| Summary: | 博士 === 國立臺灣大學 === 商學研究所 === 99 === As the emergence of Internet and information technology, it is much easier for consumers to search and share their product evaluations online. In the online review literature, consumer reviews have significant impact on consumer purchase decisions and product sales. Chevalier and Mayzlin (2006) suggest that, the sales of the online sellers who invite users to post personal product evaluations on their websites may higher than them would be without the provision of customer reviews. The objective of this research is to describe the dynamics of number of reviewers and their reviews on the website, and identify the factors which could explain the differences in frequency and continuity of their online review behaviors.
We attempt to examine the topic from an online seller’s perspective by utilizing Pareto/NBD model, which has been successfully applied in customer base analysis over past two decades. Furthermore, the current study uses the framework — incorporating time-invariant covariates into the Pareto/NBD model (Fader and Hardie 2007) — to capture the factors that impact on online review behaviors.
We collected our data using the online reviewers and their reviews available through Amazon.com. Summarizing results from an empirical study on 4,240 reviewers, we conclude our main findings as follows: the average rating of reviews and the feedback from other consumers have positively effect on both reviewing frequency and continuity. Moreover, the average length of reviews positively affects reviewing frequency but negatively affects continuity. Finally, the results suggest that, reviewers who posting reviews with real name have higher frequency but lower continuity than anonymous reviewers.
|
|---|
