Subscribe to DSC Newsletter

Techniques for solving targeting problem with very low response rate

Hi
I have a population of customer in which 180 out of 5,000,000 is buying a particular product X. Now business wants a list similar to 180 who have the propensity to buy X from the sample so that they can target in future campaigns.
Now as the response rate is low so I am skeptic of using traditional look alike model in this scenario.
Can anyone suggest any technique which can be modeling or data analysis in order to solve this problem?
Thanks for the help in advance!!

Regards
Biswajit

Views: 217

Comment

You need to be a member of AnalyticBridge to add comments!

Join AnalyticBridge

Comment by Idielle Walters on May 10, 2011 at 7:39am
You can use oversampling, but with 5,000,000 subscribers it might be best to first build a segmentation model for the subscriber base.  You will gain a lot of insight into your subscriber base by identifying different segments.  You can then start by looking at the segments in which the 180 subscribers are located.  Sometimes the behaviour/demographics of a spesific segment will indicate whether the subscribers in that segment will have a preference for the kind of product you try to sell.

On Data Science Central

© 2019   AnalyticBridge.com is a subsidiary and dedicated channel of Data Science Central LLC   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service