Delivery Apps for Restaurants: Step by Step, Pricing and Alternatives 2026

When does the delivery app fall short?

The app falls short when every order comes in with a negative contribution margin and you make up for it by selling more, because at that point volume stops being growth and turns into a leak.

The proof sits in the platform's statement: add up the commission, the payment processing fee and the co-funded promotions, and if the total takes more than the dish leaves after ingredients and packaging, the app is working against you. Other signals arrive more slowly. The customer who reorders through the app doesn't know your name, only the platform's, and the day it changes the algorithm or raises the fee you have nobody to call. At Masterestaurant that review is done dish by dish, never with the menu average, because the average hides exactly the three or four dishes that lose money on EVERY order.

Alternative 1: a separate price for the app

Raising prices on the app is the alternative with the lowest switching cost and the first one I recommend, because it requires no in-house delivery and no second kitchen, just a separate price list. It is what most operators already do: 83 % of limited-service operators in the U.S. charge more for dishes ordered through third-party apps than in the dining room or for takeout, according to the National Restaurant Association's comments to the FTC (2026). It suits owners whose dining room works and who use the app as a secondary channel, with a short menu they already master. The effort fits in one afternoon of menu engineering: recalculate every dish with the commission built in and drop the ones that can't carry it. The downside is real, since a higher price lowers conversion on the screen, so raise your highest-margin dishes first and measure for two weeks before touching the rest.

Alternative 2: direct orders through your website or WhatsApp

A direct channel gives you back the customer and the commission, but it charges you in logistics and in customer acquisition, and it only makes sense for owners who already have a base of guests who search for them by name. Think of a neighborhood taquería with loyal regulars in a short radius: there a menu on the website, a WhatsApp number and two in-house drivers can absorb a good share of repeat orders. The switching cost is heavy at the start. You pay for delivery even when no orders come in, and you have to get customers to change a habit, which almost always takes an incentive on their first direct order. Transparency plays in your favor, since in December 2024 the FTC reached a 25 million dollar settlement with Grubhub for hiding the real cost of its service, per Arnall Golden Gregory's analysis (2026), and diners who learn what they really pay on the app appreciate a clean price on your own channel.

Alternative 3: a virtual brand built for the app

A virtual brand is the way out for owners with idle kitchen hours who want to sell on apps without dragging along the dining room's menu or prices. It works because off-premise demand already outweighs dine-in across much of quick service, and that demand buys on a screen. The new brand is designed the opposite way from the dining room, packaging and travel time first, the recipe after, with a few dishes that share ingredients with the existing operation so inventory doesn't double. Its switching cost is moderate: it needs no second location, but it does need photography, a separate account on each platform and a station that doesn't get in the way of table service. The risk is dispersion. If nobody in the kitchen owns the virtual brand, the app order and the dining room ticket fight over the same burner at peak hour, and both lose, the guest at the table and the one on the screen.

Alternative 4: takeout with counter pickup

Counter pickup is the cheapest alternative of all and the most overlooked one, because it removes the delivery commission and lets the customer pay with their own time. It fits restaurants in office districts or high-traffic streets, with dishes that go out in under fifteen minutes and a counter that already exists. The change costs little: a takeout menu on your website, a pickup schedule the kitchen can actually keep and decent packaging. What it doesn't give you is reach. Someone who has never walked past your door won't discover you this way, and there the app always wins. That is why the sensible combination uses the platform so people find you and the counter so they come back, with a flyer in every bag inviting them to pick up their next order at a price the app can't match. That flyer, done well, is the most profitable piece of your entire packaging.

Should you be on one app or several?

Be on the app that dominates your city and add a second one only once the first is already leaving margin, because each extra platform doubles the kitchen's load in tablets and in reconciliation.

Concentration varies by country. DiDi Food held 38 % of monthly active users of delivery apps in Mexico in 2024, while in Argentina PedidosYa had 61 %, according to Sensor Tower (2024). In a market as concentrated as Argentina's, exclusivity has almost no bargaining value; in a fragmented one like Mexico's, a second app can open zones the first one doesn't reach. And here is the trade's real tension: more apps bring more orders but less control. The answer is to manage them as a portfolio with rules, with the same app price list on all of them, one person who owns the weekly reconciliation and the authority to switch off any app that doesn't cover your margin within a quarter.

When NOT to change your model?

Don't change your model if your dishes already absorb the commission and the app brings you customers the dining room would never reach, because in that case staying is the right call and moving only costs you money.

In the U.S., platforms charge a typical commission of 15 to 30 percent per order, plus delivery and payment processing fees, according to Wharton Magazine (2025), and some operations with app pricing and a short menu do well inside that range. What would happen if one of them launched in-house delivery out of pride? It would pay drivers during empty hours, lose the platform's visibility and, within a few months, sell less with a higher fixed cost. Diego F. Parra sets it as a rule of the Masterestaurant method: you switch channels when the cash register asks for it, NEVER when your anger at the commission does. Open this month's statement and decide with it in your hand.

What changes when delivery is designed before you sign up?

The difference between before and after is not which app you pick, it is who controls price.

Total expenses at the average US restaurant rose 36% between 2019 and 2026, according to the National Restaurant Association, so the cushion that once absorbed a sloppy commission is gone and every commission point now comes straight out of contribution margin. Mexico City already has 1,200 ghost kitchens operating on delivery apps, per Canirac data reported by Milenio (2026), all fighting for the same screen as your restaurant. A ghost kitchen or virtual brand is born with app pricing, app packaging and an app menu; a dine-in operator has to redesign those before competing on equal terms. Here is the paradox: the app that brings you new guests also trains your regulars to order through a channel where you pay commission. The Masterestaurant rule settles it with no middle ground: the aggregator is where guests MEET you, your own channel is where they COME BACK, always within each platform's rules.

What changes when delivery is designed before you sign up — in practice?

Relying on one platform is a board-level risk. In Argentina a single app holds 61% of monthly active delivery users (Sensor Tower, 2024), and a channel that heavy sets the terms.

Picture your main platform adding five commission points tomorrow: if, for example, half your sales run through it, those points come straight out of profit because rent and payroll do not drop, and the price fix arrives after the cash crunch. The decision tree fits in four questions. Does the dish absorb the commission with food cost under 32% of net price? If not, redesign before signing up. Do you already have repeat guests? Then open your own ordering channel in month one. Is there a local marketplace with neighborhood customers? Use it as a second leg and measure it for four weeks. Do daily orders in a short radius justify a staff driver? Only that density makes in-house delivery cheaper than outsourced.