Why do shoppers abandon online carts? What a Journal of Consumer Research study found
Carts full of treats get abandoned more than carts full of practical items, because indulgence feels harder to justify. Across almost 15 million items and four experiments, one fix stood out: recommend something practical.
Carts with more indulgent items are abandoned more often. Recommending a practical add-on cut abandonment from 4.61 to 3.64 out of 7. Here is how to use it.
Marketing in 3 · Issue 6 · Week of 16 February 2026
More than 70% of online carts are left behind. Photo: James Watson on Unsplash.
The short answer
Partly because of what is in them. Carts with a higher share of indulgent (hedonic) items, such as chocolates, scented candles or novelty T-shirts, were more likely to be abandoned than carts of practical (utilitarian) items. The whole cart feels more indulgent, and buying it feels more guilt-inducing. Recommending a practical add-on reverses this: shoppers saw the cart as less indulgent, felt less guilty and were less likely to abandon it, even if they added the extra item.
Key takeaways
In a four-item cart, swapping practical items for indulgent ones raised the likelihood of abandoning it from 4.41 to 5.07 out of 7, and raised guilt from 2.66 to 3.71.
Field data from two online stores, with almost 15 million items bought or abandoned, showed the same pattern: the higher the share of indulgent items, the more often carts were abandoned.
Recommending a practical item ("One more item you need") cut abandonment of a cart holding one treat from 4.61 to 3.64; recommending another treat cut it only to 4.18.
The practical recommendation worked even when shoppers assumed the item was added, which roughly doubled what they would pay.
What the researchers did
More than 70% of online shopping carts are abandoned, by industry estimates, and the lost revenue runs into
billions a year. Most fixes target the checkout: hidden costs, slow delivery, long forms. Liat Hadar, Yael
Steinhart and Yaniv Shani (Tel Aviv University) and Gil Appel (George Washington University) looked instead at
what is in the cart.
They separated hedonic products, bought for pleasure (chocolate truffles, a scented candle, a decorative pillow),
from utilitarian products, bought for a practical job (protein bars, emergency candles, a water bottle). Decades
of research show that indulgent purchases are harder to justify and make people feel guilty. The question was
whether a cart works the same way: does a cart with more treats feel like a guiltier purchase as a whole?
Field data
14.5 million items in 9.8 million sessions at a multicategory online store (2019–2020) and 179,472 sessions at an outdoor gear store (2017–2019)
Experiments
4 online experiments with 1,670 US adults, plus 2 more in the web appendix; one preregistered
Measures
Real purchase or abandonment in the field; likelihood of abandoning the cart and each item, guilt, and how indulgent the cart felt in the experiments
Classification
Product categories rated hedonic or utilitarian by two coders, with a generative AI model resolving disagreements
What they found
In both stores, carts with a higher ratio of hedonic to utilitarian items were more likely to be abandoned, after
accounting for cart value, number of items, items viewed, session length and time of day.
The experiments showed why. Shoppers imagined an Amazon cart of four items at about $10 each. With three or four
indulgent items, they rated themselves more likely to abandon it (5.07 out of 7) than with one or none (4.42 and
4.41). The more treats, the more indulgent the whole cart felt and the guiltier people expected to feel buying it,
and that chain explained the abandonment. The same held for each item: a practical item was more likely to be
dropped when it sat in a cart full of treats. Value for money did not change, so it was not about price.
Likelihood of abandoning a four-item cart, by number of indulgent items
Likelihood of abandoning a four-item cart, by number of indulgent items
Group
Value (out of 7)
0 indulgent, 4 practical
4.41
1 indulgent, 3 practical
4.42
3 indulgent, 1 practical
5.07
4 indulgent, 0 practical
5.07
Study 2, 601 US adults, imagined Amazon cart of four items at $9–11 each, 1–7 scale.
Guilt, not necessity
One alternative is that practical items are simply needed, so people keep them. A third experiment ruled this
out. A cart with a tote bag plus two indulgent items (faux succulents and a card game) was more likely to be
abandoned (5.32) than the same tote bag plus two neutral items (snack bars and a phone stand; 4.73). The
difference came from the indulgent items, not from losing a necessity.
Guilt followed the same pattern: shoppers expected to feel guiltier buying carts with three or four indulgent
items (3.56 and 3.71 out of 7) than carts with one or none (2.73 and 2.66).
Expected guilt about buying the cart, by number of indulgent items
Expected guilt about buying the cart, by number of indulgent items
Group
Value (out of 7)
0 indulgent
2.66
1 indulgent
2.73
3 indulgent
3.56
4 indulgent
3.71
Study 2, 601 US adults, 1–7 scale.
The fix: recommend something practical
Because shoppers judge the cart as a whole, what a store recommends can change how the cart feels. Shoppers with a
scented candle and chocolate truffles in their Walmart cart who were shown protein bars and an exercise T-shirt
under "People who buy these products frequently buy these items as well" felt less guilty (2.85 against 3.32) and
were less likely to abandon the cart (4.38 against 4.86).
A second test compared recommendations directly. With a tea sampler in the cart, recommending a water bottle
("One more item you need…") lowered abandonment to 3.64, from 4.61 with no recommendation. Recommending a box of
chocolates ("One more item to indulge yourself…") helped far less (4.18). Only the practical item made the cart
feel less indulgent. It worked whether or not shoppers assumed the bottle had been added, even though that meant
paying about twice as much.
