Rappi delivery strategy: the traditional method against the Masterestaurant method

Chapter 1 — The delivery channel is not an extension of the dining room: it is another business unit

Cost Rappi as a separate business unit, with its own contribution margin and its own break-even point, or the channel will eat the dining room's profit without ever showing up in your monthly books. ActiveMenus puts the total effective cost of third-party delivery between 30% and 40% per order once you add commission, packaging, transport waste, co-funded promotions and the extra work of a kitchen dispatching two flows at once. The U.S. Bureau of Labor Statistics places sector labor cost between 25% and 35% of revenue, and that percentage does not fall because the order arrived through an app. Add both bands and you will see that a dish sold at the same menu price on the platform can leave the kitchen with no margin at all. The accounting error always comes before the pricing error. Because market concentration turned the commission into a private tax on sales, and a tax is not negotiated: it is costed.

Chapter 2 — Why did a dominant aggregator's commission stop being negotiable?

Grand View Research reported that Asia-Pacific holds more than 41.0% of the global online food delivery market (2024), and Statista documented iFood capturing 87% of e-food bookings in Brazil (2024).

In the United States the pattern repeats sharply: Earnest Analytics measured a 60.7% share for DoorDash at the close of 2024, against 26.1% for Uber Eats and barely 6.3% for Grubhub. When three platforms split a market that way, the individual restaurant loses every lever at the negotiating table. The practical consequence is uncomfortable but clear: stop fighting for the commission point and fight for the food cost point, the packaging and the assortment, which do depend on you. Growing in orders while losing profit is a perfectly possible outcome, and today it is the most common ending for a badly costed digital channel. DoorDash grew 20% year over year in Marketplace GMV during 2024, according to its full-year financial results, and that figure runs in the press as proof that delivery is doing well.

Chapter 3 — The aggregator's GMV climbs while the restaurant's EBITDA falls

It is doing well for the platform. For the operator, with an effective cost of 30% to 40% per order documented by ActiveMenus, every additional sales point moved onto the app subtracts contribution margin while adding kitchen load. A restaurant routing 25% of its revenue through an aggregator without repricing is financing a third party's growth with its own working capital. And the app dashboard will never show you that subtraction, because it does not calculate it. The same commission produces four different effects depending on the size of the business, and that is precisely why a single recipe fails. Below 500 thousand dollars a year, with labor cost at the ceiling of the 25% to 35% band reported by the U.S. Bureau of Labor Statistics, the channel only survives with a reduced menu and a differentiated app price. Between 500 thousand and 1 million, routing 25% of sales through an aggregator at the 30% to 40% effective cost measured by ActiveMenus means giving up 30 to 40 thousand dollars of margin a year if you do not reprice.

Chapter 4 — How the arithmetic changes by annual revenue band

Above 1 million, dedicating a production line to the channel becomes genuinely feasible. Over 5 million, the operator negotiates packaging and promotional co-funding with volume. And above 10 million, the decision is no longer price but a proprietary virtual brand. A celebrity restaurant or a large-format themed venue above 5 million dollars a year faces a problem the small operator does not have: its product does not travel, but its brand does. The signature dish that justifies a high ticket in the dining room arrives cold, badly plated and stripped of ritual, and the guest blames the brand, not the courier. With the 30% to 40% effective cost per order calculated by ActiveMenus applied to a high ticket, the absolute value surrendered per transaction is enormous. These formats must build a deliberately different parallel menu, with product engineered to survive thirty minutes in transport, and use the channel as customer acquisition toward the dining room.

Chapter 5 — The high end loses too: the celebrity-chef restaurant and its own costs

Charging for delivery of a dish that degrades is the most expensive way to destroy a positioning that took years to earn. The app assortment must be designed from scratch around transport and margin criteria, because copying the full menu is the costliest mistake in the entire digital operation. Select the dishes that survive the trip and still leave contribution margin after the 30% to 40% effective cost per order that ActiveMenus documents, and pull the rest without sentimentality. The logic of scale is proven in other markets: Momentum Works measured Grab holding 53.9% of Southeast Asian food delivery (2024) and Foodpanda contributing 15.8% of regional GMV, worth 2.7 billion dollars; in both cases the profitable operators run short menus. First decide what you will NOT sell through the app. Then set the price. Doing it the other way around is what produces eighty-item menus that dispatch badly and earn worse.

