Applying Quantitative Marketing Techniques to the Internet

Quantitative models have proved valuable in predicting consumer behavior in the offline world. These same techniques can be adapted to predict online actions. The use of diffusion models provides a firm foundation to implement and forecast viral marketing strategies. Choice models can predict purchases at online stores and shopbots. Hierarchical Bayesian models provide a framework to implement versioning and price segmentation strategies. Bayesian updating is a natural tool for profiling users with clickstream data. I illustrate these four modeling techniques and discuss their potential for solving Internet marketing problems.