Results driver - change in offer or the differences between test & control files

GMcB

New Member
#1
Im new here and have a basic query that I would appreciate comment on please.

I’m splitting 100% of a purchased acquisition targets file into test and control, testing a new offer to half the targets with the other half receiving the established offer as the control.

How can I ensure that it’s different offers driving any differences in response rate and not the differences in the make up of the two files?

Do take a Random sample & t-test key variable/s to make sure they are not significantly different between files? (Generally 4 variables contribute equally to an overall targeting score)

Or

Can I sort the file by this targeting score and take the test population from every second record and the control as the remaining?

Although the control will have a higher score both files should not significantly differ on this key variable. T-test just to make sure.

Thoughts, Suggestions, questions?
 

Karabiner

TS Contributor
#2
My suggestion would be that you post this to a more appropriate sub-forum, like "Statistics".

With kind regards

Karabiner