Analytics & Trade Conversion

You cannot improve what you cannot see. The investment app was tracking events, but the structure was messy enough that the data underneath it couldn't be trusted to make decisions on. Before I could move any numbers, I had to fix the numbers themselves.

Fixing the foundation

I owned the event tracking end to end and rebuilt it on a clean taxonomy: a scalable event structure that meant a funnel chart actually reflected what users did. Getting that right lifted tracking accuracy by 40% and made everything downstream trustworthy.

Then, the conversions

With the data honest, I ran funnel analysis to find where users dropped off on the way to a trade, then built targeted nudges at those exact moments. Not blanket prompts, specific interventions at specific leaks.

The order mattered. The conversion lift wasn't a clever growth trick; it was what became possible once the analytics could be trusted enough to point at the real problem.