How To Win With Third-Party Delivery Apps
RESTAURANT TIPS
Third-Party apps can drive a lot of business, but what is the true cost? Here's a plain-math teardown of what a delivery order really costs a pizzeria, why a promoted order can lose money, and the one break-even calculation that tells you whether or not more in-house delivery might make sense for you.
The TL;DR
The Number on the Ticket Isn't the Number You Keep
The third-party delivery apps can act very tricky behind the scenes, always using one main strategy: they never show you the whole bill. A $30 order hits the tablet and your brain files it as a win. What you actually got was a gross number with four or five deductions hiding behind it. The commission they take is the only deduction you know by heart, because that's the number they quoted you.
But the commission is just the first layer. Promotions they charge you for, refunds you ate. Paid placement you bought to stay visible. In-app prices you raised to claw the commission back, which quietly cost you orders. Spread across five screens, it's tolerable. Added up on one, it's not. So let's add it up.
What's Actually Stacked Behind the Commission Line
Start with the part everyone knows. Here's the headline commission by platform and tier, the number you get quoted before anything else loads on top of it.
| Platform | Entry tier | Mid tier | Top tier |
|---|---|---|---|
| DoorDash | 15% | 25% | 30% |
| Uber Eats | 20% | 25% | 30% |
| Grubhub | 5% | 15% | 20%* |
*Grubhub's tiers are marketing commission. Add roughly 10% more if Grubhub also runs the delivery, rather than in-store pickup.
Now watch what loads on top of that headline fee. Four layers most owners never add up:
- The marketing-tier trap. The algorithm openly favors higher tier customers, so shops buy up into Plus or Premier just to stay visible. The rate most pizzerias actually operate at is the top of the range, not the bottom. They're paying 30% and calling part of it advertising.
- The Uber One squeeze. In March 2026, Uber raised two of its three tiers and started charging Plus restaurants the full 30% on every order from an Uber One member. Members are the heaviest orderers, so for most shops the "25% plan" is a 30% plan in practice.
- Co-funded promotions. Free-delivery and buy-one-get-one offers the app runs are partly paid by you, and you're often opted in by default. You fund the discount and pay commission on the discounted total.
- The quiet one: menu price inflation. To claw commission back, owners raise their in-app prices. That suppresses their own demand and teaches their best customers that their pizza is expensive. You pay a fee to make your own product less competitive.
The commission rate is what they quote you. The stack is what you actually pay, and you've never seen it added up in one place.
The Order That Loses Money
Take a normal order and hold it constant. Two pizzas, $30, the kind you fill every night. This is obviously a hypothetical scenario, not a full profit-and-loss analysis. First, what's left after you actually make the food, before the channel takes its cut:
| Order total | $30.00 |
| Food cost (about 30%) | −$9.00 |
| Incremental labor and packaging (about 25%) | −$7.50 |
| Contribution before channel cost | $13.50 |
You've got $13.50 to cover the delivery costs and still make money. Now run the same order through each channel and see what you keep.
| Channel | Channel cost | You keep |
|---|---|---|
| Carryout / direct pickup | Card processing, about 3% (−$0.90) | $12.60 |
| Direct delivery (own or flat-fee driver) | About $6 to $8 | $5.50 to $7.50 |
| Third-party at 30% | Commission −$9.00 | $4.50 |
Same pizza. On the third-party order the app keeps almost as much of the cash as you do, for handing you an order you'd half-earned already.
Then it gets worse. Here's a promoted single pizza, $18, on a 30% tier, with a co-funded free-delivery offer running. This is the order owners screenshot and forward to each other.
| Order total | $18.00 |
| Food, labor, and packaging (about 55%) | −$9.90 |
| Commission at 30% | −$5.40 |
| Co-funded promo (free delivery / discount) | −$3.00 |
| Take-home | −$0.30 |
Turn It Into a Strategy, Not a Grievance
Here's where the math has to stay honest. Direct ordering isn't free. The apps charge a percentage of every order, forever. Your own channel costs a flat amount per month. At low volume, the apps win. Past a certain point, the percentage becomes the most expensive thing on your books.
That point is lower than you think. Every dollar you move off a 30% app onto your own channel at about 3% processing keeps 27 cents in your pocket. If direct ordering runs you an illustrative $200 a month, you break even at about $740 in moved revenue. That's a couple dozen orders. Everything after that is margin.
The usual objection is that in-house delivery is a headache: dispatching drivers, tracking runs, figuring out who's back. That's true if you're running it off a whiteboard. It's not true with a system built for it. Adora puts driver dispatch and tracking (read about it here) inside the same POS that runs your counter, so taking delivery back in-house is an upgrade, not a second job.
See where your delivery margin is actually going.
We'll use your real numbers and show you how much of your channel you could be owning instead of renting.
