Abstract The authors develop and estimate a model,of online buying using clickstream data from a Web site that sells cars. The model,predicts online buying,by linking the purchase,decision to what,visitors do and to what,they are exposed,while at the site. Predicting Internet buying poses several modeling,challenges. These include: (1) online buying,probabilities are usually low which can lead to a lack of predictive and explanatory power from models, (2) it is difficult to effectively account for what Web users do, and to what they are exposed, while browsing a site, and (3) because online stores reach a diverse user population across many competitive environments, models of online buying must account for the corresponding user heterogeneity. To overcome these hurdles, the authors decompose the user’s purchase process into the completion,of sequential,Nominal,User Tasks (nuts) and account,for heterogeneity,across visitors at the county,level. Three major,user tasks required to complete,a purchase,are: (1) completion of product configuration, (2) input of complete personal information, and (3) order confirmation,with provision of credit card data. These tasks are managerially,meaningful,and correspond,to considerable,visitor loss. Using a sequence,of binary probits estimated,through Bayesian methods, the authors model the visitor’s decision of whether or not to complete each task for the first time, given that the visitor has completed the previous tasks at least once. The propensity for completing,one task is also allowed to affect the propensity for completing subsequent,tasks. Results indicate that visitors’ browsing,experiences,and navigational,behavior,are pre- dictive of task completion for all decision levels. The use of interactive decision aids, the