Likelihood of abandoning a cart holding a tea sampler, by recommendation shown
Likelihood of abandoning a cart holding a tea sampler, by recommendation shown
Group
Value (out of 7)
No recommendation
4.61
Treat suggested (chocolates)
4.18
Practical item suggested (water bottle)
3.64
Study 4B, 457 US adults, 1–7 scale. Labels shown: "One more item to indulge yourself" and "One more item you need".
Marketing in 3 steps
Ready. Set. Market.
1
ReadyDecide and prepare
Tag products as treats or practical. Add a hedonic or utilitarian tag to each product category. The authors used two human coders and a generative AI model to settle disagreements; a quick internal rating works as a start.
Watch indulgent carts in your analytics. Track abandonment by the share of indulgent items in the cart. Treat-heavy carts are a natural target for basket reminders and guilt-reducing messages.
2
SetBuild it
Steer recommendations toward practical items. When a cart is mostly indulgent, show useful add-ons such as accessories, refills, basics or care products, rather than more treats. Keep the cart's share of indulgent items low if the suggestion is added.
Use practical framing in the label. "One more item you need" outperformed "One more item to indulge yourself". Labels that signal usefulness help the cart feel justified.
3
MarketLaunch and measure
Discount the treat, not the basics. The authors suggest that discounts framed on the indulgent item can ease the guilt of buying it, drawing on earlier research on mixed bundles.
A/B test it. Compare your current recommendation logic with a version that favours practical items for treat-heavy carts, and measure checkout completion and basket value.
What to expect
Fewer abandoned carts. In the experiments, a practical recommendation reduced abandonment likelihood by about 0.5 to 1.0 points on a 7-point scale.
Bigger baskets, not smaller. The effect held even when shoppers assumed the recommended item was in the cart, so the fix can raise order value as well as conversion.
Small shifts in real data. In the field data the effect was consistent but modest, alongside the usual drivers such as price, delivery and browsing without intent.
Low cost. It uses the recommendation engine you already have; only the selection logic and wording change.
Where it does not work
Imagined carts. The experiments used hypothetical carts of $10 items; the effect of recommendations on real checkouts was not tested in the field.
Correlational field data. The store data show an association, not cause, and the share of indulgent items was low (13% and 2% of items).
Wording may matter. Labels such as "items you need" might make shoppers see themselves as sensible buyers; neutral labels like "related items" are untested.
Online only. Whether the same happens with physical trolleys in store is unknown.
Questions marketers ask
Why do customers abandon online shopping carts?
Besides checkout problems such as extra costs, one reason is the cart's contents. A Journal of Consumer Research study found that carts with a higher share of indulgent items are abandoned more often, because the whole cart feels indulgent and buying it triggers guilt (Hadar et al., 2026).
How can product recommendations reduce cart abandonment?
By suggesting practical items. When a cart held an indulgent item, recommending a practical add-on cut the likelihood of abandoning it from 4.61 to 3.64 out of 7, while recommending another treat cut it only to 4.18.
What is the difference between hedonic and utilitarian products?
Hedonic products are bought for pleasure and enjoyment, such as chocolates, scented candles or decorative pillows. Utilitarian products are bought to do a practical job, such as protein bars, a water bottle or trash bags.
Does adding more items to the cart increase abandonment?
Not if the added item is practical. Shoppers who assumed a recommended practical item had been added were still less likely to abandon the cart, even though the total price roughly doubled.
Do people feel guilty about online shopping?
Yes, especially for indulgent purchases. In the study, expected guilt rose from 2.66 out of 7 for a cart of practical items to 3.71 for a cart of indulgent items, and guilt explained the higher abandonment.
Consumers hold their own theories about what data an AI uses to recommend products, and these beliefs shape whether they take the advice. Across a pilot and four studies, people distinguished three kinds of data and were less accepting when the data seemed to threaten their individuality or its processing felt unacceptable. How you explain the data behind a recommendation can matter as much as the recommendation itself.
Across six experiments, people shared the least personal information when asked whether a firm could keep their data, and more when offered the option to delete it; in one study the delete option produced 54% more shared data than the keep option. Keep requests leave the action with the firm, which makes people feel vulnerable. Offering visible delete controls can increase, not reduce, what customers share.
Source
Hadar, L., Steinhart, Y., Appel, G., & Shani, Y. (2026). The effect of online cart composition on cart abandonment. Journal of Consumer Research. https://doi.org/10.1093/jcr/ucag002
This brief is written for practitioners from the published article. Open access (CC BY-NC 4.0): free to read at the publisher. Numbers are as reported by the authors; read the article for full methods.
Facing all-White staff, 50% of non-White customers chose a digital kiosk, against 36% with non-White staff. What it means for service design and AI options.
A "recommended for you" weight-loss refill lowered ratings of the pharmacy from 3.79 to 3.02 out of 7. When personalization embarrasses customers, and what to do.
Bruno Schivinski is Associate Professor of Marketing at Gdańsk University of Technology. He studies how consumers engage with brands in digital and social media.
