Retailers are being urged to check what AI shopping assistants tell customers before Black Friday, after research found that nearly half of surveyed AI shoppers had abandoned a purchase over conflicting product information.
The Adobe for Business survey of 1,011 US consumers who use AI assistants to research purchases found that 49% said they had abandoned a purchase after information from an AI tool did not match the brand’s website.
The findings, published on October 5, also showed that 59% said inaccurate AI information reduced their trust in a brand, even when the brand had not produced the answer.
Thomas Phillips, co-founder and CEO of DTC SEO Agency, explains that being recommended by an AI assistant is only useful if the information helps customers make an informed purchase.
“An AI assistant could recommend your product and still give someone a reason to leave when they reach your website. If it tells them shipping is free and they discover a charge, they have to decide which information to believe.
That customer may have been ready to buy. Now they are checking the basics, and every unanswered question gives them another reason to look elsewhere.”
With retailers preparing their Black Friday promotions, Phillips recommends three checks to identify discrepancies that could undermine a sale.
1. Check the details that could change a buying decision
Start with a small selection of best-selling products and ask AI assistants the questions customers would ask before ordering. These could cover sizing, materials, compatibility, delivery costs or whether an item can be returned after opening.
Record the answer, the date and any sources provided, then compare the information with current product pages and policies. Repeat the checks to establish whether an error keeps appearing.
“A mention in an AI answer tells you very little on its own. Check whether it describes the right version of the product and gives the customer accurate information about what they are buying,” Phillips explains.
For a small retailer, start with the products that matter most to your sales. You do not need to inspect your entire catalog to find a problem worth fixing.”
2. Review old offers before launching new discounts
Old promotional pages, outdated shipping thresholds and inconsistent return policies can create confusion before a customer reaches the checkout.
Retailers should review previous offers alongside their planned Black Friday promotions, making it clear when each offer applies, which products qualify and what conditions customers need to meet.
“Imagine someone arriving expecting last year’s discount because an old offer page is still easy to find. Even if your current promotion is clear, they are starting with an expectation you cannot meet.
Put the relevant terms close to the offer. A customer should not have to open three separate pages to work out what they will pay.”
3. Find where incorrect information is coming from
When an AI answer contains an error, check any cited pages to understand what may be causing it.
The source could be an outdated retailer listing, contradictory information on the brand’s website or an unsupported claim generated by the assistant. Identifying the source helps retailers decide what they can correct.
“Correct information you control and request updates where another website has the details wrong. You cannot guarantee what an AI tool will say, but you can remove avoidable contradictions from the information customers use to check its answer,” Phillips adds.
Before Black Friday, ask what a shopper might believe about your product before they arrive. Then check whether your website gives them a clear, accurate answer at the point they are deciding to buy.”
