Meg's hypothesis was that in the past, she may have been overstating the online contribution but completely missing email marketing's contribution to retail. It helps that REI is very much a direct marketing and catalog company, and so traditional RFM models make sense.
The results are impressive. Using this new metric, she doubled the program performance of email program (email at REI is now 2x the size it was when they started testing). Another benefit of this new measurement methodology is that campaigns that may not perform well based solely on online sales actually may have performed significantly better when measured in incremental lift compared to retail or catalog. So individual campaigns look very different when measured by incremental lift vs. cookies sales.
Brava to Meg! What a great example of how worthwhile it is to make sure that email is measured in proper context with business rules - Meg will hopefully now be able to get the attention and resources that her email program deserves!