Labour cost percentage: the number that eats profit after food
How to calculate restaurant labour cost the right way, what ranges are normal, and the schedule changes that move the number without cutting service.
Food cost gets the meetings. Labour cost quietly takes the rest. On many sites it is the larger of the two, and the one managers feel least able to touch — because cutting people feels like cutting service.
You do not start by cutting people. You start by measuring the hours you are already paying for empty tables.
The definition
labour cost % = total people cost ÷ net sales × 100
Total people cost is not the wage line. It is:
- Base wages and salaries
- Overtime
- Employer’s insurance and tax
- Service-charge or tip share you treat as cost
- Agency / daily labour
- Manager salaries (or a fair share of them)
Leave any of those out and the percentage looks healthy while cash leaves through the side door.
A monthly example:
| Line | Amount |
|---|---|
| Wages and salaries | 210,000 |
| Overtime | 28,000 |
| Insurance and tax | 42,000 |
| Daily labour | 15,000 |
| Total people cost | 295,000 |
| Net sales | 1,000,000 |
| Labour cost % | 29.5% |
What “good” looks like
Approximate, and not a rule:
| Type | Common range |
|---|---|
| Quick service / counter | 22% – 28% |
| Casual dine-in | 28% – 35% |
| Full service, high touch | 32% – 40% |
A site at 38% with a full dining room and a waiting list is in better shape than one at 26% with three people standing still from 3 to 6 pm. Compare yourself to yourself, same weekday, week over week.
Split the number or you will fix the wrong team
A single labour percentage is a blended lie.
| Split | Why |
|---|---|
| Kitchen vs floor | Prep can be early; service cannot |
| Hour of day | 4 pm is not 8 pm |
| Salaried vs hourly | Overtime hides in one and not the other |
If the kitchen is at 18% of sales and the floor is at 16%, your problem is not “staff”. It is a dining room staffed for a rush that has not started.
Where the extra points usually live
- Coverage in the dead zone. Two extra people from 3 to 6 pm, six days a week, is a full-time salary for sales that do not exist.
- Overtime as a habit. If the same closer is on overtime four nights a week, the roster is short — not “busy”.
- Clock-in drift. Ten minutes early, every shift, every person, is a percentage point by month end.
- Managers on the wage line. If the owner works sixty hours and takes no salary, the percentage is fiction. Put a market wage in, even if you do not pay it out, so the business is honest.
- No sales-per-labour-hour. Headcount without output is a feeling.
net sales ÷ labour hourstells you whether the team is producing.
A four-week way to move it
- Week 1: Get a clean number. Pull attendance and payroll into one total people cost. Split kitchen and floor.
- Week 2: Overlay sales-by-hour. Mark every hour whose labour cost exceeds 40% of that hour’s sales.
- Week 3: Cut or shift those hours first — not Friday night. Open later, close a section, or move prep earlier.
- Week 4: Cap overtime: any hour beyond the roster needs a reason, the same day.
When clock-in, shifts and overtime already feed payroll — as they do in GoSufra — the percentage is a weekly number rather than a surprise after payday. You manage the roster against yesterday’s sales, not against last month’s argument.
Labour is not a cost you minimise. It is a cost you time. The restaurants that keep the number steady are the ones that stop paying for an empty room.
Run all of this from one system
POS, kitchen, inventory, recipe costing, staff and accounting — connected, and free to start.
Create your free account