Kitchen operations · Dark kitchens
On the apps they are five restaurants. Behind the tablets: one brigade, one prep list, one extraction system. The disciplines that make a brand family work - and the shared infrastructure that can take it all down at once.
The model
On the delivery apps they look like five different restaurants - a wing brand, a smash-burger brand, a katsu brand, a dessert brand, maybe a 'healthy bowls' brand for the lunch algorithm. Behind the tablets they are one cookline, one brigade and one extraction system in an industrial unit. The multi-brand dark kitchen is a listings strategy wearing a chef's jacket: each brand is a menu engineered to capture a different search on the platforms, while the kitchen underneath is engineered for overlap.
The overlap is the whole game. A viable brand family shares a protein and prep base - the same fried chicken becomes wings under one banner, burgers under another, loaded fries under a third - shares equipment, and shares the same station skills, so one chef can cook across all of them without retraining. What multiplies is the order volume and the menu surface area on the apps. What must not multiply is the prep list, the stock holding or the equipment line, because the unit has no space for it and the margins have no room for it.
Running it
Multi-brand kitchens that work prep as a single operation: one production list of shared components, portioned and labelled once, drawn down by every brand. The brands differentiate at the finish - sauce, seasoning, build and packaging - in the last ninety seconds of the ticket. The moment a brand demands its own dedicated prep stream, it has to earn its place with volume, or it is quietly killing the economics of the other four.
Tickets from three apps land simultaneously, so the line cannot be organised brand-by-brand. It is organised by process - fry, grill, build, pack - with every brand's dishes mapped onto the same stations. The build-and-pack station carries the complexity: packaging by brand, sauces by brand, and the discipline of never letting Brand A's branding touch Brand B's bag, because on the apps those are different restaurants with different ratings to lose.
However many storefronts the tablets show, the food business is one registration with one hygiene rating pathway and one HACCP plan - and one EHO inspection that covers the lot. Allergen control gets harder with every menu added: the same fryer serving five brands means a cross-contact matrix that spans all of them, kept current every time any brand tweaks a recipe. A brand family is only as strong as the single kitchen's paperwork.
Five brands on one line means five revenue streams hanging off one extraction system, one gas interlock and - in multi-operator sites - ductwork shared with neighbouring units. A delivery kitchen frying twelve hours a day loads grease at the heavy end of the TR19 Grease bands, where indicative cleaning intervals run to around three months, and when the extract chokes or the interlock drops the gas, every brand goes dark at once. The same shared-infrastructure logic drives the utility bills, which is why communal energy costs in dark kitchen sites deserve as much scrutiny as the rent.
Scaling honestly
The economics reward adding brands - each one is a nearly-free storefront - right up until the kitchen's constraints bite. The signals are physical: the fry station running at capacity through the whole evening peak, pack times stretching past the platforms' prep-time targets, the walk-in too full to rotate stock properly, the line too hot because the extraction was never sized for this duty. A brand added past that point does not add revenue; it degrades the ratings of every brand already trading, and on the platforms ratings are the shelf position.
The honest operators run the arithmetic per station before every launch: what does this menu add to fry, grill, build and pack at peak, and what does it share? If the answer is mostly shared components and idle station capacity, launch it. If it is a new protein, a new piece of equipment and another degree over the line - the kitchen has already answered.
Questions
As many as share its prep base and fit inside its station capacity at peak - commonly three to five. The limit is physical: when the fry station saturates, pack times slip past platform targets, or the extraction cannot hold the line temperature, an added brand degrades every existing brand's ratings rather than adding revenue.
The food business operating the kitchen registers with the local authority and carries the inspection outcome - however many storefronts appear on the apps. One HACCP plan, one allergen matrix and one EHO visit cover the whole brand family, which is why the paperwork has to span every menu at once.
Delivery-led kitchens tend to sit at the heavy end of TR19 Grease's indicative bands - around three-month intervals for 12-to-16-hour cooking days - because frying-heavy menus load grease fast. The interval should be set by a measured survey of the actual system, and in shared or multi-operator sites, agreed across everyone using the ductwork.
Phoenix Duct Clean · by the numbers
We clean and certificate dark-kitchen extraction to TR19 Grease - overnight, around your platform hours, with the paperwork shared sites and insurers ask for.