Masterestaurant Dark Kitchen Index LATAM 2026: delivery leaves a 6.8% operating margin, not the 20% foodtech promises

Finding 1 — What operating margin does a LATAM dark kitchen actually earn in 2026?

A well-run LATAM dark kitchen closes 2026 with a single-digit operating margin —between 5% and 8%, around 6.8% at the midpoint— not the 15-20% the foodtech pitch suggests.

The reason is pure arithmetic. Aggregator commission plans charge 15%, 25% or 30% of the ticket depending on service tier (CloudKitchens, 2024), and that toll comes out BEFORE food cost and packaging, not after. The market itself does grow: global online food delivery moves USD 1.51 trillion in 2026 with a 6.24% CAGR through 2031 (Statista, 2026). But market size is not operator margin. Diego F. Parra says it plainly: foodtech sells market growth; the kitchen's cash register lives on what reaches EBITDA after the toll, and that is one digit, not two. The pie is huge; the slice that lands in the kitchen is thin. The aggregator commission is a dark kitchen's biggest margin drain because it is charged on the gross ticket and BEFORE food cost and packaging.

Finding 2 — The aggregator commission is the biggest margin drain

DoorDash, to cite the public reference, offers restaurants plans of 15%, 25% and 30% (CloudKitchens, 2024); in delivery-only mode that range is paid on every single order, no exceptions. Compare it to a physical restaurant: dine-in sales pay no commission and margin is defended face to face. At Masterestaurant we have seen it again and again —a kitchen celebrating a high average ticket without noticing that 25 points of it left through the channel before buying a single ingredient. With food cost at 30% and packaging at 4%, a 25% commission plan leaves just 41 cents of every dollar for payroll, rent and profit. The unit economics does not forgive that math. A physical restaurant keeps dine-in sales commission-free, while a delivery-only dark kitchen hands 15 to 30 points of its ticket to the channel on every order (CloudKitchens, 2024). That is a structural difference, not a matter of operating efficiency.

Finding 3 — Why does the physical restaurant keep margin the dark kitchen hands to the channel?

A restaurant with a dining room collects 100% of the ticket when the guest eats at the table; the ghost kitchen never sees that scenario because it has no table.

Its entire revenue runs through a third party that takes its cut up front. That is why comparing the two models by average ticket misleads: you must compare what lands in the register after the toll. The ghost kitchen market will keep growing —Research and Markets projects USD 142.5 billion by 2029— but growing volume without fixing that channel leak scales the problem, it does not solve it. Direct ordering is the largest and worst-used margin-recovery lever in dark kitchens: 58% of customers prefer to order through the restaurant's own app or website (NCR Voyix, 2024), and yet most ghost kitchens live 100% inside the aggregator. Every order moved from the third-party channel to the owned channel recovers 15 to 30 points of commission at once (CloudKitchens, 2024), falling straight to EBITDA.

Finding 4 — Direct ordering: the margin lever almost everyone leaves on the table

In a model with a 6.8% operating margin, shifting just a third of sales to the direct channel can double net profit without selling a single extra dish. The lever is not in marketing or in volume; it is in the channel mix. The operator who treats the aggregator as customer acquisition —not as their only sales channel— is the one who survives 2026. Scaling a dark kitchen without first fixing the unit economics multiplies losses instead of gains, even though capital deploys faster than in a physical restaurant. The global dark kitchen market is projected at USD 171.3 billion by 2033 (Global Growth Insights) and cloud kitchens at USD 203.72 billion by 2033 (Grand View Research); that size attracts capital that rewards volume. But volume is an amplifier: it multiplies what you already have. If each order loses margin to the channel commission, opening five more kitchens simply produces five times that leak.

Finding 5 — Does scaling a dark kitchen solve the problem or multiply it?

Diego F. Parra sums it up with cash-register logic: first fix the costing sheet and the channel mix in one unit, then replicate.

Scaling a broken model is the most expensive way to discover it was broken. Delivery market growth does not translate into operator margin growth, and confusing the two is the costliest reading error in foodtech. Statista projects Mexico's online delivery at USD 18.27 billion by 2029 (Statista, 2024) and the global market at USD 1.79 trillion by 2028 (Statista), while Uber Eats gross bookings already reached about USD 74.6 billion in 2024 (Uber, Form 8-K SEC). These are enormous figures, and they describe the size of the channel, not the profitability of the kitchen selling through it. The foodtech pitch uses pie size to justify valuations; the kitchen owner must measure how much of that pie lands in their EBITDA. This analysis is a synthesis of serious public sources with a consultant's reading: Masterestaurant did not audit a sample of its own, and each figure belongs to the cited source.

Finding 6 — Where is margin defended in each model?

In a physical restaurant margin is defended in the dining room; in a dark kitchen it is defended in the costing sheet and the channel mix —that is how Diego F.

Parra sums it up after two decades inside both the kitchen and the cash register. That line is not a slogan: it is the operating criterion that separates the profitable ghost kitchen from the one that fails while scaling. In the dining room, the server recovers margin through upselling and experience; in delivery-only that spring does not exist, so the only defense is disciplined costing —food cost below 32%, controlled packaging— and moving every possible order to the owned channel, where 58% of customers already prefer to order (NCR Voyix, 2024). Some 76% of US operators believe technology gives them a competitive edge (National Restaurant Association, 2024); in dark kitchens that edge only materializes if the technology cuts the channel commission, not if it merely raises volume.

Finding 7 — The differences that decide the margin

The aggregator commission (15-30%, CloudKitchens 2024) is a dark kitchen's biggest margin sink: it comes off BEFORE food cost and packaging, not after. Brick-and-mortar keeps dine-in sales commission-free; the delivery-only dark kitchen hands 15-30 points of its ticket to the channel on every order. The 58% of customers who prefer direct ordering (NCR Voyix 2024) is the margin-recovery lever most dark kitchens leave on the table. The dark kitchen scales capital faster (global market to USD 171.3B in 2033, Global Growth Insights), but scaling volume without fixing the unit economics scales losses. Diego F. Parra puts it plainly: in brick-and-mortar you defend margin in the dining room; in the dark kitchen you defend it in the costing sheet and the channel mix.