Delivery only restaurants: when they pay off

What is a dark kitchen on delivery apps?

A dark kitchen on delivery apps is a production kitchen with no dining room and no walk-in customers that sells only through ordering platforms, under one or several virtual brands.

What defines it is the channel, not the premises, because all of its demand comes in through an app screen and depends on how that app's algorithm ranks it against the other kitchens in the area.

That is why it should be read as a complete business model, with its own break-even point, and not as a cheap branch: the owner trades the rent of a busy street corner for a commission the platform charges on every sale, and that swap shapes almost everything else, from the short menu to the size of the crew and the way you find out whether the month closed in the black.

The four components that hold the operation up

Four pieces hold a dark kitchen up, and if one fails the others cannot make up for it. The first is the kitchen itself, designed to produce on a line with few stations and a short menu that comes together in minutes, because every extra minute at the pass is an order that arrives lukewarm and a review that pushes you down in the app.

The second is the set of virtual brands, which share ingredients and equipment so inventory turns fast. Packaging comes next, and in this model it is part of the product: it goes into the direct cost of each order with the same weight as the protein.

The fourth, the one almost nobody draws into the business plan, is the relationship with the platform (contract, commission, paid visibility and payout terms), which decides how much money actually reaches your account and in which week.

How to calculate whether a dark kitchen makes money?

A dark kitchen makes money when the contribution margin per order, multiplied by the month's orders, exceeds its fixed costs, and that math gets done before you sign a lease.

For example, if your average app ticket is 260 pesos, the dish costs you 78 (a 30 % food cost, under the 32 % ceiling the method sets as the MAXIMUM), packaging 12 and your contract's commission another 78, you keep 92 pesos per order. With 95,000 pesos a month in rent, payroll and utilities, break-even sits around 1,033 orders, a little over 34 a day, every day of the month.

If the kitchen ships 25 a day today, the conversation is no longer about marketing. It is about price, menu or contract, in that order, because raising the ticket moves the math faster than chasing orders with promotions.

The mistakes people make reading the term

The most frequent confusion is putting the platform commission inside food cost, which makes the dish look expensive when the real problem sits in the channel. The commission is a variable selling cost and belongs in the contribution margin, while food cost measures only what it takes to produce the dish.

Another common slip is spreading rent and payroll across the dishes, something we forbid at Masterestaurant because those expenses belong to break-even and, once spread, they hide how many orders are missing. Gross app sales also get mistaken for available cash: the platform pays out after deducting commission, promotions and adjustments, and a kitchen can bill a lot and still run out of cash at month end.

And the most expensive mistake is believing the app brings you your own customers, when in fact it lends them to you, and tomorrow they will see your competitor on the same screen.

How much real demand is there for eating off-premises?

There is plenty of demand for the format, and the strongest figure is in limited service: 83 % of its traffic in the United States was takeout, drive-thru or delivery in 2024, according to the National Restaurant Association.

In the same report, the association notes that 58 % of those operators say off-premises sales weigh more in their total sales than in 2019, which describes a settled habit rather than a passing fad. On the diner side, MediaPost, reviewing the 2025 study, reports that 37 % of American adults order restaurant delivery at least once a week. But demand is not margin.

That frequent customer is also the one who compares most inside the app, and tends to reward whoever arrives fast at the price shown first, not necessarily whoever cooks best.

The paradox: competing with restaurants that have a dining room

A dark kitchen competes in the app against restaurants that already pay their rent with the dining room, and that asymmetry decides who survives a price war. In US full service, off-premises traffic reached 30 % in 2024 (National Restaurant Association, 2025), and for that restaurant every app order is incremental revenue that only has to cover direct cost. You, on the other hand, need every order to also pay for a share of the whole kitchen.

Think about what would happen if your neighbor with a dining room launches a two-for-one from Monday to Thursday: they absorb it with dining room margin, you lose volume, the algorithm drops you for selling less, and by the third month your break-even point is further away than on opening day. The way out is to sell what the dine-in restaurant cannot deliver well, not to match its promotions.

When it is worth it, according to the Masterestaurant method?

It is worth it when your numbers survive the worst commission contract you could sign and still cover break-even with the volume the area already proves.

Diego F. Parra sets it out at Masterestaurant in a fixed order: first the dish costing with its food cost ceiling, then contribution margin by channel, and the location decision last. He admits that for years he did it backwards, starting with the location and fitting the menu to the rent.

Market size alone decides nothing either, because Ken Research estimates the delivery and dark kitchen market in Spain at approx. The market is huge, and a badly costed kitchen still loses money inside it. Before you sign, calculate how many daily orders your break-even point demands and compare that with what your best brand sells today.

Dark kitchen vs restaurant: where the money is made and lost

The model's paradox resolves with simple arithmetic: a dark kitchen saves the dining room, the storefront and most of the floor staff, and in exchange pays the platform a cut of EVERY sale, so the rent it doesn't pay the landlord it pays the app, scaled to volume. It wins with high, steady volume and loses on a few expensive orders.

Diego F. Parra frames it as a Masterestaurant rule: commission is a variable channel cost and lives in contribution margin, never in food cost, while labor, rent and utilities go to break-even, never to the plate. Demand is real. The National Restaurant Association's 2025 report, as covered by MediaPost, puts nearly 75% of US restaurant traffic off-premises, and 37% of US adults order restaurant delivery at least once a week.

Dark kitchen vs restaurant: where the money is made and lost — in practice

But demand is not margin, and people ordering on their phones does not mean they order from you at a price that leaves something after commission. Run it the other way. What happens if the platform raises its cut a few points, or ties visibility to a permanent promotion?

In the example above, each point on a $30 ticket removes $0.30; five points remove $1.50, contribution drops to $9 and break-even jumps from about 1,714 to 2,000 orders a month, with nothing changed in your kitchen. A single-channel operator has no leverage in that conversation; one who built direct repeat business does.