The Jam Study, Retested: 24 Flavors, 68 Percent Fewer Purchases

Proof, not claims: classic persuasion triggers, re-tested live with digital twins. → See all 23 purchase triggers at a glance.

The takeaway: Iyengar & Lepper (2000) showed in their famous jam field study that a tasting booth with 24 flavors attracted more passersby, but ten times fewer of them actually bought than at the booth with just 6 flavors. In our re-test with digital twins, the twins bought a jar on average in 2.7 of 10 passes with 6 flavors, but only 0.9 of 10 with 24 flavors, a drop of roughly 68 percent. But: only 4 of 20 justifications even mentioned the specific flavor count at all (all four in the 24-flavor condition), and not a single one named a specific jam flavor, by the strict, pre-registered criterion, that's an 80 percent confabulation rate. The finding stands on shakier ground than the 68 percent alone would suggest.

Sheena Iyengar and Mark Lepper describe a trigger that nearly every marketing guide cites today. We re-tested it with digital twins.

What is choice overload, exactly?

Few findings from consumer psychology have made it into everyday marketing thinking as thoroughly as the idea that too much choice paralyzes buyers instead of exciting them. The short version, "fewer options sell better", shows up in guides, keynotes, and product-page playbooks, almost always with the same reference: Sheena Iyengar and Mark Lepper's jam shelf.

Stylized recreation: a dense jam shelf with many jars next to a clear shelf with few jars

Iyengar & Lepper (2000, Journal of Personality and Social Psychology) set up a real tasting booth in a supermarket near San Francisco. On some days the booth displayed 6 jam flavors, on others 24. The result became a classic: the larger booth attracted more passersby (60 percent stop rate versus 40 percent), but among those who stopped, nearly one in three bought at the small booth (30 percent), versus almost no one at the large booth (3 percent). More choice drew people in, and put them off at the decisive moment.

The research picture since then is not quite as clean as the marketing quote makes it sound, though. Scheibehenne, Greifeneder & Todd (2010, Journal of Consumer Research) pooled 63 conditions from 50 studies with a combined 5,036 participants, and found an average effect of practically zero (D = 0.02). Choice overload, in other words, is not a reliable main effect but depends heavily on the context, the product, and the audience being tested.

How did we test this with digital twins?

A side-by-side comparison, where one twin sees both shelves at once and picks between them, would have been the wrong move here. The moment a twin reads "shelf with 6 flavors" next to "shelf with 24 flavors," it stops judging its own purchase behavior and starts judging the abstract number instead, and the folk wisdom that "too much choice overwhelms" is so well known that a direct comparison would almost inevitably have produced a clean but low-value result. Iyengar and Lepper themselves, notably, didn't design their original study as a comparison test either: each passerby in the real study saw only one of the two shelves, never both.

We rebuilt that setup. Two sub-panels from our twin panel each saw only one condition: a jam shelf with 6 explicitly named flavors, or one with 24 (the 24-flavor list contains the same 6 flavors plus 18 more, exactly as in the original, where the larger selection was an extension of the smaller one, not a completely different list). Across two waves with the assignment swapped, each twin ultimately saw exactly one condition, splitting the panel cleanly across both shelf sizes.

The Method: 10 digital twins (DACH consumer panel, ages 25–60) were asked to imagine a tasting booth with 6 or 24 jam flavors, described as text, not shown as an image. Each twin saw only one of the two shelf sizes, but in two different listing orders across two waves, yielding n = 10 twins per condition (rather than the 10 twins × 2 runs in the same condition usual elsewhere in the Trigger Lab). Three values were collected: how appealing the booth seemed (1–10), how many of 10 passes would result in an actual purchase, and how many would result in the decision being postponed. The panel responded in German; quotes are translated.

Of the 20 responses, 7 (6-flavor condition) and 8 (24-flavor condition) respectively supplied usable numbers. The rest stayed at general statements without a concrete figure. Important for interpreting the result: only 4 of the 20 justifications explicitly named the actual flavor count, and not a single one named a specific jam flavor, a sign that much of the response likely stems from a general attitude toward "lots of choice" rather than genuine engagement with the shelf shown. And unlike Iyengar and Lepper, who counted real purchases with real money, our twins report a stated purchase intention across ten imagined visits, not tracked purchases.

The panel: 10 digital twins*

Beate Hofmann, 58
Project manager twin* · Stuttgart · University degree

“I'm Beate Hofmann, a project manager from Stuttgart. Since my divorce I've found new stability with a new partner, and even though I'm living with chronic back pain and an active cancer diagnosis, I stay active with daily exercise and feel deeply satisfied with my life.”

What makes this twin distinct: I hold strong private religious beliefs without attending church, I'm deeply skeptical of politics and the economic situation, and I guard my data so carefully that I'll pass up a discount rather than share it.