What should labor cost be in a restaurant? The formula and a worked example

There is no official benchmark for what labor should cost a restaurant, and most percentages quoted online don’t say where they come from. The useful number is your own: total labor cost, including wages, overtime, payroll taxes and benefits, divided by sales for the same period. Track it every week, put sales per labor hour next to it, and compare with your own past. This post gives the formula, a worked week with every assumption stated, and the ways to bring the number down without making service slower.

Is there an industry standard for restaurant labor cost percentage?

Short answer: not one we could verify. We looked, and we won’t quote a range we can’t trace to a named source.

Search “what should labor cost be for a restaurant” and you get confident percentages, mostly from software vendors’ blogs, mostly without a survey, a sample or a date behind them. A benchmark without a source can’t tell you whether it describes a fine-dining room in Manhattan, a café in a small town or a fast-food chain. Using it to judge your own restaurant is guesswork with a decimal point.

What the sourced data does say is that margins are thin and labor is hard to keep. In the National Restaurant Association’s 2026 report, “42 percent of operators reported their restaurant was not profitable” the year before (NRA, Feb 2026). The US median wage for waiters and waitresses was $16.94 an hour in May 2025, tips included (BLS). And accommodation and food services had a quits rate of 4.2% a month in 2025, about twice the 2.2% across all private industry (BLS JOLTS, Table 22).

So the honest benchmark is your own trend. If your labor percentage rose three points since spring and sales didn’t, that tells you more than any number on a vendor’s blog.

How do you calculate labor cost percentage?

Short answer: total labor cost for a period, divided by total sales for the same period, times 100.

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

The formula is easy. Counting the same things every week is the hard part. Decide once what goes in, write it down and stick to it.

Use sales before tax and before tips, for exactly the same days as the labor.

What is sales per labor hour, and why use it too?

Short answer: sales divided by hours worked. It shows which shifts carry too many or too few people, which a weekly percentage hides.

Sales per labor hour = sales ÷ hours worked

A weekly labor percentage blends a quiet Tuesday lunch with a packed Saturday night. Sales per labor hour, worked out shift by shift, pulls them apart. Count hourly staff only, so salaried hours don’t blur it, and compare the same shift across weeks: this Tuesday lunch against the last four Tuesday lunches.

A worked example: one week at Casa Olivo

Short answer: in this example, labor is $10,427.20 against $32,000 of sales, or 32.6%.

Casa Olivo is a fictional US restaurant (an example, not a customer). Every pay rate, hour count and sales figure below is an assumption chosen to show the arithmetic, not a benchmark.

The assumptions.

• Sales for the week: $32,000.
• Servers: 180 hours at $10 an hour paid by the restaurant, before tips from guests.
• Hosts and bussers: 60 hours at $15 an hour.
• Cooks: 240 hours at $19 an hour.
• Dishwashers: 60 hours at $15 an hour.
• Salaried manager: $1,150 a week.
• Payroll taxes and benefits: 12% on top of all pay.

$10,427.20 ÷ $32,000 = 32.6%. Front of house (servers, hosts and bussers) is 9.5% of sales; back of house is 19.1%; the manager is 4.0%.

Sales per labor hour for the week: $32,000 ÷ 540 hourly hours = $59.26.

Now two days from the same week:

Tuesday’s people are each carrying much less in sales than Saturday’s. That doesn’t automatically mean Tuesday is overstaffed: a restaurant needs a minimum crew to open at all. But it is where to look first, and it is invisible in the 32.6%.

What moves the number?

Short answer: sales move it more than most owners expect, because most of the labor stays the same when sales dip.

Two lessons. A quiet week raises the percentage by more than three points without anyone working an extra minute. And a fair raise for the whole team costs about the same, in percentage terms, as that quiet week. If you’re judging a manager on labor percentage, judge them on hours scheduled against hours needed, not on a number that sales can swing on their own.

What is prime cost?

Short answer: food and drink cost plus total labor cost. It puts your two biggest controllable costs in one number.

Prime cost = cost of goods sold + total labor cost

Prime cost % = prime cost ÷ sales × 100

It is useful because the two trade off: buying prepared ingredients raises food cost and lowers kitchen hours, and making everything from scratch does the opposite. Watching both together stops you from “saving” labor by moving it into the food line. As with labor, we found no sourced benchmark we would quote; track your own.

How do you reduce labor cost without hurting service?

Short answer: cut hours that don’t match demand and tasks that don’t need a person, not the people your busiest hour depends on.

1. Roster to the hours when orders actually arrive. Pull sales by hour from your till for the last four weeks of each weekday, and build the rota around them, not around last year’s rota. To set how many servers each shift needs, work out your section size per server.
2. Keep the people you’ve trained. Every departure costs hiring and training time; in our server turnover worksheet it comes to about $1,070 per departure under stated assumptions.
3. Cross-train for the shoulders. A host who can run food, or a bartender who can take a section, lets you open quiet shifts with one fewer person.
4. Remove tasks, not people. Walking to the till, re-explaining the menu and fetching the check are minutes that add nothing for the guest. Our list of jobs to take off servers’ plates goes through nine of them.
5. Watch overtime weekly. It is often the cheapest hour to cut, and the first sign that the rota doesn’t match the week.
6. Don’t cut below the busiest hour’s needs. Fewer hours on a busy night shows up as slow service, lost covers and tired staff who then leave. That is a cost too; it just lands on a different line. Our 9-point diagnosis of slow service shows how to find where a short rota hurts.

Where do server minutes go?

Short answer: a lot of them go to taking orders and handling payment, according to one small study.

A Cornell report timed servers in one restaurant. Server time per table fell from 10 minutes 16 seconds to 6 minutes 30 seconds when guests paid on a tabletop device, and to 3 minutes 30 seconds when they ordered and paid on it (Susskind and Curry, Cornell, 2016, p. 6). It was one restaurant with small samples (63, 33 and 7 tables), and the co-author worked for the device vendor, so read it as a hint, not a rule.

Freed minutes are not saved wages unless you change the rota, and most owners shouldn’t on a busy night: those minutes are better spent on guests. Where they can matter is the quiet shift, or the night someone calls in sick.

Where does WaiterPilot fit?

Short answer: we don’t claim it cuts labor cost. It moves some tasks to the guest’s phone and adds one of its own.

What it does: guests order and ask menu questions by chat from their phone, and call a waiter or ask for the check from the table. Staff work from a main board, a kitchen screen and a floor screen on phones or tablets they already have. The owner page shows busy hours, which can feed point 1 of the list above.

What it adds: WaiterPilot has no POS connection, so if you use a till, someone rings orders into it by hand. Staff also confirm each table’s first order. Count those minutes honestly if you’re weighing it up.

It costs $200 per location per month after 30 days free, with 0% commission. To compare that with your own numbers, the cost calculator works through flat fees against commission. For the full price and what’s included, see pricing.

What to do next

1. Write down what goes into your labor cost, using the table above, and keep it the same every week.
2. Work out last week’s labor percentage and sales per labor hour.
3. Do the same for each shift of the last four weeks, and circle the three lowest sales-per-hour shifts.
4. Change one thing on one of those shifts, and compare the same shift next week.