As Capital One moved to adopt passkeys—a highly secure, passwordless authentication method—we faced a massive, ecosystem-level challenge. The business objective was clear: deploy this technology quickly to reduce the millions of dollars spent on password resets and fraud remediation. However, the operational constraint was equally strict: we had to scale it slowly and intuitively enough to prevent a massive spike in confused customers calling the support center.

Designing for Identity is inherently high-stakes. My role wasn’t just to design screens; it was to act as the primary architectural partner between Product, Engineering, and Risk, ensuring we didn’t compromise user trust for the sake of speed.

From day one, I broke down the traditional “waterfall” handoff. I initiated a shared Slack channel and a unified kickoff, partnering directly with Product to co-author channel scaling PRDs. I wanted to ensure that UX metrics and user-centered hypotheses were baked into our strategic requirements before a single wireframe was drawn.

I needed to work with a handful of tech teams and product leaders as they all specialized in a specific channel. To prepare for our syncs and leadership shareouts, I utilized Claude and our internal Skills to rapidly prototype, research industry standards, troubleshoot customer experience issues, and iterate on leadership feedback. This allowed us to spend less time on documentation and more time debating our high-level scaling strategy.

Identity projects are heavily governed. After MVP launch of passkeys on sign-in web, our Cyber, Risk, and Legal teams strongly advocated for dense, jargon-heavy disclosures to cover every possible edge case of passkey creation.

As a strategic partner, I recognized that dumping compliance text on the user would instantly kill conversion. Instead of accepting the mandate as a blocker, I negotiated a middle ground. I utilized progressive disclosure techniques and strict visual hierarchy—breaking up intimidating text blocks with clear typography and structured layouts. We were able to meet compliance requirements while successfully preventing user drop-off.

Shortly after our initial web launch, I proactively partnered with our Business Analyst to review our drop-off funnel metrics in Tableau. I identified a critical bottleneck: our registration numbers were stagnating because the passkey prompt lived on the pre-authenticated sign-in page.

I didn’t wait for Product to create a ticket. I facilitated a brainstorming session with Product and Tech and pitched a strategic pivot: a post-authentication intercept. Using Figma and Gemini, I architected a singular, scalable landing page designed to adapt dynamically across any channel. This systemic approach prevented our engineers from burning hours trying to retrofit passkeys into legacy flows, ensuring ecosystem-wide consistency.

To validate this new approach, I ran tests on our new designs and content variants with our research partners in dScout to gauge which value propositions resonated most strongly with our customers. The insights from this research didn’t just inform a single screen—I leveraged them with our content designer to deliver a comprehensive content strategy framework and playbook for the entire passkeys ecosystem. Furthermore, I established a rigorous A/B testing plan for our production launch, ensuring we could continuously optimize our value props, imagery, and messaging combinations to drive the highest possible conversion.

Because Identity relies on exact keyboard behaviors and real-time data inputs, static mocks are rarely sufficient. To drive engineering consensus, I built high-fidelity HTML prototypes using Claude, hosting them directly within our internal Identity GitHub repo.

Because these prototypes utilized our actual coded design system, I could hand off work to the front-end team knowing it used actual code, and to the back-end team with a clear map of necessary API endpoints. Furthermore, to mitigate technical latency when the OS passkey prompts triggered, I designed precise micro-interactions (e.g., dynamic loading spinners transitioning from “Waiting for browser prompt…” to “Signing you in…”). This masked the API lag, anchored the user, and prevented premature abandonment.

Influence isn’t just about what you build; it’s about what you choose not to build. Two weeks before our native intent governance review, Product proposed expanding the release to include two subsequent phases of the passkey roadmap.

Recognizing this as severe scope creep, I pushed back. While I validated Product’s enthusiasm, I anchored our conversation directly to the original, agreed-upon PRD. I explicitly warned leadership that rolling out untested, unvalidated experiences in the high-stakes Identity space posed a massive company reputation risk. By reframing the conversation around brand safety and data-driven phasing, I successfully convinced Product to drop the unvalidated additions. This protected our engineering timeline and ensured our foundational MVP launched exactly on track.

By treating design as a strategic, cross-functional partnership rather than an execution arm, we delivered an experience that massively impacted the business’s bottom line:

• Adoption: Onboarded over 1 million users to passkeys in under 9 months.
• Success Rate: Increased overall sign-in success from 89% to 98%.
• Business Impact: Drastically reduced password reset volume, saving significant operational capital while introducing a cutting-edge standard of security.