Schedule a Demo →Why This Hits Harder in 2026
This math would sting in any year. It's worse now because everything underneath it is moving the wrong way. In a mid-2025 industry survey, 91% of operators reported higher food costs year over year and 89% reported higher labor costs. Labor commonly runs 25% to 35% of sales for a delivery-heavy pizzeria, and wages keep climbing. Cheese alone, the input that ultimately sets your food cost, swung roughly 24 cents a pound inside a few weeks.
Stack a 30% channel cost on top of that and the room left over is thin. Operators who'll cite a 7% to 15% net margin on paper often describe a 2% to 3% reality after labor and commission. When your prime cost is already running 55% to 65% of sales, the delivery tax isn't an annoyance. It's the difference between a profitable month and a flat one.
What You Can Do About It
You don't fix this with a grand strategy. You fix it by making your own channel the obvious choice for the customer's second order. Four moves, none of them expensive:
- Put a card in every third-party box. A small insert with a real reason to order direct next time, a code, a free side, a better price. The app gave you the introduction. Use the box to ask for the relationship.
- Make direct cheaper than the app. If your in-app prices are inflated to claw back commission, your own site should clearly beat them. Customers do the math. Reward the ones who order directly from you.
- Own first-party ordering and the data behind it. Run online ordering you control, where the customer's name, order history, and contact info are yours, not rented back to you in an app dashboard. That's what lets you bring them back on purpose.
- Treat the apps as paid acquisition, and measure them like it. A new customer off DoorDash is fine if you convert them to direct. If a customer orders from the app three times and never comes to you, you didn't acquire a customer; you're renting one.
The Apps Aren't the Mistake. Leaving the Customer There Is.
Don't read any of this as "quit the apps." That's the wrong call. Third-party delivery is a customer-acquisition channel with a brutal cost of acquisition. Used right, it gets you found by people who'd never have found you otherwise. The mistake isn't using it. The mistake is leaving the customer there.
Let the apps introduce you once. Then move that customer onto your own channel for every order after, because the second order on your own channel is where the margin lives. Online orders already run 15% to 20% higher in average ticket than phone orders, and carryout is beating delivery on customer satisfaction. The economics and the experience point the same way: get them once on the app, keep them in your own system.
If you want a robust system built for managing your own customers and drivers, that's the conversation Adora was built for.
People Also Ask:
"The headline commission runs 15% to 30% on delivery depending on the tier, with DoorDash at 15/25/30, Uber Eats at 20/25/30, and Grubhub's marketing tiers at 5/15/20 plus roughly 10% more for delivery. But that headline is only the first layer. Once you add the higher tier most shops buy to stay visible, co-funded promotions, paid placement, and refunds the platform tends to grant the customer, the real per-order cost commonly lands between 30% and 40%. The number you were quoted is almost never the number you actually pay."
"Yes, and it happens more often than owners realize. Take an $18 single pizza on a 30% tier with a co-funded free-delivery promotion. Food, labor, and packaging eat roughly $9.90, the commission takes $5.40, and your share of the promotion takes another $3.00. That leaves you about 30 cents in the hole before you've counted a single fixed cost. You made the pizza, paid to advertise it, and paid to give it away. Promoted low-ticket orders are where the math most often turns negative."
"Above almost any real volume, yes. The honest comparison is variable versus fixed. Apps charge a percentage of every order forever, while direct ordering costs a fixed monthly amount. Moving a dollar off a 30% app to your own channel at about 3% processing saves roughly 27 cents. Against an illustrative $200 a month for direct ordering, you break even once you move about $740 a month to direct. A shop doing $10,000 a month in app delivery pays around $3,000 in commission, so shifting even a third of it puts about $1,000 back in pocket every month."
"No. The apps are a customer-acquisition channel with a high cost of acquisition, not an enemy to be cut off. They get you found by people who'd never have found you otherwise. The mistake isn't using them, it's leaving the customer there. The smart play is to let the app introduce a customer once, then move that customer onto your own ordering channel for every order after, where the margin lives. Quitting the apps outright leaves money on the table. Owning the relationship after the first order is what protects it."
"Don't count on it. There's no federal commission cap, and the push for a 15% national cap has stalled and isn't close. The pandemic-era emergency caps in cities like New York, San Francisco, and Chicago proved caps are enforceable, but what exists today is a thin and inconsistent state and local patchwork. Meanwhile the trend is moving the other way: in March 2026 Uber raised two of its three tiers. Waiting for regulation to rescue your margin isn't a plan. Owning more of your channel is."
Sources & Data
Commission tiers: DoorDash merchant pricing, Grubhub pricing and fees. March 2026 Uber Eats fee changes and the Uber One surcharge: Restaurant Business and Restaurant Dive. Food and labor cost survey (91% / 89%): Restaurant365 Midyear State of the Industry, 2025. Block cheese pricing and 2025 volatility: Pizza Today, tracking the CME block cheddar market. Commission cap landscape and the New York settlement: Restaurant Business and Crain's New York. Online versus phone order ticket size: Thrive POS. Per-order cost breakdowns are illustrative models, not audited figures.
