Own Channel vs Delivery Apps in Restaurants: Myth vs Reality

How much does a delivery app actually take from every sale?

Between 25% and 35% of every sale, plus tax, packaging costs, and sometimes an extra fee for paid placement inside the app.

In 2026 the delivery market in Latin America moves more than $32 billion a year, yet the average restaurant selling through third-party platforms nets only 3% to 6% profit on net sales — half of what dine-in leaves behind. Rappi, Uber Eats, and DiDi Food charge commission on gross revenue, not on profit, so the more you sell, the more you hand over in absolute dollars. I see it every month in the P&Ls I review: the owner celebrates that 'delivery is growing' without noticing that growth is eating the entire net margin. As a consultant, the first question I ask at Masterestaurant is simple: how much of that sale actually reaches your pocket after commission, packaging, and logistics? Owners almost never have that number on hand — and that is warning sign number one.

The myth of 'without the app I don't exist in delivery'

It's false: apps are an acquisition channel, not a loyalty channel, and their real cost per new customer exceeds $4.50 USD once you deduct the mandatory promotions the platform requires. When a customer reorders on Rappi, the app shows 'sponsored' restaurants first — and you can disappear from that screen unless you pay extra for placement, even if that same customer has already eaten your food three times. That's the mistake I see over and over, in neighborhood spots and chains alike: they confuse order volume with a customer relationship. The platform owns that relationship, not the restaurant, because the customer data — phone number, order history, preferences — stays on the platform's side. On your own channel (an ordering website, WhatsApp Business with a catalog, or your own app) the customer comes back because they have your link or your number, and reactivating them costs almost nothing. That's the real difference between renting traffic and owning your customer base.

The case: a restaurant with an $18 ticket and 400 monthly orders

A casual-dining restaurant with an average ticket of $18 USD and 400 monthly app orders was billing $7,200 USD a month in delivery, losing between $1,800 and $2,520 in commission alone (25% to 35%), before packaging and packing staff. Working with the Masterestaurant method, we first set up a WhatsApp Business catalog with integrated payment and a direct-order landing page, at a technology cost of $180 USD monthly plus a 10% logistics commission through a local delivery partner — not an owned fleet. In month one we migrated 22% of app volume to the owned channel; by month three, 41%. Net profit across the full delivery channel (apps plus owned) rose from 4.8% to 11.3% of sales, without lowering total order volume. The initial investment in technology and staff training — about $650 USD — was recovered in 47 days of operation, measured in avoided commission.

Why percentage commission punishes the top sellers hardest?

Because it isn't a fixed cost: it's a percentage of gross revenue, so the more you bill through the app, the more you hand over in absolute terms.

A restaurant billing $20,000 USD a month on Uber Eats pays between $5,000 and $7,000 in commission alone — before kitchen, staff, and packaging — and that money comes straight out of what should be profit. I always compare it to food cost: a dish with a 32% food cost (the maximum I recommend at Masterestaurant) already leaves just 68% gross for everything else; subtract another 30% in delivery commission and the real operating margin on that app-sold dish can drop to single digits before payroll and rent. That's why an owned channel isn't a tech trend — it's a cost-structure decision. Every percentage point of commission you avoid is a point that goes straight to your break-even, not to the platform's account.

The price control that apps won't let you keep

It's an illusion: platforms impose 'guaranteed minimum discount' promotions as a condition for appearing in top search positions, forcing the restaurant to subsidize discounts it never chose. I've seen restaurants forced into 2-for-1 deals or 20% discounts just to keep visibility, absorbing that cost on top of a margin already cut by the base 25% to 35% commission. On your own channel, you set the price: if you run a promotion, it's because it works for you, not because the app's algorithm is holding your exposure hostage. That autonomy shows up in real numbers — in the case study, the average ticket on the owned channel held at $18.40 USD, while the average app ticket dropped to $15.90 USD because of mandatory promotions. Getting price control back translates, in practice, into recovering two to three points of gross margin per transaction.

What does it really cost to build an owned channel

Between $80 and $600 USD monthly in technology, depending on whether you use WhatsApp Business with a catalog, a simple ordering website, or an integration with platforms like Otter or Flycart, plus an 8% to 15% logistics commission if you use an owned fleet or a local delivery partnership. That's a fraction of what a 25% to 35% platform commission costs on total volume. In the case study, the recurring monthly investment was $180 USD in technology plus 10% in logistics — meaning that to sell the same $7,200 USD monthly, the owned channel's total cost was roughly $900 USD versus the $1,800 to $2,520 USD that went to app commission alone. That difference — between $900 and $1,620 USD a month — is profit that used to stay with the platform and now stays in the restaurant's cash register. That's the number every owner should run before deciding where to spend their next marketing dollar.

Should I drop delivery apps completely?

No: the right call is almost never 'apps or no apps' — it's how much relative weight you give each channel based on your ability to run both with discipline.

Apps still work well for acquiring new customers in areas where your brand has no recognition, and for catching last-minute demand at peak hours where your own channel lacks the installed logistics capacity. The mistake isn't using apps; it's depending on them 100% without building an owned channel in parallel that captures and retains that customer after the first order. At Masterestaurant we recommend a channel-mix target — typically 60% owned / 40% apps for restaurants with an established local brand, adjustable by category and market. Measuring net profit per channel, not just sales volume, is the discipline that separates restaurants that grow with margin from those that grow revenue and run out of cash. Pull today's report of commissions paid to each app over the last 90 days and compare it, dish by dish, against your real food cost — not against the menu price.

The concrete action for this week

If average commission exceeds 28% and your food cost already sits in the 28% to 32% range, you have a structural margin problem no app promotion will fix. The next step is opening a WhatsApp Business channel with a catalog this same week — it costs under $50 USD monthly to launch and needs no development — and starting to migrate, with a direct incentive, the customers who already know you. Diego F. Parra repeats this in every Masterestaurant diagnostic: the owned channel doesn't replace apps overnight, but every order you migrate is one where you decide the price, the margin, and the customer relationship. That's the only metric that matters at month's end. The commission is not a fixed cost — it's a percentage of gross revenue, meaning the more you sell on the app, the more you give away in absolute terms. A restaurant billing $20,000 USD monthly on Uber Eats hands over $5,000 to $7,000 in commission alone — before paying kitchen, staff, and packaging.

The 4 Differences That Hurt Your Cash Flow Most

The apps own your customer, not you. When someone reorders on Rappi, Rappi shows 'sponsored' restaurants first — and you may not be one of them unless you pay for additional positioning. With your own channel, the customer returns because they have your link, your WhatsApp number, or your app — and reactivation cost is nearly zero. Price control is an illusion on apps: platforms impose 'guaranteed minimum discount' promotions as a condition to appear in high-visibility sections. 40% of active restaurants on Rappi Colombia reported in 2025 participating in at least one mandatory promotion that pushed their margin into negative territory. An own channel demands your own marketing muscle — and that's the real cost owners underestimate. WhatsApp Business, email, an updated Google Business Profile, and some local paid ads are the minimum investment to generate demand without the app. Owners who don't allocate a customer acquisition budget ($150-$400 USD/month) typically fail within the first 60 days and blame the channel, not the execution.