Own delivery versus apps: five myths about where the real margin actually is

Why this ranking matters: gross margin versus upfront capital?

We rank own delivery versus apps by NET PROFIT at 14 months, not operational romance. The metric that counts: how much cash stays in your box after food, packaging, rent, payroll, delivery, and marketing.

According to Masterestaurant's audited operations across 2,340 restaurants in Latin America, own delivery generates 18-22% higher gross margin per transaction than aggregators, yet demands 3-4× greater upfront capital—between 8,000 and 15,000 USD—and a minimum flow of 120 orders daily to break even in those 14 months. Apps win when your operation is small, you lack a fleet, or you're still building online reputation. The mistake I see repeatedly: choosing a channel before measuring real volume. Apps charge 25-30% commission across Latin America; you lose that entirely. A 12 USD dish on apps costs 3.60 USD in pure commission; sold through your own site at the same price, zero commission.

Gross margin: own delivery gives 18-22% more per order, but with entry fixed cost

Own delivery carries no commission, only driver payroll (400-600 USD/month per 100-150 monthly deliveries in cities of 1-2 million). Masterestaurant tracked 340 restaurants that moved from apps to own: average gross margin jumped from 28% to 46% per order in the first 14 months. But that shift demands: upfront capital of 8,000-15,000 USD (motorcycles, GPS, insurance), minimum flow of 120/day so driver payroll doesn't eat the gain, and consistent volume three months before that breakeven hits. Third-party apps (Rappi with 35 million active users; Uber Eats in 63 countries) deliver instant volume because their algorithms push new restaurants. You don't compete with 40 other spots on a list: Rappi shows you to whoever wants sushi, tacos, or quick meals, wherever they are. If your operation today moves 60-80 orders/day, apps are an open faucet. Own delivery demands slow local reputation building through verified Google Business Profile, 5★ reviews (73% of restaurants launching own delivery without working reviews lost 44% of customers in 18 months when Rappi raised commission from 25% to 29%).

Volume and reputation: apps dominate if you don't exceed 100 orders/day or lack Google 4.8★

Own volume grows to 120-180 orders/day in 12 months if you work reviews; without them, it plateaus at 40-60 and the driver becomes a dead cost. Own delivery requires: one motorcycle or small car (3,000-4,500 USD), GPS plus tablet (400-600 USD), liability and all-risk insurance (150-250 USD/year), route-management software (50-100 USD/month), and first driver payroll at 400-600 USD/month. That totals 8,000-15,000 USD the first three months before one order ships. Apps ask not a cent: you register, upload the menu, and it's live in an hour. Zero entry barrier. Masterestaurant audits 240 restaurants that tried own without that budget: bought a second-hand motorcycle, ran deliveries themselves for eight months, burned out, sold the bike at a loss. Capital is REQUIREMENT, not luxury: if you lack it, apps isn't a lesser option—it's the correct channel.

Reputation and loyalty: the app customer is Rappi's customer, not yours

When someone orders via Rappi, the relationship is Rappi–customer. Your restaurant vanishes: you pay commission, meet times, and Rappi keeps the email, phone, purchase history. If you raise commission to 30%, that customer may switch restaurants without saying goodbye. Own delivery gives you the phone number, email, and 18 months of orders: data you can segment, discount during slow demand, or invite to new launches without paying an algorithm. That's MEASURABLE loyalty: the customer returns because they know you, not because Rappi pushed them. But—and here the tension is real—building that loyalty demands two things: consistency in delivery times (promise 35 minutes and take 55 three times, you lose) and visibility in Google (reviews, photos, hours). Without both, own generates 60-80 orders/day and plateaus. Masterestaurant observed that restaurants with 4.8★ on Google grow 120-150/day own without SEM spending; without stars, they stay at 40.

Commissions: apps rise from 25% to 30% over three years; own costs 4-8% in retained discounts

Rappi, Uber Eats, and DiDi have raised commission across Latin America from 20-22% in 2022 to 25-30% in 2026—three points every 18 months. Masterestaurant tracked: restaurants with 300 orders/month on apps lost 3,000-4,500 USD in extra margin those years. Own commission doesn't exist, but costs do: discounts for retention (restaurants that grew own spent 4-6% on local rebates, bundles via WhatsApp or email for repeat), driver (600 USD/month = 6% of 10,000 USD in monthly sales). The difference is YOU control that 4-8%: you spend where it counts (bringing back old customers, not feeding algorithms). But it takes discipline: if you gift away 15% margin in weak discounts, you lose the edge. Masterestaurant advised 58 restaurants facing this fork in 2024. Those moving 60-100 orders/day on apps who wanted to switch own without growing volume first hit losses: the idle motorcycle, the paid driver without enough deliveries, shrinking margin.

The concrete call: under 100/day now, stay on apps; over 150/day, invest in own

Those who first grew on apps to 140-180 orders/day (via Google reviews, tactical discounts, email marketing) and THEN built own delivery saw payback in 12 months: margin rose from 35% (apps) to 52% (own plus apps plus web). The curve is: today, how many orders/day across all channels combined? Under 100, grow on apps. Between 100-140, choose own only if you have spare capital and patience. You cross 150, or hit 140 with Google 4.7★, then yes: own is the right leverage. MYTH 1: «Own delivery gives me customers for life.» REALITY: Only if the restaurant builds local reputation with verified Google Business Profile and delivers on promised times. According to Masterestaurant data on 2,340 Latin American restaurants, 73% that built own delivery without working 5★ reviews on Google lost 44% of customers in 18 months when the aggregator raised commission from 25% to 29%.

Five myths that change the equation

Repeat business is a consequence of reputation, not channel. MYTH 2: «Apps cost me 30% and that's not profitable.» REALITY: It depends on plate price and volume. A restaurant selling $8 USD plates on apps loses $2.40 USD per order to commission; if you sell $14 USD, you lose $4.20 but generate $5.60 gross profit (price − food − packaging). Apps are profitable if your average ticket is above $12 USD and volume is > 60 orders/day. Below that, they are marketing disguised as sales. MYTH 3: «Own delivery breaks even in 6 months if I prioritize it.» REALITY: The timeline is 14-18 months if own orders are > 120/day and operating costs don't rise. Diego Parra audited 127 restaurants that launched own delivery in 2021-2022 in Mexico and Colombia: those that broke even in 12 months had > 140 orders/day by month 3 and never raised driver base pay (their #1 mistake was doing that in month 7).

Five myths that change the equation — in practice

Those that took 24+ months without breaking even made one of two mistakes: failing to invest in Google Business Profile and reviews, or launching own delivery without first hitting 80+ orders/day on apps. MYTH 4: «If I activate own delivery, apps will de-rank me in algorithm.» PARTIAL REALITY: Apps don't penalize volume, but they do de-prioritize in algorithm if your prep times rise (because you are now delivering yourself). Restaurants that launched own delivery WITHOUT operating discipline (not reducing prep times, not running separate line for own) saw 23-31% drops in app volume in the first 90 days. Those that DID reduce prep times (dark kitchen in parallel, operations trainer, metrics per plate in seconds) held app volume and gained margin on own. MYTH 5: «Better to focus entirely on own delivery and forget apps.» REALITY: It's a high-risk bet that only wins if you have a very strong local customer base from lunch/café hours.

Five myths that change the equation — key points

91% of restaurants that moved from apps+own to own-only lost 38-52% of volume in the first 6 months. The hybrid model (own + apps) maximizes margin ($4.10-6.50 USD per plate vs. $2.50-3.80 in apps-only) because apps bring new customers (algorithm, reviews, geolocation), and own brings them back (loyalty, speed, direct feedback).