A Dual Process Model of the Influence of Recommender Systems on Purchase Intentions in Online Shopping Environments

• Published in: Journal of Internet commerce Vol. 22; no. 3; pp. 432 - 453
• Main Authors: Xu, Lina, Roy, Abhijit, Niculescu, Mihai
• Format: Journal Article
• Language: English
• Published: Binghamton Routledge 03.07.2023; Taylor & Francis Ltd
• Subjects: Avoidance; Consumers; Decision making; Electronic commerce; Labels; Low and high involvement; online purchase intention; recommendation label; Recommender systems; risk avoidance; social proof
• ISSN: 1533-2861; 1533-287X
• DOI: 10.1080/15332861.2022.2049113

Whereas much research has looked at how recommendation systems influence online purchase intentions, this article illustrates the dual process model by which they occur. Using two studies, we fill the research void in interactive marketing by demonstrating how the dual processes of social proof and risk avoidance mediate the impact of recommendation labels on consumer decision-making contingent upon their level of involvement. Study 1 (n = 73), used a mixed-subjects design with a college student sample to demonstrate that both types of recommendation labels will lead to higher purchase intentions in an online setting. Most importantly, it provides evidence for the main effect of our theoretical model across different product categories. Study 2 (n = 160) provides support for our remaining four hypotheses by demonstrating the underlying process through which recommendation labels have a two-fold effect on purchase intentions. Specifically, the recommendation label increased the risk avoidance effect for high-involvement consumers and enhanced the social proof effect for low-involvement consumers. In both cases, the recommendation labels increased purchase intentions. Implications of our findings for theoretical and practical contributions and future directions are also explored.