Article ID: | iaor20021812 |
Country: | United States |
Volume: | 31 |
Issue: | 2 |
Start Page Number: | 90 |
End Page Number: | 108 |
Publication Date: | Mar 2001 |
Journal: | Interfaces |
Authors: | Montgomery Alan L. |
Keywords: | marketing, internet, e-commerce |
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 for implementing versioning and price-segmentation strategies. Bayesian updating is a natural tool for profiling users with clickstream data. A key challenge for practitioners of Internet marketing is to extract value from the huge volume of data that can be collected. These techniques illustrate how this information can be leveraged to create better decisions.