You hand a $35 order to a DoorDash driver. By the time the deposit lands, you’ve kept somewhere around $24 — and if you ran a promotion that week, less. The “15% commission” on the marketing email was never the real number.
DoorDash and Uber Eats charge restaurants 15–30% commission per delivery order in 2026, plus payment processing, optional advertising fees, and packaging. Once promotions and refunds are netted out, the effective cost on a typical delivery order lands at 30–40% of the order total. Pickup commission is far lower — 6–10% — and both platforms now offer commission-free ordering on your own website for just a processing fee. This guide breaks down every line, does the math on a real order, and shows you exactly where your margin is going.
This is the pillar of our delivery-economics series. If you’d rather skip to solutions, the companion guides cover how to take online orders without paying commission and who actually owns the customer data in each model.
How much does DoorDash actually take from restaurants?
DoorDash runs a three-tier “partnership plan” model. The tier sets your commission rate and how much marketing the platform does on your behalf. Here’s the 2026 rate card, straight from DoorDash’s own merchant pricing pages — and note the rates differ between the US and Canada.
DoorDash US commission tiers (2026)
| Plan | Delivery commission | Pickup commission | Intro offer | What you get |
|---|---|---|---|---|
| Basic | 15% | 6% | 0% for 7 days | Smallest delivery radius, minimal marketing |
| Plus | 25% | 6% | 0% for 30 days | Wider delivery area + DashPass eligibility |
| Premier | 30% | 6% | 0% for 30 days | Largest reach, Sponsored Listings, 20-order guarantee |
Source: DoorDash Merchant Pricing (US). The Premier tier includes an order guarantee: accept fewer than 20 orders in a month and DoorDash refunds that month’s commission, per the same pricing page.
DoorDash Canada commission tiers (2026)
Canadian operators pay a meaningfully different (and on the entry tier, higher) rate card:
| Plan | Delivery commission | Pickup commission | DashPass orders | Intro offer |
|---|---|---|---|---|
| Basic | 20% | 10% | — | 0% for 7 days |
| Plus | 25% | 8% | 27% | 0% for 30 days |
| Premier | 29% | 8% | 29% | 0% for 30 days |
Source: DoorDash Merchant Pricing (Canada). Premier merchants who spend CA$100+ on marketing can claim a CA$50 monthly rebate.
The single most important line here is the gap between delivery and pickup commission. A Canadian café on the Basic plan pays 20% to have DoorDash deliver an order, but only 10% — half — when the customer walks in to grab it. That spread is the whole game, and we’ll come back to it.
How much does Uber Eats charge restaurants in 2026?
Uber Eats raised its US marketplace fees effective March 12, 2026 — so any older breakdown you find is now wrong on the entry tier. Here are the current US rates from Uber’s merchant pricing page and the fee-change notice.
| Plan | Marketplace (delivery) fee | Pickup fee | Notes |
|---|---|---|---|
| Lite | 20% | 7% | Raised from 15% → 20% in March 2026 |
| Plus | 25% | 7% | +5% on orders from Uber One members |
| Premium | 30% | 7% | Top placement, no Uber One surcharge |
| Self-delivery | 15% | 7% | You provide the driver; Uber provides the storefront |
Sources: Uber Eats Merchant Pricing (US) and Restaurant Dive’s reporting on the Uber Eats marketplace fee increase. The March 2026 update also lifted pickup fees from 6% to 7% and pushes Plus-tier Uber One member orders to an effective 30%.
Two operator-grade details people miss:
- The Uber One surcharge is real margin. On the Plus plan, an order from an Uber One subscriber costs you 30%, not 25%. Uber One members are exactly the high-frequency customers you most want to keep — and they’re the most expensive to serve through the marketplace.
- The “validated in-store pricing” condition. That 7% pickup rate applies when your Uber Eats menu prices match your in-store prices. If you mark up the menu on the app and don’t validate, pickup can run higher.
What’s the real take-rate? (The fees nobody advertises)
Commission is the headline. It is not the bill. Here are the other lines that come out of — or pile on top of — every marketplace order.
1. Payment processing. On marketplace orders, DoorDash and Uber Eats bundle card processing into the commission (good). But on their commission-free direct products, you pay it separately — about 2.9% + $0.30 per order (more below).
2. Advertising and Sponsored Listings. Both platforms sell promoted placement on top of commission. The Premier/Premium tiers fold some ad spend in, but most operators who want volume end up buying additional Sponsored Listings or running “Buy One, Get One” promos that come straight off the top. Independent restaurants routinely report effective costs of 30–40%+ once promotions are counted, per industry analyses like Rezku’s 2026 third-party fee breakdown.
