Few phrases carry as much assumption as “dark kitchen.” People picture a delivery-only unit on an industrial estate, a single app-native brand and riders waiting outside. It’s a real thing, and for a certain kind of business it’s the right description.

For a large part of the market, though, it’s the wrong word entirely, and is steering a lot of larger operators away from the right business decisions.. Because when a multi-site restaurant group or a scaling caterer takes dedicated production space, they’re sending a signal to their team, customers and stakeholders. That signal is that quality comes first, achieving this by doing things in house, specialising and digging deep into your value proposition. Value built around investment in your own space so your business owns the end to end manufacturing of your products. That’s a factory and there’s nothing dark about it. 

What “dark kitchen” came to mean

The phrase, along with “ghost kitchen” and “cloud kitchen,” came out of the food delivery boom during Covid. It came to mean a specific model: production with no storefront, built to fulfil delivery orders, often a single concept, usually associated with startups, apps and the gig economy.

That model exists and it’s legitimate. But the term has carried all of that baggage with it, impermanence, delivery dependence, a slightly speculative, startup feel, and then been stretched to cover any production that happens out of public view. So a multi-site restaurant running an audited central production operation ends up lumped in, with a one-brand delivery startup. 

What larger operators are actually doing

Large brands are using production and prep kitchens to make their own products. Rather than outsource or buy in, companies are building their own spaces and dedicated teams to make their products better. This vertical integration is a product of quality and cost control.

A multi-site restaurant group centralising its prep is protecting consistency across its restaurants, rather than outsourcing stocks, dressings or chips, they are developing their own kitchens to self supply these high volume, speciality ingredients themselves. Not only does this improve product quality, but it also allows teams to move from R&D to roll out, completely flexibly, improving product range all while improving gross margin. 

A caterer scaling into a production unit is building capacity. A central kitchen means a single team can service multiple contracts from one site. Again, improving quality control, while creating efficiencies. Specialised kitchens can invest in equipment and talented chefs, that might dig into a contract caterer’s bottom line if they were replicating that set up in every office or premises that they served. 

Neither type of operator is a dark kitchen. It’s central production. These operators have their own words for it, central production unit, commissary kitchen, production kitchen, and they describe what’s actually happening: making product at scale, consistently, as the operational base of a growing business. These aren’t delivery fulfillment nodes. They’re production infrastructure.

Why the word matters

Words frame decisions. Call a production facility a “dark kitchen” and you import the whole frame: something impermanent, delivery-dependent, potentially speculative. That framing makes a significant, strategic decision: Where and how to produce your product at scale, seem like a punt on a trend. It pushes operators to under-rate it and to treat core production infrastructure as a temporary fix rather than the asset it is.

Call the same thing a central production unit, or production infrastructure, and you see it: a capability that drives consistency, efficiency and growth, a deliberate part of how the business scales. The same four walls, completely different decision, because the word changed what the operator thought they were deciding.

That’s why this isn’t just about tidiness of language. The wrong word leads to the wrong decision, or no decision at all, by making core infrastructure sound like a delivery-startup experiment.

The old model and the new one

There’s a simple way to hold the distinction.

“Dark kitchen” belongs to one idea: a place where delivery orders get made. It’s defined by the channel, delivery, and it’s a product of a particular moment in how food reached people.

“Production infrastructure” belongs to a different and older idea: the base a business produces from in order to scale, whatever its channels. Restaurants, retail, catering, delivery, wholesale, it supplies all of them, because it’s about production, not about any one way the food is sold. Growing businesses have always needed this. What’s new isn’t the need, it’s that the infrastructure can now be rented rather than built, which is a shift in how you get it, not in what it is.

Read: What the cloud did to computing is now happening to the kitchen

Read: Should we build a kitchen?

The word you’re using

So if you run a growing operation and you’re thinking about dedicated production space, it’s worth noticing the word you’re using, because it’s shaping how you’re thinking.

You’re almost certainly not looking for a dark kitchen. You’re looking for production infrastructure: a base to make your product at scale, consistently, that supplies every channel you sell through and grows with the business. That’s a strategic asset, not a delivery experiment, and it deserves to be thought about as one.

The industry’s casual vocabulary hasn’t caught up with what central production has become for operators running at scale. Call it what it is, and you’ll weigh the decision more accurately.

Thinking about production at scale? Talk to us about what that actually looks like →