This project was a crucial lesson in the hierarchy of user feedback. Relying on the opinions of our existing, successful users would have led us down a path of optimizing features that our churning new users never even reached. The real problem wasn't in the core product, but in the first 60 seconds of the user journey. By analyzing thousands of session recordings of users who left, we bypassed subjective feedback and observed objective behavior. This data-driven approach was the only way to uncover the crippling effect of decision paralysis and pinpoint a critical, system-level bug that was invisible to both our support team and our established user base. It cemented the principle that to solve for early churn, you must observe what users do, not just listen to what the survivors say.