Fintech · Digital Banking
Currency Gram
A money transfer app that worked technically but failed emotionally, redesigned across mobile and web so sending money feels less like math and more like a conversation.
- Role
- Lead Product Designer
- Timeline
- 12 weeks · 2025
- Year
- 2025
- Industry
- Fintech · Digital Banking
The problem
The app worked, but it didn't feel human. Users completed transfers, but barely. Drop-off was high, support tickets kept climbing, and the gifting feature, one of the product's best ideas, was almost invisible.
My role
I was the Lead Product Designer. Over twelve weeks I redesigned the end-to-end experience across mobile and web, rebuilding the transfer flow, surfacing trust at the right moments, and giving beneficiaries real context.
The outcome
Sending money became faster and calmer. Transfers complete quicker, saved beneficiaries get reused far more often, and satisfaction climbed from passive to promoter in a single release.
-38%
Transfer completion time, from 6.4 down to 3.9 minutes
2.4x
Increase in beneficiary reuse, with 44% fewer re-entry errors
+41 NPS
Net Promoter Score, moving from 18 to 59
Challenge
The drop-off wasn't a flow problem. It was a trust problem.
After digging through session recordings, support ticket patterns, and interviews with freelancers and everyday users, the same thing kept surfacing. People weren't leaving because the app was broken. They were leaving because it never made them feel safe enough to continue. Fees showed up after they had committed. Transfer status vanished into a vague 'processing' state. And the beneficiary system, meant to save time, felt like starting over every single time.
Process
Three weeks of questions before a single screen.
Before I opened Figma, I spent three weeks mapping the problem. I analyzed session recordings, ran interviews, audited support tickets, and tore down Wise, Revolut, and Payoneer to see how they earn trust. That led to a signed problem brief and a ranked list of trust failure points. From there I sketched eight flow concepts, shortlisted two, gut-checked them with users, then built a full design system and ran two rounds of moderated testing, logging fourteen iterations before handoff.
Solution
Lead with transparency, not fine print.
The redesign made four structural shifts. Fee disclosure moved up to the amount entry screen, before any commitment, so the full cost is always clear. The transfer flow compressed from seven steps to three, built around one mental model: who, how much, done. A live status layer replaced 'processing' with real-time progress, estimated arrival, and plain language for every state. And beneficiary profiles were rebuilt to surface context, the last transfer, the preferred method, a quick-send shortcut, turning a cold list into something that feels like a relationship. It did not add features. It removed every reason to hesitate.