Owned Channel vs Delivery Apps: Which Method Keeps More Margin in Your Restaurant in 2026

The commission no one calculates correctly

The mistake I see over and over when auditing kitchens in Bogotá and Medellín is that owners lump commissions together with ingredient costs on the same accounting line. The result: food cost looks like 31%, when the real net margin is actually around 3% to 5% of gross billing. Over three years and more than 60 operations reviewed, 70% of owners believe a profit figure that already had the platform's cut taken before they were even paid. The difference between understanding that number and ignoring it is the difference between scaling the business and surviving month to month.

When your own channel starts beating the apps?

The real threshold is not arbitrary:

once a restaurant exceeds 350 monthly orders and the average ticket passes $22,000 COP, a direct channel — WhatsApp with a payment link, a restaurant website with a payment gateway, or a proprietary ordering system — generates between $1.1 and $1.5 million additional pesos of margin every month, without changing the menu or raising prices. The math is straightforward. The same order through a direct channel, with a payment gateway at 2%–3%, leaves $21,340 to $21,560 net. Multiplied by 400 orders, the gap exceeds $1.7 million pesos per month. In 2026, with direct-channel tool costs between $150,000 and $400,000 COP per month, break-even arrives in 45 to 60 days of real operation.

The case: a Bogotá restaurant, from app-dependent to mixed channel

The diagnosis was immediate: 78% of orders arrived via Rappi and Uber Eats, with an average commission of 26%. In the action phase, Diego F. Parra and the Masterestaurant team activated a direct channel via WhatsApp Business with a Wompi payment link, migrating a meaningful share of the habitual app traffic in 8 weeks through direct communication with the customer base already captured on the platforms. The result by month 3: 168 orders through the direct channel, 312 through apps, net margin of 7.9% ($3.16 million). Tool investment was $280,000 COP per month; the incremental gain was $1.48 million.

Data the apps never share — and what that silence costs

Rappi, Uber Eats, and DiDi Food do not share the customer's phone number or email. That sounds like a technical detail; in practice it is a dependency trap. Without customer data there is no remarketing, no real loyalty program, and no way to measure cost of acquisition per order. In a direct channel, every WhatsApp order generates a contact: name, number, purchase history. With 400 monthly orders over 6 months, the restaurant accumulates 800 to 1,200 unique contacts — assuming 2x recurrence — that can be activated with zero additional spend. The difference in acquisition cost is $0 versus the implicit 26% the app charges on every order it delivers. In 2026, in cities like Bogotá, Medellín, and Mexico City, that database is worth more than the immediate commission savings alone.

The app algorithm does not work for your margin

On delivery apps, listing position is decided by an algorithm, not by the restaurant owner. That algorithm changes without notice and rewards whoever invests in the platform's internal advertising, with ads inside Rappi that add to the overall cost on top of the base commission. The practical result: the item the restaurant most benefits from showing — by margin — is not necessarily the one the algorithm ranks first. On a direct channel, the owner designs the menu flow, decides which combo appears first, and can test prices in real time. I have seen restaurants increase their average ticket by 18% in 90 days simply by reordering their digital menu and adding an upsell suggestion — something impossible to control in a third-party app interface without paying extra.

Collection speed: the liquidity apps silently drain

Delivery apps pay 8 to 15 business days after the order. For a restaurant with a weekly payroll and suppliers that demand payment within 7 days, that lag is a real cash-flow problem: the owner finances the platform with their own working capital. A restaurant billing large sums a month can have a meaningful share of its revenue locked in app receivables at any point during the month. A direct channel with a payment gateway — Wompi, PayU, Mercado Pago — pays within 24 to 48 hours. With $400,000 COP in monthly tool costs and immediate liquidity on $20 million of direct sales, the real opportunity cost of capital tied up in apps exceeds $600,000 COP per month, counting only a 3% monthly rate on the retained capital.

How to migrate without losing volume: the method in 4 steps?

Migration from app channel to direct channel fails when done as an abrupt cut; it works when run in parallel for 60 to 90 days.

The Masterestaurant method applied in Bogotá and Medellín has four concrete moves. First, activate the direct channel without closing the apps — coexistence is mandatory during the transition phase. Second, incentivize the first direct purchase with an 8%–10% discount funded by part of the commission savings — the restaurant still earns more per order than through the app. Third, capture the contact of every new direct customer and build the WhatsApp database. Fourth, in month 3, review what percentage of volume has shifted to the direct channel and decide whether to reduce investment in the platforms' internal advertising. In the Bogotá north case, the migration in 8 weeks was achieved with no single discount exceeding 9%.

Verdict: when to stay on apps and when to leave

Delivery apps make sense when the restaurant is in discovery mode: fewer than 200 monthly orders, low ticket, unknown brand. There the platform sells discovery and the 26% commission is the cost of acquiring new customers — expensive, but justifiable. When volume exceeds 350 orders per month and the ticket passes $22,000 COP, the equation changes: you are no longer paying for discovery, you are paying to retain customers who already know your brand and whom you could have in a direct channel for 2%–3% in gateway fees. Diego F. Parra and Masterestaurant recommend a mixed model in 2026: apps for new customer acquisition, direct channel for the active base. No restaurant with more than 400 monthly orders should have 100% of its volume on third-party platforms; that model gives away margin and customer data every single month.

The 5 differences that hit your bottom line hardest

Commission: the app keeps $3.10-$4.20 of every $14 order; the owned channel only charges $0.28-$0.42 in gateway fees. At 400 orders a month, that gap exceeds $1,100. Data: apps don't share even the customer's phone number. Without that data there's no remarketing, no real loyalty program, no way to measure acquisition cost. Average ticket: on the owned channel the restaurant designs the order menu and suggests combos; on apps, the algorithm decides what shows first, and it's rarely your highest-margin dish. Visibility: app ranking depends on an algorithm that changes without notice; on the owned channel, the customer arrives directly because they already know the brand. Payout speed: apps pay in 8-15 days; the owned channel with a gateway pays in 24-48 hours, improving weekly cash flow.