Yellowcard

Yellowcard is Africa’s largest regulated crypto exchange, operating across 20+ markets with over two million customers. I was the sole designer on the consumer product, owning everything from onboarding flows to the design system across iOS, Android, web, and iPad.

This case study covers a specific problem we discovered after launch: we had two million customers, but most of them weren’t doing anything. Getting them to actually use the product became the most important design challenge I worked on there. The nudge system that came out of it moved deposit conversion 44% and buy conversion 82%.

Two million people had signed up. Most of them were sitting still. They had accounts but hadn’t deposited. They had deposited but hadn’t traded. They had traded once and never came back. The product worked, the numbers just weren’t moving.

We ran a survey to understand why. The findings were pretty clear: 85% of our customers hadn’t finished KYC verification. Most hadn’t made a first trade. And only 14% actually considered themselves crypto-savvy. The rest were curious but lost. In AAARRR terms, we had acquisition handled. Activation was completely broken.

People were avoiding the product not because they didn’t trust it, but because they didn’t know where to start. Crypto has a language problem. Wallets, gas, KYC, stablecoins: none of it maps to anything familiar. We were dropping people into a foreign country and expecting them to find their own way.

The opportunity was education, delivered in context, at the exact moment someone needed it, not more features. If we could meet customers where they were and teach them one thing at a time, we could move them from passive signups to active traders.

The research gave us a long list of possible interventions. We needed to cut it down fast. We ran an effort/value mapping exercise to figure out what would actually move the metrics without requiring months of engineering work.

The things that rose to the top were all about guidance: helping new customers understand what to do next, surfacing the right information at the right time, and making the product feel less like a trading desk and more like something approachable.

The solution came out of an internal hackathon. We called it the Nudge Project: a system of contextual, in-app education that surfaces at the right moment instead of overwhelming new users upfront. Actual guidance woven into the product, not a tutorial or an onboarding checklist.

The idea was to meet customers at different stages of their journey and give them one clear next step. Somebody who just signed up needs different guidance than somebody who deposited but never traded. The nudges had to be specific to where a user actually was.

We redesigned the entire first-time experience for what we called Tier 0 customers: brand new, never traded, probably intimidated. The old flow dropped you into a full home screen with no direction. The new one gave you a clear set of steps to complete before anything else.

The home page architecture changed to lead with those steps. Walkthrough the home page, connect a wallet, complete KYC, make a first trade. One thing at a time, with enough context that you understood why each step mattered before doing it.

The home screen is where most people land and then leave. We restructured it to be more useful for someone who doesn’t know what they’re doing yet. Instead of showing everything at once, we prioritized what was relevant to where a customer was in their journey.

New users saw their progress steps front and center. Returning users who hadn’t traded saw a nudge toward buying. People who had made trades but gone quiet saw suggested articles and a persistent prompt to explore more of the product.

The nudge screen itself was designed to feel like a helpful suggestion, not a notification or an upsell. It shows up in context, explains one concept clearly, and gives the customer a specific action to take. Copy mattered a lot here. We tested and revised the language on almost every card.

The quick insights list let customers browse educational content at their own pace, without it feeling like homework. Articles linked out to deeper explanations when someone wanted more. The goal was always the same: reduce the distance between “I don’t know what to do” and “I just did it.”

The nudge system shipped and the numbers moved in a real way. Deposit conversion hit 44%. Buy conversion hit 82%. KYC completion rate went up to 30%. These were customers who had been sitting in the product doing nothing for months.

The bigger takeaway was that the drop-off wasn’t about trust or fees or the product being hard to use. It was about people not knowing what they were supposed to do next. Once we answered that question in the product, they moved.

The nudge project taught me that activation is a design problem before it’s a marketing problem. Every passive user is a question the product didn’t answer. Before you try to bring people back from outside, check what the product itself is failing to tell them.

I’d also push for the effort/value mapping exercise earlier in any discovery process. It compressed weeks of debate into a single session and gave the team a shared view of what was worth building. That alignment is more valuable than most deliverables.