| When | September 20, 2001 6:30pm - 8:30pm |
| Where | The University of Texas at Austin McCombs School of Business (also called GSB) |
| Corner of Speedway and 21st Street (across from the main library and Gregory Gym) | |
| Room GSB 3.104 | |
| Feature | Leon S. Lasdon, PhD, Professor, University of Texas at Austin; Timothy L. Smith, MS, Research Scientist, Trajecta, Inc. |
| 6:30 - 7:00 | Refreshments and Networking |
| 7:00 - 8:30 | Predictive Modeling and Robust Optimization in CRM |
Customer acquisition,
customer retention, dynamic pricing and cross-selling are all customer
relationship management (CRM) problems that lend themselves
to solution via
optimization. What individual offers should a business give to each of
their prospects in order to build a portfolio of customers that is profitable
over the long-term and also satisfies near-term operating constraints?
How can you use optimization to find the right product and price for each
of your customers? That is the challenge of one-to-one marketing. In this
session, Lasdon and Smith will show how one vendor uses data mining and
predictive modeling to generate financial metrics that feed a robust optimization
model for CRM problems. These methods are appropriate whenever customer
data exists to support predictive modeling of key customer behaviors such
as purchases, payments, cancellations and charge-offs. The focus in this
talk will be on CRM projects that optimize credit card
portfolio management
for American banks. There will also be a demo of Trajecta's Virdix
optimization software.