3. Refunds and chargebacks. When a customer reports a missing item or a cold latte, the credit often comes out of your payout, not the platform’s pocket. One refunded order can erase the margin on several good ones.
4. Packaging. Delivery requires cups with lids, sleeves, bags, and seals that dine-in doesn’t. Call it $0.50–$1.50 per order in consumables you wouldn’t otherwise buy.
5. The customer-side fees you don’t see (but your customer does). DoorDash adds a service fee of roughly 10–15% of subtotal (≈$4 minimum), a delivery fee around $1.99–$5.99, and a small-order fee on tickets under ~$12, per DoorDash’s consumer fee page and breakdowns like Splitty’s 2026 DoorDash fee guide. You don’t pay these — but they inflate your customer’s total by 40%+, which suppresses repeat ordering and trains people to see your food as expensive. That’s a hidden cost to your brand even though it never hits your ledger.
A model: the “True Cost of a Delivery Order”
Advertised commission tells you what one line costs. To see what an order actually costs you, net out everything. Here’s a simple, repeatable formula:
Effective Take-Rate = (Commission + Ad/Promo share + Refund allowance + Incremental packaging) ÷ Order subtotal
And the dollar version — what you actually keep:
Net Margin per Order = Subtotal − Commission − Promo − Packaging − Food (COGS)
Plug in real numbers and the gap between the sticker rate and the true rate becomes obvious. The headline says 25%. The model usually says 33%+.
There’s a second cost the formula above can’t capture, and it’s the expensive one: the customer relationship. When an order comes through DoorDash, you don’t get the customer’s name, email, phone, or order history — the marketplace does. You can’t email them a slow-Tuesday offer, you can’t see that they used to come weekly and stopped, and you can’t bring them back without renting them again. Every marketplace order is a customer you paid to acquire and immediately gave away. (We go deep on this in who actually owns the customer data.)
Worked example: a $35 CAD order, marketplace vs. direct
Let’s make it concrete. A Toronto café gets a $35 CAD order. Food cost (COGS) is 28% — a realistic café number — so the food itself costs $9.80. Assume a modest 5% allowance for promos/refunds amortized across orders and $1.00 of incremental delivery packaging.
Scenario A — DoorDash Plus plan (25% delivery commission):
| Line | Amount |
|---|---|
| Order subtotal | $35.00 |
| − DoorDash commission (25%) | −$8.75 |
| − Promo/refund allowance (5%) | −$1.75 |
| − Incremental packaging | −$1.00 |
| − Food cost (COGS, 28%) | −$9.80 |
| = Net margin | $13.70 |
| Effective take-rate (fees only) | ≈33% |
Scenario B — the same $35 order through your own app or website (direct):
| Line | Amount |
|---|---|
| Order subtotal | $35.00 |
| − Payment processing (≈2.9% + $0.30) | −$1.32 |
| − Food cost (COGS, 28%) | −$9.80 |
| = Net margin | $23.88 |
Same food. Same customer. The direct order keeps $23.88; the DoorDash order keeps $13.70. That’s $10.18 more per order — about a 74% lift in what you take home — and on the direct order you also keep the customer’s contact info to bring them back for free.
Run it across volume: a café doing 300 delivery orders a month at $35 is leaving roughly $3,000/month — about $36,000 a year — on the table versus owning even half of those orders directly. For most independent cafés, that figure is the difference between a profitable year and a flat one.
Is there a commission-free option on DoorDash and Uber Eats?
Yes — and this is the part the marketplaces bury. Both platforms will let you take orders at 0% commission through your own channels, charging only payment processing:
- DoorDash Storefront / Commerce Platform: 0% commission; you pay a payment processing fee of 2.9% + $0.30 per order (Boost/Pro packages) or 3.3% + $0.30 with the Starter package, per DoorDash’s commission explainer. In Canada, the Commerce Platform is likewise 0% commission + 2.9% + $0.30.
- Uber Eats Webshop: a 2.5% order processing fee + $0.29 per order, with no marketplace commission, per Uber Eats’ merchant pricing.
- Uber Direct: white-label delivery using Uber’s couriers on orders you own, starting at $7.99 per delivery (same source).
These products prove the core point: the 15–30% you pay the marketplace is for discovery and the customer relationship, not for the technology. The order-taking tech costs roughly a card-processing fee. When a customer already knows you, paying a quarter of the ticket to “find” them is pure leakage.
The catch — and it’s a real one — is that Storefront and Webshop still live inside the marketplace’s ecosystem and branding. You’re renting their checkout. Which raises the obvious question: if the order-taking technology only costs a processing fee, why not own the whole channel — the app, the brand, the customer list — outright?
