How to Reduce Labor Cost in a Restaurant (Without Cutting Service)

The labor hours you can cut without guests noticing are the ones above what each part of the day actually needs: the extra people on a slow afternoon, the overtime that came from an uneven schedule, the replacements you hire because people keep leaving. The hours you protect are the ones guests feel, like the floor at 7 p.m. on a Friday. To find the difference, schedule to a forecast of demand by daypart (a block of the service day, such as lunch or dinner), set a minimum crew for each one, and count your savings in the dollars that actually stop going out.

Labor is half of prime cost, the number at the center of the guide to reducing restaurant costs without hurting service. It's also the line whose share of sales has climbed. In the National Restaurant Association's 2025 operations data, salaries and wages including benefits were a median 36.5% of sales at full-service restaurants in 2024 and 31.7% at limited-service restaurants, against roughly 33% and 28% on average in the association's 2010, 2013 and 2016 reports.

How to calculate labor cost percentage

Formula: Labor cost percentage = total labor cost ÷ total sales × 100, for the same period. Total labor cost = hourly wages (including overtime) + salaried pay + employer payroll taxes + benefits + workers' comp.

The most common mistake is counting only the hourly payroll. Restaurant365 puts it plainly: labor cost "includes salaries, payroll taxes, employee benefits, overtime pay, and any other costs tied to your workforce." The employer share of payroll tax alone is 6.2% of wages for Social Security and 1.45% for Medicare, per the IRS. Across leisure and hospitality, BLS compensation data for June 2026 put benefits at $4.08 an hour on top of $17.25 in wages and salaries, or about 24 cents per wage dollar.

Worked example: Say a restaurant does $30,000 in sales in a week. Hourly wages are $6,900, including $300 of overtime. Manager salaries for the week are $1,800. Payroll taxes, workers' comp and benefits add $1,740. Total labor is $6,900 + $1,800 + $1,740 = $10,440, or 34.8% of sales. Counting hourly wages alone would show 23.0%, a number that looks great and means nothing.

Enter your own week below. The calculator shows labor as a share of sales next to food cost and the combined prime cost.

What is a good labor cost percentage?

Published targets and what operators actually report are two different things. The table puts them side by side.

Two lessons. The median restaurant runs above the targets you read online, so don't panic if you do too; work on your trend. And labor separates winners from losers: in Citrin Cooperman's report, top performers, those with EBITDA of 17.5% of revenue or more, ran labor 9.48 points lower than low performers, those under 8.5%. As an example, on a restaurant doing $2 million a year that gap is $189,600.

1. Schedule to forecast demand, by daypart

A schedule copied from last month staffs for last month. A forecast staffs for next Tuesday. Here's the process:

1. Pull sales by hour from your POS for the same weekday over the last four to six weeks. Note holidays, events and weather that skewed any of them.
2. Group the hours into dayparts (for example lunch, afternoon, dinner and late) and forecast sales or covers for each.
3. Set a labor target per daypart as sales per labor hour (SPLH): sales divided by the hours your staff worked. Take it from your own best-run weeks, not from a blog.
4. Set a service floor for each daypart: the smallest crew that can open safely, keep ticket times and greet every guest. Never schedule below it.
5. Convert the forecast to hours: forecast sales ÷ SPLH target, but never below the floor.
6. Stagger start and end times so people arrive as the rush builds and leave as it fades, instead of everyone starting at 11 and leaving at close.
7. Compare forecast with actual every week and adjust the next schedule.

Worked example: Say a restaurant forecasts $4,000 in sales for a Tuesday and sets an example target of $50 in sales per labor hour. The table shows the hours each daypart needs against its current schedule.

In this example, lunch and dinner don't change at all. The savings come from the afternoon and late dayparts, where the current schedule carries two extra people: 9 hours a day.

What those 9 hours are worth depends on what actually stops being paid. BLS data put the average employer cost of an hour worked in leisure and hospitality at $21.33 in June 2026, of which $4.08 was benefits. That average is useful for budgeting, but it isn't the cash you save by taking one hour off a schedule. You save the wage of the person whose shift got shorter, plus the payroll taxes and workers' comp that rise with wages. Benefits such as health insurance usually don't fall because a shift ended two hours early, and if the hours belong to a salaried manager, nothing changes at all.

Worked example: Say the trimmed afternoon and late hours are worked by hourly staff earning $16 an hour, and payroll taxes and workers' comp add about 10% to that. Each hour cut saves about $17.60, so 9 hours save about $158 a day. If the same pattern holds on four slow weekdays, that's about $32,900 over 52 weeks, without touching a peak shift. Using the $21.33 loaded average instead would show about $39,930, which overstates the cash saving in this example. Use your own wage rates and tax costs.

