Delivery commissions for food entrepreneurs: traditional method vs Masterestaurant method

The commission you don't see is eating you alive

Between 25% and 35% of the sale price: that is what platforms charge. Rappi runs 28-33%; Uber Eats, 25-30%; iFood, 27-32% depending on the visibility plan contracted in 2025-2026. Run the numbers on a $7 USD dish with a $2.10 food cost: $4.90 gross margin in-house, yet after a 29% commission the app remits $4.97 and margin lands at $2.87, down 41%. The trap is not the percentage in the contract but the cumulative effect across the whole product mix. Launch the channel without recalibrating prices and menu, and the platform is no sales ally. It is a silent partner quietly keeping nearly a third of every dollar the customer pays. More than 12,000 dark kitchens operated across Latin America in 2025, per Euromonitor, and the count keeps climbing in 2026: setup runs $8,000-$15,000 USD against $60,000-$80,000 for a traditional location.

Dark kitchens: the model that amplifies commission risk in 2026

The risk is structural. When 100% of revenue flows through platforms and kitchen rent sits near $800 USD a month, one mismanaged commission turns a projected 8% EBITDA into a real 4% loss, because there is no dining-room cushion. What if the app raises its rate two points? A pure dark kitchen has nowhere to absorb it: pass the increase to the on-screen price or swallow it whole. Hence the 2026 trend among sophisticated operators: multi-brand kitchens, two or three concepts on one line, spreading exposure and negotiating rates on consolidated volume. Copying the dine-in menu to the platform without touching a single price: that pattern repeats across Colombia, Mexico and Peru, and Diego F. Parra finds it in nearly every kitchen he reviews. What follows is mathematically inevitable. Real cost per dish climbs to 45%-50% once the commission lands, cash flow compresses, and delivery ends up subsidizing its own losses with working capital.

The mistake Diego F. Parra sees over and over in Latin American kitchens

The Masterestaurant method starts backwards: at a 30% commission and a 30% real food cost target after commission, the dish cannot cost more than 21% of the counter price, and that math defines which products earn a spot on the delivery menu and which stay off it. Plain cost algebra, applied before the first product ever goes live. Nothing more. Delivery Price = Dish Cost ÷ (Target Food Cost × (1 − Platform Commission)). At a $2.10 USD cost, a 30% target and a 29% commission, the app price must be $10.05 USD, 41% above the $7.00 dine-in price. Does that differential scare customers off? Elasticity studies from 2025 in the Colombian and Mexican markets say no: willingness to pay on delivery runs 30%-45% higher than in-house for fast food and comfort food. Even so, raising prices overnight is not the move. What works is building a value proposition for the channel that actually earns the difference: portions and packaging consistent with the price the customer sees on the screen, presentation included.

Commission negotiation: 2026 trends and real levers

Negotiating fees stopped being taboo in 2026. Rappi Ads can cut the effective commission from 32% to 26% when ROAS clears 3.5x; Uber Eats launched 'Socios Plus' in Colombia and Mexico, at 22%-24% for restaurants above 400 monthly orders with a rating of 4.7 or better; iFood applies 12%-18% in Brazil and its expansion markets for kitchens inside its own production centers. We crossed the three schemes and the clause that matters shows up: the commission locks for 12 months, with a review that triggers only if order volume drops more than 20%. Read that fine print twice. The operator's one real window of negotiating power opens before signing the contract, never after growing inside it. Fewer SKUs, more margin. We resisted that idea for a while: cutting the menu sounds like selling less. Masterestaurant's analysis of 47 dark kitchen ventures in Bogotá, Medellín and Mexico City (2024-2025) says otherwise: under 12 delivery SKUs meant an average ticket 18% higher and an effective food cost 6 percentage points lower than menus above 25 SKUs.

Product mix and delivery menu: fewer SKUs, more profit

We saw it too in the Peruvian dark kitchen from the case on this page: 38 products on Rappi and $58 of net cash on $5,200 sold; with 9 products, an 11% margin inside 60 days. A short menu standardizes, cuts waste to 3%-5% and concentrates demand where the margin lives. The 2026 trend has a name: the 'high-rotation delivery menu'. 8-12 products built for the channel, food cost between 18% and 22%, packaging that still looks right after 30 minutes on the road, names that rank inside the app's search box. WhatsApp Business with a catalog, proprietary payment links and per-shift couriers bring the order back to a direct channel at zero commission, or at a controlled logistics cost of roughly $1.50-$1.80 USD per delivery. Between 20% and 35% of delivery volume migrates to that channel within the first 90 days among restaurants that built it with Masterestaurant, lifting average gross margin from 12%-15% to 22%-26%.

The direct channel as a strategic counterweight to platforms

Here sits the delivery paradox: to sell without the platform you first have to grow inside it, because the app brings the new customer and the direct channel keeps them. Active retention resolves the tension: a $0.60-$0.90 USD discount for ordering direct, WhatsApp replies under 3 minutes, and a confirmation that includes the estimated delivery time. Miss one of the three and the customer goes back to the app, even at a higher price. Your delivery break-even is not the dining room's; calculate it separately or you are flying blind. At a 29% commission, a dark kitchen with $800 USD in monthly fixed costs (kitchen rent, utilities, a basic staff of 1.5 people) needs roughly $5,100 USD in monthly platform sales to cover itself, assuming 28% food cost and no internal ads. Add $300 USD of platform advertising and the threshold rises to $6,200 USD.

Break-even in delivery: the number the entrepreneur never calculates

That figure, not order count or average ticket, is the KPI Diego F. Parra puts on the dashboard of every venture he advises. Below the threshold, more volume means more loss; above it, every extra dollar falls straight to margin. The math does not forgive. It does not lie either. Absorbing the commission as an unavoidable discount and chasing volume: that is the traditional method. Selling more at negative margin only widens the hole, and each extra Rappi order sinks the operator a little deeper. The Masterestaurant method takes commission as a fixed input (say 30%) and engineers backwards: for a 30% effective food cost after commission, the dish must cost ≤21% of the dine-in price. That math decides which products enter the delivery menu. Channel-specific pricing is the pillar separating both methods. Masterestaurant calculates: Delivery Price = Dish Cost ÷ (Target Food Cost × (1 − commission)). A $1.74 dish with a 29% target and a 30% commission lists at $8.57.

Key differences between managing delivery commissions the traditional way vs the Masterestaurant way

The traditional method leaves it at $6, the dine-in price, and effective food cost jumps to 43% once the commission lands. The menus diverge too. Masterestaurant is surgical: 12 products at most, all under 30% food cost before commission, items that travel well over 20-30 minutes and carry a high average ticket. The traditional method pushes all 40-80 dine-in items into the app, dispersing production and multiplying waste. The proprietary channel is the structural difference. Masterestaurant builds it from day one (a WhatsApp Business catalog, a website with an order button, 0% commission) and uses paid platforms to acquire new customers, not to retain them. The traditional method lives on a single app, exposed to algorithm changes, fee hikes or exclusion.