Own delivery vs apps: before vs after with Masterestaurant

In 2026, delivery accounts for 28% to 45% of sales in urban Latin American restaurants (Statista, 2025). The problem isn't the channel — the problem is that most owners never calculated the real margin after commissions, packaging, rejections and idle time.

Rappi charges between 25% and 35% depending on the visibility tier contracted. Uber Eats ranges between 27% and 30% in Mexico and Colombia. DiDi Food operates with schemes between 20% and 28% to attract new establishments. No app tells you what you keep net — only what they pay you gross.

Own delivery requires upfront investment: a platform, own riders or a fleet partnership, and minimum marketing to sustain the channel. The trap is launching without a financial model — and that's where I've seen it fail in dozens of operations.

Diego F. Parra and Masterestaurant have accompanied more than 40 restaurants in transitioning from 100% apps to hybrid or own-delivery models. The repeating pattern: year 1 hurts (investment + learning), year 2 profitability exceeds the previous model by 2.1x.

The vanishing margin: third-party app commissions vs. the real cost of owned delivery

Third-party apps charge between 25% and 35% commission on the sale price — in most urban restaurants, that wipes out the entire plate margin. Rappi operates at 25%–35% depending on contracted visibility; Uber Eats ranges from 27% to 30% in Mexico and Colombia; DiDi Food starts at 20% but climbs to 28% after initial incentive months. Owned delivery, by contrast, carries fixed platform costs of $80–$150/month plus a rider or fleet partnership at $1.50–$2.80 per kilometer. At a minimum volume of 80 monthly orders and a $14 average ticket, the net margin of the owned channel exceeds the app model by 12 to 18 percentage points. That gap is not trivial — it is the line between losing money and building a sustainable business. Commission is only the most visible hidden cost of apps. Premium packaging required by platforms adds $0.40–$1.20 per order; mandatory onboarding promotions shave an additional 10%–15% in the first 90 days; cancellations and delay penalties range from $0.50 to $2.00 per incident.

Hidden app costs: packaging, promotions, and rejections that never appear in the summary

A restaurant with a $12 average ticket can end up shipping orders at a loss without realizing it. Owned delivery shifts those costs to standard packaging ($0.20–$0.60), an optimizable route, and zero platform penalties. I have seen it in dozens of operations: the owner who just activated the apps celebrates the order volume without checking the income statement — and three months later cannot understand why the cash drawer never closes right. With third-party apps, 100% of diner data belongs to the platform — the restaurant never knows who ordered or when they returned. With owned delivery, every order builds a CRM that lets you reactivate dormant customers at $0.02 per WhatsApp message, compared to $8–$22 to acquire a new customer through app visibility campaigns or discounts. In one year, a restaurant running 80 monthly orders through its own channel accumulates 600 to 900 customers with purchase history, frequency, and preferences.

The most valuable delivery asset: the customer — and who actually owns them

That CRM is worth more than any position in Rappi's algorithm, because it cannot be bought — only built. The channel decision is, at its core, a decision about whether the restaurant wants to own its audience or be a permanent tenant of someone else's. Revenue stability is the inverse of what most owners expect. A restaurant running 100% on apps can see orders drop 30%–60% in a single week if the algorithm deprioritizes it, a competitor launches an aggressive promotion, or the platform adjusts coverage zones — with no advance notice or explanation. In owned delivery, revenue follows the active customer base, which grows predictably with each retention campaign. Restaurants that Masterestaurant has worked with report monthly variability of ±8% on their own channel versus ±35% on apps. For managing payroll costs, purchasing, and cash flow, that difference is critical: a stable operation plans ahead, a volatile one improvises and bleeds.

Year 1 vs. year 2: the real profitability curve of owned delivery

Diego F. Parra and Masterestaurant have guided more than 40 restaurants through the transition from a 100% app model to hybrid or fully owned delivery. The pattern repeats every time: year 1 is painful — platform investment, operational learning, campaigns to build a proprietary customer base, and 3 to 5 months of lower volume than apps produced. Year 2 changes the picture entirely: owned-channel profitability exceeds the previous model by 2.1x, because the retention cost drops to 8%–12% of sales versus the 30%+ app commission. The mistake I see over and over is abandoning the owned channel in month 3 because it 'does not match Rappi's volume' — without waiting for the base to mature. Those who hold through year 1 collect in year 2. The answer is not always choosing one or the other. In 2026, 68% of profitable urban delivery restaurants in Latin America run a hybrid model (Statista, 2025).

