On two consecutive Saturdays, shoppers at Draeger's, an upscale grocery store in Menlo Park, California, found a tasting booth stacked with British jam. Wilkin & Sons, "Purveyors to Her Majesty the Queen," made 28 varieties. The researchers pulled the familiar ones like strawberry and raspberry and set out the 24 exotic flavors that remained, priced at 4 to 6 dollars a jar, with a 1-dollar-off coupon for anyone who stopped.
Some hours the table held all 24 jams. Other hours it held 6. The displays rotated hourly, counterbalanced across the two days, so the only meaningful difference was how many jars sat on the table. Sheena Iyengar and Mark Lepper ran the study, and the Journal of Personality and Social Psychology published it in 2000 under the title "When Choice Is Demotivating: Can One Desire Too Much of a Good Thing?"
The big table won the crowd. Of the 242 shoppers who passed the 24-jam display, 145 stopped - 60 percent. Of the 260 who passed the six-jam table, 104 stopped - 40 percent. Measured by foot traffic, the long shelf was the better ad.
Then the register settled the argument. Of the 104 who stopped at the six-jam table, 31 redeemed their coupon for a jar over the following week - nearly 30 percent. Of the 145 who stopped at the 24-jam table, 4 did. Three percent. Same store, same brand, same coupon. The only variable was the length of the shelf.
The crowd is not the register
The interesting part is not that the long shelf sold less. It is that it impressed exactly the people it was built to impress and still lost. The visitors at the big table did not even browse harder: they tasted an average of 1.5 jams each, against 1.38 at the small one - no significant difference, in the paper's own test. The 24-jam display gathered a crowd and sold almost nothing to it, and my explanation is responsibility. Pick wrong out of six and you had bad luck. Pick wrong out of twenty-four and you should have known better. A wide shelf converts attention into deferral, because deferral is the only self-protection available to a person standing in an aisle.
Retail has quietly bet the same way. Aldi stocks roughly 1,400 products against the tens of thousands in a typical supermarket, in stores of about 10,000 square feet where a Walmart supercenter averages 178,000. More than 90 percent of what Aldi sells is its own label - one ketchup, not a wall of them. Scott Patton, the chain's chief commercial officer, told CNBC the curated shelves let shoppers skip the overwhelm of a vast ketchup aisle, and framed the trade without apology: "You might have to shop at another store and that's actually OK with us. What we want our customers to do is shop Aldi first." The bet is paying. Aldi held 2.8 percent of the US grocery market as of early October 2025, a rounding error next to Walmart's 21, but its store traffic rose more than 50 percent from 2019 to 2024, and in 2025 its visits grew 8 percent while the sector managed 3.1. Shoppers are not choosing Aldi despite the short shelf. They are choosing it because of it.
The jam study is famous, and famously contested, so here is the honest version. In 2010, Benjamin Scheibehenne, Rainer Greifeneder and Peter Todd pooled 50 choice-overload experiments covering just over 5,000 participants and found the mean effect essentially zero. In 2015, Alexander Chernev, Ulf Bockenholt and Joseph Goodman pooled 99 observations from 53 studies, 7,202 participants, and found a real overall effect after all - but conditional on four things: options that are hard to compare, a decision that is genuinely difficult, a shopper unsure of what she wants, and a goal of simply getting it over with. Both findings hold at once, and together they are more useful than any slogan about less being more. Assortment is not a moral. It is a setting. When the job of the moment is a crowd - a launch, a fair booth, a feed - the 24-jam table is the right table. When the job is a sale, cut it down.
One caution about my own claim: none of this is an argument for thin copy. Every jar on the six-jam table still needs its full story told, and a short menu of half-explained options loses to a short menu of complete ones. The lever is narrower - how many options earn a seat at the same table - and teams rarely set it on purpose. It gets inherited from the category, the competitor or the CMS, and nobody ever puts it on the agenda.
The short shelf, on purpose
That is why I like a leaderboard more than a directory. Bid Spotlight's front page is one ranked table: bid for a spot, the higher bid sits higher, every paid entry ships with a promo video, and the top ranks get featured in the daily social posts. The full directory of every brand still exists for browsing, the way the aisle of more than 300 jams still exists at Draeger's. But the page that sells is the short one. As I write this, the front board holds three names - every early board is that small - and I have stopped apologizing for the number, because a shopper landing on it faces a six-jam decision instead of a three-hundred-jam one.
If your product is sitting somewhere in a 300-jam aisle, come take a seat at the short table - the top spots on the board are open for bids.