0
$\begingroup$

0

I have a theoretical / stats related question. I've run the package in R and it's easy. Question: How can I know the incremental impact of Campaign #1 during a period when Campaign #2 was also active during the same time period and in the same markets as Campaign #1?

r statistics

$\endgroup$
2
  • $\begingroup$ Do you have any data for a period with only Campaign #1 or only Campaign #2, or do the two campaigns overlap entirely? $\endgroup$ Mar 17, 2022 at 22:34
  • $\begingroup$ @BryanKrause completely overlap... $\endgroup$
    – AdilK
    Mar 18, 2022 at 13:28

1 Answer 1

0
$\begingroup$

You cannot do this with any statistical model or software. You do not have two interventions, you have one intervention "Campaign 1+2". They are perfectly correlated and not possible to separate.

Your only hope would be if you could apply some prior domain-specific information to the problem. For example, if Campaign 1 said "buy oranges" and Campaign 2 said "buy apples", you could perhaps separate them by looking separately at apple and orange sales, rather than a composite "fruit sales", and using an assumption that a specific campaign for apples doesn't impact other fruits.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.