Personalized content, engagement, and monetization in a mobile puzzle game

Abstract

Digital technologies have reduced the cost of collecting detailed information on consumer characteristics and behavior. Despite the large literature on the consequences of using these data to personalize prices, little is known about content personalization. Using detailed player-level data from a mobile puzzle game and a novel structural model of player behavior, we investigate the effects on revenue of personalizing game difficulty using observable player characteristics. Our results show that, while average difficulty across players is successfully set by the game developer to maximize revenue, personalization can further increase revenue by 71%. Personalized difficulty leads to an overall increase in player engagement and, consequently, revenue generation in the form of in-app purchases. Although the largest relative increase in revenue comes from the smallest spenders, most of the absolute increase in revenue comes from a further increase in spending by the largest spenders.