AI: shifting into meaningful shopper interaction

The history of retail is littered with examples of high-tech wizardry that never really took off or failed to achieve the scale or impact that proponents initially envisioned. One of the biggest examples during my career is RFID, a technology that was supposed to create an omniscient supply chain that would lead to untold efficiency and enhanced shopper experience.

For some fairly fundamental reasons, this never really took off in the grocery sector and RFID’s usage within food retail is so far limited to use in non-food categories and some fairly high-level supply chain vigilance. In other areas, the technology has been a much longer, slower burn than initially anticipated. A case in point would be electronic shelf labels: around since the early 1990s, it is only now that their adoption is gaining any sort of mainstream momentum in some countries, held back by their cost, laxer regulations on price integrity and low wages making their relative cost even more of a disincentive.

Being a cynical Luddite, I was readying myself for a lot of the ongoing AI hype to be similarly misplaced – overegging the impact and underestimating the time it would take. However, an increasing amount of evidence is mounting that my cynicism hight have been slightly unjustified.

Perusing investor relations materials from major retailers is an enjoyable aspect of this job and – as such – I found myself leafing through the annual slides from Australian food retail rivals Coles and Woolworths. There are plenty of commonalities between the two competitors and one was that they both directed some investor attention towards the progress they are making in exploiting AI – in both the shopper-facing and back-office spheres.

In the back-office arena, Coles called out the impact of AI on forecasting, space and range optimisation and inventory and store decisions, with future endeavours set to include end-to-end supply chain intelligence. Woolworths, meanwhile, cited the benefit of AI on software development, marketing and pricing, promotions and ranging. These chime with what we’ve been hearing from other retailers around the world in terms of AI use cases that have already started getting some traction.

On the shopper-facing side, both retailers called out some solid progress. Coles highlighted personalised experiences, in-store object verification and detection and the optimisation of customer contact centres as areas that AI Has already benefited, with conversational shopping, discovery, transacting and post-purchase support as items that are firmly on the agenda.

Woolworths had some interesting progress to share: AI has been powering enhanced online experiences and the growth of agentic shopping tools including Olive (the retailer’s virtual customer assistant) and Smart Basket; a personalisation engine delivering millions of customised offers to members every week; and customer service agents (resolving over seven million customer queries in 2025/26).

Woolworths continues to introduce new and appealing features from the shopper vantage point: a recent move was the addition of a feature that enables shoppers to scan a handwritten shopping list and transform it into a digital shopping list that can be used online or in-store – precisely the sort of tangibly useful innovation that shoppers are hoping to see.

While I reserve the right to maintain a healthy degree of scepticism over the full benefit of AI, especially from the shopper perspective, the Australian market leaders are demonstrating that the impact of AI could be quicker, more sizeable and more shopper-friendly than I first gave it credit for.