Rappi delivery strategy: before and after you read the real numbers

What does selling through Rappi actually cost in 2026?

Selling through Rappi costs between 18% and 30% of order value plus VAT on that commission, as of September 2026, and where you land has less to do with your negotiating skill than with the commercial plan you signed and the city you operate in.

The Chapinero grill house that opened this case was billing 41 million pesos a month through the app, paying 27% commission, co-funding promotions and absorbing packaging and cancellations: it kept 1.9 million in profit, while the dining room, selling less, left nine times more. The arithmetic is simple and brutal. If your dish runs 32% in raw material and the platform takes 27%, you have committed 59 of every 100 pesos before touching payroll, rent, utilities or packaging, and decent delivery packaging weighs between 3% and 6% of the ticket.

What each commission tier includes: 18%, 24%, 30%?

The low tier, 18% to 21%, almost always means you supply the courier or the customer picks up, and it is the only bracket where a 32% food cost still leaves margin without raising prices.

Between 22% and 26% sits the standard affiliated-restaurant deal with platform fleet included, a service fee charged to the diner and organic presence in search; there you already need 15% of digital-menu markup to get back where you started. The high tier, 27% to 30% plus co-funded campaigns, buys visibility: a category banner, featured placement, discount coupons where the platform pays part and you pay the rest. Nobody tells you at signing that the extra 3% of exposure gets paid with volume that also pays commission, so the added weight loops back on itself.

One price across channels is the sector's costliest mistake

Putting the same number on the physical menu and the digital one hands over a silent discount equal to the full commission, financed entirely by the owner. A 38,000-peso dish in the dining room leaves roughly 26,600 in gross margin at 30% food cost; the same dish at 38,000 on Rappi with 25% commission leaves 17,100. You lost 9,500 pesos per order without anyone charging you to your face. The measurable fix is pricing 15% to 22% higher on the digital menu: that dish goes out at 45,000, commission takes 11,250, and margin returns to 22,350. This deceives nobody, it recognizes that the channel carries a different cost structure. Colombian law does not forbid differential pricing by channel, and the three big apps allow it in their menu configurator.

Five factors that move your real rate

Five variables move the effective cost of the channel, and only two of them belong to the platform. Contracted commission first, with a 12-point spread between the best and worst plan. Who supplies the fleet, second: taking it on yourself cuts 6 to 9 points of commission, though it adds a fixed courier cost that only pays off above 25 daily deliveries. Third, packaging weight, which climbs to 6% of the ticket on a hot dish with sauce on the side when a dining room pays zero. Fourth, shrinkage from cancellations and returned orders, running 2% to 4% of gross sales in kitchens with no dispatch-time control. Fifth, promotional co-funding, the most treacherous of all, because it shows up as a discount rather than a commission and ends up adding another 4 to 8 points during aggressive campaign months.

Why commission belongs at the top of the P&L?

Commission is a variable cost of sale, not an administrative expense, and that is why it sits above contribution margin rather than down among the utility bills.

Every peso arriving through Rappi is born with 18 to 30 cents already committed; the same peso arriving through your Google Maps listing and your own payment gateway is born with 3.5 cents committed. That gap is structural, not commercial, and it decides whether the channel funds the operation or bleeds it. Diego F. Parra insists at Masterestaurant on opening a channel-level P&L from month one: channel gross sales, minus commission, minus the real food cost of those dishes, minus packaging, minus shrinkage, equals channel contribution margin. Run that exercise honestly and half of all restaurants discover their star channel contributes less profit than the dessert counter.

How to negotiate and optimize without fighting the platform?

Real negotiation does not happen on the phone with the account executive, it happens on your menu. Start by cutting the digital menu to 25 or 30 dishes that travel best and leave the highest contribution margin;

a 60-dish delivery menu multiplies shrinkage and sells no more. Build combos with drink and side, which lift average ticket between 18% and 24% without touching the commission percentage. Ask in writing for the own-fleet plan and compare it against your real courier cost over 60 days before deciding anything. Use packaging as a repurchase vehicle: a printed code worth 10% off on your own channel converts between 4% and 9% of orders. And audit the weekly settlement against your POS, because differences from cancelled orders go unrecoverable after 30 days if nobody claims them.

What if the owner walks away from Rappi entirely?

Leaving the three big apps means giving up neighborhood discovery demand, and that is a cost almost nobody calculates before deciding. Follow it all the way through:

pull your listing and you lose between 30% and 45% of delivery volume in the first 90 days, recovering only half of that if you already held a customer database with phone numbers and a direct-ordering habit. Those who never built one simply lose the sale. The aggregator is your competitor and your acquisition channel at once, charging you for the customer it takes from you, and the way out is not picking a side. It is pricing differently per channel, treating Rappi as acquisition spend with a cost per new customer that you actually measure, and pushing repurchase toward your own 3.5% gateway.

The size of the market you are negotiating against

Meal delivery in Latin America will pass 39 billion dollars by 2027 according to Statista (2024), and that figure explains why platforms have little incentive to lower commissions on their own. In Brazil, iFood hit 100 million orders in a single month during August 2024 and runs with more than 380,000 partner establishments across over 1,500 cities, per its own 2024 institutional data. In Spain, Glovo holds close to 31% of the market and Just Eat 26%, according to Ken Research (2025). A neighborhood restaurant's individual bargaining power against that scale is essentially nil, and accepting it is the healthy starting point: you will not move the commission, you will move your price, your menu and your direct repurchase rate. That part is yours to run.

The differences that actually move cash

Commission is not an administrative expense, it is a variable cost of sale, which is why it belongs at the top of the P&L and not the bottom: every peso arriving through Rappi is born with 18 to 30 cents already committed, while the same peso arriving through your Maps listing is born with 3.5 cents committed to the gateway. That structural gap, not sales effort, decides whether the channel funds the operation or bleeds it. Single pricing across channels is the most expensive mistake among restaurants that have spent three or four years on delivery aggregators. Putting the same number on the printed card and the digital one feels fair, but it hands the app customer a silent discount equal to the full commission, financed by the owner. A 60-item menu on the app does not sell more: it sells worse. Rappi ranks by conversion, and a long catalog dilutes the algorithm's signal, stretches prep time and multiplies waste on slow-moving inventory. Operators who cut down to 12-18 references lift listing conversion by 20% to 35% in the first quarter.

The differences that actually move cash — in practice

Packaging is the line nobody budgets and everybody feels: between 1,100 and 2,800 pesos per order depending on category, which on a 32,000-peso ticket lands between 3.4% and 8.7% of the sale. Stack that on commission and you are already at 38% before buying a single gram of protein. Direct orders do not compete with Rappi on price, they compete on repeat purchase. A customer who discovers you on the app and returns through your Google listing costs you once and pays you many times; the one who always returns through the app costs you every time. All the local SEO work, the 5★ reviews and the geo-targeted ads exist to move that percentage. A dark kitchen with no brand of its own is an elegant financial trap: rent goes down, dependence goes up. A ghost kitchen billing 95% through aggregators owns no commercial asset, it holds a contract-manufacturing deal with a variable fee. The virtual restaurant only earns its keep when 25% or more of its orders arrive with no intermediary.