Background

Government services have undergone extensive digital “transformation.” Now, we look to bridge gaps between related services and reimagine the experience from the perspective of life events, rather than as a paperwork filing process. 

As the largest integrated health care network in the US, the Veterans Affairs (VA) provides health care and a wide-variety of benefits and services to millions of Veterans, Caregivers, and Veteran families. As part of an embedded multidisciplinary strategic VA sprint team, I worked to improve the application experience for the Program of Comprehensive Assistance for Family Caregivers.

This project’s success was recognized by the CTO of the VA and was called out in a Secretary of the VA press conference in which it was stated that the VA is utilizing research and applied human-centered design to improve the experience of getting benefits for Veteran Caregivers.

Discovery

I joined this product team about 6 months after this form’s digitized debut. Since the team was transitioning from MVP to iterative improvement, the roadmap hadn’t yet been set. To help inform our roadmap and overall strategy, I began Discovery research. This included user research, analytics review, and stakeholder workshops. My goal was to better understand how the application was performing for the humans who use it, both the Veterans and Caregivers applying for the program, as well as for the VA staff who process these applications. Based on identified pain points, we built out our roadmap and prioritized our initiatives.

The first initiative we tackled was the “Sign as Representative” feature, the section where applicants (Caregivers and Veterans) elect to have a representative sign the application on their behalf (and upload required documentation), or not. This section had the highest drop-off rate, and the success rate for applicants who chose to have a representative sign on their behalf was only 33%-  resulting in delays in application results and ultimately delays in access to benefits. Additionally, it was creating extra processing burden for VA staff.

Challenge

This initiative was particularly challenging because the types of documents that can be accepted varies widely and depends on the state and tribal laws in which the user is located. There was not a specific form we could link to or request, but instead we needed to support users to make the right choice and identify the right documentation for their specific situation, ensuring that the policy and program requirements were met.  

In order to ensure burden reduction for both the users and the VA staff, we set a goal, well above the industry standard for document upload acceptance, at 75%.

Approach

The range of documents that we saw uploaded and subsequently denied was just as varied as the types of documents that the program could accept. We hypothesized that users did not have sufficient contextual cues to know which documents they needed to provide or whether or not their specific situation even required them to upload a document at all. 

We conducted mental model research to gain a deeper understanding of how users perceive and interact with this section in the application itself and in similar scenarios that might require proof of legal representation, eg. at a bank or pharmacy. Additionally, we conducted stakeholder workshops to gather the necessary policy requirements and align with the broader VA organizational objectives.

The solution

The research indicated that the existing feature wasn’t facilitating the legal requirement that the document specifically give medical decision-making powers to the representative, the program requirement that the documentation is optional, or appropriate Veteran/Caregiver context to make the right decision for their circumstance and provide the right document if necessary. The mental model research gave us insight into how Representatives naturally decide what documentation to provide, making the decision based on what it enables them to do and not necessarily about a specific document name or type. This meant that the document examples we had been providing were actually confusing, not helpful. In this situation, the policy requirement was that the documentation should give authority to make medical decisions on behalf of the Veteran, but the way the information was presented, applicants had the impression that the document should give legal power of attorney, but not specifically medical. We needed to reframe this feature from the user’s point of view.

We reenvisioned the Sign as a Representative flow, design, and copy to not only align with what was legally required, but to align with how Veterans and Caregivers think about this type of process and document. Working to:

• Visually tie the document upload to the signature, making it clear that having a representative sign was an option, not a requirement for the program.
• Rewrite copy to clarify that this representative document needs to give legal authority to make medical decisions specifically and reduce fear that by uploading the document the Veteran’s rights to make decisions wouldn’t be affected.

We implemented these changes iteratively, while simultaneously running usability testing on the prototypes.

Before

Outcome

One month into implementation we reached a 60% acceptance rate of applications with uploaded representative documentation. This is a nearly 100% improvement in 30 days, well above industry standard for upload document features in general, but more importantly, a huge improvement for the Veterans and Caregivers applying for this program and the VA staff who process the applications.

We continued to work on the copy and are now consistently reaching over 70% acceptance month-over-month.

Deliverables/artifacts

Links will open in a new tab and you will no longer be on dene.land.

• Product roadmap
• Mental model study
  • Research plan/recruitment requirements
  • Conversation guide
  • Research report & Presentation
• Usability testing
  • Research plan/recruitment requirements
  • Conversation guide
  • Research findings report & Presentation
• User flows (use case definition)
• Wireframes and prototypes
  • Live Application
• Implementation strategy
• Technical specifications