Sector · SocialFi · Your graph, your rails

SocialFi

600KFarcaster IDs
Fiatjaf · 2020Nostr launched
SufficientDecentralisation

The social graph as a public good rather than a private moat. Identity, content, and following relationships that travel with the user across applications. Whatever app you switch to, your network and history come with you.

Sufficient decentralization means a developer can build an app that competes with us, using only the protocol, and we can't stop them. Dan Romero · Farcaster docs, 2023
descend

SocialFi is the bet that protocols beat platforms when the network effect can be ported. Dan Romero and Varun Srinivasan — both Coinbase alumni — launched Farcaster in 2021 with the sufficient-decentralization design: identity on-chain, casts in hubs, anyone can build a client. Warpcast is the reference client. Frames in early 2024 turned casts into mini-apps and made Farcaster the first social protocol that shipped genuine native interaction. The user count is small and the signal-to-noise is the highest in crypto social.

Lens (Stani Kulechov's other project, originally on Polygon, now on its own L2) took the opposite approach — full on-chain social graph as NFTs, Profile NFTs and Follow NFTs as first-class primitives. Hey is the canonical client. Then there's Nostr — Fiatjaf's protocol from 2020, relay-based, no chain, no token, growing fastest in the Bitcoin-native crowd because Jack Dorsey funded it after leaving Twitter. World (formerly Worldcoin) added proof-of-personhood via the orb, which is either dystopian biometrics or the only credible bot-resistance layer, depending on who you ask.

The Friend.tech moment in August 2023 is the cultural reference. Racer's app turned X profiles into bonding-curve-priced shares, did $50M in fees in three weeks, and collapsed inside six months. The lesson the survivors took: financialization of social capital works as a stunt and breaks as a substrate. The protocols that lasted (Farcaster, Lens, Nostr) chose composability and identity over speculation. The next round of SocialFi will be agents posting alongside humans on the same graph.

Your graph survives the platform you got it on.
The seminal text

The founding document.

Sufficient Decentralisation for Social Networks
Varun Srinivasan · Farcaster · 2022
A sufficiently decentralised network is one where two users can find each other and communicate, even if the rest of the network wants to prevent it. Identity, the social graph, and the messages themselves must each be portable across clients without permission. Anything less is a platform pretending to be a protocol.
Sufficient decentralisation →
Decentralized social, finally legible

Six attempts at owning the social graph.

Social is the hardest application for decentralization because the network effect is the product. Farcaster bet on sufficient decentralization with hubs. Lens bet on NFT-based composability. Nostr bet on relays and minimal protocol. Friend.tech bet on bonding curves and learned what happens when financialization eats sociality. World ID added the proof-of-personhood layer the others need. None has won. All of them are more interesting than the centralized incumbents pretend.

Sufficient decentralization
Farcaster
2021

Dan Romero and Varun Srinivasan's Farcaster takes the 'sufficiently decentralized' approach — accounts (FIDs) live on-chain on Optimism, but messages flow through hubs (currently mostly Snapchain, the migration target from the original Hubble nodes). Frames (January 2024) turned posts into interactive mini-apps — mint NFTs, vote, transact, all inside the cast. Warpcast was the reference client; clients can read the same data. The bet: enough decentralization that no single party can ban you, simple enough that the UX competes with web2 social.

NFT social graph
Lens Protocol
2022

Aave's Stani Kulechov spun up Lens (Polygon, May 2022) with an NFT-based primitive: every profile is an NFT, every post is an NFT, every follow is an NFT. The thesis was composability — DeFi protocols could read social graphs, social actions could trigger smart contracts. Lens V2 simplified the model; V3 (2024) shipped on its own Layer 3. Hey (formerly Lenster) is the canonical client. Different bet from Farcaster: heavier on-chain, more composable, fewer users. Same animating question — who owns the social graph.

Relay protocol
Nostr
2020

fiatjaf's Nostr is the most minimal of the three — a protocol, not a platform. Identity is a secp256k1 keypair (the same Bitcoin uses). Messages are signed JSON events broadcast to relays. NIPs (Nostr Implementation Possibilities) extend functionality without coordination. NIP-57 zaps wire Lightning payments into posts. Damus (iOS), Amethyst (Android), Primal — all read the same events. The cost: no recovery if you lose your nsec, and discovery is harder than centralized feeds. The aesthetic: closer to email than to Twitter.

