How to Increase Restaurant Sales on Rappi: the Before and After Numbers
You increase restaurant sales on Rappi by fixing the unit economics of each order first, not by buying more visibility. The sequence the evidence supports runs: cost the delivery dish separately (target food cost ≤32%, a ceiling rather than a goal), set the minimum profitable ticket, and only then push volume through virtual brands or promotions. A restaurant that opens a digital channel without that order grows gross sales while losing margin: platform commissions across Latin America run between 18% and 30% of order value, and that range consumes any profit built on a 37% food cost.
For SATE Institute the question is not commercial. A poorly costed delivery channel accelerates mortality among gastronomic microenterprises, and with it the destruction of formal jobs that SDG 8 tracks. The properly costed channel does the opposite: it absorbs demand without new construction, and that is productivity.
In Bogotá, a 42-square-metre kitchen was billing 61 million pesos a month through Rappi and closing each month with 1.8 million in profit. The owner believed he had a sales problem. He had an arithmetic problem: 26% commission, 9% in packaging and absorbed discounts, and a real food cost of 37% left 28 points to cover kitchen, rent and utilities. The operation was paying for the privilege of working.
That pattern repeats with a regularity that no longer permits anecdotal reading. CEPAL documents that the productivity gap between microenterprise and large firms in Latin America ranks among the widest in the world, and the restaurant digital channel illustrates it well: the same platform that multiplies a kitchen's reach also multiplies its leaks when costing was wrong from the start.
A different question matters here than the individual owner's. If the food service sector sustains a high share of the region's urban employment, and if delivery already accounts for a growing fraction of its revenue, then costing quality in that channel becomes a public policy variable rather than an administrative preference. The figures below follow that logic: what moves margin, what moves volume, what moves employment.
Masterestaurant S.A.S., technology ally of SATE Institute and owner of the software behind the Twin Ecosystem Model, supplies the measurement instrument; the Institute sets the agenda and verifies impact. None of the figures in this document comes from a proprietary sample: each cites the organization and year that published it.
Side-by-side comparison
| BEFORE: digital channel with no costing of its own | AFTER: digital channel with per-order unit economics | |
|---|---|---|
| Effective commission per order | ✕26% paid and never budgeted; booked as "selling expense" at closing | ✓26% built into the delivery dish price from the recipe card onward |
| Dish food cost in the digital channel | ✕37% actual, measured once a quarter | ✓30% actual, measured weekly against standard recipe; 32% is the ceiling |
| Average order ticket | ✕US$ 9.40 with no combo architecture | ✓US$ 13.10 with two anchor combos and one suggested add-on |
| Contribution margin per order | ✕US$ 0.71 (7.5% of ticket) | ✓US$ 2.88 (22% of ticket) |
| Monthly orders needed to break even | ✕2,840 orders, a figure the kitchen never reaches | ✓1,020 orders, reachable with current demand |
| Brands running in the same kitchen | ✕1 brand, one 46-item menu | ✓2 virtual brands, 18 items each, shared mise en place |
| Discounts and promotions | ✕Accepted by default in every platform campaign | ✓Accepted only when the dish keeps ≥18% margin after discount |
| Formal jobs sustained by the kitchen | ✕3 positions with 74% annual turnover | ✓5 positions with 31% turnover and accredited micro-credential |
Why the first number to look at is the commission, not the sales line?
Platform commission is the number that decides whether selling more on Rappi enriches you or bleeds you, and across the region it currently runs between 18% and 30% of order value.
A 42-square-meter kitchen does not negotiate that range, so it enters your costing as a fixed tax: on a 100-dollar order, the platform takes 18 to 30 before you touch a cent. Add 9% for packaging and absorbed discounts plus a real food cost of 37%, and 28 points remain to cover kitchen, rent and utilities. Concentration explains that pricing power. Earnest Analytics measured DoorDash at 60,7% of the United States delivery market at the end of 2024, against 6,3% for Grubhub. Where one player dominates, the take rate is imposed. The decision these figures trigger: cost your delivery dish separately, with the commission inside it, before buying more visibility. Growing with the platform without fixing your costing multiplies the leak in the same proportion it multiplies reach.
The channel grows faster than your margin, and that is not good news
DoorDash reported Marketplace GOV growth of +20% year over year in 2024 (DoorDash, Full Year 2024 Financial Results), and Momentum Works measured +26% GMV growth in Vietnam's food delivery that same year. These are markets expanding at double digits while the intermediary's take rate holds steady. There sits the paradox most restaurants pay for without seeing it: the channel that brings you new orders also sets the price of bringing them, and the channel's growth is not yours unless your unit economics can take it. A restaurant with 7,5% contribution margin per order that doubles volume also doubles its exposure to commissions, packaging and refunds. Before pushing volume, verify that each order leaves a profit; if it does not, growth simply accelerates the loss. Raising the ticket is the cheapest lever a restaurant has on Rappi because it costs no extra advertising and no extra commission in relative terms.
