Disconfirmation Effect on Online Rating Behavior: A Dynamic Analysis

This research studies the effect of “disconfirmation”—the discrepancy between the expected and experienced quality of the same product—on the behavior of consumers leaving product reviews in an online setting. Specifically, we posit that an individual’s pre-purchase expectation is driven by the product ratings she observes and the credibility of the review system she perceives at the time of purchases. Upon product consumption, the individual obtains post-purchase evaluation and uses this information to confirm/disconfirm with her expectation. We extend the Bayesian learning framework to model the individual perception of the system credibility as how closely the ex-ante expectation matches the ex-post evaluation. Incorporating such dynamic perception in raters’ decisions allows us to better understand the chance of consumer engagement in sharing product experience over time. We estimate the proposed model using a rich data set comprising compete purchasing and rating activities made by 361 individuals on an e-commerce website. Our results show that consumer participation in leaving product reviews is subject to the disconfirmation effect. The further an individual’s post-purchase evaluation deviates from her pre-purchase expectation, the more likely she is to express her own opinions about the product. Our results also show that online consumers tend to be less active with respect to posting reviews as they perceive the system to be more precise over time. From simulations, we highlight 1) the dissimilar effects of disconfirmation and experienced quality on consumer rating behavior; 2) various evolutionary patterns of product ratings over time; and 3) the effect of fake, inflated ratings on subsequent rating entries.