Overview
Project: VA Dependent Verification Digitization (Form 21-538)
Role: Design Researcher
The Mission: To reduce the $7M in annual benefit overpayments caused by outdated dependent data. By digitizing a manual 8-year verification process into an easy annual digital touchpoint, we aimed to protect Veterans from the burden of debt.
The Problem
When a Veteran’s household changes (e.g., a child ages out of benefits) and the VA isn't notified, overpayments occur. These funds must eventually be paid back, creating financial stress for Veterans.
A previous attempt at a global "one-click" pop-up was rejected for being too disruptive and failing to reach Veterans who don't visit the site frequently.
The Solution
To reach the Veterans most likely to have out-of-date dependents, we pivoted to an off-platform entry point with an email notification directing users to VA.gov. This alongside the abbreviated verification form makes verification a seamless annual habit rather than a daunting legal requirement.
Why One-Click Wasn't the Answer
The initial product vision was a "one-click" verification. However, research and stakeholder reviews revealed a friction point: Compliance vs. convenience.
Fitting enough context into a one-click solution for the Veteran to understand the ask and what they're agreeing to in completing the form proved challenging. Research showed that few Veterans understood what "verification" meant in this context. What's more, Veterans found the language around the ask vague and the tone too "optional," leading to low engagement.
Finally, users don't read—they skim. In testing, users bypassed copy about how to edit dependents because, in striving for a one-click experience, each screen had become too dense.
Pivoting to One-Thing-Per-Page
To resolve the conflict between a one-click experience and informed consent, I moved away from the one-click vision toward a One-Thing-Per-Page strategy. By slowing the user down and employing progressive disclosure, we increased user speed and accuracy. Breaking the process into digestible chunks ensured they understood what they were agreeing to and understood how to accomplish edge-case tasks.
Rather than inventing a "mini-form" pattern (an option that came up in design governance review), I used the existing VA.gov form pattern that users already know and trust. This helped to ensure the project could launch sooner and we could start gathering data on whether our solution is having an impact on the $7M metric.
Aligning Teams and Tasks
This project required navigating a complex network of government stakeholders:
Design Governance: I navigated a rigorous review process, using research data to advocate for the Veteran's mental model when client leadership was hesitant.
The Site-Wide Content Team: We originally thought the email notification would be handled by VA.gov's content team, meaning we would have greater access to analytics and quick changes. We also had to align with Site-Wide Content to ensure messaging about the new verification feature appeared wherever relevant across the platform.
Veterans Benefits Administration: Toward the end of design, we learned the email notifications were actually owned by the VBA. This meant we had to nail down messaging both within the email notification and on-platform without the availability to make quick edits. We also needed to establish a cross-agency relationship to have regular access to both email data and overpayment metrics.
Tracking Success
The project was handed off with a launch plan that included a clear framework for how to measure success:
Primary Metric: Reduction in the $7M annual overpayment total
Secondary Metric: Increase in "Update" form completions (indicating Veterans are catching inaccuracies during the review process).
Qualitative Metric: Positive shifts in Medallia survey sentiment regarding "ease of maintenance."
Reflection
If given more time, I would have bridged the gap between the VA.gov platform and VBA communications. My research suggested that well-timed communications around life events (like a child’s 18th birthday) would significantly improve Veterans' experiences.
The project reinforced that "simple" doesn't always mean "short." A longer, clearer form is often more effective than a "fast" one that leads to errors and misunderstanding.