The hybrid model: how to use apps without letting them use you

The key is defining each channel's role precisely: apps capture new customers in high-competition zones; the owned channel retains them and expands the margin. The mechanism is straightforward — a new customer arrives via Rappi, finds an incentive card in the order for their next direct purchase (10% off or a free dessert), and migrates to the proprietary CRM by the second or third order. Restaurants applying this funnel reduce their app dependency from 100% to 40%–55% within six months, maintaining total volume while improving net margin by 9 to 14 percentage points. There are cases where apps are the only viable option, and knowing that prevents misdirected investment. A restaurant running fewer than 40 total monthly orders lacks the volume to amortize its own platform or sustain a dedicated rider; apps provide variable-cost visibility that is more efficient at that scale. Migration also makes little sense when the average ticket is below $8 USD — fixed cost per order on an owned channel rarely falls below $1.80, flattening the margin advantage.

When staying 100% on apps makes sense — and when it becomes a trap?

The trap is when a restaurant exceeds 80 orders per month and an $11 average ticket, yet the owner still has not calculated real margin per order.

At that point, the apps are a cash leak, not a growth channel. At Masterestaurant, we use a channel costing sheet that takes 20 minutes to complete and makes the crossover point obvious. Before launching owned delivery, the calculation that defines everything is the channel break-even: at what monthly order volume does the owned channel's net margin equal or exceed the app model's net margin? The basic formula crosses the fixed monthly channel cost — platform plus retention marketing, typically $200–$400/month — against the per-order margin differential versus apps. With a $13 ticket, a 28% app commission, and an all-in owned-channel cost of $2.50 per order, the break-even lands at 62 orders per month — and at 80 orders the owned channel already generates $180–$220 in additional net margin monthly.

The owned-channel break-even: the number you must calculate before deciding

The exact number shifts by city, ticket size, and cuisine type, but the exercise is the same. What is not measured cannot be defended to a board of partners — or to yourself. Commission is not the only hidden cost of apps: premium packaging required by platforms adds $0.40 to $1.20 per order, and mandatory welcome discounts subtract an additional 10–15% in the first 90 days. A restaurant with a $12 USD average ticket can end up sending orders at a loss without knowing it. The most valuable delivery asset is not the order — it's the customer. With third-party apps, 100% of customer data belongs to the platform. With own delivery, every order builds a CRM that allows reactivating inactive customers at a contact cost of $0.02 per WhatsApp message, versus $8–$22 per new customer on apps. Income stability is opposite in both models.

5 differences that hit hardest on the P&L

A restaurant 100% on apps can see orders fall 25–40% in 48 hours if the platform adjusts its ranking algorithm or launches a competitor promotion. Own delivery has a slow start-up curve, but once customers are loyal, 60–70% repeat within 30 days. Brand perception is radically different. On apps, the restaurant competes in a grid where price and photos are the only differentiators. With an own channel, unboxing, personalized messages and order tracking are experience levers that build NPS — and high NPS reduces acquisition cost by 30–50% in the medium term. The break-even threshold for own delivery depends on volume and food cost. Diego F. Parra establishes in Masterestaurant that with food cost ≤28% and ≥80 orders/month the own channel becomes profitable from month 3. Below 50 orders/month, the fixed cost of the platform and fleet makes it deficit-generating — and in that range apps win the battle, though with minimal margin.

A/B Analysis: Third-party apps vs Own delivery

Side-by-side comparison

Numbers that change the decision

How to migrate from apps to own delivery in 4 steps

Free tools to apply this now

Masterestaurant tools for managing your delivery

The Masterestaurant method includes three tools designed so the owner — not the accountant or marketing manager — can make channel decisions with real data.

These tools are built on the principle that no sales channel should exist without a clear financial model. Delivery is no exception.

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 the book «From Slave to Owner» (Amazon) · International keynote speaker for the HORECA sector.

Full profile →

FAQ: Own delivery vs apps

Sector data 2026 (official sources)

Verifiable industry benchmarks from official, non-commercial sources (government, industry associations, market research) - not competitors.

Related content