Uber Eats vs DoorDash vs Grubhub: the honest math, 2026.

So read this as a decision, not a leaderboard. The question on your desk is never “which platform wins” in the abstract — it is which one to keep, which to drop, and whether a given order type belongs on a platform at all, given the restaurant you actually run. For most independents the verdict lands the same way: keep DoorDash, weigh Uber Eats against your local discovery surface, and drop Grubhub first. That single line is the whole answer; everything below is the walk that gets you to your version of it — branch by branch, with the fee math read from each platform’s public merchant documentation and not one number invented. The headline rate is roughly the same across all three. What separates them is the layer of incidental fees stacked on top, the cap on which orders you can actually opt out of, and what each platform does to your direct-channel conversion. This is the side-by-side I wish someone had handed me in 2022 when I first started keeping spreadsheets.

You are not picking a favorite platform. You are deciding which discovery tax is worth paying, for which orders — and which one you can stop paying tomorrow without losing a single guest.

The same $42 ticket, three platforms

The example uses a $42 ticket — a representative independent-restaurant order on a third-party platform, the same anchor used in the DoorDash margin walk. Your ticket will be different. The percentage spread is what carries the argument, not the dollar amount.

The DoorDash row is repeated from my earlier DoorDash margin walk. The headline numbers are about the same across the three platforms. The story is in the spread — eighteen point four to fifteen point nine percent — and where that spread comes from.

Where the spread actually comes from

The base commission isn’t the variable to focus on. Each platform offers a 15%/25%/30% tiered structure (with slightly different names and inclusions). Most independent restaurants land on the 30% tier because the 15% and 25% tiers don’t include placement in the in-app discovery surfaces — you exist but nobody finds you. The 30% is the price of being findable. That part doesn’t differ much across platforms.

What differs is the second layer:

1. DoorDash adds a sponsored-listing fee when you opt in to the in-app promo system. Roughly 3% additional, applied per-order on promo-attributed orders. Opt-in — you can decline it — but the discovery throughput drops sharply if you do.
2. Uber Eats has the Pro membership fee — about $2.50/order on opted-in orders — which gets you placement in the “Pro” tier of the app. Like DoorDash’s sponsored listings, it’s opt-in but functionally required for the discovery surface to work.
3. Grubhub has a per-order “communication fee” (their language) of roughly 1.8% that’s not opt-in. It funds the order-relay system. It looks like a smaller surcharge but applies to every order, which is why the Grubhub row ends up close to DoorDash overall despite the lower promo cost.

The marketing-fee trap nobody talks about

This is the part operators miss most. The 30% commission is on the menu price, not the take-home. You can’t deduct food cost or labor from it before the percentage applies — it’s 30% off the top. Most operators’ mental model treats it as a marketing cost, which is wrong; marketing budgets compete against profit, not against revenue. The three platforms are aggregators in the classic sense: discovery surface for one side, margin tax on the other.

The math: at a typical 12% operating margin and a 30% platform commission, every third-party order has to drive 250% incremental revenue compared to the same order at the door for the platform spend to break even on a marketing basis. It almost never does. The customer would have come anyway 80% of the time. The platform is, structurally, a 30% margin tax for ~20% incremental revenue. That’s the trap.

Which one to drop, if you’re going to drop one

For most independents the answer is Grubhub. Three reasons. One: the order volume per Grubhub channel is the lowest of the three in nearly every DMV market I’ve measured, by a factor of three to five. Two: the kept-percent at 17.6% is in the middle of the pack, so dropping it costs the least kept margin per lost order. Three: Grubhub has the highest rate of phantom-listing complaints in this market — restaurants showing up on the platform without their consent or with stale menus — which is a brand-control problem on top of the margin problem.

The harder call is dropping Uber Eats. The kept-percent is the worst of the three. But Uber Eats has the highest first-order conversion to a return customer in this market, and the in-app cross-pollination with Uber rides means the platform brings discovery volume that the other two don’t. If your restaurant is in a high-foot-traffic urban node where Uber rides are a meaningful customer-flow source, Uber Eats is probably the one to keep.

Before the walk, put the two ends of the decision side by side — the platform you almost always keep against the one you almost always cut first. Everything in the tree below is a variation on choosing between these two columns.

I’m on DoorDash and Uber Eats at one restaurant and DoorDash only at the other. Grubhub I dropped two years ago and have not regretted. The dropped Grubhub volume came back to the restaurants through other channels within a quarter — almost entirely to direct — suggesting the Grubhub orders were displaced, not net-new.

So don’t take the verdict on faith — walk the spine. Three questions, top to bottom; the first one that names your real situation hands you the move and stops you.

What this changes about your channel mix

If you’re on all three: pull a Grubhub partner statement for the last 90 days. Calculate the contribution margin (revenue minus the full third-party fee stack minus your variable cost per order). Compare to the direct-channel contribution on the same restaurant. If Grubhub is contributing less than 8% of total third-party volume and your direct channel has any pulse at all, the move is straightforward. Pause, watch the channel mix for a quarter, then delist if the pause held.

If you’re on two: the question is whether to drop to one. The math says yes most of the time, but the political cost of phantom listings (Grubhub especially) and the operator time needed to police your menu across two platforms vs one are real, just less measurable. I keep DoorDash on the restaurant because the kept-percent is the highest. I drop the second platform restaurant-by-restaurant based on the local discovery surface.

Don Goldstein is a restaurant operator and runs Muntin Digital. The three-platform comparison above runs against partner statements dated April 2026.