SSA's first post-COVID cross-platform validation program — restoring evidence-based governance, and trust across 60+ federal digital products.
As UX Research Lead, I designed and led SSA’s first post-COVID, cross-platform design-system validation initiative, producing 25+ evidence-based pattern updates, a new framework for governance, and three new institutional processes for ongoing validation.
For five years after COVID halted in-person testing, SSA's design system kept shipping — components, identity fields, accessibility-critical patterns — all running on assumption across a system supporting 200M+ annual interactions. The foundation couldn't be trusted. For users navigating disability applications under financial or health stress, even minor friction is a barrier to completion.
That program delivered: 25+ evidence-based pattern updates, a severity×confidence framework that turned findings into governance decisions, and three institutional processes built to keep the design system validated long after the work ended.
SSA's design system is the shared foundation for 60+ federal applications, built on USWDS — the federal design standard. USWDS sets the compliance baseline. But compliance is not the same as validation. A pattern can pass every technical check and still fail the real people using it.
By 2024, five years of unvalidated patterns had accumulated across desktop and mobile. Ownership existed but validation didn't — there was no mandate, and no mechanism to flag what had never been tested. It didn't become urgent until it became a risk: FY25 modernization milestones required mobile to be the premier service channel, and the foundation those milestones were being built on had never been validated with a real user.
How might we validate untested design-system patterns — across desktop and mobile — while rebuilding internal confidence and capacity for continuous usability validation?
I established a cross-platform validation framework for the design system that integrated usability research, governance, ResearchOps practices, and practitioner training.
Patterns were prioritized by user risk, tested across desktop and mobile, and translated into validated design guidance—enabling teams to continuously improve patterns across 60+ applications.
No process for who owned pattern decisions or when re-validation was required.
Patterns in production never validated for WCAG compliance or cognitive load.
Sensitive identity fields deployed without testing with affected communities.
FY25 mobile-first milestones built on an unvalidated foundation.
Two-phase cross-platform validation paired with ResearchOps infrastructure — producing validated patterns, prioritization frameworks, and governance processes that outlast any single research cycle.
The Axure prototype was structured as an end-to-end Disability benefits application journey.
Participants were given a specific scenario: you are a dual citizen living abroad who needs to apply for disability benefits. You go to the SSA website, find the link to apply, and begin.
I chose this scenario intentionally — one of the patterns under test was Address (International), and we needed participants to have a realistic reason to interact with it. A dual citizen living overseas provided that naturally: international address entry wasn't a forced task, it was just part of applying.
15 patterns prioritized with the product owner based on governance risk and portfolio deployment breadth — each required to have a real-world use case meeting an identified user need and business goal before entering the validation queue.
Each pattern scored on two axes to determine governance action.
Cross-platform testing surfaced patterns that failed for different reasons — emotional discomfort that persisted regardless of platform, and behavioral breakdowns that diverged sharply by device. Both required different governance responses.
Same design. Same tasks. Desktop users recovered from file upload errors independently — mobile users could not. Layout compression on small screens obscured the error source. Governance decision: design mandate, not deferred fix.
Patterns that passed every functional metric but caused emotional friction, coercion, or trust erosion — consistently across both platforms.
"I support inclusivity, but this question caught me off guard. I wasn't expecting it here, and it made the experience feel less comfortable."
This quote captures the core tension of the entire program: a user who values inclusive design still experienced discomfort — not because of the intent, but because of the absence of a graceful exit. Emotional usability isn't about content; it's about control.
Likert helpfulness 4.38 desktop / 4.26 mobile — functional success masked emotional harm. No graceful opt-out existed.
Consistent negative Likert scores regardless of device. Platform did not mediate the experience — the design pattern itself caused harm.
Users completed the field but emotional safety was low. Absence of rationale created distrust in the larger application.
Patterns where users failed to complete tasks or required facilitator intervention — most commonly diverging between desktop and mobile.
"I could see something was wrong — there was a red icon — but I couldn't tell which file it was. I didn't want to delete the wrong one and have to start over."
This moment — paralysis at the point of error — is the exact failure the research was designed to surface. The participant wasn't confused by the concept of uploading a file. They were stopped by a color collision and a layout that worked on desktop but broke under real conditions on mobile. Task completion metrics would have called this a pass.
Error messages not linked to source filename on mobile. Layout compression prevented recovery. Highest severity signal in the program.
Missing "Today" label and inconsistent button order (Cancel before Save) violated OS conventions on mobile. Desktop users were not affected.
The research program built the organizational scaffolding that would continue operating — and validating — long after the initiative ended.
This infrastructure didn't run alongside the research — it was the mechanism that turned findings into governance decisions and governance decisions into standards that shipped.
The design system and organizational outcomes below are the same story at different levels — validated patterns that shipped, and the governance infrastructure that will keep them validated.
The most important outcome wasn't validated components — it was a culture of evidence. Research at the design-system level has disproportionate impact: one validated pattern reaches millions. An unvalidated one carries millions of small harms.
The findings this program produced — the 54% mobile fix rate, the platform-consistent emotional harm from inclusive identity fields, the patterns that passed on desktop and failed on mobile — were only possible because each structural decision below was built and defended to protect data integrity.
| Decision | How I Led It | Takeaway |
|---|---|---|
| Testing Format | Replaced guerrilla library testing with remote moderated testing — deploying infrastructure already built for SSA Mobile App Exploratory Research | Trading a familiar method for a compliant one unlocked a more diverse participant pool.Institutional constraints, reframed, become organizational advancement. |
| Study Instrument | Removed SUS entirely; rebuilt task structure around sub-scenario tasks that isolated each pattern within the workflow | Applying a familiar instrument to the wrong unit of analysis produces precise but meaningless data.Method selection must match what you're actually measuring. |
| Research Eligibility | Established a pattern eligibility gate — real-world use case meeting user need and business goal required before entering the validation queue | Testing patterns without a justified use case is curiosity with a budget.Research purpose is as important to protect as research rigor. |
| Study Scope | Held two-phase structure — desktop and mobile as separate studies with sample sizes optimized per platform using the 5×2 rule | Cost: longer timeline and less per-device precision. Return: cross-platform evidence that held under governance scrutiny.Reliable findings later beat fast findings built on a compromised design. |
| Requirements Definition | Discovery sessions at program start defined success metrics and research questions per pattern before any testing began | Without a definition of pass before the study runs, findings can't drive decisions.Define what success looks like before scope is set. |
Leadership had already delegated full research design authority to me — trust built through prior delivery, including the SSA Mobile App Exploratory Research, which shaped the agency's mobile strategy, informed its first universal mobile application, and was presented to senior executives and the CIO.
The product team came in expecting the pre-COVID playbook — and every major structural decision required holding a better one. The decisions weren't approvals I sought. They were directions I set, communicated transparently, and defended with rationale so leadership stayed informed and could redirect if needed.
If I were starting again, I would establish the severity×confidence framework and governance escalation path before the first study — not during it. The framework worked, but it had to earn its legitimacy while findings were already in motion. Agreeing on what "critical" meant and who owned a pattern decision before any data existed would have made the outcomes structural rather than situational.