As of 30 October 2026, Coles Group reported first quarter revenue of 11.2 billion dollars, up 2.2 percent, with its supermarket segment ahead 2.7 percent on a comparable basis. Woolworths reported 19.4 billion dollars in the same period, up 4.7 percent, and headline inflation ran at 4.3 percent in the twelve months to September. One line broke the pattern: e-commerce grew 22.9 percent, roughly ten times the rate of the total business.
The gap between two baskets
The two growth rates describe different populations. A supermarket segment growing 2.7 percent against 4.3 percent inflation is losing ground in real terms, because the average basket is worth less after price is stripped out. E-commerce growing 22.9 percent is doing something else entirely: it is taking share from the store, from the competitor, or from both. The size of the gap is the story, and it is wider than a single quarter of promotion activity can explain.
Coles describes customers as value-conscious, shopping across multiple retailers and leaning harder on promotions and loyalty programmes. That description fits a shopper who has not stopped buying but has started splitting the basket. Splitting is exactly the behaviour that shows up as weak like-for-like sales in stores and strong growth online, because the online order is the one that gets consolidated while the top-up shop moves to whoever is cheapest that week.
What 22.9 percent is actually made of
Promotions and loyalty do the recruiting
Loyalty schemes and targeted offers are the cheapest way to move a shopper from a competitor's store into your own app, because the recruitment cost is a discount the retailer would partly have given anyway. That is why promotional intensity rises when households split their baskets. It is also why e-commerce growth of this size rarely arrives with a margin story attached: the channel is being bought as much as it is being earned.
The comparison with overseas operators is instructive. Kroger told investors this week that only one in seven of its store customers shops online with the company, and that around 40 percent of people who start signing up for online shopping never finish. Coles reporting 22.9 percent growth suggests a different position in the adoption curve, closer to the steep part where the constraint is capacity rather than persuasion.
Baskets are being split, not abandoned
A shopper who splits a weekly basket between two retailers still spends the same money; the retailer just sees a smaller average transaction and a shorter intervals between visits. That pattern depresses like-for-like sales without depressing category demand, which is precisely the combination Coles reported. It also explains why promotional depth keeps rising even as headline inflation cools, since each retailer is defending a share of a basket rather than the whole of it.
For brands selling through both channels, the implication is that a single national sales number will increasingly mislead. The same shopper may buy the category online at full price and in store only on promotion, and the two behaviours need different responses. Retailers that only measure channel totals will keep discovering the split late, usually when a competitor's app has already captured the recurring order.
Reading channel shift without a panel
| Indicator | Where it shows up | What it tells you |
|---|---|---|
| Channel growth spread | e-commerce versus store like-for-like | Whether share is moving or demand is growing |
| Sign-up completion rate | App and account registration funnel | Whether friction is capping adoption |
| Promotional depth | Average discount on promoted lines | Whether growth is being bought |
| Basket frequency | Orders per household per month | Whether baskets are splitting |
None of these indicators requires a consumer panel. The channel growth spread comes from the segment note, sign-up completion from the retailer's own funnel data if disclosed, promotional depth from shelf and app price collection, and basket frequency from loyalty programme reporting. Put together they describe the same shift a panel would show, at a fraction of the cost and with a shorter lag.
The industry trend worth naming is that e-commerce is no longer a separate business inside a grocery retailer. It is the same shopper arriving through a different door, priced differently, with a different cost to serve. Retailers that still run it as a channel project rather than as the primary relationship will keep reporting strong online growth alongside weak store comps and treat the two as unrelated facts.
How to instrument this before Christmas
Set a weekly read on the spread between online growth and store like-for-like, and treat a widening spread as a signal to check assortment overlap rather than to add promotions. If the same lines are on promotion in both channels, the online order is cannibalising the store trip and the growth number is partly an accounting transfer.
Then check the funnel rather than the outcome. A retailer growing 22.9 percent while only completing 40 percent of sign-ups has a queue of shoppers who already chose it and were lost to friction. Fixing that queue is cheaper than buying new customers, and it shows up in the next quarter's comp rather than in a marketing line item.
FAQ
Why is e-commerce growth of 22.9 percent significant next to 2.2 percent total growth?
Because it is roughly ten times the rate of the total business, which means share is moving between channels faster than the category is growing.
Does strong online growth mean shoppers are new to the retailer?
Not necessarily. Much of it is existing customers consolidating a split basket, which shows up as online growth alongside weaker store like-for-like sales.
What is the cheapest indicator of channel shift?
The spread between online growth and store like-for-like growth from the segment note, which needs no consumer panel and updates every reporting period.
How does sign-up completion change the picture?
It shows how much growth is capped by friction. A retailer completing only around 40 percent of sign-ups has demand it is failing to convert, not demand it needs to buy.
