I am trying to figure out how to track engagement using a Firebase A/B test and I am a bit lost. Here's the scenario
I am building a chat app. An user becomes activated once they send the first message (logged as first_message). From here, I want to test a new feature that would increase the engagement in the app. I currently think of engagement in terms of messages sent (logged as sent_message)
Here's how I would see this:
Create a new A/B test where the activation event is first_message and the goal metric is sent_message. I expect the variant with the best engagement to produce the largest amount of sent_message events
But that doesnt seem to be what happens. It seems that Firebase A/B tests work in terms of conversions? They look at whether a user first hit first_message, and then if they hit any number of sent_message. If the user sent any number of messages, Firebase considers it a conversion, and counts it as 1 conversion
Surely, I must be missing something here. A/B tests built around the number of fired events seems like a fundamental feature, there is no way this was not considered while building the platform. Am I going about it wrong?
Thanks!