Restaurant KPIs: The Six Numbers a Café Owner Should Track

Most café owners track one number: what's in the bank. It's the worst possible early warning, because it tells you what happened weeks ago, after rent and salaries have already gone out and after the supplier cheque has cleared. By the time the balance looks wrong, whatever caused it has been running for a month. The six numbers below are the ones that move first — and each of them points at a different decision you can actually make.

Why your bank balance is the wrong thing to watch

A bank balance mixes everything together: takings, a rent cheque that cleared, a bulk coffee order, a VAT payment, the deposit you paid on a new fridge. Two cafés with identical balances can be in completely different health. One is growing and holding stock; the other is quietly losing margin and hasn't paid its suppliers yet. Neither owner can tell from the balance alone which one they are.

What follows isn't a dashboard you buy — it's six figures you can pull weekly and write in the same place every time. The point isn't the number on any given week. It's the direction it has moved over the last four.

1. Average ticket — what one customer is worth

Total sales divided by number of orders, for the same period each week. Use the figure excluding VAT, so you're measuring your own revenue rather than the 5% you're collecting for the FTA. It is the fastest number to move and the cheapest to fix: a barista who stopped asking about the pastry, a second coffee that used to go on the same ticket, a menu change that quietly removed the item most people added.

Read it against transaction count, never on its own. Average ticket rising while orders fall usually means you've lost your cheapest, most frequent customers — which looks like good news for exactly one week. If you have just changed prices, the two numbers together are the only honest read on how it landed; raising prices without losing regulars goes through what to watch and for how long.

2. Covers — how many people you actually served

Count orders, by day and by hour. Not revenue — bodies. This is the number that tells you whether a quiet month is a demand problem or a spend problem, and they need completely different responses. Fewer people through the door is marketing, location or a competitor. The same people spending less is menu, pricing or service.

Broken down by hour, covers also settle arguments about opening times and staffing that otherwise run on instinct. Most cafés have at least one hour on the roster that has not covered its own wage bill in months, and nobody has looked because the day as a whole is fine.

3. Food cost percentage — what your menu costs to make

Cost of the goods you sold, divided by sales excluding VAT. It is the number most likely to drift without anyone doing anything wrong: a supplier raises milk by a dirham, portions grow slowly over a season of new staff, waste creeps up in the summer. None of those announce themselves, and all of them show here first.

Don't chase an industry benchmark — build a four-week baseline of your own and manage against movement from it. If you want the rule-of-thumb band as a sanity check rather than a target, food cost percentage covers where it comes from and why it misleads as often as it helps.

4. Labour cost percentage — the other big share

Total wage cost for the period divided by sales excluding VAT. Rent aside, this and food cost are the two lines big enough to decide whether a café makes money, and labour is the one you can adjust weekly. The failure mode is almost never overstaffing in general — it's a rota built for a rush that has since moved an hour later, or a weekend shape applied to a weekday.

Read it alongside covers by hour, not by itself. A high labour percentage in a genuinely busy week means something different from the same percentage in a dead one. Our guide to staff scheduling for a small café works through turning hourly sales into a rota rather than the other way round.

5. Repeat rate — whether anyone comes back

Of the customers you can identify, how many bought more than once in the last ninety days. This is the number nobody tracks and everybody assumes. You will only ever see it for the share of customers who give you a name — a loyalty stamp card, a phone number for a large order — so treat it as a sample, not a census, and compare it with itself over time rather than with anyone else's figure.

It matters because every other number on this list is cheaper to improve when customers return. Winning back a regular who already likes the coffee costs nothing; finding a new one costs whatever your marketing costs. If the sample is small, building a customer list covers how to collect names without making people feel harvested.

6. Cash conversion — whether profit reaches the bank

Profitable cafés run out of money regularly, and it's rarely mysterious. Card takings settle a few days behind, stock is bought before it sells, rent goes out in lumps, and a corporate order gets paid in sixty days. Once a month, put your profit for the period next to what actually landed in the account and account for the gap.

The gap is usually one of three things: money sitting in stock, money sitting with an aggregator or acquirer, or money owed to you by someone slow. All three are fixable, but only once you have named which one it is. If most of it is stock, par levels and counting is where the cash went.

Where MidaOne fits

Four of the six come straight off the till if the till is recording properly. MidaOne shows total sales, orders and average order on the reports page, sales by day and by hour for covers, and a sales summary carrying cost of goods, gross profit and margin, expenses, waste and net — so food cost is read rather than reconstructed. The customer list shows total visits and the last visit date per customer, which is what you'd count a repeat rate from. Labour and cash conversion come from your payroll and your bank statement; no till knows those, and any system that claims to is guessing.

Frequently asked questions

Six numbers is a deliberately short list. You can hold six in your head, write them on the same sheet every Sunday, and notice within a fortnight when one starts moving the wrong way — which is the entire point. A dashboard with thirty metrics gets checked twice and then never again, and the café that had a food cost problem in March finds out about it from the accountant in June.