Most published restaurant margin figures are American. They assume a tipping culture that subsidises wages, different rent structures and different tax treatment, and they are consistently optimistic when applied to an Australian venue paying award rates with penalty loadings.
This is what the numbers actually look like here, and which of them are worth your attention.
What Australian restaurants actually make
Net profit — what is left after everything, including the owner being paid properly for the work they do — typically lands in these bands:
- Independent full-service restaurant: 3–8%
- Café: 5–12%
- Pizzeria or high-volume casual: 8–15%
- Fine dining: 0–5%, and frequently the lower half
A venue turning over $1.2 million at 6% net is making around $72,000 before the owner's own wage is properly accounted for. That is the reality of the industry, and it is why the difference between 4% and 9% is not a rounding error — it is whether the business is worth owning.
Reading a restaurant P&L
A hospitality P&L has a specific shape. Reading it in the right order tells you where a problem is within about two minutes.
Revenue
Split food and beverage. Beverage carries a much better margin, so the ratio between them moves your blended result substantially. A venue at 70/30 food to beverage is a different business from one at 55/45, even at identical turnover.
Cost of goods sold
Food and beverage costs, calculated on opening stock plus purchases minus closing stock — not simply on invoices. Skipping stocktake means this line is an estimate, and an estimate here makes every number below it an estimate too.
Gross profit
Revenue minus COGS. Should sit around 65–72% for most formats.
Labour
All of it: wages, superannuation, workers' compensation, leave provisions, payroll tax if you are over the threshold. Counting only the wages line understates real labour cost by roughly 15–20% and is the most common self-deception in the industry.
Prime cost
COGS + labour. This is the number to watch. Most viable Australian venues sit at 60–65% of revenue. Above 70% and the business is usually not paying its owner properly, whatever the individual lines say.
Prime cost matters because it is the part you control weekly. Rent is fixed for years; ordering and rostering are decisions you make every week.
Fixed costs
Rent, insurance, utilities, subscriptions, accounting, marketing. Rent above about 10% of turnover is a structural problem that operational excellence will not fix.
The levers, in order of speed
Fastest: menu mix
Costs you nothing and moves within a fortnight. Shifting sales towards higher-contribution dishes through position, description and staff recommendation is the quickest available gain in most venues. This is menu engineering, and it is underused precisely because it does not feel like doing something.
Fast: beverage attachment
Beverage margin is roughly double food margin. Moving from 25% to 32% beverage share of revenue changes the blended gross profit materially, and it is a floor training question — offering the second glass, a matched wine, an amaro after dessert — not a purchasing one.
Medium: portion and waste control
The gap between costed and actual food cost is waste and portioning. Closing a five-point gap on $400,000 of food revenue is $20,000 straight to the bottom line. It takes scales on the section, a written spec, and a manager who checks — which is why it is medium rather than fast.
Medium: rostering to the actual trade curve
Most venues roster to a shape they inherited rather than to their sales-per-hour data. Overlapping a shift by two hours on the wrong days, six days a week, is a five-figure annual cost. The data is already in your POS.
Slow: price
Effective and risky. A 5% price rise across a menu drops almost entirely to the bottom line, which on a 6% net margin is close to doubling it. It is slow because it should be staged, tested on a section at a time, and paired with something the customer can see changing. Raising prices while everything else stays identical is noticed.
Structural: rent and format
If rent is 15% of turnover, no amount of costing fixes it. The honest answers are renegotiation, increasing turnover in the same footprint — more covers, an extra daypart, functions — or accepting that the site is wrong.
What to look at weekly
Monthly P&Ls tell you about a problem after it has cost you a month. Four weekly numbers:
- Sales versus the same week last year — trend, not weather.
- Labour as a percentage of sales — the fastest-moving controllable line.
- Food cost from a weekly stocktake on your top ten lines — not the whole store, just the ten that matter.
- Covers and average spend — because falling revenue from fewer covers and falling revenue from a lower spend are different problems with different fixes.
Four numbers, twenty minutes a week. It is not sophisticated, and it is more than most independent venues do.
Common questions
What is a good profit margin for a restaurant in Australia?
Net margins typically run 3–8% for independent full-service restaurants, 5–12% for cafés, 8–15% for pizzerias and high-volume casual venues, and 0–5% for fine dining. American figures are consistently optimistic here because they assume a tipping culture that subsidises wages.
What is prime cost in a restaurant?
Cost of goods sold plus total labour, as a share of revenue. Most viable Australian venues sit at 60–65%. Above 70% the business usually isn't paying its owner properly. It matters more than food cost alone because it's the part you control week to week.
How do I improve my restaurant's profit margin?
In order of speed: shift the menu mix towards higher-contribution dishes, lift beverage attachment, close the gap between costed and actual food cost through portion control, roster to your actual trade curve, then stage price rises. Rent above roughly 10% of turnover is structural and won't be fixed operationally.
What should be in a restaurant P&L?
Revenue split by food and beverage, cost of goods sold calculated from stocktake rather than invoices, gross profit, fully loaded labour including super and workers' comp, prime cost, then fixed costs. Counting only the wages line understates real labour by about 15–20%.