JoinLU is an expert-intelligence B2B platform connecting companies with vetted specialists.
Role
UX Designer
Product
Website
Areas
Design, Strategy
Duration
2023 - 2024
PROBLEM STATEMENT
If JoinLU’s value is clear
but the process isn’t,
what’s missing?
Defining the direction together.
We didn’t jump straight into design.
We took a step back to run a competitive analysis to get alignment on JoinLU's goals and how they want to show up in the market.
Making sense
of the landscape.
A side-by-side comparison made one thing clear: competitors are far more explicit about process, credibility, and next steps.
Those gaps became immediate opportunities for JoinLU to stand out.
Uncovering the why through user research.
Our competitive analysis made one thing obvious: clarity breaks down early.
To uncover why, we focused on what users initially think, how the site shapes those impressions, and where the experience disconnects most.
And as we listened to users,
a clear pattern started to emerge…
Finding structure
admist the chaos.
Before changing anything, we laid it all out.
We mapped the site’s structure and information architecture to get a clear baseline of what existed and how it was organized.
With the full flow visible, we layered in user feedback to see exactly where things started to break down.
Pinpointing and addressing the issues.
We focused first on the opening stages of the journey. When the full flow was visible, it was clear where the experience failed users.
We used those moments to restructure the journey into something clearer, calmer, and easier to move through.
The prototype, of course.
IMPACT & REFLECT
Looking back at it all…
MORE OF
I tested across multiple surfaces of the product, but in hindsight, I would have narrowed early testing to the highest-impact user flows (e.g., discovery → connection → follow-up). This would have allowed for deeper insights per flow rather than broader but shallower feedback.
CHANGES
I would have validated the information architecture earlier with users through tree testing or card sorting. Doing this sooner would have surfaced mental-model mismatches before we moved into higher-fidelity designs.