In a quiet acceleration of retail’s digital evolution, autonomous AI shopping agents are beginning to interpose themselves between consumers and the point of purchase, changing how loyalty and personalisation operate. Major grocery groups have already moved from experimental chatbots to agentic systems that can plan meals, interpret images, assemble baskets and adjudicate offers on a shopper’s behalf.
According to RetailBiz, Woolworths in January became the first Australian...
Continue Reading This Article
Enjoy this article as well as all of our content, including reports, news, tips and more.
By registering or signing into your SRM Today account, you agree to SRM Today's Terms of Use and consent to the processing of your personal information as described in our Privacy Policy.
retailer to deploy Google Cloud’s Gemini Enterprise for Customer Experience, upgrading its Olive assistant from a service bot into a proactive shopping agent able to process text, voice and images and preserve conversational context across browsing, buying and aftercare. Industry reporting shows Olive will be able to, for example, identify when a cart falls short of a threshold for a promotion and recommend a small add-on to trigger a discount. Woolworths is now among international retailers experimenting with agentic AI to streamline digital commerce; ComputerWeekly and Marketing-Interactive noted the move was unveiled at NRF 2026 and framed as a step toward a more intuitive, task-capable assistant.
A parallel initiative in Canada underlines the trend. Loblaw has launched a PC Express shopping app inside ChatGPT, enabling conversational meal discovery and ingredient curation, with product suggestions drawn from local store assortments and the option to add selected items to a cart for fulfilment, according to Loblaw’s announcement reported by GlobeNewswire and GroceryBusiness. The integration is presented as a way to meet shoppers within the conversational interfaces they already use and to make grocery planning more efficient and personalised.
These early deployments illustrate a fundamental operational challenge for loyalty schemes. Research cited by RetailBiz shows PayPal Australia found 48 per cent of Australians have used AI assistants to search for products and 78 per cent expect AI shopping tools to become commonplace. Adobe data cited in the same analysis reported e-commerce traffic from generative AI platforms jumped 693 per cent in November–December 2025 versus the prior year. When an automated agent evaluates whether to redeem points, apply a member-only price or switch to a competitor mid-transaction, the speed, clarity and machine-readability of a loyalty system determine whether a programme remains visible and valuable.
Historically, loyalty relied on human memory and intermittent engagement, card scans, app visits, manual offer activations, creating friction that protected underlying programme economics through unused benefits. Agents remove that friction. They do not forget or overlook offers; they compute the best financial outcome in real time. If a loyalty engine cannot respond to an agent’s query during the decision window, the agent will simply optimise for the next-best price or offer, effectively rendering slower programmes invisible.
That shift exposes a shortfall in many systems marketed as “personalisation”. As RetailBiz argues, a great deal of contemporary personalisation is driven by batch segmentation, refreshed daily or weekly, an approach adequate for human shoppers but insufficient for agentic commerce. Google’s roadmap for agentic consent, which would allow customers to grant agents ongoing access to member pricing and personalised offers, heightens the stakes: API responsiveness and unambiguous, machine-readable rules become competitive levers.
Providers of loyalty and personalisation technologies must therefore meet exacting requirements. They need rule sets that are explicit and consumable by machines, not buried in prose. They must deliver sub-second responses at scale to support decision-making during peak loads. They must verify balances, tier status and offer eligibility in real time so agents can trust the outcomes they return.
Eagle Eye, the vendor represented by Jonathan Reeve in the lead article, positions its platform around those capabilities. According to the company, its Google Cloud-powered architecture supports issuance and redemption of highly personalised offers in real time across channels, including in-store, and can validate and apply bespoke discounts with sub-250 millisecond response targets at heavy load. The firm frames real-time personalisation not as an add-on but as foundational infrastructure that both humans and machine agents will query continuously.
For retailers reworking their loyalty stacks, two practical assessments emerge from this context. First, systems should be able to detect a shopper’s immediate context, location, local conditions, items already in a digital basket, and generate a single-customer personalised offer in the instant that context changes. Second, platforms must validate and apply constrained or conditional discounts, for example, a $5-off-$50 offer available only to specific members, within a fraction of a second while enduring multiple-times-peak traffic. Failure on either front risks ceding transactions to faster, more machine-friendly rivals.
The broader implication is clear: visibility to autonomous agents will become a new battleground. Programmes that remain reliant on delayed synchronisation or human-triggered processes are at risk of irrelevance in an environment where agents routinely make purchase choices on behalf of consumers. As retail architectures adjust, real-time personalisation will be the backbone that enables offers to be both meaningful to shoppers and actionable for machines.
Source: Noah Wire Services
