How to increase restaurant sales on Rappi: what the channel really costs and the pricing method that keeps margin

What does it actually cost you to sell a dish on Rappi?

A $38,000 order on Rappi leaves the restaurant somewhere between $22,400 and $25,800 before food cost even enters the picture, and that is the number you must price your digital menu against.

As of September 2026 in Colombia the subtraction runs like this: channel commission between 18 % and 30 % depending on the plan you signed —third-party platform commissions run 15 % to 30 % nominal, with an effective cost reaching 30 %-40 % per order, according to Food On Demand 2026—, payment gateway between 2.8 % and 3.5 %, packaging from $1,600 to $2,900 per order, and co-funded discounts that rarely fall below 5 %. Add it up: 32 to 41 percentage points of the sale price vanish before the first potato. The Chapinero owner billed 41 million in March and lost 2.1 million in operating profit to that arithmetic, not to weak sales. Three channel price tiers exist and each one buys something different.

What each digital menu price tier buys you?

The first, salon price with no markup, is what roughly 70 % of independent restaurants use when they join delivery, and it only works if your dine-in food cost already sits below 24 %, which almost nobody achieves.

The second, a markup of 8 % to 15 %, covers packaging and gateway but leaves commission unfunded; treat it as a sixty-day bridge while you measure the channel's real ticket. The third, a markup of 18 % to 25 % over salon, is the only one that absorbs those 32 to 41 structural points and returns a contribution margin comparable to the dining room. A dish priced at $28,000 on the table gets published at $33,500 on the app: the delivery customer pays for convenience and does not have your physical menu sitting beside them for comparison. Push only the dishes whose food cost stays under 30 % INSIDE Rappi, commission and packaging already deducted, and in practice eight or nine survive out of a forty-item menu.

The eight dishes: why a short menu earns more

That pruning does four things a full menu cannot: it concentrates purchasing into fewer inputs and gives you leverage with the supplier, it cuts peak-hour waste by 3 to 6 points because no mise en place sits dying while it waits for an order that shows up twice a week, it raises assembly speed —and thirty seconds saved per order across an eighty-order peak is forty minutes of line cook you stop paying for—, and it reduces cancellations from prep time, which on Rappi punish the merchant's ranking. A long menu in the delivery channel is a labor cost dressed up as variety. Five variables decide whether your markup should be 18 % or 25 %, and you want them measured before you publish. Your commercial plan with Rappi weighs 6 to 12 points: the merchant who negotiates volume pays near 18 % while the one who signs up self-service pays as much as 30 %.

The five factors that move your channel price

Average ticket moves another 4 to 7 points, since packaging is a fixed cost per order and $2,300 on a $22,000 ticket hurts three times more than on a $65,000 one. Average distance to the customer affects your cancellation rate and product temperature. Menu mix defines the channel's loaded food cost. And co-funded promotions, almost always split with the restaurant, take another 3 to 9 points. Rappi operates in 9 countries and 350 cities with over 500,000 registered partners (Rappi, operating report 2024): you are one among many, and the negotiation rests on your volume, not your charm. Negotiating with Rappi works when you put volume data on the table with a date attached, never through an open-ended call. Ask for the commission review once you have three consecutive months above 400 monthly orders, and bring your month-over-month growth printed; the genuinely negotiable range, as of September 2026, moves between 4 and 7 points against the entry rate.

How to negotiate commission and switch off what does not pay?

Second move: turn on the automatic shutdown rule per dish, which means reviewing contribution margin by item every Monday and deactivating in the app any dish that drops below 30 % loaded food cost two weeks running.

Third, negotiate promotional co-funding on specific dishes rather than across the whole menu. In Colombia, Rappi reports more than 30,000 partner merchants and 7 million orders per month (La República / Rappi, 2024); that demand already exists and you do not need to buy it with a discount. I got this wrong for years: I used to recommend entering the aggregators with the full menu to test demand. Testing demand in a channel that keeps a quarter of the sale is not testing, it is subsidizing.

The mistake that turns growth into decapitalization

Walk the scenario all the way out: if you double your Rappi orders at salon prices with a negative margin of $1,400 per order, going from 600 to 1,200 monthly orders does not improve your cash, it multiplies the loss from 840,000 to 1.68 million a month, and it raises your kitchen payroll on top because the peak now demands one more cook. The channel's paradox reads like this: selling more there becomes profitable only after the price has been corrected, and never before. Commission is not a marketing expense justified by exposure; it is a direct REDUCTION of the sale price and it belongs in the dish spec sheet, right next to the avocado. The criterion we apply at Masterestaurant, and the one Diego F. Parra holds his delivery clients to, has three conditions and none of them is optional. First: the dish spec sheet includes commission, gateway and packaging BEFORE the digital menu price is set, with a target food cost of 26 % to 30 % already loaded.

The Masterestaurant criterion for entering the channel

Second: a short menu of eight to ten items, assemblable in under four minutes, reviewed every thirty days against real margin. Third: a channel price 18 % to 25 % above salon, published from day one rather than patched in later, because raising prices on a digital menu that already carries reviews costs you reputation. With those three in place, that same Chapinero restaurant closed July 2026 with 37 million billed —four million less than March— and 4.6 million in operating profit. Less revenue, more cash. Open your Rappi report for the last thirty days and calculate one single figure: contribution margin per order, which is published price minus commission, minus gateway, minus packaging, minus food cost. If that figure does not clear $9,000 on an average ticket of $38,000, you do not have a marketing problem, you have a pricing problem.

What to look at on your dashboard next Monday?

Some context helps you place the business:

the drive-thru share of QSR orders fell from 83 % in 2020 to 65 % in 2025 (Intouch Insight 2025), the cloud kitchen market is projected at USD 248.10 billion by 2035 (Precedence Research 2025), and iFood closed 2024 with 55 million active customers (iFood 2024). Delivery is not going away. What decides who survives is whether the channel price was set in the kitchen or copied from the dining room. Commission is not a marketing expense, it is a direct reduction of the selling price, which is why it belongs inside the recipe card before you set the digital menu price, never after the sale happens. Selling more on Rappi with a dine-in menu produces the opposite of what the owner wants: each additional order consumes cash, so growth accelerates the problem instead of diluting it. A short menu outperforms a full one because it concentrates purchasing, cuts waste and raises assembly speed, and assembly speed on the channel is a real labor cost almost nobody books.

The differences that decide whether the channel pays

A higher channel price does not scare the delivery customer away: that customer weighs convenience against the cost of leaving home, not against a physical menu they never see. Packaging for a hot dish traveling 4.5 kilometers is not dining-room packaging, and that $900 to $2,100 gap per order gets paid either in margin or in reputation. A discount co-funded with the aggregator and one funded by the restaurant alone look identical to the customer and are opposites on the P&L; confusing them is the costliest error in the category.