A Study of Consumer Switching Behavior

This paper describes an empirical study of the dynamics of consumer switching behavior across major Internet portal Websites. The study methodology is based on a multinomial Logit-Markov model framework that characterizes consumer switching-behavior. We propose a tractable Maximum Likelihood procedure to estimate the switching model’s parameters on the basis of online panel data. This empirical study highlights the potential managerial implications of the model in the context of E-commerce. In particular, the results of the study are shown to provide potential managerial insights regarding the strength of a portal Website, relative to its competitors, in terms of its ability to attract and retain visitors, as well as the relative vulnerabilities of competing portals from which visitors are drawn. In addition, the study examines the effect and impact of causal factors (e.g., visitor gender, past Internet usage, log on time, and time spent on previous sites), on portal Website switching behavior and loyalty. [