Bonding curves / failed
Friend.tech (lessons)
2023

Friend.tech (Base, August 2023) put a bonding curve on access to a person's chat — buy 'keys' on a creator, get into their group chat, sell keys at a higher price as more buyers arrive. Volume exploded for weeks. The protocol generated tens of millions in fees. By mid-2024 it was effectively dead — financialized attention has a half-life shorter than the curve's mathematical lifetime. The token launch (FRIEND) accelerated the wind-down. The lesson is structural: not all social interactions tolerate being priced.

Proof-of-personhood
World ID
2023

World ID is the credential layer for sybil-resistant social. The Orb scans an iris, hashes locally, issues a verifiable credential. Apps integrate via SDK to gate features behind 'one human, one action.' Worldcoin's pitch to socialfi: spam, bots, and AI-generated accounts kill open networks; PoP is the missing primitive. The integration work matters more than the token — Discord, Reddit-style apps, and emerging social networks gain a deduplication primitive that doesn't require KYC. Privacy critique is real but the design is more careful than headlines suggest.

Multi-protocol clients
Phaver / Hey clients
2022-2023

Phaver (Lens-native, mobile-first, multi-chain) and Hey (Lens web client, formerly Lenster) showed that protocol-and-client separation actually ships. A user picks a client, the client reads the same on-chain (or relay) data, switching costs are low. Phaver added support for multiple social protocols inside one app — Lens, Farcaster, others — a meta-aggregator on the decentralized graph. The pattern matters: in centralized social, the client and the protocol are the same company. Here they're not, and the user experience finally caught up to the architecture.

Working set

Projects we actually watch.

Conviction is stated as conviction; you decide what to do with it. Tiers below — Core, Conviction, Watch, Speculative — reflect how much of FRQNCY's attention each project currently earns, not a recommendation to buy.

FAFarcasterNONostrFriend.techFRIENDFriend.techHEHeyLELens Protocol
29.04 Farcaster monthly active casters cross 200K. Frames v2 launches with native swap. Lens V3 surpasses 5M profiles. Worldcoin orb count past 8M. The protocols are moving from 'experiment' to 'distribution channel.' desk
FA
Farcaster
· Optimism
conviction
SocialFiDecentralized Social
A sufficiently decentralized social protocol. User-owned identity, portable social graphs, and an open feed anyone can build clients on — Warpcast is one of many.
Why FRQNCY watches thisFarcaster is the credible alternative to platform lock-in. If social networks belong to users, their graph should travel with them.
NO
Nostr
· Protocol
conviction
SocialFiDecentralized Social
Notes and Other Stuff Transmitted by Relays. A simple open protocol where keys are identity and anyone can run a relay — no chain, no token, no platform.
Why FRQNCY watches thisNostr is the minimal viable decentralized social layer. The simplicity is the point — it survives platforms because it doesn't depend on one.
Friend.tech
Friend.tech
FRIEND
watch
no website Crypto
HE
Hey
watch
no website Crypto
LE
Lens Protocol
watch
no website Crypto
PH
Phaver
watch
no website Crypto
WA
Warpcast
watch
no website Crypto
Worldcoin
Worldcoin
WLD · Ethereum
speculative
IdentityAI
A global identity and financial network using biometric proof-of-personhood to distinguish humans from AI online.
Why FRQNCY watches thisAs AI grows, proving you're human becomes essential. Worldcoin asks how we preserve human agency in an age of intelligent machines.
A practice

Five small things, repeated.

Conviction is theatre without practice. Five steps that turn the thesis above into something the body actually does, not just something the mind agrees with.

i
Get a Farcaster account and post for 30 days.

Use Warpcast. Engage with at least 5 channels. The signal-to-noise is the lesson.

ii
Run a Nostr relay or use multiple clients.

Damus, Amethyst, Iris. Notice that your identity (npub) survives every client switch. That's the point.

iii
Build a Frame.

Farcaster Frames are HTML-ish. Ship one in an afternoon. Ten million wallets can interact with it.

iv
Verify your humanity at a World orb.

Then read the criticism. Form your own view on biometric proof-of-personhood with the device in your hand.

v
Bridge your social graph.

Export Farcaster follows, import to Lens, post the same content. Compare engagement. The portability is the alpha.

Two doors. Pick one.

The Crypto hub is the index of all sectors and the freedom-technology frame they share. The Fund is what happens when the same conviction gets put to work on behalf of the network.

The graph belongs to you, not the platform.
Identity is a primitive. Posts are payloads. Clients are interchangeable.
Sufficient decentralization is a competitive moat against your own founders.

Own the protocol.

FRQNCY · Crypto