Choice overload: why more options can mean fewer sales.

In 2000, the psychologists Sheena Iyengar and Mark Lepper published a study in the Journal of Personality and Social Psychology titled "When Choice Is Demotivating: Can One Desire Too Much of a Good Thing?" They set up a tasting table in an upscale supermarket and stocked it, on alternating days, with either 6 jars of jam or 24. Everyone who stopped could sample as many jars as they liked, and everyone who stopped received a coupon worth a dollar off any jar they chose to buy.

The two tables produced very different numbers, and the difference ran in a direction most retailers would not predict. Sixty percent of shoppers stopped to look at the 24-jam table, against 40 percent at the 6-jam table, a clear win for foot traffic if you are only counting who stopped. Of the shoppers who stopped at the 24-jam table, though, only about 3 percent used their coupon to buy a jar. At the 6-jam table, about 30 percent did. A shopper who paused at the smaller display was roughly ten times more likely to leave with a jam in hand.

Why walking away feels easier than choosing well

The mechanism is not laziness. Comparing 24 jams means holding two dozen sets of attributes in mind at once, sweetness against price against fruit content against whether the label looks like something worth bringing home. Working memory strains under that load, and the strain grows faster than the value of each extra jar added to the table.

A second cost compounds the first. The shopper is weighing the jars in front of her alongside the version of the future in which she picked the wrong one while a better jar sat two inches to the left. Regret aversion is one of the most consistently observed patterns in decision research, and it sharpens as the number of forgone alternatives grows. 24 jars generate 23 ways to have chosen badly. 6 jars generate 5.

Faced with rising comparison cost and rising regret on top of it, the brain has an exit that a smaller set of options does not make as tempting: leave without choosing. Walking away from a table and abandoning a half-finished sign-up form are the same move wearing different clothes: a decision, made efficiently, that this is not the moment to choose. The shopper has judged, correctly by her own accounting, that deciding right now costs more than it is worth.

Attraction and conversion moved in opposite directions

The two outcomes in this experiment moved in opposite directions from a single cause, and that is the detail worth sitting with. The large table drew the bigger crowd, exactly what a marketer watching foot traffic would want to see. It converted that crowd at roughly a tenth of the rate the small table achieved. A retailer watching only foot traffic at the 24-jam table would have called it the clear winner. A retailer watching the till would have shut it down by lunchtime.

More choice pulled in more lookers. It produced far fewer buyers.

Choice overload is easy to miss in practice for exactly this reason. The number most businesses check first, whether that is footfall or click-through rate, is the number a wide assortment inflates. Conversion, the number that pays the bills, moves the other way. A manager reading only the stop-rate data from Iyengar and Lepper's experiment would have recommended widening the range further. The coupons said otherwise.

Where the pattern shows up beyond the jam table

The pattern travels well beyond jam. A supermarket on a busy strip in Kingston or Port of Spain that stocks eighteen variants of one snack brand across three shelf facings is running its own 24-jam table every day, without measuring it. A restaurant with a laminated twelve-page menu is asking a hungry table to do the same comparison work that pushed 97 percent of the 24-jam shoppers to leave without buying anything. A telecom operator with nine postpaid plans, each differentiated by data allowance and rollover terms, is asking a customer at the counter to hold nine sets of trade-offs in mind before picking one. A bank offering six current account tiers, each with its own fee schedule and minimum balance, is asking the same thing at the branch counter.

Removing options has a real cost of its own. Genuine variety in taste and budget is real, and a business that prunes its assortment down to one option per category loses the customers who wanted precisely the option that got cut. Structure is what the jam data recommends: a clearer route through the choices that remain, more than a shorter list for its own sake.

What reduces the comparison burden, in the assortment and product work we measure, is a clearer route through the choices that remain more than a reduction in the choices themselves. Defaults do a lot of that work without a customer ever noticing: set a sensible plan or account as the one she gets unless she actively chooses otherwise, and anyone without a strong preference is spared the comparison entirely, while anyone with a preference can still override it. A bundle that pre-packages a sensible combination of features performs the same service in another form, and a product clearly marked as the most popular choice gives everyone else an anchor to measure the rest of the shelf against, instead of comparing nine plans with no starting point. Each technique leaves the full range on the shelf and only changes how much work a customer has to do to get through it.

Testing an assortment instead of guessing at it

Assortment decisions are usually made on instinct, or on whichever range a category buyer has always stocked, because testing them has felt expensive or vague. It does not have to be. Two measures translate the jam finding into something a retailer or bank can act on before a shelf reset or a plan redesign ships.