2. Stop overtime before it happens

The federal rule is simple and expensive. Under the Fair Labor Standards Act, covered employees "must receive overtime pay for hours worked in excess of 40 in a workweek at a rate not less than time and one-half their regular rates of pay," and the Department of Labor adds that "averaging of hours over two or more weeks is not permitted." A short week doesn't cancel a long one. Some states and cities add stricter rules, such as daily overtime, so check your state labor department too.

Worked example: Say a cook earning $20 an hour works 46 hours. The six overtime hours cost $30 each, or $180, against $120 at straight time. Those six hours still need working, so the saving from moving them to someone under 40 is the premium: $60 a week for one person, or $3,120 a year if it becomes a habit, plus the payroll taxes on it.

Most overtime is a scheduling problem, not a busy-week problem. To catch it early:

• Show each person's scheduled hours on the schedule itself, and cap scheduled hours a little under 40 to leave room for a shift that runs long.
• Check hours worked mid-week, on Wednesday or Thursday, and move remaining shifts to people under 40.
• Tie your timeclock to the schedule so early clock-ins and late clock-outs need a manager's approval.
• Look for the one person who can't be replaced on a shift. That's a cross-training gap (lever 3), and it's where overtime keeps landing.

3. Cross-train so slow shifts can run lean

The service floor in lever 1 is easier to hold with fewer people when each of them can do two jobs. A host who can run food and bus tables, a server who can work the bar, a prep cook who can work the line and a dishwasher who can prep let you staff a slow afternoon with three people instead of five, and cover a call-out without paying someone overtime. Pay a small premium for each certified skill. It costs less than the overtime and the extra bodies it replaces, and it gives good people a reason to stay.

4. Cut turnover, the labor cost that isn't on the schedule

Every person who quits takes their training with them. Restaurants and hotels lose people at nearly twice the rate of the private sector as a whole: in 2025, BLS turnover data show an average monthly quits rate of 4.2% in accommodation and food services, against 2.2% for all private industry.

Each replacement has a price. Black Box Intelligence, drawing on payroll and HR data from large restaurant brands, puts the hard cost of replacing an hourly employee at about $2,706 (separation, replacement and training), a non-GM manager at about $11,940 and a general manager at about $17,651. It reports hourly turnover of 110% in limited service (the rolling 12 months to the third quarter of 2025) and about 92% in full service.

Worked example: Say a restaurant has 30 hourly employees and turns over 100% of them in a year: 30 replacements at $2,706 each is $81,180. Cut turnover to 70% and it's 21 replacements, or $56,826, a difference of $24,354. Part of Black Box's estimate is time your managers and trainers spend rather than checks you write, so treat that as an estimate of cost avoided, not all of it cash, before counting the slower service of a crew that's always new.

What keeps people, cheapest fixes first:

1. A real first two weeks: a written onboarding plan, a named trainer and check-ins on day 7 and day 30.
2. Predictable schedules, posted early, with enough hours for the people who want them.
3. Managers who manage people, not only the pass. Train them for it.
4. A path up: trainer, shift lead, kitchen lead, with pay that rises at each step.
5. Pay at the market rate. Below-market wages are the most expensive way to save on labor.

5. Publish schedules early and change them carefully

A schedule posted two weeks ahead lets staff plan their lives and lets you see overtime and gaps before they happen. In some places it's also the law. In New York City, for example, fast food employers must give workers their schedules 14 days in advance and pay premiums for schedule changes and "clopenings." Before you build a habit of sending people home early or calling them in at the last minute, check the scheduling rules where you operate. A better forecast is the cheaper fix.

6. Use technology that removes hours, not hospitality

Good technology takes hours out of work no guest sees:

• Scheduling tools that build shifts from a sales forecast and warn you about overtime before you publish.
• A timeclock tied to the schedule, so early clock-ins need approval.
• Online ordering and kiosks that take orders without a person keying them in.
• Kitchen display systems that pace tickets so the line doesn't need an extra body to expedite.
• Invoice and inventory automation that gives managers back admin hours.

Be careful with tools that replace the parts guests come for. A tablet that takes the order at a full-service table can cost more in check size and repeat visits than it saves in labor.

Kutlerri.ai's Prime Cost Optimization Agent, currently in beta, works on this lever: it tracks labor cost alongside food cost and matches labor to demand.

7. Claim the FICA tip credit if you qualify

This one cuts labor cost without touching a single shift. According to the IRS, "certain food and beverage establishments use this form to claim a credit for social security and Medicare taxes paid or incurred by the employer on certain employees' tips." If your staff earn tips, ask your accountant whether you qualify and whether you've been claiming Form 8846 every year.

What not to cut

Every lever above protects the guest-facing minimums. Cuts that break them raise ticket times, comps and bad reviews, and the lost sales can push labor as a percentage of sales up even as labor dollars fall.

Labor is one half of the picture. The guide to what prime cost is and how to lower it shows how labor and food cost add up, and the guide to reducing food cost in a restaurant covers the other half. To check your numbers each week, use the prime cost calculator.