Location
260-055, Owen G. Glenn Building
Start Date
12-15-2014
Description
We investigate the impact of several recommender algorithms (e.g., Amazon.com's “Consumers who bought this item also bought”), commonly implemented in e-commerce, on sales volume and diversity using a field experiment on a top retailer website. Sales volume refers to the number of products purchased while sales diversity refers to the market share distribution of purchased products. For sales volume, we show that different algorithms have differential impacts with a widely used algorithm having no impacts. For sales diversity, we find that collaborative filtering in particular cause individuals to discover and purchase more variety of products but push each individual to the same set of popular titles, leading to concentration bias at the aggregate level. Managerially, our results inform personalization and recommendation strategy in e-commerce. Academically, this paper adds to the emerging but controversial topic of the impact of recommenders on sales volume and diversity for several algorithms with field experiment
Recommended Citation
Lee, Dokyun and Hosanagar, Kartik, "Impact of Recommender Systems on Sales Volume and Diversity" (2014). ICIS 2014 Proceedings. 40.
https://aisel.aisnet.org/icis2014/proceedings/EBusiness/40
Impact of Recommender Systems on Sales Volume and Diversity
260-055, Owen G. Glenn Building
We investigate the impact of several recommender algorithms (e.g., Amazon.com's “Consumers who bought this item also bought”), commonly implemented in e-commerce, on sales volume and diversity using a field experiment on a top retailer website. Sales volume refers to the number of products purchased while sales diversity refers to the market share distribution of purchased products. For sales volume, we show that different algorithms have differential impacts with a widely used algorithm having no impacts. For sales diversity, we find that collaborative filtering in particular cause individuals to discover and purchase more variety of products but push each individual to the same set of popular titles, leading to concentration bias at the aggregate level. Managerially, our results inform personalization and recommendation strategy in e-commerce. Academically, this paper adds to the emerging but controversial topic of the impact of recommenders on sales volume and diversity for several algorithms with field experiment