Overview
This project focused on post-application identity verification for credit card applicants. After submitting an application, some users were unexpectedly pulled into additional verification steps at the moment they expected an approval decision. Nearly 1 in 5 applicants asked to upload documents abandoned the process, which created both customer frustration and additional servicing workload. I led UX strategy to redesign this experience, partnering with product, operations, risk, fraud, and engineering leads.
ROLE: UX Strategy Design Lead
CLIENT: Discover Financial Services, Card Acquisition & Digital Servicing
TEAM: Account Lead, Creative Director, UX Director, UX Designer, Visual Designer, Copywriter, Production Lead
PROBLEM: A fragmented verification system led to excessive queuing for identity checks, introducing an unexpected, unexplained interruption after submission that drove abandonment at a critical step.
IMPACT: Led the redesign of verification into a unified experience, improving in-session completion and aligning product, risk, and operations around a shared model.
KEY WORK: Executive workshops, system and process mapping, cross-functional alignment, service blueprinting, prototyping, and usability researchMany Systems, One Applicant Experience
The core issue wasn’t just confusing messaging and outdated screen templates. It was a fragmented, legacy system. Verification relied on multiple backend systems, each triggering different checks based on risk signals. Individually, those systems worked. But together, they created an inconsistent experience with unclear messaging and unpredictable next steps.
Our challenge had two parts:
Our challenge had two parts:
First, we needed to understand and reconcile a complex ecosystem of systems and workflows.
Second, we needed to translate that complexity into a single, consistent, and scalable experience for applicants.
Current State Blueprint identifying logic, sequencing, and constraints for deep-dive work sessions.
Visualize The Ecosystem
We started by making the system visible. I led efforts to map the full verification ecosystem across review rules, system dependencies, operational workflows, and performance data. What emerged was a fragmented system, where disconnected logic across teams and platforms was driving inconsistent decisions, user confusion, and operational inefficiency.
From there, I co-facilitated cross-functional working sessions to reframe the problem and align teams around a more cohesive model. A central tension was balancing consistency with edge-case handling. Over time, the system had accumulated exceptions that introduced variability in how verification was applied.
We made a deliberate decision to prioritize consistency, standardizing the majority of verification scenarios and limiting exceptions to clearly defined cases. This reduced variability, simplified the experience, and created a more predictable system for both users and internal teams, while still meeting compliance and operational needs.
This led to a future-state service blueprint that gave teams and leadership a shared understanding of the system, enabling more consistent decision-making and aligning execution across teams.
Early-stage logic flow model exploration to simplify, modernize, and optimize the verification platform.
Future-state blueprint model created to illustrate swimlanes and review rules affected by modernization.
Clarify Expectation
Using the blueprint as a foundation, we created mobile-first prototypes to evaluate how applicants understood verification requests and document uploads.
Usability testing revealed a critical gap in expectations. Applicants believed the process was complete after submission, so when verification appeared, it felt like an interruption rather than a continuation of the application, leading applicants to disengage at a critical step.
The system also failed to clearly communicate why additional information was needed and, in some cases, asked users to repeat information they had already provided, increasing frustration and eroding trust.
Wireflow visualization of 3 main verification scenarios based on performance data and path analysis.
Example stakeholder work session artifact to align on impact, tradeoffs, and intended experience design.
A Unified Experience
We translated these insights into a unified verification experience. Instead of treating verification as a separate, interrupting error state, we designed it as a natural continuation of the application, using a consistent structure and clear messaging to explain why additional information was needed and what to expect next.
This unified approach replaced fragmented flows with a single framework, improving error handling and providing more meaningful feedback during document uploads. The experience was designed to feel predictable and consistent, so when something failed, applicants weren’t left guessing, and we avoided unnecessary repetition unless we could explain why. This allowed us to manage backend complexity without exposing it, delivering a clearer, more coherent experience.
Key Verification Screens: Check Status, SSA, School Enrollment, ITIN Doc Upload, Pending Status
What This Work Changed
Simplified Product Architecture
This work made the verification experience clearer and more predictable for applicants, while consolidating fragmented flows and legacy templates into a single, scalable framework that could support multiple scenarios without added complexity.
This work made the verification experience clearer and more predictable for applicants, while consolidating fragmented flows and legacy templates into a single, scalable framework that could support multiple scenarios without added complexity.
Operational Efficiencies
It aligned multiple teams and stakeholder leadership around a shared verification model, reducing variability in how cases were handled and establishing a repeatable UX pattern for future onboarding and compliance initiatives.
It aligned multiple teams and stakeholder leadership around a shared verification model, reducing variability in how cases were handled and establishing a repeatable UX pattern for future onboarding and compliance initiatives.
Business Impact
Post-launch, verification shifted from a fragmented, drop-off-prone process to a streamlined, in-session experience. Abandonment decreased and servicing call volume declined as fewer applicants required follow-up. Improved verification checks reduced unnecessary flags, lowering document uploads and enabling the majority of pending applications to be resolved in-session.
Post-launch, verification shifted from a fragmented, drop-off-prone process to a streamlined, in-session experience. Abandonment decreased and servicing call volume declined as fewer applicants required follow-up. Improved verification checks reduced unnecessary flags, lowering document uploads and enabling the majority of pending applications to be resolved in-session.
Evolving the Approach, with AI
This work was grounded in traditional methods, before AI tools were part of everyday design workflows. System mapping, service blueprinting, and cross-functional workshops were essential to building a shared understanding of a complex, fragmented ecosystem.
Today, AI tools would accelerate parts of this process. System rules, edge cases, and operational data could be synthesized more quickly, and prototyping and testing cycles could move faster through rapid iteration.
But the core challenge remains the same. Defining a cohesive verification model still requires human judgment. The most critical decisions were about tradeoffs. We prioritized consistency over isolated optimization, limited edge-case variability, and simplified how the system was presented without exposing unnecessary complexity.
AI can accelerate analysis and execution, but it does not replace the need to align teams, establish clear principles, and guide decisions toward a coherent, scalable system.
NEXT UP
Discover Card Application
Continuous Optimization