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

Finding 1 — What is the real operating margin of a delivery-only dark kitchen?

The median delivery-channel operating margin was 6.8% across the 312 dark kitchens Masterestaurant audited between 2023 and 2026, not the 15-20% the foodtech pitch promises.

The range runs from 1.4% to 13.2% depending on segment and country. The arithmetic rules: the aggregator commission takes 22% to 30% of the ticket, real food cost sits at 30-34%, packaging adds 4-9% and delivery waste another 3%. Added together, those lines leave a thin margin that dries up at the first dip in orders. Diego F. Parra reviewed it kitchen by kitchen: operations that bill well and have no cash at month-end. The 6.8% figure is not opinion; it is the median of real P&Ls, not a summary of outside reports the platforms never publish. Costing a dish for delivery with dining-room logic is the #1 error that bankrupts dark kitchens, because it ignores 7 to 12 margin points that only exist outside the dining room.

Finding 2 — Why costing a delivery dish with the dining-room recipe kills the kitchen

In-house, the server defends the ticket and there is no packaging or aggregator commission. In delivery, packaging adds 4-9% to the dish cost, waste from badly assembled or rejected orders hits 3%, and commission eats the margin the floor used to protect. Masterestaurant re-costs every dish with its specific packaging, its expected waste and the commission already subtracted before setting price. A dish that runs 34% food cost in-house can end at 44-46% effective cost in delivery if no one recalculates it. The aggregator menu price has to absorb that gap, or the cash register pays for it. In delivery, margin per order decides the cash far more than the number of tickets. The math is direct: 94 orders at a 6.8% operating margin leave less money than 58 orders at 17.4%. The pure-aggregator model optimizes volume because the algorithm rewards rotation, but each extra order at minimum margin barely covers its own packaging and waste.

Finding 3 — Volume versus margin: why 58 orders beat 94

The Masterestaurant method optimizes margin per order: it lifts the margin from 6.8% to double digits by re-costing the dish and moving demand to the owned channel. Across the 312 audits, kitchens that prioritized ticket count over margin showed high sales and flat or negative profit. Chasing volume on a channel where the aggregator charges 22-30% per order is running faster on a treadmill that carries you backward. A ticket that leaves no margin is not a sale; it is cost in disguise. Owning the customer is what separates the dark kitchen that survives from the one that closes in 18 months. Without an owned channel, the kitchen is hostage to the aggregator commission —28.4% effective in the audited median— and to its visibility algorithm, which can hide you from one month to the next. With 60% of sales through the owned channel (WhatsApp, web, direct order), the weighted effective commission drops from 28.4% to 11.2%: 17 points that go straight to operating profit.

Finding 4 — Owning the customer: 17 commission points that go to the cash register

Among the 312 dark kitchens, those that passed 50% owned sales showed margins of 12-14% against the 3-5% of the pure-aggregator ones. Diego F. Parra insists the aggregator is an acquisition channel, not a partner: it exists so people discover you, not so you live off it. Building an owned base is the only lever that moves 17 points at once. Packaging and delivery waste add 7 to 12 points to the dish cost and are the two lines dark kitchens discover too late. Packaging represented 4-9% of the dish cost in the audits: boxes that survive 30 minutes of transport, seals, cutlery and bags cost more than a dish that only crosses the kitchen. Waste from badly assembled, spilled or rejected orders hit 3% of sales. In-house that waste does not exist: the dish is served and eaten within meters. Masterestaurant costs packaging by SKU, not as a diluted general expense, because a $6 burger with $0.80 of packaging loses 13 points that an average hides.

Finding 5 — Packaging and waste: the two lines nobody costs in time

The method's hard rule: if packaging and waste are not in the dish spec, the food cost you see is fiction. Re-costing a dish for delivery starts from the dining-room food cost and adds the three channel-specific lines before setting the aggregator price. First, real packaging per SKU: 4-9% of cost depending on the dish. Second, expected channel waste: 3% on that dish from rejected or badly assembled orders. Third, the effective aggregator commission subtracted from the menu price, not from cost: 22-30% of the ticket. A dish at 32% food cost in-house can climb to 42-46% effective cost in delivery; the aggregator price must rise to hold the target margin, not copy the dining room. Masterestaurant sets a distinct price per channel in 100% of the kitchens it audits. The expensive error is a single price for in-house and delivery: one of the two always loses, and in delivery the loser is the cash register.

Finding 6 — What the dark kitchens that reached double digits did differently

The dark kitchens that moved from 6.8% to double-digit margins did three concrete things, measurable across the 312 audits. One: they re-costed each dish with packaging and waste in the recipe spec, recovering 7-12 margin points that dining-room costing hid. Two: they moved demand to the owned channel until passing 50% of sales, cutting the effective commission from 28.4% to 11.2%. Three: they stopped chasing volume and pruned negative-margin dishes, even the most ordered ones. The inverse pattern is just as clear: those stuck in pure-aggregator, dining-room specs and ticket chasing averaged 3-5% margin and dry cash at close. There is no magic lever: it is arithmetic applied dish by dish. Diego F. Parra sums it in one action: re-cost your best-selling dish this week with its packaging and its commission, and you will see whether it truly leaves cash.

Finding 7 — Where they truly differ

The pure-aggregator model optimizes order VOLUME; the Masterestaurant method optimizes MARGIN per order. In delivery, 94 orders at 6.8% leave less cash than 58 orders at 17.4%: the math favors margin, not ticket count. Re-costing the plate for delivery is the dividing line. Pure-aggregator cooks with the dining-room spec sheet and discovers too late that packaging and waste ate 7-12 margin points. Masterestaurant re-costs each plate with its packaging, expected waste and commission already subtracted. Customer ownership decides survival. With no own channel, the dark kitchen is hostage to the aggregator's commission and its visibility algorithm. With 60% own-channel sales, the effective commission drops from 28.4% to 11.2% —17 points that go straight to cash.