Return on web site visit duration: Applying web analytics data

Much research on internet marketing and e-commerce shares a great interest in a key performance metric for e-tailer success online: the shopper's time spent on the web site, commonly known as 'web site visit duration'. Visit duration has been used as a useful proxy to measure web'stickiness' because it enhances visitor-to-buyer conversion rate and often leads to online loyalty. The motivation of this study is the recognition of web site visit duration as an important metric to e-commerce success, and also the relative paucity of research on theoretical frameworks that explain visit duration. This study draws on the two-stage choice model literature and posits that shopper visitation to an e-commerce web site is a function of the choice decision processes. We construct a model to decompose online shoppers' decision for visit duration into two choice stages. In Stage 1, shoppers are explorative with less effortful visitation behaviour (eg using search engine to reach a site). In Stage 2, shoppers engage in more effortful and depth visitation activities (eg viewing more Web pages). We empirically test our model using observed web analytics data from 94 UK e-stores to unveil the decision process mechanism for Stage 1 and Stage 2, respectively, and also the transition between the two stages (eg bounce from the web site), which is lacking in extant literature because of its unobservable nature.