How much do delivery apps charge restaurants? (2026 rates)

Short answer: in 2026 the marketplaces list commissions of 15–30%, and the list price is only the start. DoorDash publishes 15/25/30% tiers (6% pickup in the US). Uber Eats quietly reset its schedule in March 2026: the entry tier rose from 15% to 20%, and surcharges on member orders push effective mid-tier rates to 30% on a big slice of volume. Add sponsored placement, self-funded promotions and service fees, and industry measurements land the true cost at 30–40% per order, more than the net margin of almost any dish. Below: the rate tables, a worked $40 order against a direct channel, a calculator for your own numbers, and the hybrid strategy that uses marketplaces without being owned by them.

The published rates (and the 2026 changes)

Two structural notes. The tiers are visibility auctions: the commission buys ranking, radius and marketing inclusion, so the “cheap” tier often produces few orders, nudging venues up the ladder. And the 2026 Uber Eats change is the direction of travel: platform take rates have ratcheted upward for a decade, because once a venue’s volume depends on the app, the platform sets the price of that dependence.

Why the real rate runs 30–40%

• Sponsored listings: increasingly the price of first-page visibility, typically several points of sales.
• Promotions: platform-pushed discounts (“20% off”) are funded from YOUR payout, on top of commission.
• Service and processing layers: order fees, tablet fees, integration charges vary by market and stack quietly.
• The measured result: analysts in 2026 consistently find true per-order costs 5–10 points above headline commission.

One $40 order, two channels

The last row is the compounding one: the marketplace order is anonymous, the direct order joins your list, earns points, and can be brought back with a free email instead of a paid placement. Over a year of repeat orders the gap is not $10, it is a customer lifetime.

Run your own numbers

The hybrid strategy (use them, do not be owned by them)

Marketplaces are a paid acquisition channel, and sometimes a good one: discovery in a new area, filling idle capacity, reaching app-native customers. The discipline is keeping them at the acquisition job while the repeat business moves direct: mark up marketplace menus 10–20% (within each platform’s current rules) so your own channel is visibly cheaper, put loyalty and first-order offers only on direct, print the ordering link on everything you legally control, and watch the mix monthly. A venue at 40% marketplace share is busier and poorer; the same venue at 15% marketplace / 85% direct keeps the margin AND the customer file. The full migration playbook is in delivery strategy and the Uber Eats deep dive.

Mistakes that hand over the margin

• Judging the channel by the headline commission instead of the all-in rate.
• Letting platform promotions run “for visibility” without pricing what they cost per order.
• Same prices on the app and your own channel, paying 30% to undercut yourself.
• No direct alternative at all, so every regular becomes a 30% customer forever.
• Never checking the mix: marketplace share creeps up quietly until the platform is your landlord.

Frequently asked questions

How much commission do delivery apps charge in 2026?
Published tiers: DoorDash 15%, 25% or 30% by plan (6% on pickup orders in the US); Uber Eats moved its entry tier from 15% to 20% in early 2026, with mid-tier orders carrying surcharges that push effective rates to 30% on a large share of volume, and 30% on the premium tier. Grubhub and regional platforms sit in similar bands. These are the list prices; the measured all-in cost usually runs higher.

Why is the real cost higher than the listed commission?
Because the commission is only the entry fee. Sponsored placement (increasingly necessary for visibility), platform-run promotions funded from your payout, delivery-radius and service fees, card processing inside the stack, and tablet or integration charges stack on top. Industry analyses in 2026 consistently measure true per-order costs of 30–40% for venues that participate in normal platform marketing, roughly 5–10 points above the headline rate.

Are delivery apps ever worth the commission?
Yes, for specific jobs: discovery in a new market, filling genuinely idle kitchen capacity, and reaching customers who will only ever order inside the apps. The maths works when marketplace orders are INCREMENTAL (volume you would not otherwise get) at a contribution above zero, and fails when they cannibalise direct orders. The standard playbook is hybrid: marketplaces for acquisition, your own channel for the repeat orders that make up most volume.

What does a direct online ordering channel cost instead?
Software from about NZ$29 a month plus payment processing (card fees of roughly 2–3% exist on every channel), and on Plattr, commissions from just 2.5%. On a $40 order that is roughly $1.00–1.10 of platform cost versus $12 or more at a 30% marketplace rate. Direct also hands you the customer relationship: the email, the order history, the ability to bring them back for free.

How do I move marketplace customers to direct ordering?
Make direct visibly better: slightly lower prices (you can undercut your own app listing and still earn more), loyalty points and first-order offers only on your channel, faster pickup lines, and your ordering link everywhere the customer already looks (Google profile, socials, receipts, packaging). Follow the platform rules about in-order inserts, and let the maths do the persuading: most regulars switch for a cheaper, points-earning order.

Should I raise my menu prices on delivery apps?
Most venues do: a 10–20% marketplace markup partially recovers the commission, and platforms broadly tolerate it (check each platform’s current parity rules). It narrows the loss per marketplace order while making your direct channel the obviously better deal, which is exactly the signal you want price-sensitive regulars to receive. Keep your own channel at menu price and say so out loud.