Note: due to NDA, details and projects I’m allowed to show are limited to what is publicly available and based off industry research. Design decisions based off of internal research, wireframes, iterations, and build kits are restricted American Express intellectual property and cannot be shown.
US Consumer Card Apply
Experimentation Program
— PROJECT NAME
US Consumer Card CXO Program
— ROLE
Established Program
Lead Designer
— DATE
Feb 2026 – Present
Established and lead Amex’s first formal experimentation program across its highest-revenue acquisition portfolio: US consumer card apply experiences spanning proprietary and partner cards.
Accomplishments
- 26 test concepts generated
- 7 A/B tests launched
- 57% win rate
- $59.9M incremental revenue
- 400+ accessibility bugs remediated
- Established formal accessibility testing and design QA processes
Opportunity
Establish a continuous experimentation program to accelerate optimization across US consumer card application experiences, generate actionable insights, and improve application conversion to drive incremental revenue.
Building the Program
Partnered with product, marketing, analytics, UX research, content design, engineering, and legal to:
Establish a centralized test-idea intake form and prioritization process.
Create future-state mocks and prototypes, incorporating cross-functional feedback to refine the vision.
Define an incremental roadmap toward the ideal application experience.
Featured Experiments
4 winning experiments out of 7 launched tests (57% win):
- Platinum Accept Page +$12.5M revenue
- Progress Bar +$12.6M revenue
- Header +$12.5M revenue
- Remove name on card input +$22.3M revenue
Platinum Accept Page
Problem
The existing Platinum Card acceptance page was bare-bones and visually emphasized the welcome offer above all else. However, most applicants don’t qualify for the highest offer, creating a potentially disappointing experience and missing an opportunity to communicate the card’s broader value. The goal was to improve accept rates by helping applicants see value beyond the welcome offer—all within an aggressive timeline of one week for design and one sprint for development.
Old Mobile Design
Focus on welcome offer, minimal UI
Approach
For an A/B test, led a team of three designers to transform the acceptance experience, shifting the focus from the welcome offer alone to the card’s overall value while improving personalization and conversion.
- Reframed card value: Added total card value information to highlight the broader benefits of the card, particularly for applicants who received a lower welcome offer.
- Elevated the experience: Refined typography, copy, UI and emphasized Platinum Card artwork to make acceptance feel like a celebratory milestone.
- Kept actions accessible: Introduced a sticky footer in mobile with persistent access to the primary actions—accepting the card or withdrawing—without requiring applicants to scroll.
New Mobile Design
Elevated visual design, copy, core actions, and total card value
AI
Used Replit, Figma Make, and chatGPT to explore the problem space and iterate on solutions.
Execution
Conducted design QA and accessibility audits (automated, guided, and manual testing) throughout engineering development to identify issues early and enable faster remediation, followed by final audits before launch to validate implementation quality and accessibility.
Results
+$12.5M revenue
Progress Bar
Problem
The existing experience used a segmented progress tracker on desktop, while mobile displayed step labels for steps 1–3 on pre-submit pages only. Neither provided consistent progress visibility throughout the application.
My previously designed dynamic, labeled progress bar had driven $30M in incremental revenue across other markets. However, there was still an opportunity to further improve submit rates by optimizing how progress was communicated during a long application journey.
Old Desktop
Old Mobile
Approach
I explored two progress indicators that maintained consistent behavior across desktop and mobile while balancing progress visibility with user motivation.
- Labeled progress bar: Built on the previously validated dynamic design, which adjusted to conditional steps and varying application paths.
- Dot-based progress indicator: Removed the visible progress label to avoid discouraging applicants with seemingly small progress early in a long journey (e.g., 8%). A defined endpoint maintained a clear visual boundary, helping users distinguish the end of the bar in line with color-contrast accessibility requirements.
Variant B
Proven +$30M incremental revenue in 5 other markets
Variant C
Newest variant
A/B/C Test
Launched an A/B/C test comparing both designs against the existing experience, measuring their impact on application submit rate.
Results
Variant C won, generating $12.6M in incremental revenue.
- Merged new component into acquisition design system.
- Subsequently tested the design in the post-approval commercial application journey, generating $162M in spend impact and $5.6M in revenue impact.
The dot-based tracker provided enough progress visibility to keep applicants engaged without discouraging them with small percentages (e.g., 8%) early in a long application journey.
Soft Pull Messaging Test
Problem
Soft-pull reassurance was essential to driving application submits, but the lengthy messaging took up valuable screen space, particularly on mobile. Previous testing showed that reducing its visibility negatively impacted submit rates, creating an opportunity to simplify the page without sacrificing reassurance.
Old Mobile Header
Contains soft pull with no credit impact and how the process works.
Approach
I streamlined the page-one header to reduce clutter and bring more of the application form above the fold while keeping soft-pull reassurance prominent.
- Removed the redundant “Here’s how” section.
- Prioritized soft-pull reassurance.
- Increased form visibility to help applicants get started sooner.
New Header
Removed the here’s how it works section so the form is visible above the fold.
A/B Test
Tested the simplified header against the existing experience to measure its impact on application submit rate.
Results
$2.5M incremental revenue
Name on Card
Challenge
The prefilled name-on-card field had a disabled last-name input and truncated characters, triggering rage clicks. Card customization also felt out of place before applicants had been approved.
Previous Name on Card
You can you this place to talk a little bit about the image on the side.
Approach
Removed the field to streamline the first page friction and better align the application flow with user expectations.
Removed Name on Card
Results
+$22.3M incremental revenue
Accessibility & Quality at Scale
As a certified accessibility tester, established the CXO program’s first formal accessibility audit requirement, continuous testing process, and engineering collaboration workflow.
- Accessibility: Conducted 10 automated, guided, and manual audits, documenting 400+ issues using axe and provided solution recommendations.
- Issue tracking: Established Jira-based tracking for product and engineering teams to manage and resolve accessibility issues.
- Design QA: Implemented the QA process I previously developed and presented at the 2024 global Amex Design Conference.
Enterprise AI Design Systems
Innovating at Scale
Challenge
As generative AI created new opportunities for enterprise products, our design system needed to evolve to support emerging interaction patterns at scale. We needed to identify gaps in the existing component library, determine where entirely new AI components or AI-specific variants would deliver the most value, and establish reusable patterns to support future enterprise experiences.
My Role
Established and led a 21-person, cross-functional initiative, recruiting designers, researchers, content designers, accessibility specialists, and engineers to explore opportunities for AI-driven design system innovation.
The Approach
1. Audited the Existing Design System
Evaluated the existing component library for web and native to identify gaps, opportunities for net-new AI components, and opportunities to extend existing components with AI-specific variants.
2. Prioritized High-Impact Opportunities
Ranked opportunities by potential business impact and breadth of use across enterprise products, focusing the initiative on reusable solutions with the greatest potential value.
3. Organized Cross-Functional Collaboration
Established weekly workshops for the 21-person initiative, bringing together designers, researchers, content designers, accessibility specialists, and engineers. Structured sessions around collaborative ideation and assigned follow-up research and exploration between workshops.
4. Researched and Prototyped AI Solutions
Led market research, requirements gathering, and brainstorming to explore emerging AI interaction patterns. Guided iterative solution exploration using tools such as Replit and Figma Make to evaluate concepts and refine potential component patterns.
Results
Informed an enterprise generative AI chat interface supported by 35 new design system subcomponents.
The resulting components are documented in the enterprise design system with usage guidance, establishing a reusable foundation for future product adoption.