The first is decision time paired with gaze bounce-back. Eye tracking on a shelf photograph or a plan comparison page shows where attention lands and how it moves. A shopper comparing efficiently looks across the options once or twice and settles. A shopper caught in overload keeps returning to options already seen, bouncing back across the same few items without settling on one. That bounce-back pattern, measured in seconds rather than guessed at in a debrief, is a direct read on whether a layout is helping someone decide or trapping them in comparison.

The second is biometric arousal, typically heart rate variability or skin conductance, read alongside the eye-tracking data rather than on its own. Arousal by itself is ambiguous: it rises for excitement and it rises for stress, and a rising line on a chart does not say which. Paired with gaze and outcome data, a spike in arousal that coincides with repeated bounce-back across a long options list reads as stress, the body's response to an unresolved comparison, rather than the engagement a marketer might hope a big spike means. A spike that coincides with settling on an option and moving on reads as interest. The difference matters, because a menu or plan page can look identical on a click-through report and produce opposite emotional experiences for the person using it.

A live sales test still matters once a shortlist of two or three redesigned menus or shelf sets exists. A/B testing is the right tool for confirming which of a small number of finalists sells more, once you are down to real candidates, though it has its own well-documented limits for bigger brand questions. What eye tracking and biometrics add is the diagnostic step before that stage: understanding why a layout produces the sales number it does, and catching a design that would fail before it goes anywhere near a live test.

Where the finding breaks down

The jam study is not the only word on this subject, and a fair account has to say so. In the two decades since Iyengar and Lepper published, other researchers tried to reproduce the choice overload effect in different product categories. The results came back mixed: some studies found a strong effect, and plenty of others found a small one or none at all.

In 2010, Benjamin Scheibehenne, Rainer Greifeneder, and Peter Todd pooled the available studies into a meta-analysis published in the Journal of Consumer Research, titled "Can There Ever Be Too Many Options? A Meta-Analytic Review of Choice Overload." Averaged across every study in the pool, the overall effect came out close to zero, which on its own reads like it undoes the jam result.

That average hides more than it reveals. The moderating factors matter more than the headline number. Choice overload showed up reliably when the options were hard to compare against each other and the person choosing had no clear preference walking in, the classic profile of a novice facing an unfamiliar category. It showed up far less when the shopper already had a preference or enough category expertise to size up the options quickly. A buyer choosing among 40 wines she cannot tell apart is in overload territory. A sommelier choosing among the same 40 wines is not.

The honest position, given where the research has landed, is that choice overload is real, and it is conditional. It shows up reliably under some conditions and fades under others, and no business should assume the jam result transfers to its own assortment unchanged. The only way to know whether a given assortment is triggering it is to test that assortment, on that audience. If that test has not been run yet for your product line, that is the conversation to have before the next reset or redesign ships.

Frequently asked questions

What did the famous jam study actually find?

In 2000, Sheena Iyengar and Mark Lepper set up a tasting table in a supermarket offering either 6 jams or 24, with a discount coupon for anyone who wanted to buy. Sixty percent of shoppers stopped to look at the 24-jam table, against 40 percent at the 6-jam table, but only about 3 percent of the 24-jam group used their coupon to buy a jar, against about 30 percent at the 6-jam table. Fewer options outsold more options by roughly ten to one.

Why does more choice sometimes lead to fewer sales?

Comparing a large set of options is mentally expensive, and every extra option adds a new way to imagine having chosen wrong. Those two costs, the effort of comparing and the fear of regret, rise together as the option count grows. Past a certain point, the brain resolves both by deferring the decision rather than working through it, and deferring often means walking away without buying anything.

Does this mean businesses should always offer fewer options?

No. Cutting a range indiscriminately throws away variety that some customers genuinely need, and a narrower assortment is not automatically a better one. What tends to help is structure: a sensible default unless a customer chooses otherwise, or a clearly marked most-popular option that gives shoppers an anchor to compare against. These reduce the comparison burden without removing genuine choice.

Does choice overload apply to every product category the same way?

No. A 2010 meta-analysis by Benjamin Scheibehenne, Rainer Greifeneder, and Peter Todd, published in the Journal of Consumer Research, pooled the available studies and found the average effect across all of them was close to zero. The effect was strongest when options were hard to compare and the person choosing had no clear preference walking in, and it was weak or absent when the shopper already knew what she wanted or had enough category expertise to judge the options quickly. The effect is real, but it depends on the category and the audience.

How would a business actually test whether its own product line has this problem?

Two measures work well together. Eye tracking across the product range shows decision time and gaze bounce-back, whether attention keeps returning to options already seen rather than settling on one. Biometric measures such as heart rate variability or skin conductance, read alongside the eye-tracking data, distinguish a stress spike from a genuine-interest spike during a long list of options. Once a shortlist of redesigned layouts exists, a live sales test confirms which one converts best.