Where third-party delivery genuinely earns its cut
This isn’t a “delete DoorDash” argument. Be honest about what marketplaces do well, because for some orders it’s worth every point:
- Discovery. Millions of people open DoorDash and Uber Eats with no specific restaurant in mind. If you’re new, unknown, or in a high-foot-traffic delivery zone, the marketplace puts you in front of buyers you’d never reach. Stanford economist Jack Collison’s research, The Impact of Online Food Delivery Services on Restaurant Sales, found that 30–50 cents of every dollar spent through delivery apps is genuinely incremental — net-new business, not cannibalized from your dine-in.
- Logistics. They run the courier network, the live tracking, and the support queue. Building that yourself is expensive and slow.
- Trial. A marketplace is a cheap way to let new customers try you once.
The trap is the other 50–70 cents. Those are your regulars — people who’d happily order direct if you gave them a way — paying you 30% to reach a restaurant they already love. The strategy isn’t “marketplace or nothing.” It’s use the marketplace for trial and discovery, then move the relationship onto a channel you own so the second, third, and fiftieth orders keep their full margin. (For the playbook, see how to take online orders without paying commission.)
Are delivery commissions capped anywhere?
A few jurisdictions cap third-party fees, but coverage is patchy — don’t assume you’re protected:
- British Columbia (Canada): a permanent 20% cap on total food-delivery service fees under the Food Delivery Service Fee Act — the first permanent cap in Canada. The cap covers “core services,” and the act also bars companies from cutting driver pay to offset it.
- Most other Canadian provinces: the pandemic-era caps in Ontario, Quebec, Nova Scotia, and Saskatchewan were temporary and have lapsed, so full market rates (25–30%) apply again.
- United States: caps are city-level, not federal — New York City and San Francisco set permanent ~15% delivery-commission caps, but most US markets have none.
If you’re outside a capped market — which is most of Canada and most of the US — the platform’s standard tier is what you pay, and the 30–40% effective reality stands.
Audit your delivery economics this week (checklist)
You can’t fix what you haven’t measured. Block 30 minutes and run this:
- Pull last month’s payout statements from DoorDash and Uber Eats. Find your net deposit, not gross sales.
- Calculate your real effective take-rate:
(gross marketplace sales − net deposit) ÷ gross marketplace sales. If it’s above your headline tier, that’s the hidden-fee gap. - Separate delivery vs. pickup volume. Every pickup order you’re paying delivery commission on is money you can recover by steering it to a cheaper channel.
- Identify your repeat customers on the apps. These are the orders bleeding the most margin — they don’t need to be “discovered.”
- Check whether you’re in a fee-capped jurisdiction (BC, NYC, SF). If not, assume full rates.
- Price the alternative. Compare your effective marketplace take-rate against a direct channel’s cost (processing ≈3% + a flat monthly tool fee). Do the per-order math from the worked example above on your average ticket.
- Add a “order direct” nudge at every owned touchpoint — receipts, packaging, in-store signage, your Instagram bio — pointing regulars to your own ordering.
What “owning your channel” actually looks like
Once you’ve done the audit, the resolution is usually obvious: keep the marketplaces for discovery, and give your regulars somewhere to order that you control. Concretely, that means an ordering channel where you hold the brand, the customer list, and the margin — not a co-branded checkout inside someone else’s app.
For an independent café or restaurant already running Square POS, this no longer requires a developer or a five-figure build. Tany gives you your own branded iOS and Android ordering app — order-ahead pickup, self-running loyalty, eGift cards, push notifications, and web ordering — built on the Square menu and payments you already use, live in about 24 hours for $99 CAD/month with unlimited orders. There’s no per-order commission: a regular who orders four lattes a week comes through at processing cost, not 30%, and their contact details stay in your customer list so you can bring them back without paying to find them again.
The honest caveat from earlier still applies: if you have no audience yet, an owned app has no one to serve. Build awareness first — and a marketplace is a legitimate way to do that. But the moment you have regulars, paying a marketplace 25–30% to reach people who already know your name is the most expensive habit on your P&L. The math in this article is the case for breaking it.
The bottom line
DoorDash and Uber Eats charge restaurants 15–30% commission on delivery in 2026, and the effective cost — after ads, promos, refunds, and packaging — typically reaches 30–40% per order. Pickup is dramatically cheaper (6–10%), and both platforms already offer 0%-commission ordering on your own channels for little more than a processing fee — which is the clearest signal that the commission buys discovery, not technology.
Use marketplaces for what they’re genuinely good at: finding you new customers and moving the food. But every dollar of repeat business you let ride at 30% is margin you’re handing away — and a customer relationship you’re renting back month after month. Run the audit, do the math on your own numbers, and move your regulars somewhere you own.