Baskets cost 7 to 13 percent more before delivery fees, and the format still outgrows every other online channel in Australia. The comparison comes from CHOICE price checks cited in IGD's analysis of the market as of November 11, 2026; the worst case, a Woolworths order, landed 39 percent above the equivalent in-store purchase. Read together, those two facts describe the real economics of quick commerce: shoppers are not paying for cheaper goods, they are paying for a shorter wait, and the operators who can serve that wait without wrecking the shelf are the ones pulling ahead.
A dearer basket, a narrower mission
The premium is not a pricing error. A pick-and-pack order in a supermarket aisle costs more to assemble than a trolley pushed by a shopper: someone walks the aisles, someone substitutes a missing item, someone hands the bag to a rider. IGD's read of the market is that quick commerce remains a small slice of Australian online grocery but one of its fastest-growing segments, which is exactly what a service premium looks like when demand is young.
The mission is narrower than the channel's marketing suggests. Rapid delivery suits urgent top-up trips, the forgotten ingredient, the late-night remedy, while planned weekly shops still favour standard delivery and its lower fees. Treating the two as one online channel is where category plans go wrong: they carry different basket sizes, different substitution tolerance and different margin per order.
Thirty minutes is a shelf problem before it is a rider problem
Every operator promising half-hour delivery is really promising that the item is on the shelf when the order lands. That is why the constraint shows up in availability first. A store carrying 17,000 products for on-demand orders cannot treat the top 300 fast movers like the rest of the range; if the fast movers are not faced up and counted, the promise breaks in the aisle, not on the road.
Picking density beats range width
Woolworths opened an automated eStore in Western Sydney, an investment reported at AUD 20 million, to raise delivery capacity and speed. Automation helps most where order density is high and the pick path is predictable: the same thousand items, picked thousands of times. Adding range without adding pick density mostly adds walking, which is the cost line that quietly eats the service premium.
The gap nobody prices in
The 39 percent worst case is not only a fee story. Substitutions, missing items and split orders all add to what the shopper pays and what the retailer credits back. Retailers that publish a clean substitution policy and hold availability at store level keep more of the premium; those that let the shelf drift lose it to refunds and to the next app on the phone.
Three formats, three cost lines
Australia now runs three delivery shapes side by side: standard online grocery from stores and automated fulfilment centres, on-demand marketplaces run with Uber Eats or DoorDash, and dedicated rapid fleets. They look similar to a shopper and behave nothing alike on a P&L. The table below is the one worth bringing to a planning meeting.
| Format | Cost driver | Where margin leaks |
|---|---|---|
| Standard online from stores and CFCs | Pick cost and slot density | Underused delivery windows |
| On-demand marketplace | Platform commission and rider fee | Substitutions and refunds |
| Dedicated rapid fleet | Courier utilisation per hour | Idle capacity outside peak |
Coles gives a sense of how far the marketplace route can scale: its Uber Eats partnership covers up to 17,000 products with plans to extend to 500 stores, while its own fulfilment centres, built with Ocado in Victoria and New South Wales, performed well enough in their second year to turn positive at the EBITDA line before depreciation. The first route buys reach, the second buys cost control. Running both without separating their cost lines is how a channel looks busy and stays unprofitable.
What to check before adding capacity
Three checks separate an expansion that pays back from one that just adds rent. None of them require new software; they require someone to look at the data the stores already hold. Retail media spend and promotional funding should be reviewed in the same pass, because visibility inside an app is now sold the same way shelf space once was.
FAQ
Why do delivery-app baskets cost more than in-store ones?
Two reasons stack up: the products themselves are often priced higher on the platform, and a service fee is added on top. CHOICE found 7 to 13 percent before fees, and 39 percent in the worst case.
Does quick commerce cannibalise the weekly shop?
Mostly no. It takes top-up missions and urgent purchases. Planned baskets still favour standard delivery because the fee per order is lower and substitution risk is smaller.
What should a store measure first?
On-shelf availability for the fastest-moving on-demand SKUs. If those are not on the shelf, the promise fails before a rider is ever dispatched.
Is retail media spending part of the same decision?
Yes. Paid placement inside an app is the digital equivalent of shelf position, so it belongs in the same review as availability and assortment.
