I led the project end-to-end, owning design direction, scope prioritisation, user research, daily dev communication, branding and product strategy. I worked closely with the PM and engineers, aligning business expectations with design and product strategy.
I co-shaped the economy where Creators, Subscribers and Speculators each got something the others couldn't, designed every flow across the mobile-first PWA, ran user interviews and cross-checked with on-chain and Farcaster signal to ground decisions in real usage patterns. I built the brand from scratch and made the calls on what to ship, what to kill, and what to leave for the next iteration.
Back in 2024, DEGEN became a thing on Farcaster. Users could tip each other in DEGEN directly in posts and comments using simple commands, which let people monetise content and engagement peer-to-peer without leaving the social app. It boosted Farcaster's popularity and engagement alongside Frames (mini apps embedded within social posts), and AlfaFrens launched at exactly the right moment to hop on the hype train and make SocialFi more fun to use.
Who we were building for
Three audiences had to coexist on the same platform without one cannibalising the others. The product mechanics had to make each group earn or get something the others couldn't.
Creators
Content creators who want to earn money through a gated community.
Subscribers
Users who want to digest content from niche creators they trust.
Speculators
People who strategically subscribe and stake tokens to earn yield.
Distribution: we shipped a PWA
We rolled out AlfaFrens as a Progressive Web App. We wanted the feel of a mobile app, but because it was crypto-based speculation we wouldn't have been able to get it published on the App Store. The PWA approach also let us ship a desktop version, which turned out to be frequently used too.
People don't want yet another social app.
Competing head-on with established social platforms was not an option. Countless platforms had promised to decentralise content but never got it right. So how could we get it right? By focusing on one problem we could actually solve: payments and creator incentives.
Friend Tech proved it could be done
Friend.tech showed the appetite for monetised social access. But we wanted to lean further toward content and community, rather than focusing on complex pricing curves and speculation mechanics.
Streaming tokens. Pay-per-second subscriptions.
The main objective was an app that incentivises creators for good content and lets subscribers earn on top of getting access to curated chats.
Superfluid's streaming capabilities were ideal: subscribers paid for as long as they wanted to be subscribed, second by second. If a creator went silent or the content got dull, users could unsubscribe immediately, without being locked in for the rest of a month.
Users could pay gas in DEGEN (later in ETH, we used Sparks as a unit) thanks to Biconomy's paymaster, which turned out to be the way to onboard the Farcaster crowd: native to their culture, frictionless on UX.
How could we design a SocialFi app that solves one problem really well, instead of trying to rebuild the whole social stack?
Spend to subscribe. Earn $AF tokens. Stake to earn.
- 01
Log in with Farcaster
Single-tap social login using the user's existing Farcaster identity.
- 02
Subscribe to creators
Pay-per-second stream. Users can cancel anytime, the cost stops the moment they do.
- 03
Get access and earn $AF tokens
Curated channel access plus the ecosystem token as a reward for participation.
- 04
Stake $AF tokens on any channel
Users pick a creator they believe in and back them with their stake.
- 05
Earn cashback
The bigger the user's stake on a popular channel, the bigger their slice of the rewards pie.
Revenue split
Creators earn 30% of all subscriptions on their channel. Stakers earn the remaining 70%, proportional to their stake. Everyone could launch a channel, not just curated creators.
The aggressive split was deliberate, and it was one of our core strategies to pull subscribers into the app in the first place. We made a bet that creators would be fine with their 30%, because their content already revolved around crypto and they were degens themselves, wired to value the upside over a bigger guaranteed cut.
On top of that, the creators were also subscribers to other people's channels, and they earned $AF tokens they could stake on those channels to receive cashback. So the 30% headline number understated what creators actually walked away with: their direct subscription revenue plus a share of the cashback economy on the rest of the platform.
Beyond the high-level mechanism, each persona had their own journey through the product. I designed five core flows that worked together so creators, subscribers and speculators could all win without stepping on each other.
Subscribe to access chat
The subscriber flow. From browsing channels to landing inside a chat the moment the per-second stream starts. The whole flow is built so the user understands they're paying continuously, not in a lump.
