The Anticompetitive Stickiness of Default Applications: Addressing Consumer Inertia with Randomization

By Omar Vásquez Duque (Stanford University)

Preinstalled applications are pervasive in both computers and mobile devices. Recent antitrust cases in the U.S. and Europe have scrutinized the defaults’ potential to foreclose competing applications. So far, enforcers and analysts have tended to assume that default (or stickiness) effects exist, and that forced choice strategies are the best way to overcome the users’ inertia and consequently level the field among competitors.

This article uses experimental methods to test the stickiness hypothesis and the effectiveness of potential remedies. With a series of experiments ran on Amazon Mechanical Turk, this work evaluates whether current market shares reflect what consumer would choose if they were forced to choose; or whether default effects exist and why. Due to the difficulty of resembling the choice environment in which people use a search engine, this study focused on the desktop segment of the market, but the participants completed the task from both desktop and mobile devices.

When facing a task for which a search engine must be used, most of the experiment participants choose Google, and the preference for Google resembled current market shares. Nonetheless, the participants who were assigned to the Yahoo or Bing by default conditions tended to give these alternatives a try before switching to Google. My data suggest the consumers’ misperceptions about Google alternatives’ quality is an important obstacle to their market penetration. This effect is clear for Bing and not for Yahoo. Despite both having the same search algorithm only Bing benefited from a substantial increase in its perceived quality after experimenting with it. This may imply that only defaults that users regard as a high-quality option may stick in the long term. The preceding findings suggest that the main anticompetitive effect of Google’s strategy is to prevent users from exploring competing search engines that they do not already believe to be good enough. Thus, effects of choice screens are limited by the consumers’ lack of information about their alternatives.

Continue Reading…