New goal metric: event conversion
Allow to use events as goal metric to analyse impact of experiment on an event's conversion (e.g. likelihood for a player performing a certain action like completing a level, clicking on a popup etc)
Experiment is not acquiring users
Hello, I am running the experiment for several days and still 0 users aquired. What could be wrong? "Warning. The experiment is running, but not acquiring enough users. This experiment has been running for 4 days and collected data for 0 users."
Stop Acquiring Users
Allow an a/b test to continue running but no longer acquire new players
Provide users with the ability to add test device_ids / user_ids, at Organization level. A user can then assign test ids to specific variants in a/b tests. Without the id needing to be a new user/install. Thus allowing a user who wants to make sure everything is setup correctly, to add the test user_id and know that it will be added to a specific variant.
Remote Config Change Log for A/B Tests
Information on which config keys have been changed during the course of an a/b test. Global configs should not be changed while an a/b test is running as this can void the results.
Apply Winning Variant (Target Criteria)
The ability to the the winning variant config details and apply it to all players who meet the target criteria of the a/b test.