Enterprise Level Systems

The likelihood to respond / relative future value scoring method described in the Drilling Down book is based on one of the most fundamental and predictive customer behavioral scoring algorithms in database marketing.  In fact, most other predictive modeling techniques will find the core RFM variables as highly predictive of future customer value.  For example, both a top-down regression model and a bottoms-up data-mining training session will find Recency as very highly predictive of customer response to promotional efforts.

For this reason, more advanced users can take the Drilling Down methods and models explained in the book and use them to mimic the functionality of enterprise class systems that analyze customer behavior and customize the website experience based on this behavior.  All you have to do is create the loyalty scores in your customer database and use a simple Perl script to modify the website presentation based on loyalty score.  For example, you could create an anti-defection campaign where best customers with falling loyalty scores could be automatically presented with a discount or gift.

If you are unfamiliar with these systems and how they are used, you might visit the vendor sites for more detailed information.  Below is a list of resources, not in any particular order, where you can learn more about the use of predictive customer modeling in enterprise class systems.

"How To" Article:
Measuring the ROI of CRM Implementations  

Overview of Issues and Software

Interactive Week 10/2000     

PC Magazine  01/2001

Information Week  10/01

Companies

Quaero

Unica

Quadstone     

SAS e-Intelligence     

Responsys     

Broadbase (now a part of Kana)

Xchange     

Tealeaf Technology     

E.piphany     

CoreMetrics     

Onyx Software

Accrue Software     

NetGenesis (now a part of SPSS)

Data-Mining Tools