What Third-Party Delivery Really Costs a Restaurant

Most operators know the headline commission on their delivery apps. Fewer know what a delivery order actually leaves them after the commission, the packaging, the promotions they agreed to, and the adjustments that show up on the payout a week later. That gap is where third-party delivery fees quietly change a restaurant's cost structure.

This is a walk through the real cost, using the rates the major platforms publish, and a simple way to check your own numbers.

Operators are already uneasy about the channel

In comments filed with the Federal Trade Commission in May, the National Restaurant Association shared preliminary results from a survey of 830 operators. Among restaurants that offer third-party delivery, 45% said those orders are not profitable, including 47% of full-service and 43% of limited-service operators. Two in three said the average fee they pay falls between 15% and 29.9% of the order.

The same filing notes that 83% of limited-service and 66% of full-service operators charge more for items ordered through a third-party platform than for the same items in the restaurant. Menu markup is now standard practice, and for good reason, as the example below shows.

What the platforms publish

Rates change, and many operators are on negotiated terms, so treat these as a reference point and check your own contract. As published on each platform's merchant pricing page at the time of writing:

DoorDash. Three marketplace plans: Basic at 15% on delivery orders, Plus at 25%, and Premier at 30%, with a 6% commission on pickup orders across plans. DoorDash says there are no monthly or activation fees for core marketplace service.

Uber Eats. Lite at 20%, Plus at 25%, and Premium at 30% on delivery, with pickup at 7% when the restaurant has validated in-store pricing. Uber's page notes that Lite can be lower in select cities. These rates reflect a change Uber made effective March 11; its merchant notice says Lite rose to 20% and Plus orders from Uber One members are charged 30%, and restaurants with custom rates saw an increase of 3 points, not to exceed 30%. Restaurant Dive described it as the first major adjustment to Uber Eats' fees in several years.

Grubhub. Grubhub separates marketing from delivery. Its published plans carry a 5%, 15%, or 20% marketing commission, and delivery fees for using Grubhub's drivers start at 10%. Grubhub also charges an order processing fee to cover card processing and fraud monitoring. A Plus-plan order delivered by Grubhub's fleet therefore starts around 25% before processing.

Three things are easy to miss in these tables:

• The tier affects more than your cost. Plans differ in features such as guest delivery fees and subscriber access. On Grubhub, for example, higher tiers come with lower customer delivery fees, which can affect order volume.
• Promotions are extra. Discounts and sponsored listings you opt into are charged on top of commission.
• Pickup is cheaper but not free. A 6% or 7% pickup commission still applies to orders the guest collects themselves.

Local fee caps: New York City as an example

Some cities limit what apps can charge restaurants. New York City made its caps permanent in 2021, when the City Council voted to hold delivery fees to 15% and most other fees to 5% per order. Grubhub's NYC pricing page notes that the Council changed the cap by law in 2025.

As the city's Department of Consumer and Worker Protection currently describes the rules, apps may charge NYC restaurants up to 15% for delivery, 3% for transactions, and 5% for basic service. They may also charge an enhanced service fee of up to 20%, but only if they offer a plan at the basic service fee level. In other words, an app charging the enhanced fee must also offer a basic-fee plan. If you operate in New York, confirm which one you are on.

Separately, the FTC in April asked for public comment on whether a national rule is needed for fees delivery platforms charge consumers. That inquiry focuses on what guests see at checkout, not on restaurant commissions.

A worked margin example (illustrative)

Consider a hypothetical order with a $40 food subtotal, a 30% food cost, and $1.50 in delivery packaging. The figures are illustrative, ignore card processing on all channels, and stop before labor, rent, and overhead.

At the same menu price, the app order leaves about $10 less than a takeout order at the counter. A 15% markup recovers a little under half of that gap on a 25% plan. Move to a 30% tier and the markup recovers less. Add a $5 discount you funded to win the order, and most or all of the markup is gone, depending on how the platform applies commission to discounted orders.

None of this means delivery is a mistake. A delivery order can be genuinely incremental, reaching guests who would never have walked in, and it often uses kitchen capacity that would otherwise sit idle. The point is to know which case you are in, by location and by platform.

Payout timing and reconciliation

Delivery revenue does not arrive the way POS revenue does. It arrives net, in batches, on the platform's schedule.

• Uber Eats uses a weekly remittance period from Monday to Monday, initiates deposits on Mondays, and notes that banks typically take one to two days. Stores can opt into daily pay for $1 per location per day.
• DoorDash pays weekly by default, with an option to enroll in daily payouts. Its monthly statements are available by the 5th, and DoorDash notes the monthly statement will differ from the sum of that month's payouts because payout dates are not the same as transaction dates.

DoorDash describes a payout as sales minus commission and fees and marketing spend, plus amendments such as error charges and other adjustments. That structure is typical, and it is exactly why a deposit rarely matches the delivery total in your POS.

A workable weekly reconciliation, per location and per platform:

1. Pull the payout report, not just the bank deposit.
2. Match gross sales and order count to the delivery orders in your POS for the same business dates.
3. Break out the deductions: commission, marketing and promotions, and adjustments or error charges.
4. Tie the net payout to the bank deposit.
5. Review error charges weekly. Refunds for missing or wrong items are often a kitchen or packing issue you can fix.

At three locations and three platforms, that is nine reconciliations a week. This is where many groups fall behind, and where delivery quietly becomes a line on the P&L no one fully trusts. We covered a similar pattern in multi-location restaurant inventory without spreadsheets: when data arrives from several systems at different times, the reconciliation ends up in someone's evenings.

What to do this week

• Confirm your plan and rate on each platform, for each location. Negotiated rates and old tiers are common.
• Calculate contribution per order by platform using your own food cost, packaging, and average promotion spend.
• Check menu parity: are your app prices set deliberately, or copied from the in-store menu years ago?
• Reconcile one week of payouts for one location end to end. The first one will show you where the gaps are.

If delivery reporting is one of several places your numbers live apart, it may be a sign of the problem described in signs your business has outgrown off-the-shelf software. We build connected systems for restaurant groups that bring payouts, POS sales, and costs into one view, and our pricing is published and month to month.

If you would like a second pair of eyes on your delivery numbers, you can book a conversation at any time.