Restaurant Labour Cost: How to Keep Staff Wages Under 25% of Sales
Lunch is so packed the kitchen can't keep up, so you hire one more person. The next month the shop really does run more smoothly — but at month end there's less money left than before, because at three in the afternoon you have four people standing around waiting on two tables. Labour is a restaurant's second-biggest expense after ingredients, and the one that swells most easily without anyone noticing, because it costs the same every month no matter what you sold that day.
This article walks through how to count restaurant labour cost down to the last baht, turn it into a percentage of sales, see which hours are overstaffed and which are short, and use a simple formula to decide whether it's time to hire — without squeezing wages or cutting so many people that service falls apart.
Why look at labour as a percentage of sales
Is ฿60,000 a month in wages a lot or a little? You can't say until you know the sales. If the shop takes ฿300,000 a month, that's 20% and very comfortable. If it takes ฿180,000, it's 33% and there's barely any profit left. The number that tells you how healthy the shop is isn't the amount — it's the ratio:
For a small restaurant, the range most owners work to is roughly 20–30%, with 25% as a middle target. It's a guide, not a rule. Shops where customers serve themselves or turnover is fast — khao gaeng (rice and curry), noodle shops — tend to sit toward the low end. Sit-down restaurants with table service and dishes cooked to order in several steps sit toward the high end.
Always look at labour together with ingredient cost. Restaurant people call the two combined prime cost:
The reason is that the two trade off against each other. A shop that buys pre-cut vegetables or pre-marinated meat pays more for ingredients but needs fewer hands; a shop that makes everything from scratch is the reverse. If your food cost is 35%, labour should be around 25–30%, leaving 35–40% to cover rent, utilities, other expenses and profit. If you've never worked out your food cost, start with How to calculate food cost — the formula.
Step 1: Count all of your labour, not just salaries
The most common mistake is taking only the full-time salary total and concluding that labour is still low. A restaurant's real labour cost is more than that:
- Salaries and daily wages — for every regular employee.
- Part-time pay — including the occasional helper you pay in cash.
- Overtime (OT) — and pay for working on days off.
- The employer's social security contribution — 5% of wages (there is a cap on the wage base; check the current rate with the Social Security Office).
- Staff meals and benefits — food, travel allowance, uniforms, accommodation, if the shop pays.
- Bonuses and festival-season extras — averaged out per month.
- The owner's salary — if you're at the wok or minding the front all day, you're doing the job of one employee.
That last one is skipped most often. A shop that doesn't count the owner's labour sees a beautifully low percentage — but the day you're sick, want a break or open a second branch, you'll have to pay someone to replace you, and the real number shows up immediately. More on this in 12 hidden restaurant costs that quietly eat your profit.
Example: salaries of 16%, real labour cost of 28%
A one-unit made-to-order (ahan tam sang) shop takes ฿240,000 a month, opens 26 days and averages about ฿9,200 a day. It has 3 full-time staff, 1 evening part-timer and an owner who works in the shop (made-up numbers for illustration):
| Item | How it's worked out | ฿ / month | % of sales |
|---|---|---|---|
| 3 full-time salaries | 15,000 + 12,000 + 11,000 | 38,000 | 15.8% |
| Evening part-timer | ฿50 × 4 hrs × 26 days | 5,200 | 2.2% |
| Overtime (OT) | Busy days and days off | 2,800 | 1.2% |
| Employer's social security | 5% × 43,200 | 2,160 | 0.9% |
| Staff meals | 4 people × ฿40 × 26 days | 4,160 | 1.7% |
| Total staff labour | 52,320 | 21.8% | |
| Owner's salary | What a replacement would cost | 15,000 | 6.3% |
| Total labour cost | 67,320 | 28% |
Looking only at the first line, the owner would think labour is 16% and there's plenty of room to hire. Counted in full, labour is really 28% — ฿7,320 a month over the 25% target (฿60,000), or nearly ฿88,000 a year.
