Been digging into our onboarding data over the past few months and honestly some patterns are throwing me off.
Seeing weird correlations between completion rates and user segments that don’t match what I expected. Anyone else notice their analytics telling a different story than assumptions?
Trust the numbers not your gut feelings
The unexpected patterns usually point to something worth investigating further rather than dismissing.
I always cross reference completion data with retention numbers from those same segments because sometimes lower completion rates actually bring in users who stick around longer.
Might be worth looking at what actions those surprising segments take after onboarding finishes.
Break down the segments by acquisition channel first. Users from different sources behave differently during onboarding. Then look at device type and OS version. I’ve seen completion rates vary significantly between iOS and Android for the same flow. Time of signup matters too. Weekend signups often have different intent than weekday ones. Once you isolate these variables, the weird correlations usually make sense.
Data gets messy when you have multiple flows running at the same time. Check if any experiments overlapped.