Restaurant profit margins in Canada: the real numbers (2026)

The margin every operator quotes, and what it hides

The figure repeated across the industry is that full-service restaurants run a net profit margin in the low single digits, with quick service typically higher on thinner cheques and greater volume. Treat that as a rule of thumb, not a benchmark you are failing to hit — it varies enormously with rent, service style, and whether the owner is also working the floor.

The more useful point is that a low-single-digit net margin means a few percentage points of drift in food or labour cost is the difference between a profitable year and a loss. That is why operators manage prime cost weekly rather than reading a net margin annually.

The four margins, and which one to actually manage

These get used interchangeably in conversation and they mean different things.

• Gross margin — sales minus cost of goods sold, as a percentage of sales. Tells you whether your menu is priced correctly. It ignores labour entirely.
• Prime cost — cost of goods plus total labour, as a percentage of sales. This is the operating number. It is the one worth looking at every week.
• Operating margin — what is left after all operating expenses including rent and utilities, before interest and taxes.
• Net margin — what actually remains. Useful for the year, useless for running a shift.

Prime cost: the number that decides it

Prime cost is food and beverage cost plus all labour cost, including your share of CPP, EI, vacation pay and benefits — not just wages. Divide by sales.

The widely used target is to keep prime cost around 60 to 65 percent of sales for full service, and lower for quick service. Above roughly 70 percent, most restaurants cannot cover rent and overhead and still leave anything behind.

The two components trade off against each other, which is why managing them separately misleads. A scratch kitchen buys lower food cost with higher labour. A restaurant using more prepared product does the reverse. Neither is wrong — but only prime cost tells you whether the combination works.

A worked example

A full-service restaurant doing $80,000 in monthly sales. The arithmetic is deliberately simple so you can substitute your own figures.

• Sales: $80,000
• Food and beverage cost at 31%: $24,800
• Labour, fully loaded, at 33%: $26,400
• Prime cost: $51,200 — that is 64% of sales, inside the target band
• Remaining after prime cost: $28,800
• Rent, utilities, insurance, marketing, software, repairs, professional fees: $25,000
• Operating profit: $3,800 — 4.75% of sales

What that example tells you

Look at what a small move does. If food cost drifts from 31% to 34% — three points, which is a season of supplier increases nobody renegotiated — that is $2,400 a month. On $3,800 of operating profit, you have lost nearly two thirds of it without a single thing looking obviously wrong on the floor.

This is the whole argument for measuring weekly. A three-point drift caught in week two is a supplier conversation. Caught in the annual review, it is the year.

Where margin actually leaks

In roughly the order operators find them.

• Portioning drift — the single most common. Recipes are costed once and then executed generously, and food cost rises with nothing else changing.
• Supplier price increases absorbed silently — prices move, menu prices do not, and the gap is invisible until someone recalculates plate costs.
• Over-scheduling against actual demand — labour booked to a pattern rather than to forecast covers, so the quiet Tuesday costs the same as the busy one.
• Waste and spoilage nobody records — if it is not counted, it does not show up anywhere except in food cost, where it looks like a pricing problem.
• Comps and voids — legitimate in small volume, corrosive when untracked. Worth reviewing by staff member, not just in total.
• Delivery commission — third-party orders can carry commissions high enough to make some menu items unprofitable at the same price. Worth costing separately from dine-in.
• Per-terminal and per-screen software fees — small monthly line items that scale with your operation rather than with your revenue.

Related: online ordering without per-order commission

Menu engineering: the fastest lever you own

Repricing everything is a blunt instrument that costs goodwill. Menu engineering is the sharper version: cost every item, then sort by two axes — how profitable each dish is, and how often it sells.

That gives four groups. High profit and high popularity: protect these, make them visually prominent, never let quality slip. High profit, low popularity: promote them, move them on the menu, have servers recommend them. Low profit, high popularity: these are the ones to reprice or re-engineer, carefully, because guests notice. Low profit, low popularity: remove them, and take their prep complexity out of the kitchen with them.

Dropping a slow, low-margin dish does more than remove a line — it removes the inventory, the prep time and the waste attached to it.

How to actually track this weekly

None of the above works from memory or from a monthly accounting export that arrives three weeks late. What it needs is item-level sales data and honest labour hours in the same place, weekly.

The practical setup: costed recipes so plate cost is known, POS sales reporting by item so you can see popularity against profitability, labour hours tied to actual shifts, and a weekly prime cost calculation. That is enough. It does not require an expensive system — it requires that the numbers exist somewhere other than in someone's head.

Waitery reports sales by item and by period, with inventory and time-clock as one-tap add-ons, at $14 CAD per location per month with no per-terminal fee. Whatever you use, the test is the same: can you calculate last week's prime cost this week?

Related: restaurant management tools · Waitery pricing