Delivery riders got a pay floor today. Uber and DoorDash asked for it

This one does not touch your wage bill, so nobody will write to you about it. It touches the cost of the channel that sells your food when the customer never walks in. From today the apps must pay a delivery worker at least $31.30 an hour of engaged time, that floor steps up on 1 January 2027, and from 2028 it rises with the minimum wage every year. Nothing in the order says a word about what Uber or DoorDash may charge a venue, which is exactly the point.

The floor, and the date it moves again

The Fair Work Commission made the Interim On-Demand Delivery Employee-like Worker Minimum Standards Order on 11 August 2026. It starts today, 17 August 2026. It covers app-based workers who mainly collect and deliver "consumables", which the order defines as food, beverages and liquor, plus supermarket groceries, and it covers the platforms that engage them.

The earnings floor, per hour of engaged time, from today until 31 December 2026:

• $31.30 on foot, on a pedal bike, or on an electric bike or scooter
• $31.50 on a petrol motorcycle or scooter
• $32.00 in a car or van up to one tonne carrying capacity

From 1 January 2027 those become $31.80, $32.00 and $32.50. Then from 1 January 2028, and every year after, the rates rise by the same percentage as the National Minimum Wage increase from the Commission's annual wage review, unless the Commission orders something else.

So the January step is 50 cents an hour, about 1.6%. Small. The part that matters is that the direction is now written down and automatic.

"Per hour" means the minutes on a job, and nothing else

Read the definition before you assume this is an hourly wage. Engaged time starts when a worker accepts a job and ends when they finish it. It excludes time after a customer cancels, time on a job the worker abandons, and time lost to breakdowns, accidents or breaks.

Waiting around between jobs is not engaged time. Nor is your kitchen running late, once the platform decides it counts as non-engaged time under the process in the order.

The averaging matters too. The floor applies across an "earnings period" the platform sets, which can be up to 21 days, not per delivery. If a worker falls short over those three weeks, the platform tops it up in the next period.

Both features do the same thing: they cap how much this actually adds to a platform's cost per order. Which brings us to the strange part.

Uber and DoorDash asked for this

The order is close to a consent position. The Commission's own decision says the draft "for the most part reflects a MSO to cover on-demand work proposed by the TWU, the applicant in this matter, and supported by Uber and DoorDash", and the Panel treated it "as a matter of significance" that Uber, DoorDash, the union and a number of individual workers all supported making the order in those terms. On the evidence before it, the Panel found the order applies principally to Uber, DoorDash and the workers they engage, with one comparatively minor part of Amazon's operations the only other candidate.

The people who thought the number was too low were the riders. One delivery worker told the Commission the draft rates do not cover his costs, that around $40.00 an hour would be needed to leave a minimum casual wage after expenses, and proposed $35.00 as a compromise. Two academics supported the order only as "a first step towards full cost recovery and parity with employees performing comparable work".

That is the honest read of today: a floor set at a level the platforms could live with, agreed by the platforms, with the harder question parked.

The order says nothing about your commission

Worth stating plainly, because a lot of coverage this week will blur it. This order binds digital platform operators and the contractors they engage. It does not bind you. If you employ your own driver or a delivery hand, nothing here changes what you owe them; the award and the national minimum wage still do that job.

And there is no clause anywhere in it about what a platform may charge a venue. No cap, no notice period, no obligation to change your rate, and nothing stopping them either. Your commission is a commercial term in your platform agreement, set by a company whose regulated labour cost now has a floor under it that ratchets.

You do not get told when a cost curve changes shape. You find out when the fee schedule email arrives.

The number to work out before that email

Do this now, while it is a calm afternoon rather than a reaction.

Take your last platform statement. Divide the total commission and fees you paid by the gross value of the orders they came from. That percentage is your real rate, including promotion funding and any per-order charges, and it is usually higher than the headline number in your agreement.

Then size a move. An illustration on round numbers, not a benchmark: 40 app orders a week at $35 each is $1,400 a week, about $72,800 a year through the app. On that, every single percentage point on your effective rate is $728 a year. Three points is $2,184, which for a lot of cafes is the whole profit on the channel.

Now the useful question. At what rate would you stop? Write the number down today. Owners who have a walk-away rate in advance negotiate or exit calmly; owners who do not tend to absorb increases one at a time and call it the cost of doing business.

Put 25 August in the diary

The word "Interim" in the title is deliberate. The order will be reviewed within 21 days if the Commission publishes a notice of intent and draft order in either of two related cases covering "last mile" delivery, and the Panel was explicit that it parked "the longer-term issues of a transition to proper cost-recovery rates and competitive neutrality" for that review.

Those two cases have a hearing listed for 25 and 26 August 2026 on exactly that question, with submissions due 21 August. So the trigger for reopening today's rates is being argued next week. Nobody can tell you the outcome. You can reasonably conclude that $31.30 is a floor in both senses of the word.

Why a buyer pays less for app money than for counter money

Two cafes, same revenue, same profit. One takes a third of it through a delivery app; the other takes all of it over the counter and through its own online ordering. They do not sell for the same price, and this order is a clean example of why.

In the app channel someone else sets the price the customer sees, someone else owns the customer record, someone else decides the promotion you fund, and someone else sets your cost of sale, which they can revise with an email. A buyer looking at that revenue cannot model a lever they are not allowed to touch, so they discount it, the same way they discount customer concentration. Revenue you cannot price is not the same asset as revenue you can.

That is not an argument for switching the apps off. It is an argument for knowing which of your two revenue lines is the durable one, keeping gross margin honest on both, and spending the app channel deliberately: as paid acquisition that you convert into direct customers, rather than as growth you have quietly rented.