Different but Equal? A Field Experiment on the Impact of Recommendation Systems on Mobile and Personal Computer Channels in Retail

• 成果类型: Article
• 署名作者: Lee, Dongwon; Gopal, Anandasivam; Park, Sung-Hyuk
• 署名单位: Hong Kong University of Science & Technology; University System of Maryland; University of Maryland College Park; Korea Advanced Institute of Science & Technology (KAIST)
• 刊物名称: INFORMATION SYSTEMS RESEARCH
• ISSN/ISSBN: 1047-7047
• DOI: 10.1287/isre.2020.0922
• 发表日期: 2020
• 页码: 892-912
• 关键词: consumer decision-making; E-commerce; search costs; long tail; CHOICE; sales; adoption; FUTURE; memory

摘要

The benefits of recommendation systems in online retail contexts have received much attention in prior work. Much of this work has been conducted in personal computer (PC)-based settings, although mobile devices are becoming increasingly central to the online shopping experience. It remains to be examined if the effects of recommendation systems in retail differ across these two channels, in terms of customer-level decision outcomes. In this paper, we examine these differences in some detail, studying how product views and sales attributed to a recommendation system are different across mobile and PC-based channels. Further, we examine how the effect of a recommendation system across channels influences sales diversity, an important outcome in the retail industry. We conduct our analysis using a randomized field experiment, conducted in partnership with an online retailing firm in South Korea, where the experimental treatment is access to a recommendation system. Our results show that the use of recommendation systems enhances customer-level outcomes, such as views and sales of recommended products, clickthrough rate, and conversion. More importantly, the marginal impacts of the recommendation system are significantly higher for mobile users, indicating that the higher search costs imposed through mobile devices are more effectively reduced through recommendation systems. With respect to sales diversity, we observe that although the mobile channel leads to more diverse sales, we see no interaction effects of the recommendation system and mobile use on sales diversity. These results provide boundary conditions for the efficacy of recommendation systems in retail contexts where online sales occur across both PC-based and mobile channels. We discuss the managerial implications of these results for online retailers and conclude with opportunities for further research.