De-Risking a Website Redesign Before Launch

Usability research that gave a Fortune 100 insurer the evidence to ship — and a prioritized plan for what to fix first.

Hero image — moderated session / synthesis board

ROLE

Usability Research Lead

CLIENT

Fortune 100 insurer's institute

TIMELINE

7 weeks · 2 research rounds

TEAM

Me (research lead) + PM, designer, research partner

MY FOCUS

Study design · moderation · synthesis · stakeholder alignment

METHODS

27 sessions · SUS · severity-rated

THE CHALLENGE

A leading insurance institute had built a fast-growing content hub — webinars, thought leadership, and events spanning cybersecurity to economic resilience. But the website hadn't kept up. Users struggled to find relevant content, distinguish upcoming events from past replays, and navigate long, dense pages.

A full redesign was already underway: new IA, new visual design, a high-fidelity prototype. With launch approaching, the real question wasn't whether to ship — it was whether anyone could prove the redesign actually worked for the people who'd use it. Getting that answer wrong meant launching a Fortune 100 property on hope.

MY ROLE

I led the usability research end to end — designing the study, moderating all 27 sessions across three audience types, and owning synthesis and reporting. Beyond execution, I ran the parts of the project that required leadership:

Drove the research workstream within a four-person team, coordinating a product manager, designer, and research partner

Presented findings directly to the client and their internal stakeholders

Facilitated the prioritization sessions that decided what got fixed before launch versus what went on the roadmap — my findings artifact became the single reference point for design, content, and client teams

APPROACH

Two rounds, three audiences

Co-authored the discussion guide and ran 27 moderated sessions across two rounds, covering every core audience type — so no stakeholder could dismiss findings as “not our users.”

Quant + qual together

Paired moderated sessions with a SUS benchmark across every audience — a number stakeholders could anchor confidence to, alongside the behavioral evidence.

Owned the operation

Managed recruiting, screening, scheduling, and outreach for all participants; built the screener and the Mural board mapping exactly what was tested.

Severity-rated everything

Translated raw sessions into prioritized themes and page-level recommendations, rated by severity — so the team could triage instead of debate.

SOLUTION

The research gave the team a clear, evidence-backed picture of the redesign. The 86/100 SUS score — “Excellent” — said the foundation was solid. The sessions said exactly where it broke.

What worked: Registration was intuitive across every path. Clean visual hierarchy built trust (4.7/5). Foreshadowing key content set clear expectations.

What broke: Featured replays appeared where users expected upcoming events — disorienting them on nearly every page. Long pages caused scroll fatigue. Core publications were buried below the fold.

What no one expected: Employees engaged in three modes — personal growth, client-facing alignment, and forwarding content to upskill their teams — reframing the site from a publishing platform into an organizational enablement tool. I then ran prioritization with the client and design team, splitting findings into pre-launch fixes and roadmap items.

Synthesis / severity-rated findings board

IMPACT

SUS 86/100, benchmarked across all audiences. Dozens of pre-launch fixes shipped to navigation, hierarchy, registration, and content surfacing. A severity-rated findings artifact that became the single reference point across design, content, and client stakeholders — and a strategic reframe of what the site was for.

WHY IT MATTERS

This is what research leadership looks like in practice: designing a study that ends arguments, coordinating collaborators, and facilitating the decisions that determine what ships. The client didn't get a report — they got launch confidence and a prioritized plan.

It's also a reminder of why moderated research earns its cost. The highest-value finding — the three engagement modes — wasn't on the test plan. We went in measuring whether people could navigate and register; we came out having redefined what the site was for.

NEXT PROJECT

UPS →