Why removing an option can raise sales of the rest

A chain reduces one category from twenty-six lines to fourteen, keeping the best sellers and the extremes. Category sales do not fall. In several documented cases they rise, and shoppers report the shelf as better stocked than before, which it is not.

What everyone sees

Range is assumed to be an asset: more lines mean more chances that a shopper finds their preference. Buyers defend assortment on exactly that logic, and the logic is sound for genuinely different items. It is applied indiscriminately to items that differ only in ways the shopper cannot evaluate at the shelf.

What is actually happening

Boatwright and Nunes analysed a large assortment reduction and found sales generally held or increased where popular items were retained. Broniarczyk, Hoyer and McAlister showed that perceived variety depends more on the presence of favourites and on space allocation than on the raw count. Chernev added that near-identical options raise the cost of comparison and make deferral more likely, so cutting duplicates removes work rather than possibility.

Why it stays hidden

The finding hides because the two views describe different quantities. Range counts lines; the shopper experiences whether a decision can be completed. Nobody experiences the twenty-six-line version and the fourteen-line version of the same trip, and the abandoned purchases in the larger assortment leave no record, so the assortment looks costless from the buyer’s side.