Operators talk about revenue and think about margin, and the gap between the two is where most independent venues quietly get into trouble. So here is what the numbers look like across the Berlin venues we work with — small independents, thirty to sixty seats, mostly neighbourhood trade.
Treat these as a ruler to hold against your own P&L, not as targets. Every venue's mix is different.
The rough shape
For a typical Berlin café or small restaurant, monthly revenue somewhere between €30,000 and €60,000:
| Line | Typical range | Notes |
|---|---|---|
| Food and drink cost | 26–33% | Lower for coffee-led, higher for a proper kitchen |
| Labour, including owner | 30–38% | The number that has moved most in three years |
| Rent and Nebenkosten | 8–14% | Wildly location-dependent |
| Everything else | 10–15% | Energy, insurance, software, accountant, repairs |
| What is left | 5–12% | Before the owner takes anything meaningful |
The venues at the top of that last range are not the ones with the best food. They are almost always the ones with the tightest grip on the two big variable lines.
Prime cost is the number that matters
Add food cost and labour together and you get what most operators call prime cost. Across our venues it lands between 58% and 70%.
Below 60% and you have room to breathe. Above 68% and you are one bad month from a difficult conversation with the bank. The venues in trouble are almost never in trouble because of rent — rent is fixed and everybody knew it going in. They are in trouble because prime cost drifted two or three points over a year and nobody noticed until the annual accounts.
Two or three points on €45,000 a month is €1,350 a month. It is also almost exactly the range that better prep and staffing decisions move.
Where the leakage actually is
When we look at a venue's daily numbers over a few months, the same three patterns come up.
Over-prep on ordinary days. Not the dramatic waste — the small, invisible kind. Twelve portions too many on a Tuesday, four times a week, fifty weeks a year. Nobody logs it and it never appears as an incident, but it is usually the single largest avoidable line in the business.
Overstaffing the quiet half of the week. Rotas get built from templates, and templates come from an average week. Monday to Wednesday are consistently over-covered in most venues we see, by half a person to a person a shift.
Under-covering the exceptional days. The warm Saturday, the fair week, the day before a holiday. These are the days with the highest margin per cover, and they are the days most likely to be short-staffed and short-prepped — because they are precisely the days a template gets wrong.
That last one costs more than it looks. A cover you turn away on your best day is worth several covers on your worst.
What moves the needle, in order
One point off food cost. Achievable in most venues within a quarter, almost entirely through prepping closer to actual demand rather than by changing suppliers or portions.
One point off labour. Harder, and mostly comes from the quiet days rather than the busy ones.
Better coverage on your top twenty days. Small effect on cost, real effect on revenue, and the one operators consistently underestimate.
Price. Available, and mostly already used. Berlin has real price resistance on coffee and lunch right now.
The measurement problem
Almost every venue we start with is measuring monthly, from the accountant, six weeks late. That is fine for tax and useless for management — by the time you see that March drifted, it is mid-May and April went the same way.
The venues that hold prime cost steady are the ones looking at daily numbers weekly, in a form that takes ten minutes rather than an afternoon. Not more analysis. Just closer to the event.
Where we fit
BentoPlan is a forecast, not an accounting system — we tell you how many covers you are going to do tonight, tomorrow and the day after, and on the paid plan we size the shifts against it. Our venues report cutting food waste by up to 45% and labour by around 5%.
On a €45,000 month, that is the difference between a business that pays its owner properly and one that does not. Which is really the only benchmark that counts.