Costs
What Delivery Apps Really Cost Australian Restaurants in 2026

In Australia, Uber Eats charges 30% commission on marketplace orders it delivers, 16% if you deliver, and 8% on pickup. DoorDash charges 30% on delivery and 15% on pickup. On a $30 order delivered by the platform, that is $9 gone before food cost — which is usually more than the margin.
Those are both platforms' own published rates, read from their Australian merchant pricing pages on 7 November 2026. They are higher than most comparison articles say, because a lot of what is written about delivery commission quotes American tiered pricing that does not apply here.
Something else changed that matters more than any rate. Menulog has closed. Its homepage now carries a single line — "Menulog is closed" — and directs customers to Uber Eats. A market that had three national marketplaces has two, and the one that competed on being the cheaper option is the one that went.
This article does the arithmetic properly, explains what the cheaper-looking fulfilment options really cost you, and makes the case for a channel mix rather than for quitting the marketplaces, which is rarely the right answer.
What each platform charges in 2026
Read from each platform's Australian merchant pricing page on 7 November 2026. Rates change — Uber Eats raised several in March 2026 — so check the source before you make a decision on them.
Uber Eats Australia
| Option | Fee | What it is |
|---|---|---|
| Marketplace, Uber delivers | 30% | Listed in the app, their couriers |
| Marketplace, you deliver | 16% | Listed in the app, your own drivers |
| Marketplace pickup | 8% | Customer collects |
| Webshop, Uber delivers | 25% | Ordering site on your own domain |
| Webshop, you deliver or pickup | 6% | Your site, your drivers |
Uber also charges a $350 activation fee, excluding GST, for each location after the first. Uber Direct — their couriers fulfilling orders from your own site, with no marketplace listing — is priced as a fixed fee plus distance rather than a percentage.
DoorDash Australia
| Option | Fee | What it is |
|---|---|---|
| Marketplace delivery | 30% | Listed in the app, their Dashers |
| Marketplace pickup | 15% | Customer collects |
| Self-delivery | 15% | Listed in the app, your own drivers |
| Self-delivery, DashPass order | 12% | Order from a DashPass subscriber |
| Online Ordering | 1.75% + $0.30 | Their ordering page on your site — card processing only |
On self-delivery, DoorDash also charges a flat $9.99 per delivery if you call in a Dasher for an order you would otherwise have driven yourself. Their Online Ordering product adds a $5.50 delivery fee including GST.
Two things stand out. Pickup is nearly twice as expensive on DoorDash as on Uber Eats — 15% against 8% — which is worth knowing if collection is a real part of your trade. And both platforms price their own-site products far below their marketplace rates, which tells you how much of that 30% is paid for the audience rather than the technology.
The arithmetic on one $30 order
Round numbers, one order, 30% food cost. Follow it through each channel.
| Marketplace, they deliver | Marketplace, you deliver | Your own channel | |
|---|---|---|---|
| Order value | $30.00 | $30.00 | $30.00 |
| Commission | −$9.00 (30%) | −$4.80 (16%) | $0.00 |
| Card processing (~1.75% + $0.30) | included | included | −$0.83 |
| Food cost (30%) | −$9.00 | −$9.00 | −$9.00 |
| Delivery cost | $0.00 | your driver | your driver, or a flat courier fee |
| Left before labour and overhead | $12.00 | $16.20 | $20.17 |
The gap between the first and last column is $8.17 on a single $30 order. At fifty delivery orders a week that is roughly $21,000 a year, which for most independent venues is a larger number than any other line item they could realistically negotiate.
Two caveats, because the simple version flatters the right-hand column:
- Your own delivery is not free. A driver, their fuel, their time and the insurance all cost money, and at low order volumes a courier on a per-delivery fee is usually cheaper than employing someone. The honest comparison for self-delivery is against the real cost of a driver, not against zero.
- The marketplace brought the customer. That $9 bought reach you did not have to pay for separately. For a first-time customer it is a customer-acquisition cost, and judged as one it is not unreasonable. For the same person's eighth order it is just margin leaving the building.