If this shop's food cost is 35%, or ฿84,000, prime cost comes to ฿151,320, or 63% of sales. That's still within an acceptable range, but there's no room left for error. A month when pork prices rise or sales drop 10%, and things get tight straight away.
There are two ways to bring labour back to 25%: cut ฿7,320 of labour, or add about ฿29,000 a month in sales (roughly ฿1,100 a day) with the same team. Most shops get there with a mix of both — and the starting point is knowing which hours of the day your people and your workload don't match.
Step 2: Measure sales per labour hour
Monthly labour % tells you whether there's a problem, but not where it is. The number that points to it is sales per labour hour:
Split an ordinary weekday at the example shop into periods, counting everyone working including the owner (made-up numbers, rounded for easy maths, breaks not deducted):
| Period | Sales | People working | Labour hours | Sales per labour hour |
|---|---|---|---|---|
| 10:00–11:00 (opening prep) | 300 | 4 | 4 | ฿75 |
| 11:00–14:00 (lunch) | 4,800 | 4 | 12 | ฿400 |
| 14:00–16:00 (afternoon) | 600 | 4 | 8 | ฿75 |
| 16:00–20:00 (evening) | 3,500 | 5 | 20 | ฿175 |
| Whole day | 9,200 | 44 | ฿209 |
This shop pays ฿67,320 a month in total labour for about 1,144 labour hours (44 labour hours × 26 days) — roughly ฿59 an hour. To hold labour at 25%, it needs to sell at least this much:
Across the whole day the shop manages ฿209, a little under the ฿236 line — which matches its 28% labour cost exactly. The interesting part is the breakdown. Lunch hits ฿400, which means the team is stretched: a few more customers and orders start running late and sales get lost. The two afternoon hours, meanwhile, use 8 labour hours to sell just ฿600. That's where labour money quietly leaks out, every single day.
You don't need an expensive POS system to build this table. Write down sales every hour on a sheet of paper for 7 days, or look at the timestamps on transfer slips and order tickets, and you'll see the shape of your day.
Step 3: Staff to your sales, not to habit
Once you can see which hours are overstaffed and which are short, here are 5 things a small shop can actually do without cutting anyone's salary:
- Stagger start times — not everyone has to arrive together at opening and leave together at closing. Have one person start at 10:00 to prep and another at 11:00 when customers begin to arrive; whoever starts first goes home first. Hours per person stay the same, but they pile up where the work is, and closing-time OT comes down.
- Use part-timers for the peak only — if the shop is truly busy just 3–4 hours a day, a part-timer at ฿50 an hour for those hours costs less than half of a full-time, all-day hire.
- Turn quiet hours into prep hours — monthly-paid staff cost the same whether or not there are customers in the afternoon. Use that time to chop vegetables, marinate meat, portion dipping sauces, wash up and reset the shop. The peak goes faster, you serve more customers with the same team, and nobody has to stay late cleaning up on OT.
- Remove work that doesn't need a person — let customers pour their own water and pick up their own cutlery, drop dishes that sell a few plates a day but need their own prep, and let customers scan and order from the table themselves. Front-of-house staff get more time to serve and clear tables. See how in How to make a QR code menu for your restaurant.
- Train everyone to cover more than one station — a server who can plate dishes or fry an egg, and a kitchen helper who can come out and take orders, let the shop handle the rush and days when someone's off without carrying a spare person.
Step 4: Control OT before it controls you
OT is the most expensive labour in the shop. Thai labour protection law requires overtime on a normal working day to be paid at no less than 1.5 times the hourly wage. For monthly-paid staff, the maths is:
An employee on ฿12,000 a month working 8 hours a day earns ฿50 an hour, so OT is ฿75 an hour. Staying 2 hours late to close up every day for 26 days is 52 hours, or ฿3,900 a month per person — nearly a third of their salary. If three people do that, the shop is paying almost ฿12,000 a month in OT, more than the wages of two part-timers.
