Do online reviews still matter post-purchase?

PurposeThe influence of electronic word-of-mouth (eWOM) information, such as online reviews, on consumers’ decision making is well documented, but it is unclear if online reviews still matter in post-purchase evaluation and behaviours. The purpose of this paper is to examine the extent to which online reviews (aggregate rating (AR) and individual reviews (IR)) influence consumers’ evaluation and post-purchase behaviour by considering the valence congruence of online reviews and consumption experience (CE).Design/methodology/approachFollowing social comparison theory and relevant literature, the authors conduct an online experiment (pre-test: n=180; main study: n=347). The authors rely on a 2 (CE valence) ×2 (AR valence) ×2 (IR valence) between-subjects design.FindingsCongruence/incongruence between the valences of CE, AR and IR affects consumers’ post-purchase evaluation at the emotional, brand and media levels and review-writing behaviour. In comparison to aggregated rating, IR are more important in the post-purchase stage. Similarly, consumers have a higher eWOM-writing intention when there is congruence between the valences of CE, AR and IR.Practical implicationsThe authors demonstrate the importance of service providers continually monitoring their business profiles on review sites to ensure consistency of review information, as these influence consumers’ post-purchase evaluation and behaviours. For this reason, the authors illustrate the utility of why media owners of review sites should support the monitoring process to facilitate the engagement of both businesses and customers.Originality/valueThe authors break new ground by empirically testing the impact of online review information post-purchase seen through the theoretical lens of social comparison. The approach is novel in breaking down and testing the dimensions of post-purchase evaluation and behavioural intentions in understanding the social comparison elicited by online reviews in the post-purchase phase