The Self-Service Portal for Combined Insurance (A Chubb Company) is the primary destination for policyholders to manage their coverage — paying bills, reviewing benefits, filing claims, and tracking status. Yet for many, the journey ended before it began.
More than half of registration attempts failed, locking customers out of every other self-service feature.
As the lead UX/UI designer, I partnered with our Director of Customer Experience and Product Owner to understand why, then led the redesign to remove friction, bring the interface up to modern design standards, and navigate unforeseen obstacles to deliver a solution on time and in scope.
Following the deployment, successful registrations saw an immediate 4% increase, translating to 400 more users. To move the needle further, we are now finalizing the deployment of FullStory to analyze user behavior and inform future UX enhancements as we await critical back-end architectural solutions.
The reality we uncovered
Our discovery effort combined stakeholder interviews, call center insights, and journey mapping. We spoke with people across IT, claims, enrollment, marketing, and customer support to understand both the technical and human sides of the problem.

(Note: visual artifacts optimized for desktop viewing)
The numbers from Google Analytics told part of the story: 53.8% of attempts failed. The qualitative insights we gathered told the rest.
Confidence evaporated quickly when customers hit errors they didn’t understand. One-time passcodes sometimes never arrived. Messages didn’t explain what to do next. Frustrated users often abandoned the process entirely — some even canceling their policies.
“Lots of people call in to cancel because they can’t register to service their policy.”
– Martha, Customer Service Team Lead
The current registration flow wasn’t just inefficient; it was actively damaging the customer relationship.

We discovered that ‘Find Your Account’ and ‘One Time Passcode’ were giving users the most trouble.
(Note: visual artifacts optimized for desktop viewing)
Two design cycles behind
The portal’s framework was so outdated that our developers couldn’t use components from the design system UI kit, and the current front-end visuals were two design cycles behind — still using branding from before the 2022 refresh and lacking the newer digital design system standards that govern everything from responsive breakpoints to interaction states.

(Note: visual artifacts optimized for desktop viewing)
Patching the existing experience would only add to the inevitable tech debt, so we proposed a full modernization to the development team, aligning the registration flow with current standards.
This approach also let us draw on proven design elements, from micro-interactions (e.g. button states) to macro-interactions (e.g. selecting a way to receive a one-time passcode), that had already been tested for usability and accessibility.
The promise — and gap — in code reuse
At the same time, our global design team was building a new standardized login framework for all platforms. If we could align with it, we might be able to reuse their work for registration and reduce development effort.
I facilitated discussions between our portal development team, product leadership, and the global design system team to explore the idea. Initially, the answer was encouraging — the new framework was flexible and could be customized.
But as we dug deeper, we uncovered two blockers:
- The framework wouldn’t be ready in time for our launch
- It didn’t include any registration functionality at all.
This was a valuable discovery: it surfaced a missing piece in the global solution, and it confirmed that for our project, we’d be starting from scratch.
Changing the way we built
The development team was new to our design system and accustomed to working in silos, surfacing builds only late in the process. That approach risked mismatches between design and implementation.
I introduced a new collaboration model:
- Twice-weekly office hours for real-time feedback.
- A dedicated channel for quick design questions.
- Ad hoc reviews throughout the sprint to catch issues early.
- Onboarding to our design system documentation so developers could self-serve.
This kept the build aligned with design intent, and ensured every component — from spacing to interaction states — matched the new standards exactly.
The result
Even without the shortcut of reusing global code, we launched the redesigned registration experience on time and in scope. The new flow aligned with modern design standards, reduced confusion, and provided clearer guidance for users. It also introduced a collaboration model that’s now being used on other portal projects.
And by surfacing the registration gap in the global login framework, we influenced the direction of a future enterprise-wide solution.

(Note: visual artifacts optimized for desktop viewing)
Early signs of impact
Our discovery made it clear that the most severe failures stem from back-end data and technical architecture — issues outside the scope of this release. That insight gave stakeholders a clear roadmap for where system-level changes are most urgently needed.
Within the scope of what could be addressed through UX and UI, our redesign:
- Combined identification steps into one screen, allowing users to try another method (phone, SSN, policy ID) without restarting.
- Added guidance on where to find key details like policy numbers, tailored to different customer types.
- Introduced FAQs and helper text to clarify how to enter names and other information that must match policy records exactly.
- Improved error messaging and validation so users understand what went wrong and how to fix it.
- Updated visual design and layout to align with current standards and reduce clutter.
Usability testing confirmed these improvements: users were more confident navigating the process and better able to complete steps that previously caused avoidable abandonment. The front-end is now positioned to deliver full value once back-end fixes are in place.