AI aids product discovery, but shoppers shun checkout

Artificial intelligence is becoming a more important tool in how consumers discover products online, but its role in completing purchases remains limited, according to new survey data from Criteo.

The study found that 47% of respondents use AI to compare products, 39% to find new items and 38% to look for deals. Only 8% use it to place an order, indicating that shoppers mainly turn to AI during the research and discovery stage rather than at checkout.

The findings show that the shopping journey remains spread across multiple digital channels, even among frequent users of AI assistants. Among shoppers who regularly use AI shopping assistants, 96% also use other routes during the buying process, including search engines, social media, brand websites and retailer sites.

AI assistants are not yet the main starting point for most shopping journeys. Just 14% of respondents said they begin with an AI assistant. By comparison, 36% start on an online marketplace, 28% with a search engine and 15% on a retailer website.

Discovery Shift

The data suggests retailers and brands may need to focus on how products are surfaced by AI systems, particularly as consumers show interest in recommendations that go beyond direct search matches. More than half of respondents, 56%, said they would like product recommendations that extend beyond exact results.

This points to demand for a mix of relevance and broader exploration in online shopping, rather than a purely functional search experience. While AI may help narrow options, consumers still want room for discovery when browsing products.

Andy Stephen, managing director of retail media UK at Criteo, said the balance between recommendation and exploration will be central as AI shopping tools develop.

"AI assistants that can surprise and delight, delivering options that preserve the sense of exploration in online shopping, will become the gold standard. Far from being random, these 'relevant surprises' should be based on robust recommendation algorithms. As a result, brands will need to start marketing to agents as well as people. Success won't only hinge on visibility in feeds or search results, but also on how clearly a brand is understood by AI systems - when it's the right answer, how it fits into bundles, and whether it's trusted," Stephen said.

The comments reflect a broader shift in digital commerce as companies assess whether search optimisation and online merchandising will need to adapt to a world in which AI assistants increasingly influence product selection.

Privacy Concerns

The survey also examined consumer attitudes to sharing personal data in AI-assisted shopping. The responses suggest shoppers are selective rather than uniformly resistant. A majority, 52%, said their willingness to share information depends on the topic, while 25% said they are generally comfortable doing so if it improves the shopping experience.

Even so, certain categories of information remain especially sensitive. Consumers were most cautious about sharing personal identity or location data, cited by 57% of respondents, followed closely by payment-related information at 55%.

These results suggest that while some consumers may accept data sharing in exchange for convenience or better recommendations, trust remains a significant factor in whether AI tools gain wider use in commerce.

That caution was reinforced by findings on content reliability. The survey found that 52% of consumers want protection from fake or biased content when using AI-assisted shopping, suggesting that confidence in the accuracy and neutrality of information may be as important as the quality of recommendations.

Channel Reality

The research covered 6,379 respondents across six countries and examined digital and omnichannel shopping habits across a range of sectors. It also included analysis of inbound traffic sources from about 500 US merchants in Criteo's network.

The picture that emerges is of AI becoming part of retail discovery, but not replacing existing consumer habits. Marketplaces, search engines and retailer websites still dominate as starting points, and shoppers continue to move between channels before making a purchase.

For retailers, that means AI currently sits alongside established digital touchpoints rather than displacing them. For brands, it raises questions about how products are described, categorised and presented so AI systems can identify them correctly as they shape more purchase decisions over time.

The survey suggests the immediate commercial impact of AI in retail may lie less in automated transactions and more in influencing what shoppers notice, compare and consider before they buy.

As AI tools become more embedded in online shopping, the data indicates that adoption will depend not just on convenience, but on whether consumers see them as useful for discovery, trustworthy in their recommendations and careful with personal information.