I have 2 groups (test and control).

I want to measure the impact of my intervention in the test group.

Notes:

• This is a time-series question

• Test and control group are NOT equal in number. The test group is smaller.

• Each intervention (on the test) have happened at different time.

• The goal is to measure incrementality and not just the difference

Work done already:

• We could potentially not use the control but in our case, we want to utilize the control sample and prove incrementality
• Causal impact analysis: We can use the Causal impact analysis framework by google but our sample size has +300 samples.

I have tried Causal impact analysis and results make sense but now I want to scale the method and include a bigger sample size (+500) so i can’t do all of them manually.