Taming Tech Giants’ Algorithms: An analysis of consumers’ choices on Amazon

Abstract

Online marketplaces rely on opaque ranking algorithms that steer consumers’ choices. We study how transparency about such algorithms affects consumer behavior on Amazon’s marketplace, focusing on the Buy Box—the feature that preselects a default seller for most products. We conduct an online (framed) field experiment with Italian Amazon users, varying the amount of information provided about the Buy Box to study resulting changes in consumer search and deal selection. Informing consumers increases their willingness to inspect alter- natives and reduces reliance on the default deal, without increasing search time or effort. At the end of the experiment, most participants hold critical views of the Buy Box mechanism and are willing to pay a fee to opt out of it in favor of a customized selection service. Our results highlight how a limited understanding of platform algorithms reinforces consumer inertia and demonstrate that simple information interventions can improve welfare without altering platform design.

Keywords

JEL codes