One of AI's many promises is that it will help us do everything better. Work faster, research more seamlessly, and perhaps be more financially prudent? Well, not exactly. Research suggests that AI-assisted shopping may end up making you spend more.
According to a new study from Bayes Business School at City St. George's, University of London, shoppers who used AI-powered shopping carts spent 32% more at the supermarket than those who stuck with an ordinary trolley. The technology also resulted in people buying more items and lingering longer in the aisles.
The researchers set out to measure the impact of a digitized shopping experience on consumer behavior, and by extension, retailer profits. The tool of choice was a "smart trolley," a regular cart fitted with a tablet on the handlebar. These screens let shoppers digitize their shopping lists, receive personalized product recommendations, get in-store navigation, and pay without ever queuing at a checkout. In theory, a friendlier, frictionless trip. In practice, possibly a rather effective upselling machine.
Over a one-month period in March 2025, the researchers observed 12,418 unique shopping sessions at a well-known German supermarket chain, 9,422 of which involved a "smart shopper" using one of the AI carts.
For every session, they logged the basket value, the number of items bought, and the total time spent in store, then timestamped each trip and tied it to the day of the week and the time of day the cart was first switched on. They split days into morning, afternoon, and evening, and separated weekdays from weekends. The team even tracked what shoppers used the screens for, whether that was lists, recommendations, navigation, or just watching interesting ads.
According to the collated data, smart-cart users spent 32% more on average than non-users. They also loaded up on 25% more items (12.02 against 9.61) and spent 23% longer in store (40.40 minutes versus 32.75). Spending spikes were sharpest in the afternoons and on weekends, while the longest, most leisurely trips happened in the evening. As a bonus finding, hotter weather appeared to send everyone home sooner, shrinking trip durations across the board.
Lead author Dr. Sabrina Gottschalk, a Lecturer in Marketing at Bayes Business School, framed this as a real revenue opportunity for stores. Beyond steering customers toward offers and new products, she noted that going digital ramps up engagement and gives retailers a fresh advertising platform right on the handlebar.
“Retailers are increasingly using technology to enhance the shopping experience for consumers, and our findings suggest there may be significant revenue gains from doing so effectively. As well as guiding customers towards special offers and new products, digitalization induces increased engagement and additional advertising platforms,” she said.
However, there was a twist. When the researchers zoomed in on their most enthusiastic users, the "superusers" who racked up more than 20 screen interactions in a single trip, things no longer appeared to be perfectly aligned. These shoppers bought significantly more items and stayed even longer, yet they did not spend more money overall. Past a certain interaction peak, both spending and basket size actually started to fall.
Now, on the surface, this looks like a contradiction. How do you buy more but spend the same, or tap the screen more and end up buying less? This is common in observational studies like this, which usually can’t clearly separate causality from correlation due to some immeasurable variable. It's possible that smart-cart users spend more because of the cart. However, it could also be that the kind of person who reaches for a smart cart was always going to fill a bigger basket.
Then there is the reading that scores some points for AI. If superusers bought more items on average but did not outspend lighter users, perhaps they were making smarter choices, spotting the deals, comparing prices, and letting the AI point them toward better value. Then again, it could just as easily mean that once shopping turns into a screen-tapping game, people drift toward browsing for entertainment rather than actually buying much.
That said, the researchers see the 32% average higher spend as a clear outcome and an opportunity for retailers, whether the extra spending was driven by sneaky nudging or by AI helping consumers shop smarter. After all, spending more doesn't necessarily mean bad decision-making. It could be a save-more-in-the-long-run kind of situation, or simply remembering to get everything you actually need in one trip. Either way, cha-ching for the retailers! The researchers even suggest “enticing” customers to use the tech.
“Our research shows clear spikes in spending and consumption for those deciding to use in-store digital assistance, suggesting that retailers should offer this technology and entice customers to use it – perhaps with a promise of loyalty points or prizes,” said Dr. Yusuf Oc, co-author on the study published in the Journal of Business Research.
So, should you brace yourself the next time a tablet-toting trolley rolls your way? Not necessarily.
The carts are genuinely convenient, and nobody is directly forcing anything into your basket. However, it is worth remembering that every recommendation and cheerful prompt on the screen is engineered with the retailer's bottom line in mind, not yours. As Gottschalk cautions, the smart play is to enjoy the convenience while staying conscious of how those digital nudges shape what ends up in your cart. Bring a list, know what you came in for, and maybe let the AI handle the navigation while you keep a firm hand on your wallet.
