Restaurant KPIs: A Practical Guide to the Metrics That Show Where You Make or Lose Money

A restaurant produces hundreds of numbers a week. Most of them do not matter. A small set of them — the ones that show where the money comes from, where it leaks, and whether you have enough cash to keep going — are the restaurant KPIs worth an owner’s time. This guide explains which ones those are, how to calculate them, what benchmarks really mean, and how to review them in a weekly routine that fits around service.

It is written for owner-operators and managers of independent, single-location restaurants. Chains have finance teams; you have a POS export, a supplier folder and an hour on Monday. The framework below is built for that reality.

What restaurant KPIs are — and what they are not

A metric is anything you can count: covers, tickets, hours, invoices. A key performance indicator is a metric you have deliberately chosen because a change in it forces a decision. That distinction matters more in restaurants than in most businesses, because the data is noisy and the margins are thin. A 2-point rise in food cost on a $60,000 month is $1,200 — roughly the profit of a good week for many independents.

A useful restaurant KPI has four properties:

• It is a ratio or a rate, not a raw total. Sales of $14,000 means nothing on its own; sales of $14,000 with a 34% food cost and a 36% labor cost tells you the week barely broke even.
• You can calculate it from data you already have — POS reports, supplier invoices, payroll, a monthly stock count.
• Someone owns it. Food cost belongs to the kitchen lead; labor cost to whoever writes the rota; average check to the floor.
• It has a threshold that triggers an action. If nothing changes when the number moves, it is a report, not a KPI.

Most owners do not need 40 KPIs. They need eight to ten that answer three questions every week: Are we selling enough? Are we keeping enough of it? Will we have the cash to pay the next set of bills?

A practical restaurant KPI framework

The framework below runs in one direction: Input → KPI → Diagnosis → Priority → Action. You enter a handful of figures, the KPIs tell you which part of the business moved, you work out why, you pick the one thing that matters most this week, and you do it. Everything else in this guide hangs off that loop.

Each KPI below has its own detailed article. This page explains how they fit together.

Revenue, guests and average check

Revenue is the number every owner watches, but on its own it hides more than it shows. Split it into how many guests and how much each guest spends, and you can immediately see whether a soft week is a traffic problem or a spend problem — two very different fixes.

Average check = revenue ÷ number of guests. Some operators divide by tickets (tables) instead; either works as long as you are consistent. Track it by daypart if your lunch and dinner menus differ, because a blended number can flatten a real change. Average check moves with menu mix, pricing, upselling, portion and course structure, and the share of takeaway or delivery orders, which usually carry a lower spend per head.

Read the full breakdown in average check size for restaurants.

Food cost percentage

Food cost % = cost of food sold ÷ food sales. “Cost of food sold” is not the same as “food purchases”: it is opening inventory + purchases − closing inventory for the period. Without a stock count you are measuring purchasing, not consumption, and the number will swing with delivery timing rather than with what the kitchen actually used.

Food cost rises when supplier prices move, when portions drift, when waste and spoilage go unrecorded, when comps and staff meals are not tracked, and when the sales mix shifts toward low-margin dishes. It falls when you re-cost recipes, re-engineer the menu, tighten prep lists and buy in line with forecast covers. Beverage cost is worth tracking separately; bar margins behave differently and a blended cost of goods hides that.

See food cost percentage: formula, interpretation and control.

Labor cost percentage

Labor cost % = total labor cost ÷ sales. Total labor cost includes wages, employer taxes and contributions, benefits, and — if you draw a wage — you. Owner-operators who leave their own hours out of the number routinely believe their labor cost is 8 points better than it is, and then cannot understand why the business does not survive their first holiday.

The two levers are the rota and the sales forecast. Labor cost spikes when you staff for the week you hoped for rather than the one you got, when overtime accumulates, when training and turnover keep the floor in a permanent onboarding state, and when a slow daypart is staffed like a busy one. Sales per labor hour (revenue ÷ hours worked) is the companion metric that shows whether a rise in labor cost is a scheduling problem or a sales problem.

See restaurant labor cost: how to calculate it and what moves it.

Prime cost

Full guide: restaurant prime cost — formula, worked example and what moves it.

