This should be a very interesting session. I had a chance to hear from @Jean-Yves a few weeks ago about Uplift Modeling. Uplift Modeling seems like an important approach for attributing the specific impact of your marketing activities on customer retention. (As long as you are willing to randomly sample across you customer population)
In the work that I do with membership and subscriber programs the idea of churn is problematic. Uplift modeling and even survival analysis seem to be interesting alternatives to a simpler more traditional churn approach. What do you think?
--Tom