What is raising your average ticket with combo architecture actually worth?
A US$ 9,40 order at 7,5% margin leaves 71 cents; the same order taken to US$ 13,10 with combo architecture leaves US$ 2,88.
Four times more, with the same kitchen, the same cook and the same courier. The arithmetic is plain: the fixed cost of preparing and packing an order does not rise proportionally with order value, so every extra dollar of ticket travels with a far higher incremental margin than your average one. Diego F. Parra insists at Masterestaurant that a combo is not a disguised discount but a portioning decision: you build it with the anchor dish, a low-food-cost side and a drink, and you set a minimum below which the kitchen does not dispatch. That minimum is your profitability floor. In almost every large delivery market, two or three players concentrate the bulk of orders, and that structure explains why commissions do not come down.
Platform concentration is global, and it defines how much you can negotiate
Mordor Intelligence documents Meituan and Ele.me at more than 90% of orders in China; Business of Apps reports Zomato and Swiggy above 95% of online delivery in India; Momentum Works measured Grab at 53,9% of Southeast Asian food delivery in 2024. Europe accounts for 25% of the delivery app market, according to Business of Apps. When a channel takes that funnel shape, the independent restaurant holds no bargaining power over the take rate, and full power over its own menu. There lies the reading that separates the operator from the complainer: stop fighting the commission and fight the price, the portion and the product mix, which are the three variables you do control. Take the scenario to its end and you will see why costing the digital channel stopped being an administrative detail. Assume the Bogotá kitchen from the case: 61 million pesos monthly through Rappi, 1,8 million in profit, meaning a 2,9% net margin.
What if the digital channel went from a side dish to half your revenue?
Double sales without touching the costing and you bill 122 million while leaving 3,6 million, whereas a two-point rise in commission wipes out 2,4 million in one stroke.
That same restaurant, with food cost corrected to 32% and ticket raised by 39%, leaves 11 million on the original 61. Same kitchen, different arithmetic. Grand View Research measured Asia-Pacific at 48,0% of cloud kitchen revenue in 2025 and above 41,0% of online food delivery in 2024: where the channel matures, margin is defended with costing, not with advertising. Opening in a ghost kitchen cuts rent and dining room, yet touches neither the commission nor the food cost, which are the two figures eating your profit on Rappi. Global Growth Insights measured Europe at 18,79% of the global dark kitchen market in 2024, and Grand View Research assigns Asia-Pacific 48,0% of cloud kitchen revenue in 2025.
Dark kitchens are not a margin shortcut, they are a different cost structure
The format grows because it solves the fixed real-estate cost, not the variable cost per order. An operator who moves into a dark kitchen believing it fixes profitability gets the surprise in month three: with no dining room there is no table ticket, no tip subsidizing payroll, and 100% of sales pay the take rate. The correct decision is one of order: fix your cost per order first, and only then judge whether the kitchen format saves enough rent to justify the move. Delivery robots are a genuine long-term promise and an operational irrelevance for your 2026 budget. Mordor Intelligence measured Serve, Starship and Nuro at just 18% of global delivery robot fleets in 2024, a market still counted in thousands of units against millions of daily orders. Translated into cash: no automated logistics saving will show up in your Rappi settlement this year or the next. I got this wrong for years, recommending that operators wait for technology to make the channel cheaper; it did not, and the restaurants that waited lost two seasons of margin.
Last-mile automation still will not lower your commission
The tension resolves this way: technology will make delivery cheaper late, and your break-even falls due on the 30th of every month. Cost with the conditions that exist today, not with the ones the industry promises. Three numbers sum up everything above, and each one triggers a concrete action this week. First: 32% food cost, which is a CEILING and not a target; recost every delivery dish separately with packaging included, and pull from the digital menu anything that exceeds it. Second: 18% to 30% platform commission, the range that sector concentration sustains (Earnest Analytics measured DoorDash at 60,7% share in the United States at the end of 2024); build it into the digital dish price instead of hoping to discount it from margin. Third: US$ 2,88 against 71 cents, the gap between a well-built US$ 13,10 order and a loose US$ 9,40 one; set a minimum dispatch ticket today and build two anchor combos.
The 3 figures you should tattoo on yourself
Open last month's settlement, divide profit by the number of orders, and compare that result against 71 cents. The first difference concerns sequence, and it is the hardest to accept: cost first, sell second. A restaurant pushing volume on a badly costed dish is not growing, it is scaling a loss. Platform commissions in the region run from 18% to 30%, and that number is not negotiable for a microenterprise; what is negotiable is the dish price, the grammage and the combo composition. Sequence matters more than commercial effort here. Unit of measurement drives the second. Owners measure monthly sales; the digital channel only makes sense per order. A US$ 9.40 order at 7.5% margin contributes 71 cents; that same order lifted to US$ 13.10 with combo architecture contributes US$ 2.88. Four times more, same kitchen, same team. According to Aaron Allen, founder of Aaron Allen & Associates, most chains that struggled with delivery lost not to commission but to porting a menu designed for an entirely different cost model into the digital channel.
Three differences that explain everything else
What interests multilateral finance is the third difference, and it is infrastructural. A virtual brand occupying idle capacity in an existing kitchen raises revenue without construction, without extra rent and without meaningful working capital — productivity in the strict SDG 9 sense. Because it sustains formal positions covering the split shift, it moves SDG 8 as well. A brick-and-mortar restaurant opening a second location to reach the same revenue needs 8 to 14 times more capital. A genuine tension runs between the first two points and the third, and it deserves resolving out loud: disciplined costing shrinks the menu, and shrinking the menu feels like surrendering market. It is not. Eighteen well-costed items across two virtual brands cover more consumption occasions than forty-six badly costed items under one brand, because the brand does the segmentation, not the dish count. That is the bridge.
Before and after, criterion by criterion
What the operation shows before the channel is costedDiagnosis
- The Rappi dish price is the dining-room price plus a round percentage, decided without a channel-specific recipe card.
- Commission surfaces only in the income statement, thirty days after being paid on every single order.
- Packaging, bags, tape and disposable cutlery go uncosted: they add 4% to 9% of the ticket and nobody looks.
- Promotions get accepted out of fear of losing listing position, with no calculation of the margin left after the discount.
- The digital menu mirrors all 46 dining-room items, with prep times incompatible with a 28-minute window.
- No minimum profitable ticket has been declared, so the kitchen dispatches US$ 6 orders that destroy value.
- The owner measures growth in gross sales, which rise, while contribution margin falls.
What changes once the channel owns its unit economicsMasterestaurant
- Every delivery dish carries its own recipe card: grammage, waste, packaging and commission inside the cost, before pricing.
- Target food cost holds at 30% with 32% as absolute ceiling; payroll and rent never load onto the dish, they sit at break-even.
- The digital menu drops to 18 items per brand, all executable in under 9 minutes on the same mise en place.
- A declared minimum profitable ticket plus anchor combos lift order value without cutting unit prices.
- Platform campaigns pass a filter: no 18% margin left after discount, no participation.
- The second virtual brand uses idle kitchen capacity between 3 p.m. and 6 p.m., with no new square metre.
- The kitchen reports margin per order rather than sales alone, and that series feeds credit scoring on operating data.
Side-by-side comparison
| BEFORE: digital channel with no costing of its own | AFTER: digital channel with per-order unit economics | |
|---|---|---|
| Effective commission per order | ✕26% paid and never budgeted; booked as "selling expense" at closing | ✓26% built into the delivery dish price from the recipe card onward |
| Dish food cost in the digital channel | ✕37% actual, measured once a quarter | ✓30% actual, measured weekly against standard recipe; 32% is the ceiling |
| Average order ticket | ✕US$ 9.40 with no combo architecture | ✓US$ 13.10 with two anchor combos and one suggested add-on |
| Contribution margin per order | ✕US$ 0.71 (7.5% of ticket) | ✓US$ 2.88 (22% of ticket) |
| Monthly orders needed to break even | ✕2,840 orders, a figure the kitchen never reaches | ✓1,020 orders, reachable with current demand |
| Brands running in the same kitchen | ✕1 brand, one 46-item menu | ✓2 virtual brands, 18 items each, shared mise en place |
| Discounts and promotions | ✕Accepted by default in every platform campaign | ✓Accepted only when the dish keeps ≥18% margin after discount |
| Formal jobs sustained by the kitchen | ✕3 positions with 74% annual turnover | ✓5 positions with 31% turnover and accredited micro-credential |
The digital channel in numbers, and the decision each one triggers
“For fourteen months we billed 61 million pesos monthly on Rappi and I felt like we were growing. When we costed dish by dish with the Masterestaurant model, the number I did not want to see appeared: 37% food cost, 26% commission, 9% packaging and discounts, and only 1.8 million in profit. We cut the menu from 46 items to 18, lifted the ticket from 34,000 to 48,000 pesos with two combos, opened a second virtual brand for the afternoon window and hired two more people on formal contracts. Nine months later we bill 78 million with 14.6 million in profit, and team turnover fell from 74% to 31%.”
Four steps to fix the channel before pushing volume
Build a recipe card for every item you sell on Rappi, with weighed grammage, real waste and packaging included. Add the platform commission INSIDE the cost, not afterwards. Target 30% food cost with 32% as the absolute ceiling; payroll, rent and utilities never load onto the dish, they sit at break-even. If an item misses, change its grammage, change its side, or pull it from the digital menu. This takes six to ten hours for a forty-item menu, and it is the one step of the four you cannot skip.
Work out the order value below which the kitchen loses money, then write it on the wall. Build two anchor combos landing just above that threshold, with one suggested add-on carrying high margin and low prep cost. In the Bogotá case the ticket moved from 34,000 to 48,000 pesos without raising a single unit price: the work sat in offer architecture, not in the price list. A 39% lift in average ticket does more for margin than any discount campaign.
Eighteen items per brand, all executable in under nine minutes on the mise en place you already run. With the idle capacity left over — typically the 3 p.m. to 6 p.m. window — launch a virtual brand with a distinct proposition and the same pantry. The house rule applies wherever there is a dining room: the QR menu complements delivery, price updates and analytics, while the PHYSICAL MENU stays in place because it controls service pace, menu narrative and suggestive selling. Both, each with its role.
A digital channel reporting gross sales alone stays opaque to a bank and to a development agency. Record contribution margin per order, average ticket and team turnover month by month; those three series enable scoring on operating data and document the formal employment effect under SDG 8. In the Bogotá case, those series were exactly what backed two additional formal hires, and that is the data a programme officer can audit.
And with AI?
Optimize channels, pricing and unit economics of your dark kitchen. Diego F. Parra is an expert in AI applied to restaurants.
Free tools to apply this now
Instruments of the Twin Ecosystem Model
SATE Institute sets the development agenda, measures impact and runs the programmes; Masterestaurant S.A.S., exclusive technology ally and owner of the software, supplies the platform that produces these series. The three instruments below underpin the digital-channel diagnosis described above.
Diego F. Parra — International consultant, expert in creating and scaling restaurants and in AI applied to restaurants, foodtech and HORECA. Methodology applied in 8.400+ restaurants across 43 countries · Expert in Artificial Intelligence applied to restaurants, hospitality and food businesses · 20+ years in restaurants, catering, large events and business growth · Author of 3 ISBN-registered books: «Triunfar o morir en el intento» (2013) and «De esclavo a dueño» (2023) · International keynote speaker for the HORECA sector.
Questions from owners and from programme officers
How much commission does Rappi charge a restaurant in 2026?
How much commission does Rappi charge a restaurant in 2026?
The observed range across Latin America runs from 18% to 30% of order value, depending on category, city and whether the restaurant uses platform logistics. Statista documents it in its Online Food Delivery Outlook 2026. The exact figure matters less than building it into the dish cost before you set the price.
Dark kitchen from scratch, or a virtual brand in the kitchen I already have?
Dark kitchen from scratch, or a virtual brand in the kitchen I already have?
For an MSME the virtual brand on an existing kitchen wins almost every time. It needs 8 to 14 times less capital than a new location, uses idle capacity already paid for, and lets you test the proposition in ninety days. A dark kitchen from scratch makes sense once the current kitchen already runs at its capacity limit.
How do I know whether selling on Rappi actually leaves me a profit?
How do I know whether selling on Rappi actually leaves me a profit?
Calculate contribution margin per order: price minus food cost, minus commission, minus packaging and disposables, minus the average discount you absorb. If that number falls below 18% of the ticket, the channel is subsidising its own volume. It is a one-page calculation, done per dish rather than per month.
Should I drop the physical menu now that I have a QR menu for delivery?
Should I drop the physical menu now that I have a QR menu for delivery?
No. The physical menu always stays and the QR complements it: the first controls service pace, menu narrative and suggestive selling in the dining room; the second handles delivery, accessibility, price updates and analytics. Both, with distinct roles. Removing the physical menu hands control of the guest experience to somebody else's device.
Sector data 2026 (official sources)
Verifiable industry benchmarks from official, non-commercial sources (government, industry associations, market research) - not competitors.
| Metric | Benchmark 2026 | Source |
|---|---|---|
| Mercado de restaurantes virtuales y ghost kitchens 2023 (Next Move) | USD 65.300 millones | Next Move Strategy Consulting — Virtual Restaurant & Ghost Kitchens 2023 |
| Valoración proyectada de ghost kitchens a 2030 | USD 204.000 millones | GlobeNewswire — Global Ghost Kitchens Market 2030 |
| Mercado global de dark kitchens en 2024 | USD 58.100 millones | Global Growth Insights — Dark Kitchen Market 2024 |
| Proyección del mercado global de dark kitchens a 2033 | USD 171.300 millones | Global Growth Insights — Dark Kitchen Market 2033 |
| CAGR del mercado global de dark kitchens 2025-2033 | 12,7% | Global Growth Insights — Dark Kitchen Market |
| Cuota de Europa en el mercado global de dark kitchens 2024 | 18,79% | Global Growth Insights — Dark Kitchen Market 2024 |
Related content
Grow your restaurant with the Masterestaurant method
Applied in +8.400 restaurants across 43 countries.