Description
App Start measurements don't align with anecdotal/real-world performance and average app start time seems to be lower than we expect it to be. Our app uses a separate activity as a loading screen before our main Activity is shown, and we suspect that automated Activity tracking for the first Activity is prematurely ending the App Start span, resulting in an inaccurate measurement.
Describe the solution you'd like
- I'd like to have a way to specify which Activity should be used when determining the end of the App Start span.
- I'd like to continue to use automatic Activity tracking.
- I'd like to continue to see View Load spans for all Activities.
Describe alternatives you've considered
Refactoring our app loading sequence to remove the loading screen Activity and move to a truly single Activity architecture.
Additional context
I'm making assumptions around the behavior of the SDK and the results we're seeing here.
If there's a way for me to confirm which Activity instance is being used to determine my App Start "end" time, I'd be happy to help triage this (as there's a chance that this might be working correctly from the start)