Overview
For this redesign of VA.gov's Dependent Management interview-style form (Forms 21-686c & 21-674), I modernized the digital "Request to Add or Remove Dependents" experience to ensure parity with the PDF version of the form put out by the Veterans Benefits Administration (VBA). I also led migration to USWDS v3 components and worked to resolve critical redundancies in the form.
Role: Design Researcher
Duration: January 2024 - January 2025
A Disconnect in Documentation
Veterans rely on these forms to ensure their disability or pension benefits accurately reflect their family status. However, the digital form had fallen out of sync with the paper form maintained by the VBA. Veterans faced potential claim rejections because the digital submission format was technically out-of-date.
UX Debt
A review of Medallia survey data revealed users were frustrated with redundancies in the form.
The form asked users to only answer certain questions if they are pension recipients at several points throughout the form, increasing cognitive load for users and showing unnecessary information to disability benefits recipients.
If a Veteran wanted to add a school-aged child (18-24) but hadn't added them to their benefits previously, the form asked them to enter the dependent's identifying information twice, a redundancy carried over from the PDF forms.
I also partnered with the site-wide content team for a full content audit of the form, which unearthed, in addition to several opportunities to clarify form requirements, some potential quick wins for accessibility.
Systems-First Design
As the solo designer, I had to balance the need for a total overhaul with the constraints of a mature, strictly-governed design system.
Audit & Alignment: I mapped the new PDF requirements against the existing digital flow to identify "parity gaps."
Component Migration: I led the transition to USWDS v3 components, ensuring every interaction met the VA’s rigorous accessibility standards.
Content Chunking: I broke down dense, intimidating form sections into digestible steps to reduce cognitive load—a primary pain point for the Veteran community.
Pivoting for Constraints
During build, we discovered a limitation in the v3 component for multiple responses—it couldn't dynamically update based on external selections. This meant my solution to Veterans having to add school-aged children twice couldn't be implemented as content within the component couldn't update dynamically based on selections made in the first step of the form.
Because the design system board was reluctant to allow one-off patterns, I pivoted back to messaging-based solutions. I used clear, screen-reader-friendly copy to manage user expectation around redundancies within the bounds of a not-fully-vetted tech stack.
Turnover and Red Tape
This project was a lesson in long-term design stamina. Navigating a year-long cycle of shifting leadership and changing priorities developed my flexibility and patience. Working within a "mature" system meant improvements took time. Any improvements had to go through several cycles of review with the design governing board, especially if it meant additions to the design system.
Object-Oriented Reflections
Midway through the project, a working session with the site-wide content team sparked a major realization: we were designing a form but Veterans needed a dashboard. Research showed Veterans expected to manage dependents (Add/Remove/Edit) from a central "View Dependents" hub rather than a linear, one-way form.
While my tenure ended before launch, the project resulted in a fully up-to-date, USWDS v3-compliant design.
When this redesign began, I understood it as a quick update and chance to tackle low-hanging fruit. If I were to start over, I would advocate for a complete overhaul in the form's flow with an object-oriented approach to align with the Veteran's mental model of "managing dependents"
