How to Calculate Labor Cost Percentage (Beyond Just Wages)

When I first ran labor cost percentage for my own restaurant, I divided gross wages by sales and got 24%. Healthy by any benchmark. Six months later the bookkeeper slid the P&L across the desk and pointed at the labor line: 32%. I figured he’d fat-fingered something. He hadn’t. I’d been reading the wrong number for half a year.

The 24% was wages. The 32% was what it actually costs to employ people once you count everything. The industry calls that difference payroll burden, and most operators leave it out of their weekly tracking. That’s the eight points right there.

The formula (the version that matches your P&L) #

Labor Cost % = (Gross Wages + Salaries + Payroll Burden) ÷ Total Sales × 100

The headline version on the Toast and 7shifts blogs gives you wages over sales and stops there. Fine for staffing decisions. Worthless for understanding your P&L. The fully loaded version above is the one that shows up on your income statement, and it’s the one your accountant is looking at when he tells you margin’s thin.

What’s in payroll burden:

• FICA (Social Security + Medicare): 7.65% of wages, employer share
• FUTA (federal unemployment): 0.6% on first $7,000 per employee
• SUTA (state unemployment): 1-6% depending on state and your experience rating
• Workers comp insurance: 3-7% of payroll for restaurant industry (NCCI class codes 9082 / 9083)
• Health insurance contribution: 0-15% of wages depending on plan and your match
• Paid time off accrual: 2-5% if you offer PTO/sick
• 401(k) match: 0-3% if you offer

The all-in burden multiplier for a typical independent restaurant lands in the 25-35% range. It runs higher in California, New York, and Washington, where you’ve got mandatory paid sick on top of high SUTA and high workers comp. It runs lower in Texas, Florida, and Tennessee. A $20/hour line cook in California, with workers comp at 5%, SUTA at 4%, FICA at 7.65%, and a 4% health stipend, costs you about $20 × 1.27 = $25.40/hour fully loaded. Same cook in Texas, with lower SUTA and no mandatory paid sick: $20 × 1.18 = $23.60/hour.

So a “we’re paying $20/hour” decision is really a “we’re paying $24-27 of labor expense” decision. Build the burden multiplier into your hiring math or every wage decision you make is off by 20-35%, and you won’t even know it.

A worked example #

Casual full-service restaurant, one-week period.

Gross hourly wages: $11,400 Salaried management (weekly portion): $2,100 Subtotal wages: $13,500

Payroll burden (28% blended for this operation):

• FICA: $13,500 × 7.65% = $1,033
• SUTA + FUTA: $13,500 × 4.2% = $567
• Workers comp: $13,500 × 5% = $675
• Health stipend: $13,500 × 5% = $675
• PTO accrual: $13,500 × 2% = $270
• Total burden: $3,220 (23.9% effective)

Fully loaded labor: $13,500 + $3,220 = $16,720

Sales for the week: $82,000

Headline labor cost: $13,500 ÷ $82,000 = 16.5% (the misleading number)

Fully loaded labor cost: $16,720 ÷ $82,000 = 20.4% (the real number)

That 4-point gap is the whole story. It’s the difference between the number you use to build a schedule and the number that hits your P&L. Track only the headline and you’re understating your real labor by about 25%, every single week, until the monthly statement catches you.

Labor % vs SPLH: which one to trust #

Labor cost percentage and sales-per-labor-hour (SPLH) answer different questions, and you need both.

Labor cost % tells you the share of revenue going to labor. It’s your target number, the one you track against the P&L.

SPLH = total sales ÷ total labor hours. It tells you how much each scheduled hour actually produced, which is what you want when you’re building next week’s schedule.

Here’s why one number isn’t enough. A weekly labor % of 28% can hide a Saturday night that ran 38%, because your dead weekdays at 22% pull the average back down. The percentage smooths it out and hands you a clean-looking number while one shift is bleeding. SPLH catches that Saturday in the moment. The productivity number tanks the same week, in the same shift, before payroll even closes.

