All Posts

Restaurant KPIs Worth Checking Every Week

Industry Insights 7 min read 9 October 2026
Manager and chef talking by the window while service runs behind them

There are two common ways to run a restaurant by numbers, and both of them fail. The first is to look at nothing until the accounts arrive, by which point you are reading a report about a month you can no longer influence. The second is to buy something that produces forty metrics, open it twice, and quietly stop.

The version that works sits between them. It is five numbers, checked once a week, and it takes about twenty minutes. What makes it worth doing is that a problem found in week one costs you a quarter of what the same problem costs by the time it surfaces in the accounts.

Why Month-End Is Too Late

A monthly figure tells you what happened on average across four or five weeks. If one week went badly, the average absorbs it, and by the time you see the number the people who could explain it have worked twenty shifts since.

Weekly changes the nature of the question. Instead of asking why the month came in at 34 percent, you are asking why this week came in at 38 when the last three came in at 32, and that is a question somebody in the building can actually answer. The week is still fresh and the invoices are on the desk, so whatever caused it is probably still happening.

The Five Numbers

Food cost percentage for the week. Your purchases against your sales for the same seven days, which is a rough figure and deliberately so. It will bounce around, because a big delivery on a Thursday distorts a week in a way it does not distort a month, so read the trend across four weeks instead of any single point. What you are watching for is a step change that holds. Comparing it to a sensible benchmark is useful once, and comparing it to your own last eight weeks is useful every time.

Variance on your top-value items. This is the number most restaurants never see weekly, and it is the most diagnostic of the five. Count only the expensive lines, so the proteins and the spirits and anything easy to over-portion, and compare what your sales say you should have used against what actually moved. Twenty lines takes about twenty minutes. The gap between theoretical and actual on those items will find a portioning problem long before the food cost percentage notices it.

Purchase spend against sales. Simply what you spent with suppliers this week set against what you took. It is close to the first number and it catches something different, which is buying ahead. A kitchen that has quietly built up stock will look fine on food cost and will have a lot of cash sitting on shelves.

Waste value. The cash figure, taken straight off your waste log. One week on its own tells you little. The same product appearing in three consecutive weeks tells you where to go.

Stock value on hand. How much cash is sitting in the building. This one moves slowly, and that is exactly why it earns a weekly glance, because a slow climb over six weeks is the clearest signal there is that ordering has drifted away from what you are selling.

Then a sixth thing, which is a check instead of a metric. Look at which ingredient prices moved this week. It takes a minute and it is where most unexplained food cost movements start, particularly when a supplier has put prices up without saying so.

What Is Missing From That List

There is no labour number here, and no prime cost. Both matter a great deal to a restaurant and neither is something this article is going to pretend to cover well, because they belong to a different discipline and a different set of tools. What is above is the food and beverage cost side of a weekly check, which is the half that lives in the kitchen.

Guest-side numbers, so covers, average spend and the rest, are worth watching too. They are also usually already visible in your POS reporting, so they are not usually the thing a restaurant is missing.

What a Bad Week Looks Like

The value of a short list is that the combinations mean something.

Food cost up and variance up together points at the kitchen, so portioning, waste or something going out of the door unrecorded. When food cost rises and variance stays flat, the cause is usually prices, which is why the sixth check exists. Stock value climbing while food cost looks fine means you are buying ahead of your sales, and that corrects itself painfully once you stop. And if waste concentrates on one product while everything else holds steady, you have a storage or ordering problem on that line and nothing more.

None of those readings requires much interpretation, and that is the point. A weekly check is meant to sort problems into the right pile quickly, so the time you have goes to the one that is actually costing you.

Numbers That Look Important and Are Not

Percentages on small lines will waste your time. A 40 percent movement on something you spend 30 euros a week on is arithmetic, not a problem, and a dashboard that ranks by percentage will surface it above a two percent movement on your biggest line.

Single-week readings of anything volatile are similarly misleading. Food cost percentage in a week with a large delivery is not comparable to a week without one, so a chart with four points on it is worth more than any individual figure.

Anything nobody will act on is worth removing entirely. If a number has been on the sheet for three months and has never once changed a decision, it is costing you attention and returning nothing.

Building It Into a Week That Is Already Full

Put it on the same day every week, after the last delivery and before you order again, so the numbers arrive when a decision is still available. Keep it to one page, and keep the same page each week, because comparability is doing most of the work.

Bring somebody from the kitchen into it. A food cost figure discussed with the head chef becomes a conversation about a section, and the same figure read alone in an office becomes an email nobody enjoys.

Reading Instead of Assembling

Most of the twenty minutes above is gathering, and gathering is the part that decides whether a weekly check survives past the first month. Stockifi reads supplier invoices as they arrive and pulls sales from the POS, so purchases, sales and price movements are already in one place. Counts done on a phone give the variance and the stock value without a spreadsheet, and waste entered as it happens is priced from the recipes automatically. That turns the weekly check into reading instead of assembling, and that is the difference between a habit that holds and one that lapses in November. See how the reporting works.

Frequently Asked Questions

What KPIs should a restaurant track? On the cost side, food cost percentage, variance on high-value items, purchase spend against sales, waste value and stock on hand will cover most of what you can actually influence. Guest-side numbers like covers and average spend matter too and are usually already in your POS reporting.

How often should restaurant KPIs be reviewed? Weekly for anything you can still change, and monthly for anything you cannot. A monthly-only cycle means every problem gets found after it has finished happening.

What is a good food cost percentage? It depends heavily on your style of service and your market, and the more useful comparison is against your own recent weeks. A stable figure that suddenly steps up and stays there matters more than whether it sits at 30 or 34.

Do I need software to track weekly KPIs? No, and a spreadsheet works if someone will genuinely keep it up. The reason most weekly reviews lapse is the gathering rather than the reading, so anything that removes the assembly step is what makes the habit stick.

Twenty Minutes a Week

Pick a day, pick the five numbers, and put them on one page you keep. The first month will feel like admin and produce nothing much. Somewhere in the second, a week will come in wrong, you will know about it on the Tuesday instead of the following month, and it will pay for the whole exercise.

See where your margin is leaking