A Model of Web Site Browsing Behavior Estimated on Clickstream Data

Abstract

Using the clickstream data recorded in Web server log files, the authors develop and estimate a model of the browsing behavior of visitors to a Web site. Two basic aspects of browsing behavior are examined: (1) the visitor's decisions to continue browsing (by submitting an additional page request) or to exit the site and (2) the length of time spent viewing each page. The authors propose a type II tobit model that captures both aspects of browsing behavior and handles the limitations of server log-file data. The authors fit the model to the individual-level browsing decisions of a random sample of 5000 visitors to the Web site of an Internet automotive company. Empirical results show that visitors' propensity to continue browsing changes dynamically as a function of the depth of a given site visit and the number of repeat visits to the site. The dynamics are consistent both with "within-site lock-in" or site "stickiness" and with learning that carries over repeat visits. In particular, repeat visits lead to reduced page-view propensities but not to reduced page-view durations. The results also reveal browsing patterns that may reflect visitors' timesaving strategies. Finally, the authors report that simple site metrics computed at the aggregate level diverge substantially from individual-level modeling results, which indicates the need for Web site analyses to control for cross-sectional heterogeneity.Since the commercial inception of the Internet, the ability of Web sites to track the behavior of their visitors has been considered one of the most promising facets of the new medium. The detailed records of Web usage behavior (clickstream data) provide researchers and practitioners with the opportunity to study how users browse or navigate Web sites and to assess site performance in various ways. Therefore, the use of clickstream data to model visitors' usage of a specific site and how that usage may change with experience could produce important dividends for researchers and practitioners interested in Web site design, Web site customization, and ongoing monitoring of a site's performance.Despite the importance of analyzing how users browse a given site, to the best of our knowledge, no disaggregate clickstream model of detailed within-site browsing behavior and its potential dynamics has yet appeared in the marketing literature. In this article, we develop a modeling approach for understanding some of the basic aspects of within-site browsing behavior at the individual level. We conceptualize user navigation through an Internet Web site as a series of the following decisions: (1) whether to request an additional page (thereby remaining on the site) or to exit the site and (2) how long to view a given page on the site. We estimate a type II tobit model on the clickstream data collected by the operator of a major commercial Web site in the automotive industry. Our modeling results provide several findings about browsing behavior and a series of implications for Web managers.The existing literature on Web site mode...

Stay or leave: Applying Approach–Avoidance Theory to virtual environments

“…Web site designers should stay within the norms of other web site designs to ensure that these sites meet users ' expectations. 2 Research has found that ease of usability becomes strained when users are unfamiliar with the design or structure of a web site. 2,22 This is similar to Lutz and Kakkar ' s 25 fi nding that brick-and-mortar consumers who are overloaded with information and choices in the retail marketplace feel intimidated and often exit the environment.…”

“…9,19 To improve web site ease of use, designers should strive to maintain lower numbers of pages needed to complete a task. 2 Additionally, web sites must have clearly distinguished links that inform the user of the destination. 20 Furthermore, Huber 19 calls for the importance of web site fl ow, and suggests effective use of headlines, photos and captions, and use of text on buttons to call for specifi c actions.…”

“…Research has shown that users enjoy web sites that place more information displayed in an organized manner on one page versus information spread across multiple pages. 2,5 Additionally, redundant or non-essential information is detrimental to user satisfaction. Numerous studies have concluded that information leads to satisfaction only when it is informative, relevant, concise and meets the absorption rate for that users ' web site experience level.…”

“…7,9,13 Research shows that people prefer to browse web sites with short paths and less complex stimuli. 2,31 Conversely, affi liation decreases when people feel dominant over a web site. 1,2,30 For example, users spend less time per session as they gain control and experience within a web site 2 because they no longer feel curious to explore 31 and often become bored.…”

Using online search data to forecast new product sales

“…However, they did find that more active shoppers tend to visit more sites. Bucklin and Sismeiro (2003) model online browsing behavior and find that with repeat visits to an online site, browsers view fewer pages within the site, but the time they spend viewing each page is unchanged. Moe (2003) uses clickstream data to identify online shoppers as buyers, searchers, browsers, or knowledge-builders based on characteristics of their site visit sessions, while Moe (2006) uses clickstream data to develop a two-stage model of the consumer decision process.…”