Benchmarks

Restaurant Labour Cost Percentage: What Independent Venues Actually Spend (2026)

US full-service restaurants ran a median labour cost of 36.5% of sales in 2024. Germany's last official ratio puts restaurant personnel costs at 31.7% of turnover — in 2019, the freshest full figure that exists. Here are the verified benchmarks, the misquoted German number, what 2026 wage rules really add in Germany and the UK (including the offsets most coverage omits), and what one percentage point is worth to your venue.

Alex Riesenkampff

July 5, 2026 · Updated July 6, 2026 · 12 min read · Markdown

If you run an independent restaurant, labour is almost certainly your largest controllable cost — and in 2026 it is the one moving fastest. The freshest audited benchmark puts the median labour cost for full-service restaurants at 36.5% of sales, and for limited-service venues at 31.7%, including benefits. Those figures come from the US National Restaurant Association's operations data for 2024, and they are several points above what the same survey showed a decade ago.

Europe's official equivalent exists — but it runs years behind. The last full personnel-cost ratio for German restaurants in the EU's structural business statistics is from 2019: 31.7% of turnover. Everything since is either partial or misquoted, which is exactly why this article exists. The shape of the answer is the same in Cologne, Camden, or Chicago — roughly a third of everything you take goes to the people who make the place run, the statutory floor under that third is rising, and the difference between a venue at 33% and one at 31% is usually scheduling, not wages.

The benchmarks that actually exist

The National Restaurant Association's operations data puts median labour cost at 36.5% of sales for full-service restaurants and 31.7% for limited-service restaurants in 2024 — roughly 3.5 percentage points above the mid-2010s. It is the only large-sample benchmark with a current reference year; the NRA's own analysis is blunt about the direction: labour costs are "well above historical averages".

Two things stop you from copying the US numbers straight onto a European P&L. First, US labour costs are structurally understated compared to Europe: the federal tipped cash wage has been $2.13 an hour since 1991 (states vary), so for tipped roles a large slice of what a German or British operator would book as payroll arrives as guest tips instead. That makes the convergence in the table more striking, not less — even with the tip credit suppressing front-of-house wages, US full-service labour now runs above the last official German ratio. Labour scarcity is overriding regional wage structures. Second, service models matter more than country: the 4.8-point gap between the two US segments is bigger than most country gaps.

The more durable discipline is the one operators actually manage to weekly: prime cost. Keep cost of goods plus total labour at or below 60–65% of sales — the convention used consistently across restaurant-operations platforms like Toast, Restaurant365, and 7shifts. Labour and goods trade off (bake in-house and you swap goods cost for labour cost), so the sum is the number that decides whether your model works.

The German number everyone quotes — and what the statistics actually say

Search for the German labour-cost benchmark and you will find the same line everywhere: personnel costs are about a third of turnover, "according to DEHOGA". We checked the current DEHOGA Zahlenspiegel (IV/2025) line by line: the "one third of turnover" figure attributed to DEHOGA does not appear in the Zahlenspiegel — the attribution is wrong, even though the number itself is roughly right.

Here is what the official statistics actually say:

• The EU's structural business statistics measured German food & beverage services at 31.7% personnel costs of turnover in 2019 — €22.78bn against €71.81bn (Eurostat, NACE I56, including employer contributions). That is the last year a full official ratio exists; 2020 is COVID-distorted (34.0% on collapsed revenue), and the series then changed.
• The current series publishes only wages and salaries — 25.1% of net turnover in 2023 — without employer social contributions. Anyone citing a current official full-cost share is extrapolating.
• The trajectory since the last real reading is measured, though: labour costs in the German Gastgewerbe rose 39.6% between Q4 2019 and Q4 2025 (Destatis labour cost index), while real revenue in 2025 was still 14.8% below 2019. The Zahlenspiegel puts it plainly: "Die Kosten für Waren, Personal und Energie sind seit 2022 teilweise um bis zu 40% gestiegen."

