Archive for March, 2010

Why One Model May Not Suffice

Friday, March 5th, 2010

In today’s tough economic times, marketers are continually searching for ways to gain efficiencies and improve return on investments.  Dramatic improvements in predictive models are possible through identifying variables that may not only work differently for different segments, but might only be relevant to certain segments, and not to others.

A while back, we wrote an article that dealt with the possibility, actually the probability, that different segments of your customer file may behave differently with regard to their relationships with well known key predictor variables such as Recency, Frequency, Monetary Value and various measures of Product Purchase data.
(more…)


Rss Feed Tweeter button Facebook button Technorati button Reddit button Myspace button Linkedin button Webonews button Delicious button Digg button Flickr button Stumbleupon button Newsvine button Youtube button