Case study
Finding the leak in a real-estate funnel
Instrumented two critical funnels on a DIY real-estate platform, found a blocker hiding at the end of the flow, and redesigned the worst step.
- 2
- critical funnels instrumented
- 1
- blocking issue found and removed
The problem
Traffic was coming in, but listings and closed sales were not coming out. Everyone had a theory. Nobody had data.
What I did
- Instrumented the funnels. I added step-completion events to the two flows that mattered most: listing onboarding, and offers and counteroffers. That showed true drop-off per step instead of a single conversion number.
- Removed a blocker. The data surfaced a critical issue: a Terms of Service restriction that was only communicated at the very end of the flow, which invalidated the advisor-assisted way many users actually operated.
- Iterated on the evidence. We grouped questions into logical blocks, adjusted the UI on the steps with the highest drop-off, and redesigned the most conflictive step as a hybrid flow assisted by an internal platform advisor.
- Validated before shipping. A closed beta came before the production rollout.
Where it stands
Measuring the impact on step drop-off and new listings was still in progress when I moved on, so I’m not going to claim a number here. What this project shows is the method: measure the funnel step by step, find the leak, fix the largest one first.