The Decoy Effect and Recommendation Systems.

This article examines the impact of the decoy effect—a marketing strategy where an asymmetrically dominated item (decoy) influences consumer choice—within recommendation systems, distinguishing between personalized and nonpersonalized contexts. Through randomized controlled experiments using a movie recommendation platform, the study finds that in nonpersonalized recommendation settings, adding a decoy increases demand for the target item, consistent with traditional decoy effect findings. Conversely, in personalized recommendation settings, including a decoy decreases demand for the target item and increases the likelihood of consumers opting out of a purchase, attributed to diminished perceived recommendation quality and system reliability. These results highlight the importance of considering consumer expectations and recommendation context when designing recommendation systems, suggesting that while decoys may enhance effectiveness in nonpersonalized settings, they can be counterproductive in personalized ones.