When Capital One acquired Discover in a $35.3B deal, every debit card had to move networks. I designed the servicing experience that made a high-stakes migration feel seamless for millions of cardholders.
Customers didn't ask for this. An unexpected replacement card created doubt about a basic need. Access to their money: recurring payments, digital wallets, merchant acceptance, travel and military timing, even duplicate cards from loss or theft.
The business needed it invisible. Moving debit onto an owned network was strategically critical. But only if the experience stayed low-effort, holding down call volume, escalations and manual servicing at scale.
The happy path was easy to design. The real work was everything that happens when customers don't follow it.
The goal was always self-service in-app. Customers activate their card, update recurring payments, done. No call needed. But millions of real people hit edge cases, and those calls had to work flawlessly as a backup.
Self-service first: A tracker shows cardholders where their card is: preparing, shipped, delivered, activated. Contextual help and no surprises. Phone support second: When they call in, associates see the customer's conversion state and have a playbook for every path: lost cards, early-issuance requests, travel holds, payment failures.
I designed both paths: the self-service journey that keeps most customers happy, and the associate servicing tool that catches everyone else.
Built a timeline customers check in-app: where's my card, why hasn't it arrived, what do I do next? Each stage branches into contextual help, reducing calls by showing shipping details, address confirmation and next steps.
Enhanced the associate servicing tool to read where a customer sat in the conversion. Prevented duplicate cards and a second round of activation and payment updates.
Introduced a one-time, risk-constrained way to briefly reactivate the old card. For travel, military service or disasters, so access never lapsed.
Live in production and launched incrementally, proof that the edge cases I designed for are actively running at scale.
The strongest moves weren't screens. They were decisions about scope, risk and proof.
Seeing an early-issuance request as a replacement avoided a one-off flow. Cut tech debt and associate cognitive load.
Standardized how cards are named: network, last four, ATM vs. debit. Customers and associates heard the same thing everywhere.
Under strict confidentiality, I pressure-tested flows weekly with Bank Voice and ran Legal, Risk and Compliance reviews in place of external validation.
This work is covered by confidentiality. Happy to walk through the screens, decisions and outcomes behind them.
Let's chat