Simplifying a Complex Claims Journey

Making a claim is a multi-step process involving claim type, service provider, expenses, supporting documentation, and submission.

As part of a broader replatforming initiative, the entire Make a Claim journey was being reconsidered. I led the design of the first two steps and collaborated with the designers and cross-functional team working across the remainder of the journey.

The challenge was not simply to improve the interface. It was to rethink how members navigate a complex set of business rules while ensuring their decisions worked within the technical and operational realities of the new platform.

The existing experience asked members to first distinguish between Online Claims and Other Claims, then navigate a complex hierarchy of claim types. This created uncertainty around where specific claims belonged.

For example, a member submitting diabetic supplies such as syringes could reasonably select Medical Equipment & Supplies, while the claim actually needed to be submitted under Drugs. Selecting the wrong category could result in a rejected claim.

The impact extended beyond the member experience:

• —Incorrect or rejected claims
• —Additional processing and operational effort
• —Members needing to resubmit claims
• —Potentially increased calls to member support
• —Longer resolution times

Prior research I conducted around benefits, coverage, and balances revealed an important pattern. Members tend to think about the person before the benefit:

Research validated this mental model. I brought this insight into the replatforming work and proposed that the claims journey begin by establishing who the claim is for and which plan it belongs to, before asking members to identify the claim itself. This became the foundation for the first step of the new journey.

Before the research above confirmed it, this was a hypothesis I brought into the project. The original flow started with the plan, what are you claiming, and it read as mechanical and operational, structured around how the business organized claims rather than around the person filing one. I proposed flipping that order: establish who the claim was for and which plan it fell under first, before ever asking what they were claiming.

The flip paid off. It felt like the more natural place to start, both in testing and in how the rest of the journey could use that context, and it set the tone for a claims experience that led with the person rather than the paperwork.

Nearly everything I designed on this project lived in this second step. The claim type selected here didn't just shape what a member saw next, it determined how the claim was processed once submitted. Picking the right category was often the difference between a claim resolving automatically through the platform's adjudication rules and a claim getting routed to manual, investigative review. Getting a member to the right category wasn't just about their confidence in the moment, it was load-bearing for everything the system did with their claim afterward.

The previous experience required members to understand an internal distinction between Online Claims and Other Claims before they could find the appropriate claim. The design opportunity was to remove as much of that cognitive burden as possible while still preserving the underlying business requirements.

Some categories needed to resolve one or two levels deeper before a claim type was specific enough to route correctly. Medical equipment and supplies, for example, could go from a high-level category down to something as specific as a particular breathing aid, but only for the categories that actually needed that extra step.

The search bar carried its own share of the redesign, and it mattered enough to become its own workstream. The previous search only matched a claim type's exact name, so a member searching “mental health” found nothing useful unless they already knew to search “psychologist.” I worked with the team that owned the platform's search to rebuild it around themes and everyday language instead of exact wording, and to make sure a genuine no-match still pointed members somewhere useful rather than a dead end.

We combined quantitative data with qualitative research to understand member behaviour and identify where the existing information architecture was breaking down.

• —2 rounds of card sorting
• —2 rounds of tree testing
• —3 rounds of usability testing
• —Adobe Analytics
• —Claims submission data

The most frequently submitted claims included massage therapy, chiropractic services, eyewear, physiotherapy, psychology, drugs, and dental services. This helped us focus on the areas where improvements could have the greatest impact while still considering less frequent and more complex scenarios.

The work didn't stop at launch. A follow-up usability study validated the single-page, progressive-disclosure flow and the new search feature, and a later tree test checked newer claim expenses, things like bandages, meal-replacement drinks, and infant formula, against the existing categories. That round surfaced a genuine limit: food and nutrition-related claims didn't intuitively read as “medical supplies” to members, even when the underlying category was technically correct. Not every mental-model gap closes on the first pass, and this one went back into the list of things worth revisiting rather than getting called done.

Research showed that members often had the right intent but struggled to map that intent onto the organization's taxonomy. One participant submitting a massage claim explained:

These moments highlighted an important distinction. The problem wasn't simply finding information. It was feeling confident that the information was correct. That confidence mattered because an incorrect decision at this stage could affect every step that followed.

Card sorting surfaced a related pattern: members were more likely to associate a claim with who provided the service than with where they received it. Nursing, for instance, tested better under Health Professionals than under its original grouping, because members pictured the person providing care rather than the setting. Moving it based on that evidence measurably improved selection accuracy in later testing.

The underlying claims taxonomy and business rules could not simply be removed. The solution needed to balance:

• —Member mental models
• —Business and processing requirements
• —Information architecture
• —Technical constraints
• —Web and mobile requirements
• —Future scalability
• —Dependencies across the remainder of the journey

I explored different interaction models through iterative design and usability testing, focusing on making common claim types easier to identify while maintaining access to less common scenarios.

Category descriptions did more of that work than a first glance would suggest. In testing, a member trying to claim diabetic syringes second-guessed Medical Equipment before a description mentioning syringes under Drugs corrected their path, exactly the kind of course-correction the interface needed to support without ever showing an error.

Although I led the design of the first two steps, these decisions could not be designed in isolation. The broader journey continued through:

The first two steps acted as inputs into everything that followed. I worked closely with the designers responsible for the downstream experiences to ensure that our decisions connected across the journey and could support the different states and requirements introduced later.

This was particularly important within the replatforming effort, where the experience needed to work as part of a larger technical system rather than as an isolated redesign.

The ideal solution would have involved changes beyond the interface, including some underlying operational processes. However, not all operational changes were possible within the scope of the project. This meant using design to absorb as much complexity as possible through:

• —Information architecture
• —Interaction design
• —Content and hierarchy
• —Progressive disclosure
• —Clearer decision points

The goal was not to eliminate every possible source of confusion, but to reduce unnecessary cognitive effort and help members make better decisions within the constraints of the existing system.

The work established a new foundation for the claims journey as part of the broader replatform. Instead of asking members to understand the organization's internal claims structure, the experience was designed around the way they naturally approach the task, reducing incorrect claim selection, minimizing downstream rework, and creating a clearer path through a complex process. That showed up directly in the numbers: