Rappi delivery strategy: before vs after with Masterestaurant

Why am I losing money on Rappi if I'm selling orders every day?

Most restaurants lose money on delivery because they treat Rappi as a passive channel, not as an independent business with its own P&L.

They upload the same menu, don't adjust prices or promotions, and Rappi's commission (25-30% depending on category) eats the margin without them seeing where it fails. Rappi's algorithm doesn't reward passivity: it rewards consistent volume, speed (fulfillment ≤28 minutes), and rating (4.7+ stars). Trying to sell $8 tacos with $6.50 total cost plus $2.40 commission closes every sale at a loss. Diego F. Parra has audited 8,400+ restaurants over 20 years; 67% of those selling on Rappi never reach break-even in delivery because they didn't separate budget or fulfillment. Open a SEPARATE P&L for delivery, don't let the restaurant hide it. Calculate fixed cost (dedicated cook, packaging, compostable containers) and variable per order (ingredients adjusted for speed, not variety; Rappi commission; third-party delivery if used).

How do I know if my delivery orders are making or losing money?

Gross margin must be 50-65% before that commission. If you sell a taco for $5 with $1.80 cost, margin is 64%; after 30% commission (-$1.50), you keep $1.70 per taco.

With 40 tacos/day sold, you sum $68 daily, minus fulfillment cost ($40-50), you reach $18-28 EBITDA. Masterestaurant designs that P&L to be visible: if you don't separate budget, the restaurant hides the truth from you. That's the first step. Rappi isn't a mailbox where you upload your menu and wait for orders. Its recommendation algorithm behaves like a market: it shows more frequently to restaurants selling consistently (6-10 orders/day without gaps), delivering fast (not taking 45 minutes), and with good rating (4.7+ stars; below 4.5 loses visibility). Three levers. The algorithm LEARNS in 2-3 weeks what your pattern is: if some days you sell zero orders, it lowers your feed ranking.

What does it mean that Rappi is an algorithm, not a channel?

Those who execute those three levers well jump from local zone (your block) to long-range searches (5 km around), because Rappi knows that restaurant won't fail.

It's like a game where the physics are visible if you know where to look. The mistake is ignoring that physics exists. 60+ items on Rappi costs you: kitchen slowness, order errors, rating drop. Choose 18-24 items (bestsellers, maximum margin, maximum speed). A neighborhood restaurant with 80 dishes can sell 20 in-store and choose 18-22 for Rappi: the ones with highest turnover, controlled cost, and recipes that don't fail. Reorder those 18-22 within the app by putting at the top what sells MOST and cooks FASTEST: that order impacts clicks and speed. Rappi rewards speed; Uber Eats (another channel) rewards variety. Different strategy per channel. Diego F. Parra sees restaurants uploading 90+ items to Rappi and then surprised when fulfillment fails and rating drops from 4.8 to 4.3 in 3 weeks.

What's the real size of a delivery menu?

Less is more. Don't double prices on Rappi: clicks sink. Calculate the sale price that absorbs commission without being obvious.

If your cost in-store is $8 and you sell at $13 (38% margin), on Rappi that same dish would cost $13 + 30% commission ($3.90), sum $16.90. Don't drop it to $12 to look cheap (looks deceptive; besides, still loses margin). Better: raise that item to $14.90, absorb the $4.47 commission, and the customer sees $14.90 as fair price in app (vs $13 in-store). The difference is $1.90, which the customer doesn't see because on app they don't compare with in-store in real time. That psychology of pricing in digital channel is a lever 70% of restaurants ignore. Rappi's algorithm reclassifies every 10-14 days if your metrics improve. It takes 2-3 weeks to execute well: consistent speed (25-28 minutes from accept to pickup), 4.7+ star rating, and volume without gaps (minimum 5-6 daily orders).

When does the restaurant jump from 6 orders/day to 40+ on Rappi?

Once Rappi sees that pattern, it increases your feed frequency in adjacent zones, not just yours. Suddenly you're selling 6 orders/day (just your zone) to 30-40 (your zone + 3-4 km around).

The jump isn't linear: it's a step. Masterestaurant sees this every month in audits. The common mistake is giving up in week 2 because you don't see immediate growth; the algorithm takes time. Dark kitchen (rented ghost kitchen) speeds up if your restaurant is at max capacity and you want to separate operation completely. Cost: rent + cook + infrastructure, added. Option 2: use idle capacity of your space (cook during less busy hours, 2:30-4:30 PM and 10:00-11:30 PM). Option 3: shared kitchen (rent hours in another kitchen, split cost). For small restaurants, idle capacity is the answer until Rappi generates 60+ orders/day. At that point, reconsider dark kitchen.

What's the difference between a dark kitchen and using your restaurant's kitchen?

The mistake is renting dark kitchen without data: spend $1,200-2,000/month without knowing if those deliveries will pay that rent. Diego F.

Parra sees audits where restaurants spent on unnecessary dark kitchen and went broke in 3 months. Start in your space; measure; grow. Measure 5 numbers: (1) average orders/day (line going up or flat is the main diagnosis); (2) average fulfillment time (drop from 35 to 25 minutes = algorithm raises your ranking); (3) average rating (if it falls from 4.7 to 4.3, there's error in kitchen or packaging); (4) net margin AFTER commission (should be 35-50% depending on item); (5) customer repeat rate (how many order again vs new). Margin is what restaurants hide most: they say 'I'm selling lots' when losing money. Masterestaurant designs that metric transparent in the separate P&L. If your weekly net margin is negative, adjust price WITHIN Rappi (not the product) or remove low-margin items.

What should I measure every week to know if I'm doing well?

Repeat rate is your signal of loyalty: if >60%, product is good; if <40%, something in service or packaging fails.

<strong>Budget separation.</strong> Delivery is a standalone P&L, not a line under "other income." That way you see if it wins or loses without the physical restaurant hiding it. Fulfillment cost (dedicated cook, packaging, delivery), Rappi commission (25–30%), and sales price: the equation must close before you launch. <strong>Small, reordered menu.</strong> 60+ items on Rappi costs: slowness, kitchen mistakes, rating drops. Choose 18–24 items (bestsellers + high margin + speed). Reorder by expected volume (put fast movers at the top). Rappi rewards speed; Uber Eats rewards variety. Different strategy per channel. <strong>Commission-absorbed pricing.</strong> Don't double prices on Rappi (kills clicks). Math: if it costs $10 + 30% commission ($3), minimum price is $14.30 for 31% margin. The menu reduction (better prime cost) + volume (spreads fixed costs) solves it.

The 5 differences that separate winners from abandonment

<strong>Fulfillment ≤28 minutes.</strong> The algorithm measures it; the customer feels it. If the app says "arrives in 35 min" and delivery takes 45, you drop in rating. Use a dedicated cook + shared kitchen or dark kitchen to have capacity without stopping the physical restaurant. Time each item (how long does each dish actually take?). There's where you find slack. <strong>Weekly KPI tracking.</strong> Orders total, margin/order, rating, % of 5⭐ reviews, average time, cart abandonment. Adjust every 7 days: if rating drops, remove slow dishes; if it rises, push bestsellers. If margin doesn't hit 28%, lower product cost or boost volume (in-app ads). No measurement = no adjustment; no adjustment = abandonment in 6 months.