A ghost kitchen (also dark kitchen) is a kitchen without a dining room, producing exclusively for delivery and pickup — as a standalone concept or as a virtual second brand run from an existing restaurant kitchen. The appeal: revenue without floor space and service staff. The trap: platform commissions of up to 30% hit a business that already calculates on a knife edge. Whether delivery carries itself is decided by the contribution margin per order — not by gut feeling.
Enter order value and costs — what really sticks per order appears instantly.
Kitchen staff and fixed costs still come out of the contribution margin — below a ~30% CM share, platform delivery is rarely sustainably profitable. The comparison value shows: every order shifted to your own webshop is hard cash.
| Model | Cost logic | When sensible |
|---|---|---|
| Platform with their riders | Commission often 25–30% incl. delivery | Reach entry, test phase, extra utilisation in weak hours |
| Platform as marketplace only (own riders) | Commission ~13–17% + own rider costs | High order density in a small radius |
| Own webshop + pickup/delivery | Payment/shop costs ~3–6% | Converting regulars — the highest-margin model |
As incremental business in weak hours and as a marketing channel: often yes. As the core business: only with a delivery-optimised menu, tight cost of goods and higher delivery prices (a 5–15% markup on the in-house menu is market standard and accepted).
Without alcohol service and a dining room, trade registration plus food-hygiene duties (HACCP, instructions) usually suffice in Germany. Clarify building-law use and rider parking with the municipality beforehand.
Yes, common practice. Important: operator transparency on the platforms (imprint!), clean allergen data per brand, and don't overrun the kitchen's capacity.
Rule of thumb: below €20 the logistics eat the profit. Levers: minimum order value, bundles/menus and active upselling in the ordering flow (drinks, desserts — they sell remarkably well digitally).