theft. Without recipe cards you cannot compute the theoretical, so you cannot measure the gap — you steer blind. Recipe cards are the foundation, and they enable menu engineering: ranking dishes by popularity and margin to decide which to promote, rework or drop.

Food cost is necessary but insufficient. The real compass of profitability is prime cost: food cost (+ beverage cost) + fully-loaded payroll, divided by net revenue, with a target below 65%. An operator can "buy" a good food cost by overstaffing the kitchen, or cut payroll by buying pricier semi-prepared products. Only prime cost captures that trade-off. See our articles on restaurant financial KPIs and profitability.

In order of return: set up recipe cards and compute theoretical food cost; run a monthly valued stocktake and compute actual food cost; measure the gap and find the first leak; standardise portions; renegotiate the 20% of suppliers that make 80% of purchases, data in hand; work the menu (engineering); and track comps and waste. None is spectacular alone; together they routinely deliver 2 to 4 points — tens of thousands of euros a year on an average venue.

Food cost gets all the attention; its twin, beverage cost — drink purchases over drink revenue — is often forgotten. Yet drinks (soft drinks, coffee, wine, beer, cocktails) are frequently the highest-margin items on the menu, and a drift there is as costly as in the kitchen. The usual target is 22–28%, but it varies sharply by category: a coffee or soft drink has a very low material cost, a wine by the glass depends on the bottle price and glasses served, a cocktail on the spirit measure. A blended beverage cost hides these gaps; read it by family (hot drinks, softs, beers, wines, spirits/cocktails). Classic leaks: over-pouring spirits, draught losses, untracked comps, breakage and theft — high unit value makes drinks a common shrinkage zone. For high-volume bars, reconcile alcohol purchases against 20%-VAT drink revenue — the same cross-check the tax authority uses in an audit.

Food cost is only useful if tracked regularly and consistently. Three rhythms: monthly (the minimum, on the month-end stocktake); weekly (recommended for high volume and quick service); and continuous (via management software linked to the till and supplier orders). The tool matters less than the discipline: a stocktake done at the same moment, on the same scope and valuation, produces a comparable series over time — and it is the trend, not an isolated figure, that signals drift. A food cost moving from 30% to 33% over three months is a signal; the same 33% seen once says nothing. We set the frequency to the venue's volume and read food cost alongside payroll and cash via prime cost — see our restaurant financial KPIs article.

Food cost is a steering tool, not a year-end statistic. Compute it on consumption (with a stocktake), compare theoretical and actual, and always read it alongside payroll via prime cost. To embed this in your accounting, see our restaurant accounting support and the complete 2026 restaurant accounting guide.

Updated 3 June 2026. This article sets out a general method for steering food margin; targets are indicative and depend on your concept. Sources: INSEE, French GAAP, HCR sector benchmarks.