Almost every independent in Berlin is on at least one delivery platform, and almost nobody can tell you whether it makes them money. Not because operators are careless — because the costs are spread across four places and only one of them shows up on the invoice.
The part everyone knows
Commission on the major platforms in Germany typically runs somewhere in the region of 13% to 30% of order value, depending on whether they deliver for you or you deliver yourself, and on what you negotiated. Marketing placements and promotional discounts come out on top of that.
On a €25 order at 28%, that is €7 gone before you have bought an ingredient. If your food cost is 30%, you are handing over more to the platform than to your supplier.
That number is well understood and most operators have already run it. The three below are the ones that get missed.
The packaging line
Containers, lids, bags, cutlery, the sticker. For a main course, budget somewhere between €0.60 and €1.20 depending on how much you care about how it arrives.
On a €12 lunch that is up to 10% of the order value, and it comes straight off an item you already discounted to compete on the platform. It also never appears in food cost, because it is filed under consumables.
The kitchen time
The one nobody costs properly. A delivery order interrupts a line that is set up for plated service. It arrives on a screen with no relationship to your table timings, it needs different packing, and it lands during your peak because that is when everybody orders.
Two or three delivery tickets during a Saturday rush do not just cost you their own margin — they slow down the dining room, which is where your good margin is.
The cannibalisation question
The one that decides whether the whole channel is worth it: is delivery bringing you new customers, or is it selling to your neighbours at a discount for people who would have walked in anyway?
There is no universal answer, but there is a test. Look at what happened to your walk-in covers in the months after you went live. If they dropped by roughly what delivery added, you have converted full-margin customers into commission-paying ones.
Why you must never add the two together
This is the mistake that makes the whole picture unreadable, and most reporting tools make it for you.
A dining-room receipt might be four covers. A delivery order is one, maybe two. Adding them into one number produces a figure describing neither business — and, worse, it hides the thing you most need to see.
Delivery orders are lowest when you are busiest. Most operators pause delivery when the room is slammed. So on your best nights, the delivery number drops. Any tool that reads that as demand falling has learned exactly the wrong lesson, and it will under-prep your busiest services from then on.
Keep them apart. Forecast them apart. They have different peaks, different weather sensitivity — rain is good for delivery and bad for a terrace — and different economics.
What to do about it
Work out your true delivery margin, including packaging and a realistic share of labour. Plenty of venues find one platform is fine and another is not.
Price the channel differently. Most operators now run a delivery menu priced 10–15% above the dining room. Customers understand it.
Decide when you turn it off, in advance, rather than in a panic at half past eight. A rule — "off when we are over X covers" — is easier to hold than a judgement call mid-service.
Watch the two series separately. BentoPlan forecasts your dining room and your delivery orders as two different things, precisely because they behave differently, and it knows that a night with no delivery orders may well have been your busiest of the month.
Delivery is not a mistake. But it is a different business bolted onto yours, and it should be measured like one.