That distinction is the whole strategy, and the rest of this article is about acting on it.
What "commission-free" providers actually charge you
Every direct-ordering vendor says commission-free. The money has to come from somewhere, so check which of these it is:
- A flat software fee. The straightforward model: a monthly subscription, no cut of order value. Predictable, and the better it works the cheaper it gets per order. This is what we charge, priced per location.
- Card processing only. Honest as far as it goes. You will pay a processing fee to someone regardless — around 1.75% plus 30 cents is typical in Australia — so the test is whether the vendor adds a margin on top of it, and whether the rate is disclosed before you sign.
- A per-order fee. Commission-free in the strict sense, because it does not scale with order value. Still a variable cost, and on small orders a flat fee can be a larger percentage than the commission you left.
- Free software, paid delivery. Common and reasonable, but the courier fee is the real price and it is often charged per kilometre. Model it on your actual delivery radius rather than on the headline.
- Free, funded by upsells. The one to read carefully. Watch for mandatory payment processing at a non-disclosed rate, a long contract, or hardware you must buy.
The questions worth asking any vendor: is the fee per location or per terminal? Is the processing rate disclosed, and is there a markup on it? What is the contract length and the exit cost? Is online ordering included in the plan you are being quoted, or billed as a module?
For the equivalent treatment of POS quotes, where the same games are played with hardware and per-terminal fees, see how much a restaurant POS system really costs.
Why marketplace orders still make sense
Deleting your Uber Eats listing to save commission is usually a mistake, and the reason is that the two channels are not substitutes.
A marketplace is a discovery channel. People open the app hungry, with no particular restaurant in mind, and choose from what is there. If you are not there, you are not in that decision — and that demand does not transfer to your own site, because the customer was never looking for you. With Menulog gone, those two apps are now most of the discovery surface for delivery in Australia.
The useful way to think about it: a marketplace order at 30% is an advertisement you only pay for when it works. Compared with most restaurant marketing, that is a good deal. Where it stops being a good deal is when the same customer keeps arriving through it.
So the goal is not fewer marketplace orders. It is a higher proportion of repeat orders arriving directly, while the marketplace keeps doing the job it is actually good at.
How to move repeat customers to your own channel
Practical, in rough order of effect per unit of effort.
Put something in the bag
The cheapest acquisition channel you have. Every marketplace delivery is a paid introduction to a customer who already liked your food enough to order it. A card with your own ordering address and a reason to use it — a drink thrown in, a small discount, anything — converts a meaningful share of them. Do not make it a generic flyer; make it a specific offer with a short, memorable URL.
Price the convenience difference, not the food
Many venues raise marketplace prices to cover commission. Reasonable, and it is worth understanding that the platforms notice: Uber Eats charges a higher pickup rate when in-store pricing is not matched. Check your agreement's terms on price parity before you act, and consider the alternative that works regardless: keep prices level, and make your own channel better in ways the marketplace cannot match — a loyalty reward, a larger menu, the items you do not list in the app, collection at a time the guest chooses.
Own the in-venue ordering experience
Your dine-in guests are the warmest audience you will ever have, and they are standing in your restaurant. QR table ordering puts your own ordering system in their hand while they are eating your food, which is a far easier introduction than any advertisement. Guests who order at the table from your system already know how to order from it at home.
Make your own channel genuinely good
Most direct ordering loses to the apps because it is worse, not because it is unknown: a PDF menu, a phone number, a checkout that asks for an account. If your own ordering is as fast as the app, on your own domain, with the menu actually up to date, the commission difference is yours to keep. One catalogue should drive your online ordering, your QR codes and your POS, because the moment those are three separate menus, one of them is wrong.
This is what Zamarut Afghan found with table ordering. In their words: "The QR menu alone changed everything. Guests order from the table and we upsell drinks without a word from the staff. Our average ticket went up and we need fewer people on the floor." There are more examples on our case studies page.
