Opinion convergence versus polarization: examining opinion distributions in online word-of-mouth

Abstract

We examine how opinion distributions (i.e., opinion polarization and convergence over time) differ across product salient platforms (product platforms) versus product non-salient platforms (non-product platforms). Drawing on the theory of impression management, we hypothesize and explain when and why consumers choose to post their comments on different platforms, and how their behavior will be affected when they choose to post on online platforms. To test the hypotheses, we collected and text-mined online posts from product platforms such as review aggregator sites, discussion forums, and consumer rating websites, and non-product platforms such as microblogs. The results showed that product platforms have more polarized opinions, and exhibit more convergence in opinion across time, compared with non-product platforms. Our findings advise researchers and practitioners to pay attention to the characteristics of online platforms, and how users' perceptions of the purpose of the online platform may affect their online posting behavior.