Masterestaurant Dark Kitchen Index 2026: the real unit economics of delivery in Latin America

Finding 1 — Why does a dark kitchen fail in Latin America?

A dark kitchen in Latin America doesn't fail because of food cost, it fails because of the delivery cost stack. With aggregator commissions on the ticket, that bite hits the entirety of sales, not a residual margin.

I've seen it across dozens of operations: the operator celebrates a healthy food cost and keeps burning cash, because they forgot that the only cell on the scorecard that decides survival is the contribution margin after commission. The market pushes hard, with the region's meal delivery segment in steady expansion, but volume is not profitability. The mistake I see again and again is confusing gross sales with available cash. Every ticket that enters through the app arrives with a significant share already committed before touching the first onion.

Finding 2 — How much does the aggregator commission really weigh?

The aggregator commission is the cost that separates a viable model from one that burns cash.

The detail almost no one models correctly is that in a pure dark kitchen that commission hits the entirety of sales, because there is no dining room compensating with intermediary-free tickets. Delivery packaging and transport shrinkage usually take another bite out of that margin. From there comes the EBITDA, if any. That's why I insist to every operator I coach with the MASTERESTAURANT method: the delivery menu price is not the dining-room price, it must absorb the commission by design, not discover it at month-end close.

Finding 3 — Is platform concentration a business risk?

Platform concentration is territory risk, and in Latin America it's severe when a single app dominates the region's largest market. When a single aggregator dominates like that, commission pricing isn't negotiated, it's accepted.

This isn't trivia: it defines your real bargaining power, which is nearly zero. The pattern repeats outside the region —DoorDash closed 2024 with 60.7% of U.S. delivery (Earnest Analytics 2024) and other regional leaders show similarly concentrated shares elsewhere—, always with the same effect on the small operator. The lesson for the gastro-group leader is harsh: diversifying channels (own brand, WhatsApp, direct pickup) is not a marketing fad, it's the only lever that gives you back some control over a commission that, if you depend on a single player, is an imposed price.

Finding 4 — Does a healthy food cost guarantee profitability?

A healthy food cost is a necessary but insufficient condition for a dark kitchen's profitability.

The Masterestaurant rule sets food cost at ≤32% per dish as the maximum, and healthy hovers well below that ceiling, but without controlling the delivery stack that food cost doesn't save the margin. The arithmetic is relentless: food cost plus an aggregator commission already consumes most of the ticket before touching packaging, kitchen-station rent or payroll. And watch where you load each cost: payroll, rent and utilities are NOT charged to the dish, they go to the break-even point. The classic mistake I see is folding cloud kitchen rent into the dish cost and ending up with pricing that scares the customer without fixing the cash. First control commission and packaging; food cost, though it matters, is the second battle, not the first.

Finding 5 — Does the cloud kitchen model's scale translate into positive EBITDA?

The cloud kitchen model's scale is real but doesn't guarantee sustained positive EBITDA, with the global market set to keep growing fast through 2035 according to Precedence Research (2025).

That signals a fragmented market, heavy entry and exit, and few who achieve stable margin. They grow in number, not necessarily in profitability. Scale lowers cost per kitchen, yes, but it doesn't touch the aggregator commission, which remains a fixed percentage of the ticket. Scaling a model with negative contribution margin only multiplies the loss faster.

Finding 6 — How do you build an honest unit economics scorecard?

An honest dark kitchen unit economics scorecard is ordered around a single survival metric: contribution margin after commission, not sales nor food cost in isolation.

I build it like this with every group I advise: delivery menu price, minus food cost (≤32%, ideally lower), minus aggregator commission, minus packaging, minus transport shrinkage. What remains is the only figure that pays rent, kitchen and profit. In a Latin American meal delivery market that keeps expanding, the temptation is to chase volume; the disciplined operator chases margin per ticket. The track record of Diego F. Parra —restaurants supported across 43 countries over 20 years— confirms a pattern: whoever fails to model commission as a variable cost on every sale isn't doing unit economics, they're doing accounting illusion.

Finding 7 — What levers remain to defend delivery margin?

The levers to defend a dark kitchen's delivery margin are three, and none is cutting food cost below healthy levels. (Earnest Analytics 2024)—, pushing orders through own brand and WhatsApp where commission drops to zero.

Second: design the delivery menu with prices that absorb commission from the start, not dishes copied from the dining room. Third: attack packaging and transport shrinkage, that invisible bite that eats the margin. The consultant's verdict is simple: whoever controls the delivery stack survives; whoever only watches food cost closes with an impeccable kitchen and an empty till.

Finding 8 — The differences that decide unit economics

The aggregator commission is the cost that separates a viable model from a cash-burning one; in the dark kitchen it hits the entirety of sales, not a residual share. Platform concentration is territory risk: when a single app dominates delivery, commission pricing isn't negotiated, it's accepted. A healthy food cost (≤32% per dish) is necessary but insufficient: without controlling the delivery stack, even a sound food cost doesn't save the margin. The global cloud kitchen market will keep growing fast through 2035, according to Precedence Research (2025), but few operators achieve sustained positive EBITDA.