AI blows up shopping as we know it

By Joe Beier, EVP

Sorry to come with bad news, but have you heard that all your shopper purchase journey studies just became obsolete?

Sorry to come with bad news, but have you heard that all your shopper purchase journey studies just became obsolete?

The horse and buggy era of shopping just got lapped by a Model T. Like all else it touches, AI does not tread lightly but tends toward the transformative. Some recent retail announcements make it clear that shopping is never going to be the same.

While AI is already becoming part of the new shopping landscape – 53% report using it for product research (Adobe survey, Sept 2025) two innovations by retail heavyweights suggest that when it comes to AI’s role in shopping, we are in a transformative moment:

• Walmart announces a partnership with ChatGPT that will enable shoppers to shop on Walmart.com and make purchases, all while remaining in the ChatGPT app. No need to bounce out to another site to buy.
• Amazon’s recent launch of a “Help me decide” button, which will use AI to process myriad data on the shopper to provide a “tiebreaker” recommendation when the shopper has been comparing a lot of options.

Purchase journey studies are a vital tool for brands and retailers to understand how buying decisions are made, and of course, how to influence them. Central is an analysis of what factors are most influential in determining the final purchase decision, which means that this new shopping landscape demands a new accounting for AI’s hugely influential role in shopping.

A new shopping landscape also requires us to rethink the research methods best suited to build an actionable shopping insights foundation to meet the moment. So, how should our approach to gathering purchase journey insights evolve?

When researching emerging spaces, the “Listen/Ask” model has proven powerful:

LISTEN

• Web scraping can be a low-cost methodology to gain a basic understanding of evolving shopping behaviors, and specifically AI’s emerging role. Ideally, providing a snapshot of AI adoption, use cases, and shopper satisfaction.  Enables initial hypothesis development that can be validated in later qual or quant phases.
• Online shopalongs can illuminate in detail whether and how shoppers are utilizing AI shopping tools. Typically, a respondent would be asked to complete an online shopping task under the watchful eye of a moderator. The moderator can also ask drill-down questions when they see something interesting in the respondent’s behavior.  e. “I noticed you dismissed the ‘help me decide’ button – can you tell me why?” This method delivers a powerful mix of both the observed ‘whats’ and the underlying ‘whys.’

ASK

This new era of AI shopping has some key implications for how we need to think about any quantitative shopper journey research:

• Sample: In more stable times, purchase journey research has relied both on actual category purchasers as well as purchase intenders. This enabled larger sample sizes but was predicated on the assumption that intenders were familiar with the tools available to help them shop and could make a well-informed “projection” of how they would be likely to navigate their purchase journey. This assumption becomes clearly problematic with the availability of brand-new AI shopping tools that intenders may have zero experience with.  Suggesting that any studies done in this transitional period rely solely on actual vs intending buyers.
• AI barges into new category spaces: In traditional purchase journey studies, the focus on measuring online shopping behavior varied dramatically depending on category context. For higher involvement and more durable categories like apparel or electronics, the questions we asked drilled deeply into the specifics of online shopping, such as which specific sites were consulted and for what purposes. But for most CPG categories, presumed to not be deeply researched online, we might have just asked whether online was one of the factors influencing their decisions at all and leave it at that. With AI being such a powerful and seamless shopping tool, it is easy to project that even lower involvement consumables categories are likely to see a major uptick in AI shopping activity: (“Amazon, help me decide which brand of pancake mix I should get”) This implies that we should start building more robust and deeper batteries of questions around online shopping even for those categories where online was considered to be a minor factor.

While these moments of disruption can feel unnerving, they also represent opportunities for insights teams to step up and play a vital leadership role in their organizations. We hope these ideas can help you along your journey to thriving in the AI-enabled shopping era.

Good luck!