For deeper segment-level benchmarks (by concept, by quartile), the paid dwif/DEHOGA Betriebsvergleich exists — but there is no free, current, official full-cost ratio. That's the honest state of German benchmarking: a solid but six-year-old anchor, a measured cost trajectory, and a lot of confident misquotation on top.

What a German employee actually costs in 2026

The other number that "roughly" ruins: employer on-costs. The four core social-insurance employer shares sum to about 20.15% of gross pay in 2026 (7.30% health + 1.45% average supplemental + 9.30% pension + 1.30% unemployment + 1.80% care — TK contribution table). But that is not what you pay. Small employers also carry the U1 sick-pay levy (1.9–2.1% at the standard tier, mandatory up to 30 employees), the U2 maternity levy (0.42–0.44%, everyone), the insolvency levy (0.15%, fixed by law), and statutory accident insurance via the BGN (hospitality risk class 2.94 — around 1% of gross wages). All-in, a small German venue should budget roughly 23.5–24% on top of every gross wage in 2026; a €3,000 gross salary is really a line item near €3,715. Larger venues (over 30 staff) drop U1 and land closer to 21.5–22%.

And the floor is rising: the Mindestlohn jumps 8.4% to €13.90 on 1 January 2026, with €14.60 already legislated for 2027 (BMAS). The one piece of relief arriving the same day: VAT on restaurant food returned permanently to 7% (beverages stay at 19%) under the Steueränderungsgesetz 2025.

What 2026 adds to your payroll, by country

The statutory floor is moving in all three markets at once — but the impact is less linear than the headlines suggest.

Germany

From January 2026 the Mindestlohn is €13.90, an 8.4% jump in a single step — and every extra wage euro carries the ~24% all-in employer add-on with it. For a venue with staff at or near minimum wage, payroll rises ~8.4% on those hours before you change a single rota. The 7% food VAT is the offsetting hand the same legislation deals you; whether it nets out depends on your food-to-beverage mix.

United Kingdom

UK operators took the hit in two waves — but with a shield most coverage skips. In April 2025 employer National Insurance rose from 13.8% to 15% and the threshold fell from £9,100 to £5,000 — yet the same Budget doubled the Employment Allowance to £10,500 and removed its £100,000 eligibility cap, so a three-person full-time team still pays £0 employer NIC. Run the numbers on the April 2026 National Living Wage (£12.71/h, ~£24,785 a year full-time): each such employee generates about £2,968 of employer NIC, so the allowance absorbs roughly the first three and a half. At seven staff, the new regime costs about £10,274 net against £10,151 under the old rules — a wash. The £3.4 billion sector-wide hit UKHospitality's Kate Nicholls described — "At the Budget in October, the Government delivered an increase of £3.4 billion to the sector's 2025 annual tax bill" — is real, but it lands overwhelmingly on payrolls big enough to exhaust the allowance: multi-site groups and larger venues. April 2026 then added the second wave for everyone: the National Living Wage up 4.1% to £12.71 and the 18–20 rate up 8.5% to £10.85 — roughly £1.4 billion in wage costs no allowance offsets.

United States

No federal movement — the minimum wage stays $7.25 and the tipped cash wage $2.13 — but the market is doing the work instead: the NRA's 2026 outlook projects $1.55 trillion in sales while reporting that 42% of operators ran unprofitable in the prior year, with more than nine in ten citing labour among their most significant cost challenges. Median labour shares at record levels are the mechanism.

What one percentage point is worth

Percentages hide the money. Put your own revenue in:

For a typical independent venue doing €25,000 a month, one point of labour share is €3,000 a year — roughly a week of revenue. Two points is a decent espresso machine, every year. That is why the gap between 33% and 31% matters more than most menu decisions — and why cutting hours blindly is the most expensive way to chase it.

