How platform commission changes delivery economics

What the commission is actually charged on

A delivery platform charges the restaurant a percentage of the order, deducts it before paying out, and reports what is left. That single sentence hides three questions that decide the real cost: what the percentage is applied to, what else is deducted next to it, and which of those deductions move when you change your prices.

The percentage is normally applied to the value of the food in the basket. Around it sit deductions that are not the commission but arrive on the same statement and leave the same way: payment processing, in-app advertising, the restaurant-funded part of promotions, refunds and adjustments for orders that went wrong, and, depending on the market and the contract, packaging or service items. Tax treatment differs by country — in some the commission is invoiced with VAT on top, in others it is netted off differently — which is another reason two restaurants quoting "the same rate" can be paying different amounts.

So the honest way to think about the commission is not as a single number in a contract but as everything the platform keeps between the customer's payment and your payout. That total, divided by what customers actually spent with you, is the number that governs your economics.

The order of the lines and their exact names differ between platforms and between countries. What does not differ is the split into costs that scale with the basket and costs that do not — that split is what makes delivery economics behave differently from dine-in.

Why your actual rate is not the published one

Public price lists exist to be quoted in the press and to give a new restaurant a starting expectation. The rate that appears on your own statement is the outcome of your contract, and it drifts from the published figure for reasons that are all ordinary commercial practice:

• Service level. A platform normally charges differently depending on whether it delivers the order with its own couriers, hands it to your own driver, or the customer collects it. The same restaurant can sit on several of these at once, at different rates.
• Negotiated and volume terms. Larger accounts, groups and chains agree terms individually. Two restaurants on the same street can be on different contracts.
• Launch and campaign periods. Reduced rates for new partners, seasonal campaigns and market-entry offers are common, and they expire — often quietly, in the middle of a month.
• Country and city. Rates are set per market. A group operating in two countries should expect two different sets of terms.
• Everything charged next to the commission. Even when the headline rate is exactly the published one, advertising, promotions and refunds land on the same statement and change what you actually keep.

This is why we do not publish platform rates anywhere on this site, including in the calculator on the home page: any number we printed would be wrong for most readers, and would be quoted back as if it were a benchmark. The only rate that describes your restaurant is the one on your own invoice, and the only rate worth planning with is the effective one you can compute from it — see how to work out your own effective rate below.

What commission does to a single order

Take one order and split its costs in two: costs that are a share of the basket, and costs that are fixed per order. The commission and the promotion discount are in the first group. Packaging, the labour of preparing and packing, and the time at the pass are in the second. Delivery may sit in either group depending on the model.

Two consequences follow, and they drive most of the day-to-day decisions on a delivery menu.

First: small baskets are the dangerous ones. The fixed cost per order is roughly the same whether someone orders one item or four, but a small basket has far less gross margin to absorb it. Below a certain basket value an order stops paying for itself, and that threshold is higher on delivery than on the same menu in the dining room, because the commission has already taken its share before the fixed costs are covered.

Second: price increases are partly shared with the platform. Because the commission is a percentage of the basket, part of every price rise goes straight back out as a larger commission. You keep the rest — so raising prices still works, it simply works less than the increase suggests, and needs to be sized with the commission already priced in rather than added on afterwards.

Why a percentage of gross revenue is the wrong measure

Gross revenue on a delivery platform is not money you received. It is the sum customers paid, before the platform kept its share, before restaurant-funded discounts, before refunds, and in most dashboards before the advertising you paid for separately. Judging performance — or paying anyone a fee — on that number has three specific failure modes.

It charges you for money that never arrived. A percentage taken from gross revenue is taken from the platform's commission as well as from your own margin. The larger the commission in your market, the larger the share of that fee that is being charged on money the restaurant never saw.

It rewards the wrong kind of growth. Discount-driven volume raises gross revenue reliably. It raises the platform's commission too, raises your fixed costs per order, and can leave less in the bank at the end of the month than before the campaign started. A measure based on gross cannot tell those two months apart.

It hides mix. Two months with identical gross revenue can be completely different businesses underneath: different average basket, different share of orders on each service level, different promotion load. Only the net view shows this.

This is the same logic our own fee follows: we calculate on what is left after the platform commission and the advertising spend, never on gross revenue. The step-by-step version is on the home page.

How restaurants compensate for the commission

The commission is not negotiable for most independent restaurants, so the work goes into the lines that are. In practice there are four, in roughly this order of reliability.

