Business It News / Business Intelligence
AI shopping recommendations put product data under scrutiny
Australian consumers are already buying products recommended by AI tools, giving retailers a new reason to examine whether their product information helps shoppers make the right choice.
Nearly a third of Australian respondents, 32%, told Blue Yonder’s 2026 Consumer Retail Returns Survey they had purchased a product following a recommendation from an AI tool or assistant, such as a consumer chatbot.
Among that group, 45% said they had returned an AI-recommended purchase within the past year. The finding is indicative: it comes from a small subgroup of the 250 Australian consumers surveyed and does not establish that AI recommendations cause higher return rates. It also measures shoppers who made a return, rather than the proportion of AI-recommended purchases returned.
Even with those limits, the research raises a practical business question. If customers increasingly use an assistant to help choose products, how well does the information available to that assistant describe what the retailer is selling?
Duncan Angove, CEO of Blue Yonder, argues that the quality of that information is central to the experience.
“An AI assistant is only as good as the product information it can find, and most product data was written for a search bar, not for a machine suggesting products for someone,” he said.
“When what arrives doesn’t match what the shopper expected, it goes back in the box.”
For retailers, the implication is that product information needs to do more than help an item appear in search results. It needs to support a purchasing decision. Details such as dimensions, materials, fit and intended use can help explain whether an item suits a particular need.
Incomplete or ambiguous descriptions leave room for a mismatch between the recommendation and the product delivered. Reviewing those descriptions is a concrete step businesses can take, without assuming that adding an AI shopping feature will itself solve the problem of unsuitable purchases.
Adoption varies considerably by age. Half of Gen Z respondents and 44% of Millennials said they had purchased a product following an AI recommendation. That compared with 24% of Gen X and 10% of Baby Boomers.
The differences suggest retailers will encounter customers with very different levels of familiarity and confidence. An experience designed around enthusiastic users may not suit shoppers who still want to assess product details themselves.
Trust in AI sizing advice is also conditional. A quarter of Australian respondents said they would trust an AI tool to recommend the right size or fit to avoid a return. Another 41% would potentially trust it depending on the retailer, while 34% would not.
Among both Gen Z and Millennials, 35% would trust such a recommendation, compared with 17% of Gen X and 10% of Baby Boomers.
The retailer’s role in that conditional trust is significant. Consumers are being asked to rely on a recommendation, but the retailer remains responsible for describing its merchandise accurately and delivering the purchase experience it promises.
Businesses could use the reasons customers give for returns to identify recurring gaps in product information. If shoppers repeatedly misunderstand a feature or choose an unsuitable fit, the description and guidance deserve review. The survey does not measure the effect of such changes, but they offer a practical area for investigation.
The study was fielded by a third-party provider in September 2026 as part of an international survey of more than 5,000 consumers. For retailers considering AI-assisted shopping, its findings support a measured approach: examine the information behind the recommendation and assess the experience through to what the customer decides to keep.
