Digital shopping carts that offer features such as personalized recommendations, shopping lists and store navigation were linked to larger and longer supermarket trips in a new study.
Researchers analyzed 12,418 shopping sessions at a major German supermarket over one month. Shoppers who actively used the smart-cart technology spent an average of about 32% more, bought 25% more items and stayed in the store about 23% longer than shoppers who did not use it.
But the study, published in the Journal of Business Research, was observational. Shoppers chose whether to use the technology, so the findings cannot show that the carts themselves caused people to spend more or buy more.
That distinction matters because people planning a larger grocery trip may have been more likely to use the digital carts in the first place.
The carts were equipped with screens that could help shoppers upload lists, navigate the store, receive personalized recommendations and complete purchases without a traditional checkout.
Across the study period, 9,422 shopping sessions involved active smart-cart use, while 2,996 did not.
Average spending was €30.18 (about $35) among smart-cart users compared with €22.89 (about $26) among nonusers. Smart-cart users also bought an average of 12.02 items per trip, compared with 9.61 items, and spent about 40 minutes in the store, versus roughly 33 minutes among nonusers.
The differences were not the same at all times.
Spending differences were especially pronounced in the afternoon and on weekends, while smart-cart users tended to buy more items in the afternoon and evening. Shopping trips also lasted longer in the evening.
Researchers also looked at how people interacted with the carts.
Shoppers who uploaded shopping lists tended to have higher basket values but bought fewer items and completed their trips more quickly. In contrast, people who interacted heavily with the screen, which researchers called “superusers,” stayed longer and bought more items but did not necessarily spend more overall.
That suggests more interaction with shopping technology does not always translate into greater spending.
At very high levels of screen engagement, both spending and basket size began to level off or decline. The researchers suggested that some highly engaged shoppers may have been interacting with the technology for curiosity or entertainment rather than because they were responding to offers.
“Although technology should in theory improve shopping experiences, consumers should be wary of relying solely on devices and remain conscious of how digital prompts and recommendations shape their purchase decisions,” said lead author Sabrina Gottschalk, Ph.D., a marketing researcher at Bayes Business School.
The study does not reveal what shoppers actually bought.
Researchers measured basket value, number of items and time spent in the store, but they did not report whether smart-cart users bought more produce, snacks, prepared foods or other specific categories.
That means the findings cannot tell us whether the technology influenced the nutritional quality of shoppers’ purchases.
The results also should not be interpreted as evidence that artificial intelligence itself increased spending.
The carts included digital features such as personalized recommendations and navigation, but the study examined smart-cart use as a whole. It did not isolate which features, including any AI-driven components, were responsible for the differences in shopping behavior.
Because shoppers were not randomly assigned to smart or standard carts, other differences between users and nonusers may also help explain the results.
For example, someone planning a full weekly grocery run may be more willing to use a cart with a screen, while a shopper stopping in for a few items may have little reason to use one. The study design cannot fully separate those differences from any effect of the technology itself.
Still, the findings add to a broader body of research showing that the shopping environment can influence purchasing behavior, from store layout and promotions to digital recommendations and other cues.
The available study information does not identify a separate funding source. Two of the study authors were employed by the company that provided the shopping data and worked within its innovation team.