I was curious if anybody runs their own RFM analysis internally.  If you do, would you mind sharing your formula?  I would like to compare it to the one we are going to use to see how it compares.  Thanks!