Restaurant labor cost percentage: how to calculate it and how to actually bring it down

Labor is the half of your prime cost that fights back. Food cost sits still while you fix it; labor argues, calls in sick, and quits on a Friday. It's also the line most operators measure loosely — a number they "run about," pulled from a gut feel rather than the actual arithmetic — which is exactly why it drifts.

I spent four years as the operating partner who signed the payroll at a multi-location group, and labor was the number I watched most nervously, because unlike rent it moves every single shift. This is the version of labor cost that survives scrutiny: how to compute it properly, what's normal for your concept in 2026, and which levers actually move it — separated from the ones vendors only claim move it.

It's the companion to our food cost percentage guide from a few days ago — together they're the only two costs you truly control.

The formula — and the part everyone under-counts

Labor cost percentage is simple division:

Labor cost % = (total labor cost ÷ total sales) × 100

A restaurant doing $50,000 in sales with $15,000 of labor runs 30%. The trap isn't the formula — it's the word "total." Wages are the visible part; the fully-loaded cost is bigger:

• Hourly wages and salaries
• Overtime
• Employer payroll taxes — FICA (7.65%), FUTA, and state unemployment (SUTA), which varies by your experience rating
• Workers' compensation (for restaurants, roughly $1 per $100 of payroll)
• Benefits, paid time off, bonuses, training

A fully-burdened employee commonly costs 25–40% more than base wage. Federal data backs the scale: in the leisure-and-hospitality sector, benefit costs run around 18% of total employer compensation. If you budget to wages alone, you're under-counting your real labor line by roughly a quarter — and wondering later where the money went.

What "normal" looks like by concept

Most restaurants land between 28% and 33% of sales, but the band for your format is what matters. Published ranges by concept:

Two cautions. A fine-dining room at 38% isn't failing — high-touch service is the product, and the check average carries it. And a standalone bar at 20% versus a full-service kitchen at 32% tells you nothing except that pouring drinks needs fewer hands than cooking dinner. As with food cost, the percentage is a control metric, not a trophy — compare yourself to your concept and your own trend, not to the operator down the street running a different format.

Why you manage to prime cost, not labor alone

Labor cost read by itself lies, because it trades against food cost. Prep your sauces in-house and labor rises while food cost falls; buy them pre-made and the numbers swap. Optimize either one in isolation and you can quietly wreck the other.

That's why operators manage to prime cost — food plus labor — and the published target is roughly 60% of sales (a 55–65% band, with quick service lower and fine dining toward the top). Above 65%, there's usually too little left for rent, utilities, and profit. A kitchen running a tight food cost with an overstaffed floor still has a prime cost problem; you just can't see it until you add the two together. Track prime cost weekly, the same cadence I argued for with inventory counts, and the trade-offs become visible instead of theoretical.

The 2026 wage reality you're costing against

Labor benchmarks are getting harder to hit, and the macro picture is why. The federal minimum wage is still $7.25 — frozen since 2009, the longest stretch in the law's history — but almost nobody competes at the federal floor anymore:

• 22 states and D.C. raise their minimum at some point in 2026, pushing the average state minimum to around $11.51.
• For the first time, more workers live in $15-plus states than in $7.25 states — 17-plus states and D.C. now sit at or above $15.
• California's fast-food minimum is $20.00 under AB 1228, more than three dollars above the state's general minimum.

And the pressure isn't only the wage floor. The National Restaurant Association's 2026 outlook reports 42% of operators were unprofitable in 2025, with labor among the top-cited challenges, and a shrinking 16-to-24 population making experienced staff structurally harder to find. Rising wages plus a tighter labor pool means the same schedule costs more this year than last — before anyone works a minute differently.

The levers, ranked by the evidence

Plenty of things nudge labor cost; these are the ones with real support, strongest evidence first. I've flagged where the numbers are vendor-reported rather than independently studied.

1. Schedule to a sales forecast, not a hunch. The single biggest controllable. Building the schedule against forecasted demand — ideally your own POS sales history by daypart — is where idle labor gets cut. Industry estimates put the savings from predictive scheduling at roughly 6–10% of labor spend (vendor-reported, attributed to a Deloitte outlook); treat the exact figure as directional, but the mechanism is real: you stop scheduling four people for a two-person Tuesday.
2. Put guardrails on overtime. Overtime is labor at 1.5× with zero extra revenue. Real-time alerts that flag an employee approaching 40 hours reportedly cut overtime spend 20–30% (vendor-reported). Even discounting that, catching overtime before it's worked beats discovering it on the payroll run.
3. Reduce turnover — it's a labor cost in disguise. Cornell's foundational study put the cost to replace a frontline hospitality employee near $5,864 (almost certainly higher today), most of it lost productivity while a new hire comes up to speed. With restaurant turnover averaging around 75% — and quick service often north of 130% — you may be re-buying your entire staff every year. Schedule fairness, predictable hours, and decent onboarding are labor-cost levers even though they never show up on the labor line.
4. Let technology absorb the peaks. Self-order kiosks and tableside handhelds don't replace your team; they raise what each person can handle at the rush. Kiosk vendors report 10–14% higher average checks alongside the labor reallocation (vendor-reported — I'd weight the throughput more than the check lift). The point isn't fewer people; it's fewer people standing idle between peaks.

What your POS should be doing about labor

Here's the test: if your manager builds next week's schedule in a spreadsheet with no view of forecasted sales, you're flying the most volatile cost in the building blind. The data to fix it already exists — your POS knows exactly how sales move by hour and day. A reporting layer that pushes that forecast into scheduling, shows labor as a live percentage of sales during service (not after closeout), and flags overtime before it's worked turns labor from a number you reconcile into one you steer.

Disclosure: I work at Katalyst, and we build the integrated version of exactly that, so read the bias accordingly. But the discipline doesn't belong to any vendor: count the fully-loaded cost, benchmark to your own concept, manage food and labor together as prime cost, and schedule against real demand instead of yesterday's habit. Labor is the hardest of the two controllable costs precisely because it pushes back. That's also why the operators who run it tightly are the ones still profitable in a 42%-unprofitable year.

Related Katalyst products

• See Katalyst analytics