Delivery-specific pricing

A menu priced for the dining room and copied to a platform is priced as if the commission did not exist. The fix is to set delivery prices with the commission already inside them, item by item, rather than applying one blanket uplift across the menu. Where a platform restricts price differences between channels, the contract terms decide what is allowed — check them before changing anything.

Raising the basket rather than the price

Combos, sets, sides and drinks add gross margin to an order without adding another set of fixed costs. This is usually the most durable answer to the commission: it improves the ratio of variable margin to fixed cost per order without asking the customer to pay more for the same thing.

Modifiers and structure

How the menu is built decides what a click is worth. Well-structured modifiers, portion options and clearly separated add-ons raise the average basket, and they also reduce refunds by making the order easier to get right. Refunds are a real cost line, not an edge case.

Promotions and advertising treated as investments

Both are optional costs that must return more than they consume. The question is never "did revenue rise while the campaign ran" but "did it rise by more than the campaign cost, after the commission on the extra orders". Our longer guide on Uber Eats advertising works through the reporting side of that question in detail.

What none of these do is remove the commission. They change the shape of the order so that the commission is charged on a basket that can carry it.

How to work out your own effective rate

This takes about fifteen minutes per platform and needs no tools beyond the statements you already receive. Do it per platform and per month — the numbers are not comparable across platforms with different service models.

• Take one full calendar month, and use the platform's own invoice or payout statement rather than the dashboard summary.
• Write down gross sales for the month: what customers paid for food, before any deduction.
• List every deduction separately: commission, payment and service items, the restaurant-funded part of each promotion, advertising spend, refunds and adjustments.
• Add them up and divide by gross sales. That is your effective rate for the month — the real cost of selling through that platform.
• Repeat for the two previous months. The trend matters more than any single month, and a sudden move usually points to an expired launch rate, a promotion that stayed on, or a refund spike.
• Do the same per service level if you use more than one. A blended figure across collection and platform-delivered orders can hide a channel that is not paying for itself.

Once you have it, use it. Any decision about pricing, combos, promotion depth or minimum basket should be tested against your effective rate, not against the rate you remember from the contract you signed.

Terms used on this page

Commission

The platform's share of an order, charged as a percentage of the basket and deducted before the restaurant is paid. Set by contract, per market and per service level.

Published rate

The percentage a platform states publicly. A starting point for expectations, not a description of any particular restaurant's contract.

Effective rate

Everything the platform kept in a period — commission plus every other deduction — divided by gross sales in the same period. The number that describes what selling through the platform actually costs you.

Gross revenue

What customers paid, before commission, promotions, refunds and advertising. Reported prominently by every platform dashboard; useful for demand, misleading for margin.

Payout

What the platform transfers to the restaurant's account after all deductions. Payout periods and adjustments rarely line up exactly with calendar months.

Average basket

Gross sales divided by the number of orders. The main lever a restaurant has against fixed cost per order.

Fixed cost per order

Costs that are the same whatever the basket is worth: packaging, the labour of preparing and packing, handover time. They do not shrink when the basket does.

Baseline

Average monthly revenue over a fixed period before any work begins, agreed in writing and not changed afterwards. Growth is measured against it, so it must be fixed in advance to mean anything.

Method

Where this comes from

Day-to-day management of restaurant accounts on Uber Eats, Glovo and Bolt Food in European markets, and the platform invoices, payout statements and advertising reports of those accounts. The mechanics described here are the ones that repeat across accounts, not a reading of any single restaurant's numbers.

Period

Accounts managed through 2025 and 2026, with statements reviewed monthly. Article first published 7 August 2026; last reviewed 7 August 2026.

What is not included

No commission rates, contract terms or client figures are published here. Platform terms differ per restaurant, per service level and per country, and they are commercially confidential, so any figure printed on a page like this would be wrong for most readers. Tax treatment of platform fees also differs by country and is not covered — check it with your accountant.

How to check it

Every claim above can be verified against your own statements using the six steps in how to work out your own effective rate. If your numbers behave differently from what is described here, your numbers are the ones that are right.

Alex · Delivery Lift — restaurant growth agency with offices in Lisbon and Chicago. We manage restaurant accounts on delivery platforms by hand, inside the restaurant's own accounts, and earn only a share of the growth we brought in.

All insights → Uber Eats advertising → How our fee is calculated →