Rappi delivery strategy: real prices, commissions and the margin nobody is measuring

The steakhouse billing 41 million and losing 3.2

A steakhouse in Medellín was billing 41 million pesos a month through Rappi and losing 3.2 million on that channel, and the owner had no way of knowing because his P&L blended dining room and delivery into a single sales line. He had uploaded the dining-room menu exactly as it stood, with no channel markup, a 27% commission and 1,900 pesos of packaging per order that never entered the plate cost. Once we split the two channels using the Masterestaurant method, delivery gross margin landed at 9% against 41% for the table, and that 9% did not even cover the kitchen labor assigned to peak shifts. The big number kept growing month after month while cash got tighter, and that contradiction —more sales, less money— is the classic symptom of a channel with no price of its own. Between 15% and 22% over the dining-room price, depending on the commission you negotiated and what your packaging actually costs.

How much should you mark up your Rappi price to avoid losing money?

The arithmetic leaves no room for argument: a dish priced at 38,000 pesos in the dining room with a 30% food cost leaves 11,400 in raw material and 26,600 in gross margin;

that same dish on Rappi, at a 26% commission, hands the restaurant 28,120, and after subtracting 1,900 of packaging and 11,400 of ingredients you are left with 14,820 pesos. The gap against the table is 11,780 pesos per plate, nearly half the margin. An 18% channel markup lifts the price to 44,840, leaves 33,182 after commission and 19,882 of margin once packaging and ingredients come out, and the customer accepts it because he is paying for convenience, not for food. Charging the same in both channels means funding the aggregator's operation with your kitchen's margin. As of September 2026, in Colombia and Mexico the channel markups that hold margin fall into three ranges, and each one covers something different.

What each price range includes on the digital menu?

The low range, 8% to 12%, only works with a commission negotiated below 20% or for brands using Rappi as an acquisition window: it covers part of the commission and nothing else, because packaging stays outside.

The middle range, 15% to 18%, is where 70% of the menus we audit live: it absorbs commissions of 24% to 27%, packaging of 1,500 to 2,200 pesos per order and transport shrinkage, which runs near 2% on saucy dishes. The high range, 20% to 25%, applies to kitchens with food cost above 30%, low tickets under 25,000 pesos, or categories where thermal packaging costs more than 3,000 pesos per order. One single price for the whole menu is the most expensive decision anyone makes in this channel. Five variables decide where your markup lands, and you should measure them before touching the menu. Commission carries the most obvious weight: every commission point translates into roughly a point and a half of required markup, so going from 24% to 28% forces you to raise the price about 6 points more.

Five factors that move your channel price

Packaging comes second and gets ignored the most: 1,900 pesos on a 30,000 ticket eats 6.3 margin points, while on an 80,000 ticket it barely costs 2.4. Third, the cancellation and refund rate, which in sushi and ice cream categories runs above 3% and has to be provisioned. Fourth, menu mix: a dish at 26% food cost tolerates the channel, one at 34% does not tolerate it even with a markup. And fifth, restaurant-funded promotions, that two-for-one the aggregator proposes and that comes straight out of your pocket. The house rule does not change with the channel, and that is why it works: maximum 32% food cost per plate, and that 32% is the ceiling, never the target. For Rappi, Diego F. Parra tightens the requirement to 28% or less BEFORE commission, because the aggregator's cut behaves like a second variable cost that no supplier will ever discount for you.

Cost by channel, not cost in general

Payroll, rent and utilities do not load onto the plate —they belong to break-even— but packaging does, since it is a direct input of that order and it disappears with it. At the Medellín steakhouse, applying that discipline meant pulling eleven dishes with food cost between 33% and 39% off the digital menu, 22% of the catalog, and the average ticket climbed from 34,000 to 47,000 pesos because the categories that genuinely withstand the channel were the ones left standing. The gain did not come from selling more, it came from no longer selling what cost money to dispatch. According to the National Restaurant Association, in its State of the Restaurant Industry 2026, 42% of operators report that delivery grew as a share of their sales while their operating margin stayed flat or declined.

Why margin falls while sales climb?

