Ok, this one came up in a recent project.
Assume the customer has Recency = 15 Months .. .the customer hasn't bought anything in a long time. Assume historical spend is between $175 and $225 ... equalizing historical spend.
Now we segment the customer based on very simple segmentation variable.
- Yes = Customer clicked through at least one email campaign in the past month.
- No = Customer did not click through one email campaign in the past month.
Simple enough, right? You've got this data in your omnichannel customer data warehouse, right?
Now measure rebuy rates by channel in the next month, as well as overall. Here's an example from a recent analysis, where we look at overall rebuy rates in the next month.
- 15 Months Recency, $175 - $225, Click = Yes: Next Month Rebuy = 3.52%.
- 15 Months Recency, $175 - $225, Click = No: Next Month Rebuy = 1.20%.
Oh ... look at that!
All things being equal, the customer who clicked through an email campaign last week was almost 3x as responsive as the customer who didn't, after controlling for recency & historical spend.
Seems like you'd want to know that.
Seems like you'd act upon that if you knew that, right?
Seems like if you were a catalog brand you'd kick out a hotline catalog if the catalog was not previously mailed, right?
You can perform a simple overlay to get this information. You can use Hillstrom's QuickScores (click here) to get this information. You can use your own in-house models to obtain this information. What is most important, of course, is that you do something with the information.