About AMA

However, the key drivers of Web site visit duration are poorly understood because a myriad of person-specific, Web site–specific, and visit-situation factors interact in intricate ways to shape visit duration. In the absence of formal guidance, managers are tentatively experimenting with various approaches in their quest for longer visit durations. This is the first study to provide managers with some fundamental results regarding the relative importance of these factors and how their interaction impacts on visit behavior.

This article develops an individual-level random effects model that accounts not only for heterogeneity that arises from differences among people and Web sites but also for the effect of unobserved visit-situation heterogeneity across multiple visits by the same person to the same Web site. In a context in which customers have multiple experiences with a product, accounting for visit-situation heterogeneity is crucial because the situation of a visit by the same person to the same Web site changes from one visit occasion to another (e.g., the amount of browsing time available to a user on a given visit occasion). In such situations, the authors show how a “triple-heterogeneity” model can be developed to incorporate simultaneously not only person-specific and product-specific heterogeneity but also heterogeneity that is specific to the visit situation. The model is fit using user-centric panel data and encompasses the 50 most popular sites in a market. The authors find the following:

Some interactions between demographic and site characteristic variables reveal interesting insights. For example, sites with higher levels of advertising usually result in lower visit duration, but this is not the case for older people. The appeal of Web advertising to older people might come as a surprise to Web advertisers, who, in general, target younger males because of their higher levels of Internet use. In addition, older users spend less time than younger users on sites that have too many graphics and functionality features. Older Web users prefer a “keep-it-simple” Web site format and functionality, but the opposite is true of younger visitors.

Heterogeneity in repeat visits dominates over heterogeneity across people and Web sites. Most of the heterogeneity in duration times can be assigned to the “situation” of the visit occasion; the authors find that the situation variance explains approximately 80% of the assignable heterogeneity.

If managers have access only to demographics and Web site design variables, they should use the insights from the first set of findings to optimize their traditional segmentation strategies. Webmasters can use these results to tailor a site’s features and functionality to particular age groups to sustain longer visits, and advertisers can use them to target certain age groups. However, the second finding suggests that the greatest potential for increased effectiveness lies in the development of new segmentation and targeting strategies (as well as personalization strategies) that are dynamically based on the situation at the time of the visit, after a user enters a Web site. For example, the findings indicate that it is preferable to use path analysis methods to customize pages on the basis of the immediate history of the current visit rather than on a visitor’s personal characteristics or a Web site’s attributes.

Biography
Peter J. Danaher is a professor and is chair in the Marketing Department at the University of Auckland, New Zealand. He has also held visiting positions at the London Business School, the Wharton School, and Massachusetts Institute of Technology. He has a PhD in Statistics from Florida State University and an MS in Statistics from Purdue. His primary research interests are media exposure distributions, advertising effectiveness, television audience measurement and behavior, Internet usage behavior, customer satisfaction measurement, forecasting, and sample surveys. He has published on these subjects in journals such as Journal of Marketing Research, Marketing Science, Journal of Marketing, Journal of Advertising Research, Journal of the American Statistical Association, Journal of Retailing, Journal of Business and Economic Statistics, and American Statistician. He has consulted extensively with Telecom, Optus Communications, Unilever, ACNielsen, and other market research companies. He has also been the survey auditor for the television ratings services in New Zealand, Australia, and Ireland.

Guy W. Mullarkey works in the commercial division of Shell Australia Ltd. At the time of writing, he was a graduate student and tutor in the Department of Marketing at the University of Auckland, New Zealand. Guy’s areas of research interest include consumer choice behavior, the relationship between innovation and organizational structures and processes, and user loyalty and repeat site usage behavior on the World Wide Web. His work has appeared in Journal of Advertising Research, Journal of Marketing Theory and Practice, Marketing Intelligence and Planning, and The University of Auckland Business Review.

Skander Essegaier is Associate Professor of Marketing and is currently affiliated with the LAREQUAD research center at the University of Tunes, Tunisia. He was previously on the faculty at the Wharton School, University of Pennsylvania, and the Stern School of Business, New York University. He has also been a visiting faculty at INSEAD, France. His expertise lies in the areas of pricing, personalization, and competitive and strategic behavior. His work has been published in many journals, including Marketing Science, Management Science, Journal of Marketing Research, Journal of Applied Probability, and SIAM Journal on Control and Optimization. He also serves on the editorial board of Recherche et Applications en Marketing. Professor Essegaier has a PhD from Columbia University, an MSc from the London School of Economics, and an MSc and a BSc from the Ecole Nationale.

J Marketing Research, Volume 43, Number 2, May 2006
View Table of Contents.

Member Comments (0):

To rate or comment on articles, you must be a logged in AMA member. Click here to join