7 Numbers a Restaurant Owner Should Know Every Week

The restaurant KPIs worth checking every week are few: covers and spend per head, prime cost, the food cost gap, labour percentage, sales per labour hour, revenue per available seat hour, and waste. Seven numbers, read on the same day each week, will tell you more than a dashboard of forty.

Think of the Monday morning review as a pre-shift meeting for the business. It should take twenty minutes, and every number should lead to a decision. If a figure cannot change what you do on Tuesday, drop it from the list. The seven below earn their place, and each comes with its trap.

1. Covers and spend per head

Sales is covers multiplied by spend per head, and the two move for different reasons. Track them separately. Rising sales with falling covers means you are charging more for fewer guests, which can be healthy or can be the start of a slide. Rising covers with flat spend usually points to a discount or a menu that guests are using to trade down.

Compare each with the same week last year if you have it, and with the previous four weeks. Weather, holidays and local events distort single weeks.

The trap: celebrating total sales. A busy week can hide a drop in what each guest spends, and the reverse.

2. Prime cost percentage

Prime cost is labour plus cost of goods sold, divided by sales. One software firm's guide puts the typical range at 55% to 65% of sales, with quick service at 55% to 60% and full service at 60% to 65%. Those ranges come from the firm's own article, which cites no external data, so use them to orient yourself and no more.

The same guide suggests setting your own target from your past twelve weeks, and calculating the full figure weekly. That is sound advice. Prime cost is the one number that tells you whether the two largest controllable costs are in step with sales.

The trap: a good-looking prime cost that is flattered by timing. If invoices arrive late or stock is miscounted, the week lies. Count stock on the same day each week.

3. The food cost gap (theoretical versus actual)

The gap between what your recipes say food should cost and what it actually cost is the most useful food number you have. The 7shifts guide illustrates this with a theoretical food cost of 28% against an actual 33%, a gap that can point to waste or portioning problems. The figures are an example, not a benchmark.

A five-point gap like this example points to leakage rather than pricing: over-portioning, unrecorded waste, theft, supplier short deliveries or recipe drift. The fix starts with a recount and a look at the five highest-cost items. For the formula and ways to lower the figure, see food cost percentage.

The trap: chasing the absolute food cost percentage. A restaurant with a lower percentage can still earn less if it sells cheaper dishes.

4. Labour cost percentage

Labour percentage is fully loaded labour cost, including wages, payroll taxes and benefits, divided by sales. Restaurant365 says most operators try to stay within 25% to 35%, and also calls the operator's own historical figure the most important benchmark.

Labour rules and costs vary widely by country, so this range may not fit you at all. In some markets service charges, tips and employer contributions change what "fully loaded" means. Decide on a definition once, write it down, and keep to it.

The trap: cutting hours to hit the percentage. Short staffing slows service, which cuts covers and spend, which pushes the percentage back up.

5. Sales per labour hour

Sales per labour hour shows whether the rota matched the room. Restaurant365 notes that a low figure can signal overstaffing and an unusually high one understaffing, and that targets should be set by daypart. It gives no benchmark, so set one from your own history.

Calculate it for lunch, early evening and late evening separately. A weekly average hides the Tuesday that was overstaffed and the Friday that was stretched.

The trap: reading it without the guest. High sales per labour hour with long waits is not efficiency.

6. RevPASH by daypart

Revenue per available seat hour tells you how hard each seat works in each hour. It was proposed for restaurants by Kimes and Chase in 1998 and developed in later work by Kimes and others. A recent piece in Hospitality Net by Melissa Kalan argues for splitting the formula: service revenue divided by service hours, divided by seats. That way you can see whether a change came from seats or hours.

Imagine a 50-seat restaurant open for three hours at lunch. If lunch revenue rises but the dining room is half full for the first hour, the average hides a slow start. Calculate RevPASH in hourly slices and pair it with seat utilisation, covers served divided by capacity in that period.

The trap: pushing guests out to raise the figure. Pair it with table turn time and guest comments, so speed does not replace welcome.

7. Waste, voids and comps

The seventh number is what you threw away or gave away. Track food waste by weight or cost, and add voids and complimentary items from your till. They are the cheapest signals of a problem in the kitchen or on the floor.

Restaurant365 quotes FoodPrint's figure that up to 10% of food purchased by restaurants is wasted before it reaches guests. Treat that as a US-centred reference point, not a target. Your waste log, even on paper, is better. The ideas in ten ways to cut food waste start here.

A number you read every Monday becomes a habit. A number you read once a quarter becomes a surprise.

The trap: logging only the dramatic spills. The steady trim and the unsold prep are where the money goes.

How do you run the weekly review in twenty minutes?

Use the same sheet, the same day and the same order. Enter the seven numbers, mark each as up, down or flat against your own four-week average, and write one action next to anything that moved more than you expected. Share the sheet with your head chef and floor manager, because they hold the explanation.

If a number moves for two weeks running, stop treating it as noise. Pull the detail behind it, whether that is a supplier invoice, a rota or a menu change, and decide who will look into it by Wednesday. Small, steady follow-up beats an annual clean-up.

Do not compare yourself with a venue on a different street. Compare this week with your own last twelve.

Key takeaways

• Seven weekly numbers are enough: covers and spend per head, prime cost, food cost gap, labour percentage, sales per labour hour, RevPASH and waste.
• Treat software-firm benchmarks as a starting point; your own twelve-week history is the stronger yardstick.
• The gap between theoretical and actual food cost is more useful than the percentage alone.
• Read sales per labour hour and RevPASH by daypart, never as a weekly average.
• Define each number once, in writing, and count stock on the same day.
• Every number should lead to one action, or it should come off the list.