Restaurant Labour Cost Percentage: The Number Behind the Roster
Labour cost is the second-biggest line in the kitchen, and the only one you decide in advance. You set it on the roster on Sunday — not on the P&L on the 30th. Most kitchens find out what their labour cost was. The good ones choose it.
Ask a chef their food cost and you'll usually get a number. Ask their labour cost and you'll get a shrug, or a figure that's missing half of what an hour actually costs. That's strange, because labour is the line you have the most control over — you write it down every week before it happens. Food cost reacts to supplier prices you don't set. Labour cost is a decision you make with a roster grid in front of you.
This piece is about treating that grid as the financial document it is: what labour cost percentage really includes, what target is sane for your model, why prime cost matters more than either number alone, and how to roster to demand instead of to habit.
The formula, and the trap inside it
If you spent $9,000 on labour in a week that turned over $30,000, your labour cost is 30%. Simple — and almost always understated, because "total labour cost" is not what you paid in wages. It's what the staff actually cost you, which is meaningfully more.
The wage on the payslip is the visible part. Underneath it sit the on-costs: the legally required and structurally unavoidable additions that turn a $30/hour cook into something closer to $38–$42 of real cost. Miss these and your labour cost looks healthy on paper while the bank account quietly disagrees.
What an hour actually costs
For a kitchen in Australia, the load on top of base wages typically includes:
- Penalty rates — evenings, weekends, and public holidays carry loadings under the relevant award. A Sunday hour can cost 1.5–1.75× a Tuesday hour for the same person. Rostering a senior chef across a long weekend is a different financial decision than rostering them midweek.
- Superannuation — currently 11.5% on top of wages, rising on schedule.
- Annual leave + leave loading, personal/sick leave — accrued every hour worked, paid later.
- Payroll tax, workers' compensation insurance — a few percent more depending on state and payroll size.
Stack those and the real cost of an hour lands roughly 1.25× to 1.4× the base rate. The exact multiplier is yours to calculate, but the principle is fixed: if you cost your roster at bare wages, you are underestimating labour by a quarter or more. The legally correct rates — base, casual loading, penalty multipliers — are published by the Fair Work Ombudsman, and they are the floor, not a suggestion.
What target is actually sane
There's no universal "good" labour cost. It depends entirely on your service model — how much skilled handwork each plate needs and how much revenue each labour hour generates.
| Model | Typical kitchen + FOH labour % |
|---|---|
| Quick service / counter | 22–28% |
| Casual full-service | 28–32% |
| Fine dining (high handwork) | 32–40% |
| Bar-led / wet-heavy venue | 20–26% |
Fine dining runs a high labour cost on purpose — the product is labour. A tasting menu with twelve garnished courses can't be cheap to staff, and that's fine if the price reflects it. The mistake is judging a fine-dining kitchen against a bistro number, or letting a bistro drift into fine-dining labour without fine-dining prices.
The number that actually predicts survival: prime cost
Food cost and labour cost are each useful, but on their own each is gameable. Cut staff and your labour cost falls — while service collapses and food waste climbs. Use cheaper ingredients and food cost drops — while the dish needs more labour to make edible. The two trade against each other, which is why the number experienced operators watch is the sum.
Prime cost is the combined weight of your two largest, most controllable lines. As a share of revenue, most full-service venues need prime cost at or below roughly 60–65% to leave room for rent, utilities, and a profit. Above that, the model is fighting itself — every other line has to be perfect just to break even.
The value of watching prime cost is that it stops the seesaw. A roster cut that "saves" three labour points but adds four points of waste and comps is a loss, and only prime cost shows it. Manage the pair, not each in isolation.
Roster to demand, not to habit
Here's where the percentage is actually won or lost. Most rosters are copied from last week, which was copied from the week before. They reflect the kitchen's habits, not the week's demand. The result is a line cook scrubbing benches at 3pm on a dead Tuesday and a section drowning at 8pm on Friday.
Rostering to demand means three boring disciplines:
- Know your covers by daypart. Pull last year's same-week numbers and your booking sheet. Tuesday lunch and Saturday dinner are different businesses; they should not have the same brigade.
- Match the brigade to the curve. Stagger starts. Not everyone needs to be on at noon if the rush is at one. A 30-minute staggered start across five people is two and a half labour hours saved a day, every day.
- Cost the roster before you publish it. The labour cost of the week should be a number you read off the draft, not a surprise you discover on payroll. If Saturday's wage bill is too high, you move a shift now, not next month.
In ProChefDesk
The Roster & Labour tool totals the wage bill as you type each shift — set an hourly rate per person and the labour cost lands live, before you commit. Leave codes (OFF, annual, sick, RDO, public holiday) and department grouping keep a 12-person brigade readable; you can hide the cost column when you print the staff copy and show it on the boss copy.
Two quiet traps
The salaried-chef illusion
A salaried head chef feels "free" once the roster is full — no hourly meter ticking. That's the trap. A salaried 38-hour chef working 58 hours has a real hourly cost a third lower than their contract implies, which feels efficient — right up until they burn out, leave, and take a year of menu knowledge with them. Cheap labour that you're quietly buying on credit against someone's health is not cheap. Count salaried hours honestly; if a salary only works at 55+ hours a week, the role is underpriced, not efficient.
Productivity beats headcount
Labour cost percentage can mislead if you only ever cut. The better lens is output per labour hour — revenue per labour hour and covers per labour hour. A kitchen doing $45 of revenue per labour hour at 30% labour is healthier than one doing $30 per labour hour at 28%. Sometimes the right move is more skilled staff producing more covers, not fewer staff producing fewer. The percentage alone won't tell you that; the productivity ratio will.
What to do this week
- Calculate one true hourly cost per person — base rate × your on-cost multiplier. Write it next to each name.
- Cost next week's draft roster with those numbers, by day. Find your single most expensive shift.
- Pull covers for that daypart from last year + bookings. Is the brigade matched to the curve, or to habit?
- Move or stagger one shift. Just one. A single 30-minute staggered start, repeated daily, is real money over a quarter.
- Track prime cost, not just labour. Add food + beverage + labour as a share of revenue and watch the pair, so a labour cut that creates waste shows up honestly.
You're not trying to grind labour to the bone — a starved kitchen loses more in service and turnover than it saves in wages. You're trying to make the number a decision instead of a discovery.
The point
Labour cost percentage isn't a payroll report. It's the financial readout of a roster you wrote on purpose. Cost an hour honestly, match the brigade to real demand, watch it next to food cost as prime cost, and the percentage stops being a monthly shock and becomes a weekly lever.
"We weren't overstaffed. We were staffed for a Friday, seven days a week."
That's the most expensive habit in the business, and it hides in plain sight on a roster nobody costs before they publish it. Cost it first. The number is yours to choose.