Speculate on channel price and $AF rewards
The speculator flow. Their goal is finding the cheapest channels that pay out the most $AF tokens. We designed the filtering so each speculator could land on their own strategy without us spelling one out. Speculators routinely subscribe and unsubscribe as $AF earnings on a channel rise and fall, so the UI was built to make switching cheap. The alternative was losing them to off-app scripts hunting the same arbitrage, so we kept them in the product. From there they could stake $AF on the channels they backed to claim a bigger cashback slice from the top earners.
Create content and share with the audience
The content creator flow. Creators launch a channel in a few steps, share the channel link with their community, send messages, and watch the revenue grow in real time. Creators could also participate in staking, so if they wanted a bigger slice of their channel's revenue, they had to fight for cashback alongside the speculators.
Scarcity, vitality, presence.
Timing was everything. The app was ready, the brand was weird, and all hands were on board. To get the hype around AlfaFrens we used freshly-baked Frames to distribute access codes: like and share to spin the wheel and see if they won an invite.
It worked for some, for many it didn't, but we hit huge virality on the platform and everyone was talking about us. On top of that, every user with access could give 5 invites, which made them come up with their own distribution strategy within their friend groups and communities.
Every team member ran their own channel and engaged with the community directly. Farcaster presence was the priority. The first weeks were everything: we listened to feedback, reacted to FUD, ran fun events (e.g. best NFT contest) where users could win $AF tokens.
It took Farcaster by storm.
Once the original hype settled, we shifted from launch mode to listen and refine. I interviewed a dozen early AlfaFrens users, split into three profiles: Content Creators, Subscribers and Speculators. The objective was to understand how they actually used the app, what they were unknowingly looking for, and where the friction points lived.
Going in, I was worried that despite our carefully planned and reviewed flows, the app would still be hard to navigate because the economy underneath it was so complex. It turned out the feedback was extremely positive: users of every background understood the basic premise very well. What helped us plan next steps was the fact that all users were very vocal about what was missing. I cross-checked the interview results with signals we were receiving from other sources such as Farcaster, and they closely matched.
Results
- 01
Unidirectional chat
We launched AlfaFrens with unidirectional chats: creators saw all messages, but subscribers only saw the ones the creator publicly cited. What we worried might be too restrictive turned out to be one of the community's strongest likes, so we kept the mechanism to prevent spam and make chats feel curated instead of chaotic. - 02
Reactions and mute
Both creators and subscribers wanted emoji reactions to make the chat feel alive. Mute gave creators control without having to ban. - 03
Withdrawals
Disabled in the first weeks, and users were fine with it (!). As the hype calmed down and more withdraw requests followed, we rolled the feature out so people could cash out earnings. - 04
Additional rewards for invites
Active users could invite friends and earn an extra share of $AF tokens from their invitees' on-platform activity, turning every committed user into a distribution channel. - 05
Channel categories
As the platform grew, channels were hard to navigate. We grouped them into categories so users could find what mattered to them without scrolling the whole directory. - 06
Migration from DEGEN to ETH
Gas spiked from our volume and DEGEN's volatility made costs unpredictable. We migrated the gas and subscription token from DEGEN to ETH for more stable economics. - 07
Quality and Activity score
We added a per-channel Quality and Activity score to keep the app from being flooded with low-quality content and engagement-farming speculators.
What we killed
We added a Trading Feed inside channels alongside chat access, hoping it would deepen engagement. Users were not interested at all. We pulled it.
What we wrestled with
- 01
Explore tab recommendation logic
Big channels grew bigger, small ones were impossible to find. We reworked the reward calculation to even the odds for emerging creators. - 02
AlfaFrens broke Base
Gas spiked from our volume. We had to scramble to optimise on the chain side and rework throughput before the network caught up with us.
Am I ugly?
The mechanics and the strategy pushed me toward digital brutalism: bold colours, weird illustrations, collages. It was supposed to look fun and daring. I figured people would either love it or hate it. They loved it.
On top of that, designing an icon for a SocialFi app that could carry the vibe was a real challenge, but eventually SpikyBoi was born.
AlfaFrens was a crazy ride with a team of highly motivated professionals who found fun in the challenge and brought a new payment standard to SocialFi.
Superfluid streaming tech was put through its biggest consumer stress test to date. Around 38,000 new wallets were introduced to the Superfluid ecosystem in the process. SocialFi got a payment primitive that actually felt native to the medium.