Measure the mix, monthly
You cannot manage this without the number. Track the share of orders arriving through each channel and watch it move. If direct is not growing, nothing above is working, and you should know that in a month rather than a year. Reporting that reconciles every channel into one set of totals is what makes that a two-minute check instead of a spreadsheet exercise.
What a sensible channel mix looks like
No universal answer, but a shape that works for most independents:
- Stay on both marketplaces, and treat the commission as acquisition spend. Review it when a listing stops bringing new customers.
- Switch fulfilment where the arithmetic says so. If you already run a driver, self-delivery at 15–16% instead of 30% is the largest single saving available, and it needs no new customers.
- Push collection to your own channel. Pickup commission is 8% on Uber Eats and 15% on DoorDash, but a collection order placed on your own site is 0%. Collection customers are local and repeat, which makes them the easiest group to move.
- Make direct ordering the best way to order from you, then tell every customer who arrives any other way.
RestroManage is built for that mix: a flat monthly fee per location, no commission on order value, your own domain, and one menu behind dine-in, collection, delivery and QR orders. The only transaction cost is your payment processor's own fee, which goes to them rather than to us. Current plans are on the pricing page.
Frequently asked questions
How much commission do delivery apps charge in Australia?
As published on 7 November 2026: Uber Eats charges 30% when it delivers, 16% when you deliver, and 8% on pickup. DoorDash charges 30% on delivery, 15% on pickup, and 15% on self-delivery, falling to 12% on DashPass orders. Both also price their own-site ordering products far lower than their marketplace rates.
Is Menulog still operating?
No. Menulog has closed, and its website now directs customers to Uber Eats. Australia's national delivery marketplaces are Uber Eats and DoorDash, which means less competitive pressure on commission rates and a stronger case for building an ordering channel you control.
Is it cheaper to use my own drivers on a delivery app?
On commission, yes — 16% on Uber Eats and 15% on DoorDash, against 30% for platform delivery. Whether it is cheaper overall depends on your real driver cost, including wages, fuel and insurance. At low order volumes a per-delivery courier fee is often cheaper than employing a driver; at steady volume your own driver usually wins.
Should I raise my prices on delivery apps to cover commission?
Many venues do, and it is a legitimate response, but check two things first. Your agreement may have terms about price parity, and Uber Eats charges a higher pickup commission when in-store pricing is not matched. Making your own channel better — loyalty, a wider menu, a chosen collection time — achieves a similar result without touching the agreement.
Does commission-free online ordering really have no fees?
No commission on order value, but there is always a cost somewhere: a flat software fee, card processing, a per-order fee, or delivery charged separately. Card processing of roughly 1.75% plus 30 cents is unavoidable whoever you use. The questions that matter are whether the vendor marks up that rate, whether the fee is per location or per terminal, and what it costs to leave.
Should I leave the delivery apps entirely?
Usually not. Marketplaces are a discovery channel: customers open them without a restaurant in mind, and that demand does not move to your own site if you delist, because those customers were not looking for you. The better strategy is to keep the listing for new customers and move repeat customers to your own channel.
Rates in this article were read from Uber Eats' Australian merchant pricing page and DoorDash's Australian merchant pricing page on 7 November 2026, and Menulog's closure from menulog.com.au on the same date. Published rates change; confirm them at the source before making a decision.
Keep reading

QR Code Ordering for Restaurants: A Complete Guide
How QR table ordering actually works, what it does to tips and table turnover, and a rollout sequence that starts with two tables rather than the whole room. Includes signage, staff briefing, weak Wi-Fi and the objections worth taking seriously.

How to Choose Restaurant Management Software
A buyer's framework rather than a feature list: the six questions that actually decide it, the contract terms that trap people, what has to integrate, and how to switch systems without losing a service.
How Much Does a Restaurant POS System Really Cost in 2026?
A breakdown of what restaurant POS systems actually cost once you count hardware, monthly fees, and processing rates — and how cloud-based, hardware-free systems change the math.