How to calculate your own labour cost percentage — properly

The formula is one line — total labour cost divided by revenue, times 100 — and almost everyone feeds it the wrong ingredients. Done right:

1. Pick a clean period. A full calendar month or a fixed four-week cycle, and use the same period on both sides of the division. Mixing a calendar month of wages with a POS export that runs Monday-to-Sunday quietly skews the number.
2. Count the full cost of labour, not the wage slips. Gross wages plus everything on top — in Germany roughly 23.5–24% all-in for a small venue (social insurance, U1/U2/U3 levies, accident insurance), in the UK employer National Insurance at 15% above £5,000 per head after your £10,500 Employment Allowance is exhausted, in the US payroll taxes and benefits. Add holiday and sick accrual, bonuses, and agency shifts. In this calculation a €3,000 gross salary in Germany is a line item near €3,715 — count it that way or the benchmark comparison is meaningless.
3. Put your own hours in. If you cook, pour, or cover shifts, price those hours at what you would pay a stranger to do them. A venue that only works because the owner works free is not at 28% labour — it is at 35% with an invisible subsidy.
4. Use net revenue for the same period. Strip VAT before dividing — and in Germany food and drinks now carry different rates (7% food, 19% beverages since January 2026), so net out each category rather than applying one blended rate.
5. Then cut it by daypart. The monthly number tells you whether there is a problem; covers per labour hour by weekday and daypart tells you where it lives. That is the version worth tracking weekly.

If that sounds like an evening of spreadsheet work — it is, every month, done by hand. It is also exactly the kind of bookkeeping Super44 exists to absorb: connect your POS, share your accounts, answer a few questions in chat, and it keeps your labour share current — employer costs included, by daypart — and comes back with the specific, numbered actions hiding in it, like the rota gaps and dead hours in the next section.

When cutting labour costs you revenue: a real example

Cutting the percentage by cutting hours only works if the hours weren't earning. One neighbourhood café in Germany moved its Tuesday opening from 10:00 to 12:00 to save two staff hours — and when we measured it, Tuesday revenue had dropped from €299 to €191. Net of roughly €30 in saved labour, the change cost about €1,700 per quarter. The rota looked leaner; the P&L said otherwise. (The fix wasn't reverting blindly, either — an 11:00 compromise recovered most of the revenue at half the extra hours.)

That is the trap with labour percentages: the denominator is revenue, and revenue responds to staffing. Judge every schedule change by both lines, not by the wage line alone.

Where an out-of-line labour share actually comes from

In practice, a labour share that sits above its segment's benchmark traces to scheduling patterns far more often than to pay rates. The culprits are boringly consistent:

1. Schedules built on habit. The Tuesday rota looks like the Saturday rota because it always has. Expected covers, not tradition, should set the roster.
2. Shift-change overlap. Two full teams on the clock for 60–90 minutes a day compounds to a part-time salary a year.
3. Prep on service hours. Peak-rate paid hours spent on work the quiet morning slot could absorb.
4. Quiet days staffed like busy ones. Without covers-per-labour-hour by daypart, nobody notices — the weekly total looks normal while Tuesday afternoon quietly runs at 60% labour.

None of these need a consultant. They need your POS covers data next to your rota, week by week — tedious by hand, trivial for software.

The squeeze is real — you are not imagining it

If your labour share has crept up despite decent trading, the sector data says that is the norm, not mismanagement:

• Germany: real (price-adjusted) hospitality revenue in 2025 was still 14.8% below 2019 — the sixth consecutive year of real losses — while 2,314 hospitality insolvencies were filed January–November 2025, up 25.8% year on year against +11% economy-wide (DEHOGA/Destatis).
• UK: 3,353 hospitality businesses entered insolvency in 2025; since the April 2025 cost wave, 84% of operators raised prices, 48% reduced staffing, and 61% cut hours (Buchler Phillips; CGA/Sona survey).
• US: 42% of operators reported running unprofitable in the prior year even as sales hit records (NRA).

Rising labour is the loudest pressure, not the only one — price rises are hitting the limits of what guests will pay, and leases and energy don't flex. Which is precisely why the venues that hold their margin are not the ones that cut deepest: they are the ones that know, weekly and by daypart, which hours earn their keep. That is a data habit, and it is the whole reason we built what we built.