Rappi delivery strategy: the traditional method against the Masterestaurant method

Commission isn't negotiated: you cost it before uploading the first dish

Rappi charges between 18% and 30% of order value depending on the plan, the exposure tier and who supplies the courier, and that range, published in its own commercial terms for partners, is the only honest starting point. The traditional method uploads the dining room menu at dining room prices and finds out at month close that a dish carrying 33% food cost, plus 26% commission, plus 1,400 pesos of packaging, delivered negative contribution margin on every single unit sold. The Masterestaurant method flips the order: set the channel price first, then decide which item goes in. On a 45,000-peso ticket that's nearly 12,000 pesos of commission somebody has to pay, and if you didn't put it into the price, you are paying it out of your own pocket, order by order, with a smile.

Channel pricing: 18% to 22% above the dining room price

Charging the same on Rappi as at the table is the costliest pricing mistake in the trade, and the market already settled it: 62% of mid-sized chains apply menu price inflation according to 2025 sector reports, with markups that in practice land between 18% and 22%. The counterargument from owners who keep prices flat is that customers compare and punish. They do compare, but they compare against other restaurants that also inflated. A burger at 32,000 in the room goes out at 38,500 in the app, the diner doesn't read it as expensive because delivery and packaging are services he is genuinely consuming, and you recover 6,500 pesos that used to evaporate. Channel pricing WINS outright: matching prices only works when a restaurant negotiated commission below 15%, and that doesn't exist for an independent operator. A long menu on an aggregator doesn't sell more: it sells eight dishes and drags sixty others that stretch your prep time, bloat inventory and sink your rating the day someone orders precisely the one that doesn't travel.

Seventy items versus fourteen to eighteen

Cutting down to 14-18 items lifts conversion because users decide in under 90 seconds inside the app, and the algorithm rewards whoever accepts short times and cancels rarely. The kitchen thanks you in cash: less dead mise en place, less waste, one single plating flow for takeaway. A long catalogue only makes sense running a dark kitchen with dedicated stations, a format that in Brazil already moves USD 5,702 million in the home segment according to Global Growth Insights. With one location and a kitchen shared with the dining room, the short menu WINS. Ranking third for «burger» inside your polygon brings you four to seven times more orders than ranking fifteenth, and that order isn't luck: five measurable variables decide it —conversion, accepted prep time, cancellation rate, rating and campaign participation—. Picture your kitchen swamped on a Friday, cancelling three out of every ten orders. The rating drops, the algorithm buries you in the listing, orders fall, you compensate by joining a 30% promotion to buy back visibility, and now you sell the same volume at half the margin.

What happens when the algorithm drops you to fifteenth?

That's the spiral that sinks restaurants which technically cook well. A short catalogue prevents it at the root, because it protects prep time, the variable the other four hang from.

Take the case that repeats most often in consulting work. A fast-food restaurant with a single site, 68 items loaded on Rappi, average ticket of 41,000 pesos and 380 orders a month: it billed 15.6 million in the channel, handed over 26% in commission (4.05 million), carried 33% average food cost and closed with channel contribution margin near 9%. After trimming to 16 items, raising channel price 20% and redesigning packaging for the four dishes that accounted for 71% of orders, the ticket rose to 47,500, orders grew to 470 on better placement, and channel contribution margin moved to 21%. Same site, same kitchen, same cook. What changed was the unit of analysis, not the talent.

The aggregator is the cheapest channel to acquire on and the priciest to sell on

Here sits the paradox worth resolving, because both statements hold at once: Rappi hands you qualified demand, with purchase intent already declared, at zero fixed cost —no billboard, no ad spend, no corner lease does that— and it is simultaneously the most expensive partner per unit sold you will ever have. The bridge is neither abandoning it nor surrendering to it. Treat it as a channel with its own product: separate catalogue, separate price, designed packaging, a weekly read of which item wins and which one drains. The global delivery market goes from USD 380,430 million in 2024 to USD 618,360 million in 2030 at a 9.0% CAGR according to Grand View Research; demand is there and it will grow. The question isn't whether to be present, it's with which menu. According to Diego F. Parra, consultant and founder of Masterestaurant, uploading the dining room menu untouched to an aggregator is a kitchen decision disguised as a commercial one, and that confusion explains most delivery channels that bleed without the owner knowing where.

Why Diego F. Parra calls this a kitchen decision in disguise?

Nobody sat down to measure whether the risotto survives 22 minutes inside cardboard, or whether its 34% food cost withstands a 26% commission: it went up because it was on the printed menu.

At Masterestaurant the rule runs the other way, and it's uncomfortable: a dish enters the channel only if it survives three tests —transport, time and costing at aggregator price—. Some 76% of US operators believe technology gives them a competitive edge according to the National Restaurant Association; the edge lives in the selection criteria, not in the app. If you run a single location, with the kitchen already built and stations shared with the dining room, pick the Masterestaurant method without hesitating: 14 to 18 items, channel price 18% to 22% higher and packaging designed for the four that concentrate the bulk of orders.

What to pick according to your operating profile

If you operate a dark kitchen with a station dedicated to the channel, exclusive staff and sustained volume above 900 monthly orders, you can hold a 25 to 30 item catalogue and exploit second brands, that multi-brand model which in India already represents USD 4,500 million according to Global Growth Insights. And if your negotiated commission tops 28% while food cost never drops below 36%, the catalogue isn't the problem: that dish shouldn't exist in any channel. This week open the last 60 days of item-level sales and delete everything that doesn't reach 2% of orders. CHANNEL PRICING. The traditional method charges the table price on Rappi and assumes commission comes out of margin; the Masterestaurant method sets channel pricing 18% to 22% higher, which is precisely what the US industry calls menu price inflation and which 62% of mid-sized chains already apply according to 2025 sector reporting.

The four differences that move the cash

This is not gouging the guest: it charges for a service carrying real intermediation, packaging and transport costs that the dining-room guest never consumes. CATALOGUE SELECTION. Seventy references on an aggregator do not sell seventy dishes; they sell eight, while the other sixty-two stretch preparation time, inflate inventory and drag the rating down whenever somebody orders the one item that travels badly. Trimming to 14-18 references lifts listing conversion because the user decides in under 40 seconds and punishes choice paralysis. PREPARATION TIME AS A RANKING LEVER. Delivery aggregators penalise the placement of any restaurant that accepts an order and then runs late. Cutting the committed time from 30 to 15 minutes can move a location from twelfth to fourth inside its polygon, and that jump multiplies orders four to seven times without a peso of advertising spend. VIRTUAL BRAND AS A SECOND DOOR. The same kitchen, staff and raw material can feed a virtual restaurant with its own concept aimed at a daypart where your main brand does not compete.

The four differences that move the cash — in practice

Building a dark kitchen from scratch runs between 45,000 and 120,000 dollars according to foodtech industry reporting; launching a virtual brand on a kitchen that already exists costs menu design, packaging and photography. WEEKLY READING, NOT MONTHLY. The monthly settlement arrives once the month is already lost. A weekly per-reference board tells you on Tuesday that the prawn pasta has run three weeks at negative margin with a 9% cancellation rate, and you pull it on Wednesday.