Do Recommender Systems Manipulate Consumer Preferences? A Study of Anchoring Effects

Recommender systems are becoming a salient part of many e-commerce websites. Much research has focused on advancing recommendation technologies to improve accuracy of predictions, while behavioral aspects of using recommender systems are often overlooked. In this study, we explore how consumer preferences at the time of consumption are impacted by predictions generated by recommender systems. We conducted three controlled laboratory experiments to explore the effects of system recommendations on preferences. Studies 1 and 2 investigated user preferences for television programs, which were surveyed immediately following program viewing. Study 3 broadened to an additional context— preferences for jokes. Results provide strong evidence that the rating provided by a recommender system serves as an anchor for the consumer’s constructed preference. Viewers’ preferences appear malleable and can be significantly influenced by the recommendation received. Additionally, the effects of pure number-based anchoring can be separated from the effects of the perceived reliability of a recommender system. When the recommender system was described as in testing phase, the anchoring effect was reduced. Finally, the effect of anchoring is roughly continuous, operating over a range of perturbations of the system. These general findings have a number of important implications (e.g., on recommender systems performance metrics, preference bias, potential strategic behavior and trust), which are discussed.