Catch Social

Role: UX/UI Designer

Core Responsibilities: UX/UI Design | Content Design | Design Systems

Overview

Ghost was an encrypted wallet-to-wallet messenger that let people in crypto communities talk to each other without tying those conversations back to a real-world identity.

I worked alongside lead designer Yasir Eryilmaz, who owned the visual creative direction of the app: the living gradient that ran throughout it, the core illustrations and logo, and the onboarding and wallet views. I owned the messaging experience end to end, which covered the chat interface and interactions, reactions and replies, sticker integration, the notification system, empty states, and in-app purchase flows. I also built and maintained the Ghost design system and wrote the UX copy across the app.

Getting up to speed in an unfamiliar space

I came into this project with zero background in crypto, and in an agile startup that meant getting fluent fast. I leaned heavily on research to understand who our users actually were and where the real opportunities in the product lay, and much of what I found ran counter to what I would have assumed going in.

Most product design starts from the idea that trust comes from identity, so you reach for verified badges, real names, and complete profiles. Our users wanted the opposite of that. Anonymity wasn't something to design around; it was the whole reason they were there, and a lot of my work came down to figuring out which conventions I needed to throw out.

Conceding on verification

There was no way to truly verify that someone owned the .eth account they claimed in the app, and we struggled with this a lot. We could have designed around it with a verification flow, or gated features, or some kind of trust signal, but we decided not to. 

Any real verification would have meant linking a wallet to something outside of it, which is exactly what our users were on the app to avoid. Solving for impersonation would have broken the promise the product was built on, so we accepted the risk as the cost of the thing that made Ghost worth using in the first place.

The same question came up again when we started talking about X integration. We had a lot of conversations about how to connect accounts without opening up an easy path to impersonating well-known people in the space, and we landed in the same place: light and loose, without restrictions, much closer to how the DeFi space actually operates than to how a mainstream social app would handle it.

It wasn't a comfortable call to make. App Store constraints made this territory even harder to navigate, and I personally think it's part of why so many crypto apps from that period didn't survive. But designing something safer would have meant designing something our users didn't actually want.

Navigating notifications

End-to-end encryption usually means the server can't read a message, so it can't put anything useful into a push notification either. We solved this with an account layer, where users created an account that was only ever associated with their .eth account. That gave us something to deliver notifications against without exposing any message content or asking anyone for personally identifying information.

Building a design system for more than the design team

Two designers doesn't obviously justify the overhead of a design system, but we weren't the only people relying on it.

Yasir set the visual direction and my job was turning that into something the whole company could build from. Engineering worked out of it, marketing worked out of it, and it had to hold up across web, iOS and Android, the browser extension, and our marketing materials all at once. Once it became the shared reference point it kept the team organized, honest, and agile, which is what made the ongoing effort of maintaining it worth it.

Translating mobile into the browser extension

The extension was the hardest craft problem on the project. It had a maximized and a minimized state, and the minimized state needed to align with the mobile view, so most of the work was organizing the logic behind which screens should appear alongside one another, which ones needed to collapse, and what governed the movement between them.

What I'd do differently

The feature I most wanted was sticker integration with FrenMoji, where users could use their bespoke FrenMoji NFT as their profile pictures and their NFTs directly in chat. We had a lot of fun ideas for it, but it never made it into the shipped app because of time constraints and waning market interest in NFTs.

That's a smaller piece of something bigger I'd change. Looking back, I would have really pushed to ship faster and iterate while the app was being used. We struggled to meet the moment, and in crypto the moment is short. We spent time working toward a fully polished release when we could have been shipping into an app that people were already using and enjoying, and I'd trade some of that polish for the chance to iterate against real usage.

More Info

Contact me to request more information about this project.