Abstract/Summary

The Internet has brought disruptive changes to people’s life. Such drastic and profound changes demand much attention of research in understanding user’s behaviour, especially the interactions between consumers and businesses or products. Nowadays consumers are enabled to search relevant information of products from social media, such as product review, which is normally referred as the electronic Word-of-Mouth (eWOM). Such eWOM conveying product descriptions and individuals’ experience has shown significant influence on consumers’ purchase behaviour. Focusing on the theory and practice of realising the substantial value of the digital resources, this thesis explores the impact of eWOM on consumer behaviour and help to understand the online consumer reviews on social networking sites. Most e-commerce websites provide social networking services that allow users to add friends or follow trustworthy reviewers. However, in relevant studies, the source of eWOM, i.e. from friends or crowds, has not been fully addressed. It is thus an open question whether or not friend’s reviews differently impact consumer’s behaviour in comparison to the general crowd reviews. In this thesis, we develop a Probit model to study the impact of friend reviews and crowd reviews on the possibility of subsequent consumers’ posting behaviour. Despite the common perception that the volume, valance and variance of reviews significantly impact the likelihood of following posting behaviour, we find that such influence comes from friend’s reviews. Furthermore, a Monte Carlo simulation experiment is carried out to examine to what extent the consumers’ decision is affected by the product popularity among friends and crowds respectively. The simulation confirms the econometric analysis, and it is shown that about 75% of the posting behaviour are affected by product popularity amongst one’s friends. Such results imply the importance of trust in the process of eWOM diffusion. Despite the huge value of online reviews, the overwhelming amount of them makes vi consumers impossible to access all of them before they find the most proper product. Thus, to uncover the most helpful reviews becomes a task of both practical and theoretical significance. This thesis applies a two-wave based dataset consisting of online reviews and their helpfulness information collected at two different time points and studies the increment of helpfulness between such two time points. Though previous studies have confirmed that the helpfulness of reviews can be largely explained by the features of reviewers and reviews, we show that these features have much less explanatory power for the increment of helpfulness in the future. Furthermore, the reviewer activeness and review disclosure information are shown to be more predictive. Product recommendation network is another way to help consumers find interesting products quickly. The demand and sale of products have been shown to be highly correlated to that of their neighbours. In this thesis, we employ an empirical book recommendation network of Amazon to investigate the effect of product distance on their eWOM in terms of online review volume and rating. The analysis indicates the connectivity between books has significant influence on the WOM. To summarise, the thesis explores the interplay between consumers, products and online reviews in the context of social network and product recommendation network, and examines the factors that are contributing to the helpfulness of reviews. In theory, this thesis highlights the network effect on the formation and diffusion of eWOM. The influence of reviews on users can be enlarged through social network, and the product network also largely reshapes the eWOM of products. The developed econometric as well as computational methods also contribute to the knowledge providing new ways of studying consumer behaviour. The results reported in this thesis can inform practitioners on the design of online reviewing system, so to better aid consumers accessing information, and the possible online marketing strategies.