10-06-2014 12:42 PM
Has anyone had experience building successful incremental lift model using test and control groups? Could you share your experience on why it worked or not worked?
I tried using the incremental response node in SAS EM but when I had to validate the model using off-time campaign, the incremental response ranking totally messed up.
I guess it is because the model was quite unstable across different campaigns data but I don't know what to do next.