Here is a small problem that costs Berlin restaurants a surprising amount of money, and almost nobody notices it happening.
Saturday. The room fills at seven. By eight you are turning people away at the door, and by nine the kitchen has crossed two things off the menu. You do 120 covers.
The number that goes into your books that night is 120.
120 is not what Saturday wanted
It is what your room could hold. Demand that night might have been 150. It might have been 180. Nobody counted, because nobody stands at the door counting the people who walk off to the place across the road.
That is fine for one night. The trouble starts when that number becomes evidence.
Next Saturday, you look back at recent Saturdays and see 120. You prep for 120. You sell out again at nine. You write down 120 again. Do that four weeks running and you have taught yourself — and any software watching — that Saturday is a 120-cover night, when it has actually been a 150-cover night that you keep capping at 120.
You are not under-performing. You are under-measuring, and then acting on the measurement.
It shows up in more places than you would think
Once you start looking for it, the same shape turns up all over a kitchen's numbers:
Delivery switched off. Most operators pause delivery when the dining room is slammed. So the nights with the fewest delivery orders are often your busiest nights overall. Any tool that reads that as "delivery demand was low" has learned precisely the opposite of the truth.
A day you were closed. A blank in the spreadsheet gets read as a zero. One two-week holiday, entered as fourteen blanks, will drag that whole period's picture down for months.
Ninety minutes of an oven being broken. A perfectly good Friday recorded as a bad one, with nothing anywhere to say why.
A private hire. Enormous revenue, no walk-in trade, and a number that looks like a spectacular Tuesday you will never repeat.
What we do about it
Four things, and none of them require you to do much.
We ask. If the numbers suggest you hit your ceiling, you get a question rather than an assumption — one tap to confirm you sold out. That night then counts as "at least this busy" rather than "exactly this busy", and it stops dragging your Saturdays down.
Delivery is kept separate from the dining room. Always. They are two different trades with two different rhythms, and adding them together makes both worse.
Closed is a tick, not a blank. A day you were shut is never treated as a day you were quiet.
Your own notes count. Broken oven, private party, building site outside for a fortnight — mark it once and it stops distorting everything that comes after.
Why this matters more than it sounds
Because the error only ever goes one way. Capacity limits always make busy nights look less busy, never more. So the mistake compounds in the direction of under-ordering and under-staffing — on precisely the nights when running short costs you the most, and where the customer who got turned away decides where to go next time.
The fix is not clever mathematics. It is mostly just noticing, and then asking.