DoorDash fees for restaurants: run the month, not the order.

I sell payment equipment — Clover, Square, Valor, PAX and Dejavoo — and I build restaurant websites with your own ordering inside them at $2,000 up front, then $149/mo. I am not a marketplace, I do not resell DoorDash, Uber Eats or Grubhub, and I make money when a restaurant owns its ordering. Read the recommendation knowing exactly that.

Which is why the recommendation is not the one you might expect. I am not going to tell you to delete the apps. They do something your own website does not, and pretending otherwise would cost you customers. What I will tell you is that the same commission charged on a stranger and on a regular are two different purchases, and only one is worth the money.

Every commission figure below is reused from this site's own ordering article. Marketplace tiers and rates change — the article notes Uber Eats raised its lowest tier from 15% to 20% in early 2026 — so confirm the current rate for your plan with each platform before you build a decision on it.

In 2026 DoorDash, Uber Eats and Grubhub all run tiered plans that generally land between 15% and 30% commission per order. The catch is structural: the cheapest tier usually buries you in the app's search and sends fewer orders, so most independent restaurants end up on the 25% to 30% tiers simply to stay visible. You are not really choosing between 15% and 30%; you are choosing between 30% and being hard to find.

Put a real ticket on it. On a $40 order at 30% commission, roughly $12 is gone before you have paid for a single ingredient or a minute of labor. For a category that often runs on single-digit profit margins, that is the difference between a busy night that makes money and a busy night that just makes noise.

And commission is only the headline. Underneath sit card-processing fees, marketing and promoted-placement fees to stay near the top of the app, the discounts and free-delivery promotions the platform encourages, and the drag of extra tablets, mistakes and refunds. Stack those on the base commission and the all-in cost of a marketplace order commonly reaches 30% to 40%.

The apps are not villains. Their real product is discovery: putting you in front of a hungry stranger who has never heard of you and would never have found you otherwise. For a new place, or one down a side street, that exposure can be worth a commission for a while. Nobody should tear up a listing that is bringing them first-time customers.

The expensive mistake is paying the same 25% to 30% on the regulars — people who already love your food and are using the app out of habit, because it is the option they can see. You are not buying discovery on that order. You are renting access to somebody you already earned, and renting it forever, because the marketplace keeps their contact details and order history rather than handing them to you.

A direct channel means customers order from your own site, the order lands in your kitchen like any other ticket, and you pay a flat monthly instead of a percentage — so cost per order stops scaling with success. Commission-free does not mean free of all cost: you still pay ordinary card-processing fees, because somebody has to move the money. What disappears is the percentage cut.

That is the package I build: a restaurant site with your own ordering inside it, installable to the home screen, running on your own card processing, at $2,000 up front, then $149/mo. The customer list stays yours, which is what makes a slow Tuesday fixable — you can actually contact the people who already like your food.

Then the second lever. Every direct order still carries a card fee, as do your in-store sales. A compliant zero-cost or dual-pricing setup — legal in all 50 states when set up and disclosed correctly — offsets that cost instead of taking it out of your margin on every sale.