Restaurant Labor Cost Percentage: The Real Ranges, Not the Magic Number

What is a healthy labor cost percentage for a restaurant?

A healthy labor cost lives between 30% and 36.5% of sales, and the exact band is set by your service format, not by the owner's wishes.

The National Restaurant Association's Restaurant Operations Data Abstract 2025 put the full-service median at 36.5% of sales during 2024, and that is the number a dining room with servers, a host stand and a line kitchen should be measured against. If you run white tablecloth, a wine list and slow table turns, sitting three points above that median is no tragedy; sitting eight points above it is, because the check average rarely stretches far enough to absorb them. The decision coming out of this is concrete rather than abstract: with 36.5% as your yardstick, you already know whether your next hire fits or whether you first have to fix the productivity of the hours you are already paying for.

Format rules: why counter service and tablecloths don't compare

Comparing a counter-service café with a full-service dining room is the costliest cost-structure confusion repeated across the industry. That same National Restaurant Association source sets the limited-service median at 31.7% of sales: nearly five points below full service, and that gap does not mean counter operators are better managers. It means they sell a product that needs fewer front-of-house hours per dollar billed, with fewer steps between the register and the handoff. When an owner with table service becomes obsessed with reaching 31.7%, the usual outcome is a stripped floor, longer service times and a ruined check average, the very thing that justified the format. The right call starts by naming the format you actually operate, then demanding the benchmark that belongs to YOUR family of restaurants rather than your neighbor's.

The band that separates profitable operators from losing ones

There is an eight-point band that works like a waterline, and knowing it beats any industry average. For 2024 the National Restaurant Association reports labor cost at 34.2% of sales among profitable operators versus 42.9% among operators running at a loss: payroll is not one expense among many, it is the marker that predicts whether the year closes in black or red. That contrast carries an uncomfortable operational reading. When your payroll drifts toward the upper end of that band, the problem almost never sits in the hourly wage, it sits in how shifts are scheduled against the real sales curve by daypart. Pull the hourly sales report for the last six weeks, lay it over the hours you paid, and you will see exactly where you are buying presence that never sells anything.

Payroll and kitchen are judged together: prime cost

Judging labor cost without looking at food cost is like weighing yourself with one foot on the scale. Toast sets 60% of sales or less as the prime cost ceiling (food plus labor) for operating with room to breathe: above that line, profitability turns uphill even when every expense looks defensible on its own. Inside that ceiling sits a real trade many owners never use. Every point you cut in purchasing through menu engineering or waste control buys air to keep a stable team, and the reverse holds too, because a trained crew wastes less product. Under the Masterestaurant method the food cost ceiling per dish is 32%, and it works as a lever rather than dogma; the useful question this week is which of the two lines you can move faster with the crew you already have.

Profitable QSR and the 30% floor

Profitable quick-service operators mark today's realistic floor for the trade: 30.0% of sales in labor cost, according to 2024 National Restaurant Association data. That figure deserves a careful reading, because it describes a design model rather than an austerity model. A QSR that gets there does it with short menus, repeatable processes, kiosks or digital ordering that strip steps out of the register, and demand forecasting decent enough to avoid three people staring at each other at three in the afternoon. None of those elements belongs exclusively to a global chain. An independent restaurant can steal two or three of those ideas next week: cut the dishes that never turn, push orders to the digital channel, close the gap between the scheduled shift and the sales curve. Treat 30% as borrowed efficiency, not as a target for your tablecloth dining room.

How to read these numbers in YOUR operation: three scenarios?

Translate the benchmark to your size before touching a single shift, because the same figure calls for different decisions. Small scenario, a 40-seat spot where the owner cooks or works the floor:

your REPORTED labor cost lies downward if you pay yourself nothing, so impute your own market salary before comparing yourself to the full-service median. Mid scenario, two locations with 80 seats each and one manager per unit: the classic leak here is a middle layer duplicated per location instead of shared, and the decision is whether a single operations manager can cover both. Group scenario, four or more units with a central office: that office payroll has to be read separately from floor payroll, because blending them hides an inflated dining room and a corporate structure that grew ahead of sales at the same time.

Where these benchmarks come from and how far they go?

These numbers come from operator surveys rather than forensic audits, and it is worth saying so plainly.

The full-service and limited-service medians come from the National Restaurant Association's Restaurant Operations Data Abstract, which aggregates income statements submitted by U.S. operators; Level compiles those same NRA 2024 medians and calculates prime cost at roughly 68% of costs for full-service and 64% for limited-service. Two limits you should carry with you: first, the base is American, and across Latin America informality distorts the comparison, with a regional average rate of 47% per the ILO (2025) across all sectors. Second, every operator defines differently what belongs inside payroll. Before getting upset about your percentage, confirm whether it already includes benefits, house tips and the owner's salary.

The 2024 context and what to do on Monday

Pressure on payroll turned nearly universal, and that changes how the benchmark should be used: 99% of operators reported spending more on labor, according to TouchBistro (2024). When everyone rises, the average stops being a consolation and starts being a thermometer, because what matters is no longer that your payroll grew but whether it grew faster than your format's and whether the check absorbed it. At Masterestaurant, Diego F. Parra insists on a sequence that sorts out the mess: measure your real percentage with the owner's salary inside it, place that against the median for YOUR format, and only then choose between raising prices, redesigning shifts or trimming the menu. Start this week with the cheapest move of all: print sales by daypart for the last forty days and mark it against the hours you paid.

Four differences that change the outcome

The first difference is FORMAT, and it invalidates most dinner-table comparisons. Limited-service runs fewer front-of-house hours per dollar sold, which is why its labor median settles at 31,7 % of sales while full-service lands higher: one group is not more efficient than the other, they sell different things with different hands. Second, labor cost is never judged alone, it is judged paired with the kitchen. For full-service operators, food and labor together account for roughly 68 % of costs according to National Restaurant Association 2024 medians compiled by Level, and inside that pairing there is a genuine trade: every point you pull out of food cost buys breathing room for payroll, and the reverse holds too.

Four differences that change the outcome — in practice

Third, and this is the uncomfortable one that draws pushback in board meetings: the profitability line is NOT the industry average, it sits considerably higher. Losing operators closed 2024 with labor at 42,9 % of sales against 34,2 % for profitable ones (National Restaurant Association, 2025). Eight and a half points between breathing and drowning. Fourth is REACTION SPEED. Forecast-assisted scheduling delivers labor reductions in the 8-12 % band, a figure McKinsey documented in 2023 and TimeForge reports, and the mechanism is not magic: you stop staffing for a rush that never arrives. I got this wrong for years, recommending fixed rosters to protect service. Service did not break when we adjusted; it broke when we adjusted without a forecast.