Overview
Across the homepage, lead-gen flow, and community pages, I ran a series of A/B tests to validate design assumptions with real user behavior rather than opinion.
Problem
Several design decisions — a binary budget selector, stacked contact methods, in-progress community pages — were based on assumption, not evidence.
What I did
- Tested a search bar with care-type options above the fold on the homepage, aiming to simplify community search and improve data collection — the new bar didn't attract more input, prompting a redesign of the whole above-fold section
- Tested consolidating a sticky chat bubble and a phone number that abruptly launched FaceTime into a single persistent contact module — blocked by technical constraints, which redirected the team toward a navigation redesign instead
- Replaced a binary “below/above $2,000” budget option in the lead-gen flow with a granular slider for better data collection — paused by development constraints
- Tested a “Request a brochure” CTA on community pages still under construction, to keep users engaged while those pages caught up
Outcome
A set of validated (and invalidated) hypotheses that redirected the roadmap — several tests didn't move the needle on their own, but directly informed the navigation redesign and lead-gen flow work that followed.