For many retailers, the website used to be the starting point of the customer journey... that assumption is starting to look dated.
A growing number of consumers are asking ChatGPT, Gemini, Claude, and other AI assistants what to buy before they visit a brand site, if they visit one at all.
According to Bloomreach research of more than 4,000 consumers, 75.4% have already used an AI tool to help them shop.
For brands, issues could arise if customers begin their journey in a conversational interface, then land on a static product page or a standard search bar, as retailers may quickly lose the confidence that AI helped build.
As Christina Augustine, COO of Bloomreach, put it:
“If your consumers are starting that way, it’s so important that when a consumer goes to your site, they also can have the AI conversational experience. Because they literally just indicated, ‘This is how I want to shop.’”
That is a challenge for retail leaders. Yet it is also a sizeable opportunity.
AI Is Filling Gaps in the Shopping Journey
AI is not replacing every part of the buying journey... at least not yet.
Instead, it is proving most useful at the points where customers have traditionally had to do the hard work themselves – be that discovering where to start or building the confidence to make a purchase.
“I think there are two fundamental places that I see that AI is basically adding a capability that was an unmet need in our customer base before,” Augustine said.
The first is early discovery. A shopper may know they need outdoor furniture, a gift, or a product that meets several specific requirements, but not know which retailer, category, or product to begin with.
The second is closer to the point of purchase. Here, customers may have already chosen a retailer but still need reassurance that a product is suitable, available, and worth the money.
That is especially significant as household budgets remain under pressure. Shoppers are often not looking for more options; instead, they want help narrowing them down.
OpenAI’s recent product discovery updates point in the same direction. ChatGPT is rolling out conversational refinement, visual search, side-by-side product comparisons, and merchant product feeds – all designed to help customers assess products by budget, preferences, and constraints.
Google is making similar moves, with its AI Search roadmap including shopping capabilities and agents that can act on users’ behalf.
Elsewhere, Sprinklr recently launched LLM Insights to help brands monitor and manage how they are represented across AI-generated search results.
Retailers may not control those interfaces, but they still control much of the information that powers them.
A Better Customer Signal Than Clicks?
Traditionally, personalization has largely relied on interpreting behavior. Brands track clicks, product views, dwell time, abandoned carts, and purchase history, then try to infer what a customer wants.
