# Restaurant food cost, labour & prime cost benchmarks by segment (2026)

> What healthy food cost, labour and prime cost percentages look like across restaurant segments — from quick-service to fine dining — and how to read your own numbers against the band.

Published: 2026-06-15 · Author: Papaya

Ask two operators what a "good" food cost is and you will get two different answers — because they are running different businesses. The number that travels across concepts is **prime cost**: food cost plus labour, as a share of sales. It is the biggest slice of the P&L a restaurant actually controls, and it is where margin is won or lost.

The table above shows the quartile bands for each restaurant segment. Here is how to use them.

## Food cost sets the ceiling, labour sets the floor

Food cost and labour trade off against each other, which is why reading either alone is misleading:

- A **fine-dining** kitchen buys expensive ingredients but the plating, service and skill are what guests pay for — so food cost can look moderate while labour runs high.
- A **quick-service** operation is the opposite: cheap, standardised ingredients but a lean crew, so labour is low and food cost carries more of the cost base.
- A **bar** buys the cheapest cost-ratio product there is — a pour of spirit — so its blended food cost sits well below every food-led segment.

That is why the same 60% prime cost can be built very differently in two healthy venues. Compare each half to your own segment, not to a universal rule.

## What "on the wrong side of the median" costs you

Prime cost compounds fast. On €85,000 of monthly revenue, every point of prime cost is €850 a month — about €10,000 a year. A venue sitting five points above its segment median is quietly giving up €50,000 a year versus a typical peer, before it has done anything wrong on the menu.

That is the case for treating the benchmark as a floor to beat, not a line to relax at.

## How to close the gap

Whichever half is further from the band is the one to work first:

- **Food cost high?** Count weekly, re-cost recipes against current supplier prices, and tighten the few dishes where theoretical and actual usage diverge most. Our [food-cost guide](/blog/cutting-food-cost) walks the full playbook.
- **Labour high?** Look at scheduling against covers by day-part before you touch headcount — most labour overruns are a rostering problem, not a wage problem.

Then re-measure next month. The point of a benchmark is not the one-time comparison; it is watching your number move toward — and past — the median.