Masterestaurant Dark Kitchen Index LATAM 2026: the real net margin on pure delivery hovers near 6.8%

Finding 1 — How much margin does a dark kitchen actually leave in LATAM?

A well-run dark kitchen in LATAM leaves a net margin near 6.8% on pure delivery, not the 20-25% the foodtech pitch promises.

The reason is cash arithmetic: the aggregator charges between 15% and 35% of every ticket, and with an average ticket of USD 20-35 per order in 2025 (Lightspeed 2025), each commission point takes a good half-dollar before the kitchen sees a cent. Diego F. Parra repeats it in every diagnosis: a ghost kitchen removes the dining room but not the law of contribution margin. Global delivery grows from USD 380.43 billion in 2024 to USD 618.36 billion by 2030, a 9.0% CAGR (Grand View Research 2025), so demand exists. What does not exist is the generous margin many assume. This business is won in the subtraction, not in the headline. The aggregator replaces the dining-room rent with a commission that scales with every order, and that swap is the accounting trap Masterestaurant flags first.

Finding 2 — The aggregator is your new landlord, and it charges per order

A dining room costs the same at 100 or 300 covers; a 15% to 35% commission grows with each sale and never drops. That is why cities like San Francisco capped commissions at 15% (Restaurant Dive 2020) and New York set a permanent cap of 15% per delivery plus 5% for other services (Restaurant Business 2023): they saw the free percentage was suffocating operators. LATAM has no such cap. With Rappi reporting 35 million active users and 150 million downloads by August 2024 (Rappi 2024), the aggregator holds the power to set the rate. Diego F. Parra says it plainly: whoever fails to negotiate the commission does not set their margin, the platform sets it for them. Pure delivery's contribution margin is more fragile because there are no dining-room drinks or cross-sell to dilute the food cost. In a dining room, a soda or a high-margin dessert offsets the main dish; in a dark kitchen each order travels alone, without that cushion.

Finding 3 — Why pure delivery's contribution margin is more fragile

So a food cost tolerated at 30% in the room weighs double once the aggregator's 15-35% has already eaten the ticket. With an average online ticket of USD 24 per order in Spain (Ken Research 2025) and USD 20-35 in the US (Lightspeed 2025), there is no room for waste. Diego F. Parra insists: the food cost ≤ 32% per dish rule is the ceiling, not the target, and in pure delivery that discipline is the difference between a 6.8% net margin and operating in the red. Every point counts double here. A dark kitchen reaches break-even through order volume, not by raising the ticket, because the kitchen and staff are fixed costs covered only by many USD 20-35 orders (Lightspeed 2025). Market appetite is not the problem: in 2024 there were nearly 3 billion online delivery users worldwide (Statista 2024), with ~1.84 billion in Asia alone and ~355 million in Europe (Statista 2024).

Finding 4 — Break-even is reached by volume, not by ticket

The challenge is capturing enough of them each day to amortize the fixed base. Meituan closed 2024 with more than 770 million annual transacting users (Meituan 2024), proof that scale is reachable, but also that this is a traffic game. Masterestaurant models it this way: without a sustained floor of daily orders, the dark kitchen never crosses into profit, no matter how tuned the menu is. The physical restaurant uses delivery as a growth lever; the dark kitchen bets everything on it, and there lies the structural risk. The National Restaurant Association describes off-premise as a traffic driver: nearly 75% of US restaurant traffic is now off-premise (NRA 2025). For the physical venue, delivery adds on top of a dining base that keeps selling; for the dark kitchen it is the entire base. Statista counted more than 20,000 operating ghost-kitchen locations in the US in 2023 (Statista), a sign the model scales, but also that the competition for the same courier and the same customer is fierce.

Finding 5 — Delivery: a lever for the physical restaurant, the whole thesis for the dark kitchen

Diego F. Parra frames it bluntly: diversifying channels protects the cash; betting the company on a single aggregator concentrates the risk in someone else's hands. The physical venue has a net; the dark kitchen walks without one. A dark kitchen's menu engineering optimizes dishes that travel well and survive 25 minutes in transit, not a dining-room menu. A dish that goes soggy, loses texture, or arrives cold destroys the review, and in delivery the review is the only counter. So the design starts from the journey, not the plating. Behavioral data confirms it: 46% of US diners prefer third-party apps and order nearly 5 times a month (DoorDash via Restaurant Business 2024), so repeat business depends on the food arriving as in the photo. With Brazil's home dark-kitchen segment at USD 5.702 billion (Global Growth Insights) and India's at USD 12 billion (Global Growth Insights), the market rewards whoever masters packaging and the trip.

Finding 6 — Menu engineering: dishes that survive 25 minutes in transit

Masterestaurant sums it up: delivery margin is cooked, but it is lost in transit. Packaging is a variable cost that eats your margin if you fail to control it, because every delivery order pays for packaging the dining room never needs. Added to the aggregator's 15-35% commission and the food cost, it is the third bite out of a ticket that runs around USD 24 online in Spain (Ken Research 2025) and USD 20-35 in the US (Lightspeed 2025). Deliveroo reported GTV per order of GBP 27.6 in 2024, up just 5% (Deliveroo plc 2024): the ticket rises slowly, so savings must come from cost. Diego F. Parra insists on costing packaging as one more input of the dish, not as an invisible expense. In a business leaving a 6.8% net margin, a poorly negotiated package wipes out a full point of profit. Cash discipline is not optional: it separates the profitable dark kitchen from the one that fails with a full kitchen.

Finding 7 — How to armor the cash of a dark kitchen in LATAM 2026

To armor the cash of a dark kitchen in LATAM in 2026 you must negotiate the commission, control the food cost, and diversify the ordering channel, in that order. The commission is the biggest lever: where there is no legal cap like New York's 15% (Restaurant Business 2023), the operator must trade volume for rate. Food cost ≤ 32% per dish is the ceiling (Masterestaurant rule), and packaging is costed inside the dish. The third front is the channel: 46% of diners prefer third-party apps (DoorDash 2024), but direct ordering from your own site avoids the full commission. With the US QSR market at US$ 289.68 billion in 2024 (Business Research Insights 2024) and delivery heading to USD 618.36 billion by 2030 (Grand View Research 2025), the pie grows. Diego F. Parra always closes the same way: measure your contribution margin per dish today, before adding one more order.

Finding 8 — What changes between the two models (the consultant's read)

A dark kitchen does not save costs: it moves them. It swaps the salon rent for the aggregator commission, which scales with every order instead of being fixed. Pure delivery's contribution margin is more fragile: without dine-in drinks or cross-sell, every point of food cost and commission weighs double. A dark kitchen reaches break-even by volume, not by ticket: it takes many USD 20-35 orders to cover the kitchen and the team. The physical restaurant uses delivery as a growth lever (the NRA describes it as a traffic driver); the dark kitchen makes it the whole business thesis, with the risk that implies. Menu engineering differs: the dark kitchen optimizes dishes that travel well and survive 25 minutes in transit; the physical venue optimizes the table hero dish.