Consumers today will often consult an AI assistant when considering which products to buy. Cathal McCarthy, Chief Strategy Officer at Kore.ai, discusses how AI assistants like ChatGPT and Microsoft Copilot are reshaping how consumers discover and buy products.
Not that long ago, if you wanted to buy something online, you opened a browser and started searching. Increasingly, many consumers are starting their search with an AI assistant.
In conversations with colleagues, customers and friends, I have noticed the first instinct becoming: “I’ll ask AI.” Whether finding a hotel, comparing software or choosing a product, people are becoming more comfortable delegating parts of the decision-making process to intelligent systems.
That behavioural shift may seem subtle, but for retailers it has enormous implications.
Historically, retailers invested heavily in controlling the customer journey through search, personalisation engines, loyalty programmes and digital experiences designed to keep customers inside the brand ecosystem. Today, customers increasingly begin that journey in AI interfaces such as ChatGPT, Amazon Rufus and Google’s AI Overviews. These platforms are becoming the first point of product discovery, shaping which retailers and products customers even consider before they reach a brand’s own channels.
Consider travel. When OpenAI opened ChatGPT to third-party integrations, Booking.com was among the early companies to embrace the opportunity. Rather than treating AI as a threat to its search experience, it used OpenAI’s models to build an AI-powered trip planner that helps travellers discover destinations and experiences they may never have searched for directly.
The lesson is not how quickly the technology was deployed. It is that a company whose business depends on helping people find the right travel options recognised that discovery was changing, and chose to be present where those decisions were increasingly being made. An AI interface it did not own became another channel for demand rather than a leak in the funnel.
That distinction matters. The greatest shift is not where transactions happen, but where decisions are formed. Customers may still complete purchases on retailers’ own websites or apps, but increasingly the shortlist has already been created elsewhere.
Retailers are no longer competing only for attention. They are competing to become the recommendation an AI system returns.
From browsing to delegation
Consumers are beginning to delegate elements of judgement to AI, trusting it to interpret taste, budget, urgency, fit and context accurately enough to recommend a small number of highly relevant options. Browsing gives way to prompting, changing both how products are discovered and how retailers compete for attention.
This represents a genuine shift in control at the beginning of the customer journey. But treating it purely as a defensive challenge misses the larger opportunity.
Products are no longer competing simply to appear in search results. They need to be understandable to machines before they can be recommended to customers. Rich product metadata, consistent attributes, accurate inventory information and contextual descriptions become essential. Retailers relying on sparse descriptions or fragmented data risk becoming invisible to AI recommendation systems, regardless of product quality or brand strength.
Being recommendable increasingly depends on the quality of the information retailers provide.
Owning the moments you can
Not every response involves surrendering the front door. Some retailers are using AI to make their own experiences significantly more valuable.
IKEA’s Kreativ tool allows customers to scan their room, remove existing furniture digitally and visualise IKEA products in their own home before purchasing. At the same time, IKEA has also published an AI assistant through OpenAI that helps customers discover furnishing ideas based on room dimensions, style preferences, budget and sustainability goals.
Retailers need to deepen the experiences they control while also ensuring they appear wherever customers increasingly begin their search. Own the moments you can, and be recommendable in the ones you cannot.
Neither is possible without the right foundations: clean, structured product data, real-time pricing and availability and the orchestration needed to make that information available consistently across every customer touchpoint.
Shared trust, shared accountability
Traditionally, retailers owned the trust relationship throughout the buying journey. Today, that trust is increasingly shared. Customers trust AI platforms to guide their decisions, while retailers remain responsible for fulfilling the promise behind the recommendation.
When a recommendation results in a poor purchase, the retailer manages the return, but the AI system also learns from the outcome. Every successful accomplishment reinforces confidence in both the retailer and the recommendation. Every disappointing experience risks affecting future visibility.
In that environment, loyalty shifts as well. Customers increasingly value the system that consistently helps them make good decisions, while retailers differentiate themselves through reliable fulfilment, relevance and delivering on expectations.
Brand remains important, but how brand value is earned is evolving.
Margin expansion as fuel, not dividend
There is another effect that is easy to overlook. The same AI capabilities that reshape customer discovery also create operational efficiencies. Customer service becomes more automated. Product discovery becomes faster. Merchandising becomes more intelligent. Internal processes become more efficient.
That expands the margin. The temptation is to treat those savings as the outcome. The retailers that move furthest ahead will treat them as investment capital, reinvesting in the data quality, orchestration and infrastructure that determine whether an AI recommends them in the first place.
Margin expansion becomes the funding mechanism for remaining competitive in an AI-mediated marketplace. It becomes a flywheel rather than a dividend.
Building the foundations for AI commerce
Many organisations still approach AI as a collection of individual tools rather than as a connected business capability. But becoming recommendable is not simply a marketing challenge or a technology deployment.
It depends on trusted data, integrated systems, governance and the ability to orchestrate information consistently across customer-facing and operational platforms.
This creates an opportunity for technology partners to move beyond implementation and become strategic advisors. Increasingly, retailers need help understanding where AI creates measurable business value, modernising fragmented technology estates and building the foundations that allow AI to operate effectively across the organisation.
The work ahead
The customer journey is being restructured. Discovery increasingly begins in AI systems, making recommendability as important as visibility once was. The retailers that succeed will be those that combine exceptional customer experiences with the data, infrastructure and orchestration that allow AI to recommend them with confidence. In the age of AI commerce, success depends on becoming the answer an intelligent system gives when a customer says, “I’ll ask AI.”