So track labor % weekly and SPLH by shift. The weekly number tells you whether the whole operation is sustainable. The shift number tells you exactly where to go fix it.

For the SPLH math in detail, see Sales Per Labor Hour: The Number That Tells You Who To Cut First.

The 30% benchmark across segments #

Industry-standard labor cost benchmark for full-service restaurants: 28-32% (NRA 2024). Some specific segment medians:

• QSR / fast food: 25-30% (high check velocity, low wage rates)
• Fast casual: 28-32%
• Casual full-service: 30-35%
• Family restaurant: 30-35%
• Fine dining: 32-38% (more BOH, skilled positions, higher service ratio)
• Steakhouse: 32-36%
• Coffee shop / café: 28-34%
• Pizza (delivery + takeout): 25-30% (lean BOH, high check velocity)
• Sports bar / pub: 28-34%
• Bar-only (no kitchen): 18-24%

The latest published NRA medians: full-service 36.5%, limited-service 31.7% (NRA 2024 Industry Factbook, based on 2024 operator survey data). Both are sitting at historical highs, and it’s post-pandemic wage compression doing it. If your numbers look high against the old rules of thumb, that’s why. The whole industry moved.

State-by-state wrinkles #

Labor cost varies more by state than by concept.

• California, New York, Washington, Oregon: Higher base wages, mandatory paid sick, no tip credit. Labor cost typically runs 4-8 points higher than national average.
• Texas, Florida, Georgia: Lower base wages, tip credit allowed, lower SUTA. Labor cost typically runs 2-4 points below national average.
• Tipped employees: Whether your state allows a tip credit ($2.13/hr base federal, up to $5.12/hr credit) changes labor cost dramatically. A server paid $7.25 in a state with no tip credit costs the operator twice what a server paid $2.13 in a tip-credit state.

This is why I don’t trust segment benchmarks on their own. A California fine-dining steakhouse running 38% labor might be lean as hell. A Texas QSR at 30% might be carrying two people it doesn’t need. Same two numbers, opposite stories, and the only thing that tells them apart is where the building sits.

Where the leak usually is #

Labor over-runs are almost never about wage rates. They’re about hours. Here are the two patterns I catch most.

1. Same-time arrivals. Everybody scheduled at 4pm for a 5:30 dinner rush. So the first 45 minutes is six people standing around doing the prep that should’ve already been done, on the clock, producing nothing. Stagger them, 4:00, 4:30, 5:00, and you pull back 2-4 hours of labor a night per shift. That’s free money you’re leaving on the schedule.

2. Slow shift over-staffing. Four servers on a Tuesday lunch when two would cruise through it. Operators schedule what feels safe instead of what the forecast actually supports, because nobody wants to be the manager who got slammed short-handed. I get it. But an 8-hour weekly over-run across your slow shifts is $200-300 of fully loaded labor that produced zero revenue, and it does that every week.

The fix is forecast-based scheduling. Pull your last 4 weeks of same-day-of-week sales. Set a labor target as a percentage. Then build the schedule against that forecast instead of against habit, and stop guessing.

What this looks like in the calculator #

The labor cost calculator on this site asks for gross wages, salaries, your burden multiplier (it’ll default to a sensible number for your state), and total sales. You get back both the headline and the fully loaded percentage, side by side, with the benchmark for your concept. Run it against prime cost when you want the food-plus-labor view in one shot.

What to do today #

Pull last week’s payroll. Add a 28% burden multiplier, or the right number for your state. Divide by sales. That’s your real labor cost. If it’s more than 2 points off from what you’ve been reporting, that gap isn’t a rounding error. It’s the cost category you’ve been quietly understating, and it’s the reason your monthly P&L keeps saying “net margin lower than I expected.”

This doesn’t need a software purchase. It needs you to multiply gross wages by 1.28 every week before you write the number down. That’s the whole job.

Sources: NRA 2024 Industry Factbook, Toast, 7shifts, Restaurant365, DOL Wage and Hour Division.