KAIST College of Business Selected Publications > Faculty Research > Faculty & Research >KAIST COLLEGE OF BUSINESS
본문 바로가기 사이트 메뉴 바로가기 주메뉴 바로가기

When Seeing Helps Believing: The Interactive Effect s of Previews and Reviews on E-Book Purchases

INFORMATION SYSTEMS RESEARCH2019-12

Choi, Angela Aerry | Cho, Daegon | Yim, Dobin | Moon, Jae Yun | Oh, Wonseok

Online reviews offer consumers the indirect experience of products through others' consumption evaluations, whereas previews afford them direct experience through product trials. Although conceptual and empirical studies on the business ramifications of online reviews abound, little is known about how online previews moderate the effects of online reviews on sales. To cast light on this issue, the current research investigated the interactive effects of exposure to online previews and reviews on individual purchase decisions. We analyzed a unique two-month panel data set on 270,260 sessions that comprise clickstream data on consumers' exposure to previews and reviews and data on their subsequent purchase behaviors. On the basis of analyses underlain by a two-stage hierarchical Bayesian framework, we found that online previews positively influence individual purchase decisions. More importantly, significant interactions exist between previews and reviews, as evidenced by the decreasing positive effect of previews with increasing review volume and average review rating. In addition, previews can complement reviews when a high variance in the latter renders purchase decisions difficult. We further examined the sequence effect of exposure to previews and reviews and discovered that exposure to previews following the experience of reviews may exert a considerable positive influence on individual purchase decisions. The results from an additional field experiment and a text-based sentiment analysis reinforced the validity of our main findings by mitigating concerns regarding the endogeneity and the accuracy of the review quality, respectively. The findings provide practical implications with regard to the design of optimal strategies for releasing preview information to digital platforms.

Publisher
INFORMS
Issue Date
2019-12
Article Type
Article
Citation
INFORMATION SYSTEMS RESEARCH, Vol.30, No.4, pp.1164 - 1183
ISSN
1047-7047
DOI
10.1287/isre.2019.0857
Contact : Joo, Sunhee ( shjoo2006@kaist.ac.kr )

Faculty & Research

KCB ISSUE