Venues rarely close because of the food. They close because nobody costed the menu for six months, the chef left mid-season, and the menu is the one from opening day — with prices set two supplier increases ago.
This piece is about four numbers that get ahead of those situations. None of them needs a system costing thousands. A spreadsheet and an hour a quarter will do.
Number 1: food cost
The share of the menu price taken by the ingredients. Typical ranges depend on the format: cafés 15–30%, fast food and street food 25–40%, bistro 30–45%, fine dining 35–50%.
The figure itself says less than its consistency across the menu. A card where one dish sits at 22% and the one next to it at 48% means prices were set by feel — and that your best-selling item may be the one you earn least on.
How to work it out item by item, and what to do with the result, is in menu engineering, with a spreadsheet to download.
Number 2: labour cost
The second-largest line after ingredients, and the largest in venues with full service. What matters is not the wage but the fully loaded cost of an hour worked, compared against turnover in that same hour.
That comparison shows things a rota does not: that Tuesday afternoon costs more than it brings in, or that one more person on Friday would pay for themselves after three tables.
It also holds the most commonly ignored cost — staff turnover. Bringing a new server up to speed takes 3–6 weeks, during which service is slower and errors more frequent. At 4 departures a year, that cost often exceeds the raise that would have kept the previous person.
Number 3: table turnover
How many times during a service the same table takes new guests. At 40 covers, the difference between 1.8 and 2.2 turns over lunch is a dozen more guests a day on the same headcount and the same rent.
Turnover can be improved without rushing anyone. A guest waits three times — for the menu, for somebody to take the order, and for the bill — and each wait is 3–8 minutes at peak. The first two can be removed without shortening the meal itself; we went through it in table ordering by QR.
Number 4: the share of regulars
Winning a new guest costs several times more than keeping an existing one, and in most venues a minority of regulars accounts for the majority of turnover. Which means the marketing budget and the budget for return visits are two different things — and the second usually does not exist.
The simplest measurement without a loyalty system: for 2 weeks, count how many guests a day the floor team recognises. It is an approximation, but the movement of that number over time says more than a one-off audit.
The cheapest mechanism for return visits is not a discount. It is a reason to come back — a daily special that changes, a seasonal item worth knowing about. One condition: there has to be somewhere to see it before leaving the house, which means the Google Business Profile and a menu at a fixed address.
Where these numbers usually break
Three situations repeat across venues regardless of format:
The ingredient price rose, the menu price did not. The supplier raises in March, the menu goes to print in September — six months of selling at the old price with the new cost. On one dish selling 15 times a day with a €0.70 difference, that is €315 a month. The full calculation is in what a restaurant menu really costs.
The dish left the kitchen but not the menu. Every following guest asks, the server explains, and the table loses a minute. On a printed menu there is no way to fix it mid-service.
The international guest orders safely. Not understanding the menu, they take what they recognise — usually the cheapest, most obvious item, with no starter and no dessert. At a double-digit share of international guests, that is a visible difference in the average bill.
What these four numbers will not show
Honestly: none of these 4 numbers measures what people come back for.
A dish somebody drives across town for can have an average margin and average sales. A server guests ask for by name appears in the labour line only as a cost. A room that feels right has no column in the spreadsheet.
The numbers exist so that those things survive — not to replace them. A restaurant with excellent food cost and an empty room is not a success, it is a well-costed closure.
Frequently asked questions
What food cost is good in a restaurant?
It depends on the format: cafés 15–30%, fast food 25–40%, bistro 30–45%, fine dining 35–50%. Consistency across the menu matters more than the number itself.
How often should these be calculated?
Food cost and item profitability — once a quarter and at every menu change. Labour cost and table turnover — monthly. More often makes no sense: the data needs time to accumulate.
What comes first when time is short?
Food cost for the twenty best-selling items. Those dishes make most of the turnover, so decisions based on them come out the same as after analysing the whole menu.
Is a system needed for this?
No. A sales report and a spreadsheet are enough. A system starts to matter when you want to change prices more often than printing a menu allows.