Prime cost % = (food and beverage COGS + total labor cost) ÷ sales. Prime cost is the single most useful weekly number in a restaurant because it captures the two costs you can actually influence week to week. Rent, insurance and loan payments are decided once a year; food and labor are decided every shift.

Watching prime cost rather than food and labor separately also stops a common mistake: cutting labor so hard that waste, errors and slow service push food cost and lost sales up by more than you saved. If prime cost is drifting upward over four to six weeks while sales are flat, something structural has changed — a supplier, a menu, a rota template — and it will not fix itself.

Gross and operating profitability

Gross margin = (sales − COGS) ÷ sales. It shows how much of each dollar is left after ingredients to pay for everything else. Operating margin = (sales − COGS − labor − occupancy and other operating costs) ÷ sales is the number that decides whether the restaurant is a business or an expensive hobby.

Gross margin is a menu question: pricing, portioning, mix. Operating margin is a whole-business question: it can fall because of a rent increase you cannot control, or because of a rota you can. Track both, but make decisions on operating margin. Restaurant profit margins are structurally thin; a few points of operating margin are usually the difference between reinvesting and refinancing.

See restaurant profitability KPIs: gross, operating and net margin.

Break-even

Break-even sales = fixed costs ÷ contribution margin %, where contribution margin % is (sales − variable costs) ÷ sales. For most restaurants the variable costs are food, beverage, packaging and any commission you pay on delivery orders; rent, salaried staff, insurance, utilities minimums and loan payments are fixed.

Knowing your break-even week — the sales level below which you lose money — changes how you read every other KPI. A slow Tuesday is a problem if you are close to break-even and a non-event if you are well above it. It also tells you exactly what a price change, a new lease or an extra salaried manager does to the sales you need. Recalculate it whenever a fixed cost changes.

See how to calculate a restaurant break-even point.

Table turnover and seat occupancy

Table turnover = parties seated ÷ number of tables, per service or per day. It is a capacity metric: it tells you how hard your dining room is working. Turnover is only “good” relative to your concept — a 40-minute lunch counter and a 2-hour tasting menu both have the right turnover if the dining room is full and guests leave happy.

Where it earns its place is in diagnosing lost revenue. Low turnover with a waitlist means the bottleneck is in service flow or the kitchen; low turnover with empty tables means the problem is demand. Pairing turnover with seat occupancy (covers ÷ available seats) separates the two.

See table turnover rate: formula and how to use it.

Customer retention and repeat guests

Repeat guest rate = returning guests ÷ total guests over a period. Most independents cannot measure this perfectly — you need a loyalty program, reservation data or a payment-tokenised POS to identify returning guests — but an approximate number, measured the same way every month, is still far better than none. Regulars are the cheapest revenue a restaurant has: no marketing cost, higher average check, more forgiving of a bad night.

Retention moves with consistency more than with novelty: the same dish tasting the same, the same welcome, the same wait. A falling repeat rate is usually the first visible sign of an operational problem that reviews will describe a month later.

See customer retention rate for restaurants.

Inventory turnover and waste

Inventory turnover = COGS ÷ average inventory value for the period. It tells you how many times you sell through your stock. Low turnover means cash sitting on shelves, more spoilage and more shrinkage; very high turnover can mean stock-outs and emergency purchases at retail prices.

Waste % = recorded waste ÷ sales. Waste is only a KPI if it is recorded — a clipboard by the bin is genuinely one of the highest-return tools in the building. Track spoilage, prep waste, plate waste and comps separately for a month and the biggest category will usually surprise you.

See restaurant inventory turnover: how to calculate and improve it.

Cash flow: the KPI that ends restaurants

Profit is an opinion; cash is a fact. A restaurant can be profitable on paper and still miss payroll because supplier terms shortened, a quiet January arrived, or a large deposit went out for a new kitchen line. Two simple habits cover most of the risk: a rolling 8-week cash forecast (money in, money out, closing balance) and a rule for how many weeks of fixed costs you keep in reserve. Neither is a ratio, but both belong in the weekly review.