Rules that help: OT has to be approved in advance every time, log OT hours per person every day, and if the same time slot produces OT again and again, treat it as a sign the roster needs fixing — not as a normal cost you just keep paying.
The legal basics every owner in Thailand should know: normal working time is capped at 8 hours a day and 48 hours a week; staff must get at least 1 day off a week; pay can't be below your province's minimum wage, which is revised from time to time; and once you have even 1 employee you must register for social security. Check current rates and details with the Department of Labour Protection and Welfare and the Social Security Office. For contributions and the related taxes, read Restaurant taxes in Thailand.
Step 5: Before you hire, ask how much in sales the new person has to bring in
"We're busy, we need another person" is a decision made on feeling. Turn it into a number:
A new server on ฿11,000 a month, plus ฿550 social security, plus a ฿40 meal each day (฿1,040), costs ฿12,590 a month in full. Divide by 0.25 and you get about ฿50,400 — the shop needs roughly ฿1,940 more in sales every day for labour to stay at 25%.
Compare that with a 4-hour peak part-timer: about ฿5,460 a month including social security, which needs only around ฿840 a day in extra sales. If the shop's problem is "we can't keep up at lunch" rather than "we can't keep up all day", the part-timer is usually the better-value answer.
Before deciding, ask yourself three questions. Is the shop actually losing sales because it's short of hands — customers walking out, cancelling after a long wait, delivery apps switched off during the rush? Is it busy every day, or only on certain days? And after staggering start times and prepping ahead, is it still not enough? If the answer to all three is yes, hiring is an investment, not a cost.
Labour that's too low is a problem too
The goal isn't the lowest possible labour cost. A shop running below 15–18% is usually paying in some other way: the owner works 14-hour days without counting their own pay, customers wait so long they don't come back, and staff burn out and quit. Every time someone leaves, the shop spends weeks finding a replacement, training them and absorbing their mistakes — a cost that never shows up in the books.
Staff who know the job, stay for years and are paid fairly usually produce more sales per hour than a cheaper new hire. Good labour control means putting working hours where the work is — not paying as little as you can.
Common labour cost mistakes
- Counting salaries only — forgetting OT, social security, staff meals and helpers paid in cash from the drawer.
- Not paying yourself a salary — so the "profit" you see has the owner's wages mixed into it.
- Hiring because of one peak day — busy on Saturday, but paying the extra wages all seven days.
- Running the same roster every day — even though Monday and Friday sales can differ by nearly double.
- Treating OT as normal — without ever adding up what it costs in a month.
- Looking at labour separately from ingredients — cutting kitchen staff and switching to pricier ready-prepped ingredients may not lower prime cost at all.
If your shop is busy but there's nothing left over, labour is usually one of several causes happening at once. See the others in Busy but not profitable: the 7 most common causes.
Record all your labour and the numbers speak for themselves
Every formula in this article needs just two things: sales, and labour recorded in full. The trouble is that labour gets paid in bits and pieces — salaries by transfer at month end, part-timers in cash by the day, OT sometimes folded into salary and sometimes not — so when it's time to add it up, nobody wants to.
Bunchee makes this easier. Set up a "Labour" expense category and log every payment, whether it's a salary, a daily wage or OT. Set a budget for the labour category alone and let the app alert you as you approach the limit, see labour's share against your other expense categories in the charts, and ask the AI assistant "what percentage of this month's income went on labour?" without touching a calculator.
Restaurant labour cost isn't a number to be afraid of — it's a number you need to see. Count it in full, work it out as a percentage of sales, look at sales per labour hour and put people where the work is. Most shops find they don't need to let anyone go at all: move the working hours to the right place and labour drifts back toward 25% on its own.
Know your labour % without adding it up yourself
Bunchee is a free restaurant accounting app — income and expenses by category, budgets and alerts per category, AI receipt scanning and real-time daily/monthly summaries.