A Data Science Central Community
I am looking for guidance and help and would appreciate any suggestions you members might have:
I need to estimate Donor (Non for Profit) life time expectancy. I was trying to do it with Churn formula, Retention and also read about Survival analysis. I do not have access to any tool but R and just started learning R.
Is there a way to predict Donor/Customer Life Time Expectancy using some kind of formula? For example for a Donor that has just joined the organization or has been with the organization for a short time (for example one year) how can we predict how long he/she will stay with the organization.
Very grateful for your help.
I would say survival analysis is the right direction to go. This technique is able to deal with censored data (i.e. donors that have joined the organization for some time and have not left yet) very well. It is about modeling the life time distribution (i.e. probability density function), and thereafter you can use it to estimate the mean (or median) or ranges of lifetime that a donor fall into. One model variant (hazards model) of the survival analysis family also allows you to study how different variables (e.g. demographics and lifestyle characteristics) impact the lifetime. However, if you have not read through all the details yet (or have no software to do it), you may also wish to use a non-parametric approach. Example: using a life table. A link for reference: http://www.tiem.utk.edu/~gross/bioed/bealsmodules/lifetables.html
Thank you very much Sunny Lam.