Restaurant benchmarks: what they mean and why they vary

You will find published ranges for every KPI on this page: food cost around 28–35%, labor around 25–35%, prime cost below roughly 60–65% of sales, and so on. These are rules of thumb, not standards, and treating them as targets is one of the most common mistakes in restaurant management. They vary for structural reasons:

• Concept and service model. A steakhouse runs a high food cost and a low labor cost; a pasta restaurant runs the opposite. Quick service and full service are different businesses.
• Location costs. Rent as a share of sales can be 6% or 18%, which changes how much room the other costs have.
• Owner labor. Whether the owner’s hours are paid at a market rate or not moves labor cost by several points.
• Delivery and takeaway mix. Third-party commissions sit outside food and labor but can take 20–30% of an order’s value.
• Accounting choices. Whether beverage is inside COGS, whether kitchen supplies are counted as food, whether a stock count is done at all.

The useful benchmark is your own trend. A food cost that moves from 30% to 33% over six weeks is a signal regardless of what a trade magazine says the average is. Use published ranges to sanity-check that you are not wildly out of line; use your own history to run the restaurant.

What a restaurant KPI dashboard should look like

A dashboard is not a chart gallery. For an independent restaurant it is one page with three parts:

1. Inputs — the raw figures you enter each week or month: sales by category, COGS, labor cost and hours, covers, seats, a stock value, recorded waste, a repeat-guest count if you have it.
2. KPIs — the ratios calculated from those inputs, shown against the target you set for your own restaurant, colour-coded so the eye lands on the exception.
3. Trend — at least the last six periods side by side, because a single month tells you almost nothing in a business with this much seasonality.

A weekly review that takes about 15 minutes

The routine matters more than the tool. A version that works for most independents:

1. Enter the week’s inputs (5 minutes): sales, covers, labor hours and cost, purchases. Stock and waste monthly, or weekly for high-value items.
2. Read the exceptions (3 minutes): which KPIs moved outside their target, and in which direction.
3. Diagnose (4 minutes): for each exception, is it a sales problem, a cost problem or a mix problem? Average check down and covers flat points at the menu or the floor; labor cost up with sales flat points at the rota.
4. Pick one priority (2 minutes): the exception with the biggest dollar impact that you can actually influence this week.
5. Assign the action (1 minute): one owner, one deadline, checked at next week’s review.

Monthly, add the stock count, the break-even recalculation if any fixed cost changed, and a look at the six-month trend rather than the week.

From measurement to action

Most restaurants that track KPIs still do not act on them, because the numbers arrive without a diagnosis. Food cost is 34% — so what? The step that turns a number into a decision is asking which of its drivers moved: prices, portions, waste, mix or theft. Each driver has a different fix, a different owner and a different cost to fix. Priority comes from the size of the gap in dollars — and, where you can judge it, how much of it you can realistically recover this month.

This is the logic behind the Restaurant Performance System: a weekly input sheet feeding a small set of KPIs, each with a diagnosis of which KPI moved and what to check, a priority ranking by dollar impact, and one recommended action. It is designed for an owner with about 15 minutes a week — not for a finance team. Start with the free template above if you want to see the KPIs first.

Frequently asked questions

How many KPIs should a small restaurant track?

Weekly: six to eight — sales and covers, average check, food cost, labor cost, prime cost, and one guest metric. Monthly: add margins, break-even, inventory turnover, waste and a cash forecast. More than that and the review stops happening.

What is the most important restaurant KPI?

If you track one number, track prime cost percentage. It combines the two costs you control most directly and it moves early when something is going wrong.

How often should restaurant KPIs be reviewed?

Sales, covers, labor and purchases weekly; stock-dependent KPIs (food cost calculated from a count, inventory turnover, waste) at least monthly. Daily tracking of everything usually produces noise rather than decisions.

Do I need special software?

No. A POS export, payroll totals, supplier invoices and a spreadsheet are enough for a single location. Software becomes worth it when data entry, not analysis, is what stops the review from happening.

Are restaurant KPI benchmarks reliable?

They are useful as a sanity check and unreliable as a target. Concept, location, service model and accounting choices move every ratio by several points. Your own six-month trend is the benchmark that should drive decisions.