Hi,
We are trying to test a classical scenario where we would like test different types of controls for a mobile racing game and see how they effect new users metrics and apply the winning one..
In our experiment scorecard, we find acquisition and retention parts are important for our tests as we are trying to find out how much users are engaged with different controls.
My questions is as we understand from our daily player count, even though we would like to improve our new users' stats, we see retention values for existing users also, does it mean that Day 1 retention for existing users are the retentions after the day experiment started? (That maybe very different compared to new users' retention).
If my above assumption is right, what is the proper way for making tests that effects only newly created users to improve retention and engagement without intervention of already engaged users?
Best Regards,
Daghan