Onsite Personalized Recommendation: A Field Experiment (under review)

Research output: Working paperAcademic


Onsite retargeting is a type of targeting and product recommendation strategy typically offered on the homepage of online retailers based on customers’ prior behavior on retailers’ own websites. Onsite retargeting consists of a retargeted product that consumers previously interacted with and the recommendations based on the retargeted product. While onsite retargeting presents a fertile opportunity to address customers in a personalized manner, little is known about how its effectiveness varies depending on the choice of the retargeted product and how customers respond to the retargeted versus recommended products. In collaboration with a leading online retailer, we collected data from a large-scale randomized field experiment to examine the effectiveness of wishlist-based retargeting (retargeting based on products in consumer’s wishlist) compared to commonly used view-based retargeting (retargeting based on viewed products). Our empirical results suggest that wishlist-based retargeting is more effective than view-based retargeting in generating clicks and conversions for the retargeted (source) products. However, the incremental effectiveness of wishlist-based retargeting declines for recommended products. In addition, we find that onsite retargeting loses effectiveness as time passes, but this negative timing effect is smaller for wishlist-based retargeting than view-based retargeting. We further conduct an online experiment that replicates the findings of the field experiment and explores the mechanisms underlying the observed behaviors. This work contributes to the retargeting and personalized recommendations literature and provides meaningful implications for practitioners.
Original languageEnglish
Publication statusPublished - 2022


Dive into the research topics of 'Onsite Personalized Recommendation: A Field Experiment (under review)'. Together they form a unique fingerprint.

Cite this