That paradox has a structural explanation and it is not bad luck:

the digital channel grows on top of demand that already existed, so a good part of delivery is not a new customer but a dining-room customer who migrated, and every migration swaps a 40% margin for a 14% one. The same association measured in 2025 that nearly 75% of U.S. restaurant traffic is already off-premise, which means the phenomenon will not reverse itself while you wait. In Latin America, Rappi closed August 2024 with 35 million active users and 150 million downloads, per its own operating report, and Statista projects 147 million delivery users in the region by 2026. The channel is inevitable, the single price is not. Negotiate with volume, temporary exclusivity and data, in that order, and bring your own numbers to the table.

How to negotiate commission and optimize the catalog?

Aggregators move commission between 4 and 8 points when a restaurant clears 800 monthly orders or accepts a visibility campaign;

dropping from 27% to 22% at the Medellín steakhouse was worth 2.05 million pesos a month on identical billing, more than any promotion delivered. Second move: cut the digital catalog down to 18 to 24 references, because a 60-dish menu inside an app inflates decision time and cancellations. Third, build combos with consolidated food cost under 27% instead of flat discounts, since a combo protects margin while a discount gives it away. Fourth, charge packaging as a visible line whenever the platform allows it. And review your channel price every quarter: commissions shift, packaging shifts too, and a digital menu priced a year ago was costed against another reality. With a 30% food cost and 1,900 pesos of packaging, that 38,000-peso dish would hand you 25,840 pesos, and after 11,400 of ingredients and the packaging you would keep 12,540 of gross margin, 33% against the dining room's 70%.

What would happen if Rappi raised your commission to 32% tomorrow?

And if you never ran the channel-pricing exercise, your only way out would be raising prices 10 points at once in the middle of high season, which is the worst possible way to raise a price.

Here sits the channel's real tension: the aggregator brings you demand you never captured and simultaneously strips away the margin that would make it worth having, and the bridge is neither walking away nor standing still, it is treating every platform as a point of sale with its own financial structure. That is the Masterestaurant doctrine: commission, packaging, shrinkage, cancellation and menu engineering measured channel by channel. Open your P&L today, split the sales line in two and calculate delivery margin on its own. CHANNEL PRICE, not a single price.

The four differences that decide whether the channel funds you or bleeds you

A dish priced at 38,000 pesos in the dining room with a 30% food cost leaves 11,400 in ingredients and 26,600 in gross margin; that same dish on Rappi at 26% commission returns 28,120 to the restaurant, minus 1,900 in packaging, minus 11,400 in ingredients: 14,820 left against 26,600 at the table. An 18% channel markup closes almost all of that gap and the customer absorbs it, because convenience is what they came for. Diego F. Parra puts it bluntly: charging the same in both channels means funding the aggregator's operation with your kitchen's margin. COST BY CHANNEL, not by restaurant. The Masterestaurant house rule does not bend for delivery: 32% food cost per dish is the maximum, and 32% is a ceiling, never a target. Payroll, rent and utilities never load onto the dish — they belong to break-even.

The four differences that decide whether the channel funds you or bleeds you — in practice

What DOES enter the digital channel recipe is the full packaging, and this is where most calculations collapse: container, lid, bag, security seal, napkin and cutlery run between 600 and 2,400 pesos per order depending on cuisine, and that weight comes out of the dish margin, not out of thin air. SHORT, CURATED CATALOG. Publishing 74 dishes on Rappi does not multiply sales: it multiplies refunds, one-star reviews and waste. A catalog of 18 to 24 SKUs chosen for transport resistance and margin — what menu engineering calls the channel stars — is what holds a rating above 4.6, the mark from which the algorithm starts pushing you up the list. Risotto does not travel. Breaded chicken goes soft in twelve minutes. Premium packaging will not fix that; not publishing it will. PHYSICAL MENU AND QR MENU, each in its own role.

The four differences that decide whether the channel funds you or bleeds you — key points

I got this wrong for years recommending full digitization: the physical menu governs the dining room experience — service pace, menu narrative, suggestive selling, hospitality — while the QR is the complement, covering delivery, accessibility, same-day price changes and analytics on what guests actually browse. The correct verdict is BOTH, with separate jobs. A restaurant that killed its physical menu lost dessert upsell and saw average ticket slide; the delivery channel menu is a third thing, with its own prices, and it lives apart.