UK hospitality labour cost benchmarks (2026)

Labour cost % = total labour cost ÷ net revenue (ex-VAT) × 100, for the same period.

Total labour cost should include: gross wages, employer NIC (15% on earnings above £96.15/week from April 2025), employer pension contributions (3% minimum on qualifying earnings), holiday pay accrual (~12.07% of hours worked) and agency cover. Wages-only figures understate true cost by 10–15%.

Wet-led pubs: 22–28%. Drinks revenue is high-margin and faster per server, which keeps labour as a share of sales lower than food-led venues.

Cafés / coffee shops: 25–30%. High traffic, low ticket — the labour line is sensitive to peak-hour staffing decisions.

Full-service restaurants: 28–35%. The widest range. Casual dining sits at the lower end; fine dining can run to 40%+ once kitchen brigade and sommelier roles are factored in.

Quick-service / takeaway: 22–28%. Pre-portioned production and counter service compress labour more than table service.

Hotel F&B: 30–38%. Multiple service periods and slower covers per server push the number up.

Sales mix. A higher drink-to-food ratio usually drops labour %. A high dessert and side-order attach rate raises ticket without adding service time.

Service style. Counter and QR-order at table reduce front-of-house labour. Full table service with sommelier upsell raises it but lifts revenue per cover.

Trading pattern. Two short peaks (lunch and dinner) waste hours either side; all-day trading tends to flatten labour %.

Pay rates. The April 2026 NLW rise to £12.71/hr will add ~4% to most hospitality wage bills versus the 2025 rate. Adjust benchmarks accordingly when comparing year-on-year.

Below the typical range isn’t automatically a win. Common signs of under-staffing: rising wait-times, dipping review scores, table-turn times stretching, kitchen errors, staff turnover spiking. If the labour line is comfortably under benchmark but those metrics are slipping, you’re trading service for short-term margin.

Monthly labour % is too late to act on. Track weekly, by trading day, against forecast revenue. Most overshoots come from one or two specific shifts — usually a Monday lunch or a quiet midweek dinner — not a structural problem.

A simple weekly review (planned vs actual hours, vs revenue) catches drift before it costs a quarter.