Chapter 6 — The channel's break-even and the instruments that measure it

Calculate how many daily orders the channel needs to cover its incremental cost before you accept a single co-funded promotion, because without that figure every discount decision is a blind leap. Diego F. Parra insists on one point that organizes the whole Masterestaurant framework: the commission enters the plate costing BEFORE the list price is set, never afterward as an administrative expense on the biweekly settlement. With labor cost at 25% to 35% of revenue per the U.S. Bureau of Labor Statistics and delivery effective cost at 30% to 40% per order per ActiveMenus, channel break-even demands volume and assortment discipline at the same time. Masterestaurant S.A.S. built the instruments that translate the aggregator dashboard into verifiable unit economics. Start this week: split Rappi revenue out in your P&L and calculate contribution margin per app dish. A food-service SME growing 40% in orders while losing EBITDA does not have a private management problem: it has the exact symptom of the productivity gap that ECLAC and CAF flag in their regional agenda.

Chapter 7 — Growth without productivity: why this is a development policy matter

The formal employment that restaurant sustains, the subject of SDG 8, depends on the digital channel being profitable rather than merely voluminous, and the digital infrastructure enabling it, the subject of SDG 9, only produces development when the operator has the analytical capacity to read it. The concentration data makes it worse: with iFood at 87% of e-food bookings in Brazil per Statista (2024) and Asia-Pacific above 41.0% of the global market per Grand View Research, the information asymmetry between platform and restaurant is structural. For SATE Institute, closing that asymmetry through unit economics training is a development intervention, not luxury consulting. The first difference is accounting, and it moves the most money. The traditional method treats Rappi commission as an administrative expense showing up in the biweekly settlement; the Masterestaurant framework treats it as direct variable channel cost entering dish costing BEFORE the list price is set.

Chapter 8 — The four differences that move margin

With total effective third-party delivery cost between 30% and 40% per order according to ActiveMenus, a dish selling at 10 dollars in the dining room with 30% food cost leaves 7 dollars of gross margin; that same dish at that same price on the app leaves between 3 and 4. A restaurant in the 500 thousand to 1 million USD annual band moving 25% of revenue through the aggregator gives away 30 to 40 thousand dollars a year simply by not repricing. The second is assortment. Nobody has shown that an 80-SKU menu outsells a 24-SKU menu on an app, and platform shortlist behavior points the other way: the user decides within the first six to eight visible items. Menu engineering applied to the digital channel is not the same discipline as the dining room, because the dominant variable shifts from table turnover to product integrity in transit.

Chapter 9 — The four differences that move margin — in practice

A risotto that stars on your table service may be your worst decision on Rappi. Filtering the menu by logistics viability lifts average ticket and lowers complaint rate, and both land in next month's margin. Third comes architecture. Once the digital channel covers its own break-even and demand clusters in polygons where delivery time punishes conversion, the right answer may be a ghost kitchen or a virtual brand rather than a full second location. Grand View Research estimated Asia-Pacific at 48.0% of cloud kitchen market revenue (2025), and Global Growth Insights measured Europe contributing 18.79% of the global dark kitchen market (2024): the format is mature across two continents, with CapEx between a fifth and a third of a full-service site. But opening a dark kitchen from scratch to escape a money-losing channel only multiplies the loss. The fourth difference is customer ownership, and it decides your company's value five years out.

Chapter 10 — The four differences that move margin — key points

With Zomato and Swiggy together above 95% of India's online delivery according to Business of Apps (2025), and Meituan plus Ele.me beyond 90% of orders in China according to Mordor Intelligence (2025), the regional lesson reads clearly: where the aggregator concentrates, the restaurant that never built an owned base ends up without bargaining power. I got this wrong for years, recommending volume maximization on the app before building the direct channel; the correct order runs the other way. First you set a dependence ceiling, then you grow volume.