BizReport IT Research Library

Help | Advanced Search
What's New?
What's Popular?
Identifying Who Can Be Saved and Who Will Be Driven Away by Retention Activity: Using Uplift Modeling to Reduce Churn in Mobile Telephony
sponsored by Portrait Software
Posted:  08 Oct 2008
Published:  08 Oct 2008
Format:  PDF
Length:  10   Page(s)
Type:  White Paper
Language:  English


ABSTRACT:
As an increasing number of operators have discovered, retention activity can have negative as well as positive effects. It follows from this that it is dangerous to target retention activity primarily on the basis of estimated churn risk, whether or not this is weighted by customer value. This paper shows examples of two retention campaigns, one highly effective and profitable, and the other counterproductive and severely loss making, and demonstrated that both can be radically improved by using uplift modeling to predict savability. In both cases, this has a dramatic and positive impact on the profitability of the campaign, and also on the overall level of retention achieved by the operators.

The adoption of uplift modeling requires operators to embrace fully the use of systematically randomized control groups, and commits them to using sophisticated, modern uplift modeling methods. These are more complex than traditional methods, but believe that the demonstrable improvement in results that can be achieved more than justifies such a transition.


Authors

Nicholas J. Radcliffe
Founder, Director ,  Stochastic Solutions
Nicholas Radcliffe is founder and director of Stochastic Solutions Limited, a consultancy that focuses on customer targeting and optimization problems. He was previously a founder and Technical Director of Quadstone Limited, where he led the development of their approach to uplift modeling and numerous client engagements. Nicholas is also a Visiting Professor in the Department of Mathematics and Statistics at University of Edinburgh, where he works in the Operations Research group.

Rob Simpson
PortraitSoftware
Rob Simpson has worked in data mining for business for over 15 years following a PhD in pattern recognition, and has assisted over dozens of major corporations to generate RoI from data analysis. He has been Head of Customer Services for Portrait Software, formerly Quadstone, for the last eight years, prior to which he was working as an independent consultant. In recent years he has worked with many European telecommunications companies managing projects including churn, credit scoring and cross-sell.



BROWSE RELATED RESOURCES
Best Practices | Business Intelligence | Customer Satisfaction | ERP | Mobile Workers | Model Driven Architecture | Systems Management | Telephony Services | Telephony Systems

View All Resources sponsored by Portrait Software

Library Home | Advertise with Us | Product Library
A Service of Bitpipe