# Farcaster v/s Lens and Farcaster Deep Dive.

==An other unfinished deep dive==

## Lens vs Farcaster.
Social consumer applications like FaceBook, Twitter, TikTok and many others have seen a meteoritic growth in the last decade. This growth can be attributed to the increased internet access to people around the globe. Their growth especially leads to what is called the creator economy while also acting as a medium serving real-time information.

The growth of social consumer applications gave the parent entities of the application the superpower to influence people across countries, age, and race… along with tracking data connected to their users. With the superpower in hand, they started acting in favour of a few entities and started censoring content, individuals and organisations either by banning or shadow banning(tuning recommendation algorithms to less appear the posts of individual/organisation XYZ) while promoting other individuals and organisations or content of their favour. There are a few incidents, social consumer applications directly interfered with the likes of National Elections in various countries.

Banning on social media platforms results in a complete loss of the social value(followers, posts, activity) an individual or entity has earned on the platform. The loss of platform social value is only possible because platforms possess all the control over user's accounts and data.

## Key properties for a killer social (media) app:
Every social media application that has seen widespread growth has their unique value proposition in the category they offer. Facebook started as a platform to share posts and pictures. Instagram, serves for images. Twitter for micro blogging. Reddit for communities. Youtube for video content. TikTok for short video content.

The growth of social media apps has also been affected by the recommendation algorithms used by them, to cater content to users upon landing.

These apps cater to millions of users handling loads of network activity and volume efficiently due their centralised architecture, that drives them to take decisions that affect maximising the revenue and performance. Centralisation in deciding on infrastructure is useful, as decisions can be made quickly and infra can be set up as fast as required.

## What farcaster and lens are aiming to solve?
Both farcaster and lens protocol are aiming to solve the issue of centralisation in the space by building permissionless social consumer applications that are sufficiently decentralised while scaling to masses and handling millions of messages a day. i.e. scalable decentralised social with account and data ownership back at users.

Both Farcaster and lens protocol allows users to create accounts/user handles that can be used across their apps permissionlessly. Anyone can create apps. Apps can completely set their own rules for posts and users on them and be community specific(like tech, art, social…) or agnostic(general purpose). A good example is the blogging client t2.world on lens protocol and anonymous casting client sealcaster on farcaster.

Warp cast is the most popular client(connected app) among all on farcaster. Being a sole leader in farcaster apps gives an advantage to farcaster to have client based activity aggregation per category. Farcaster hit its mainnet in August 2023, far later than Lens protocol’s February 2022.

## Team behind Farcaster:
Farcaster is co-founded by Dan Romero and Varun Srinivasan.

Dan Romero, worked as vice president at Coinbase before cofounding Farcaster.
Varun Srinivasan, before farcaster worked at Coinbase as senior director and built soundfocus to help people with hearing loss.

Farcaster has a team of 14(12 Full time + 2 Contractors) working remotely in US Times.

## Architecture of Farcaster:


Farcaster have both onchain and offchain components. Onchain components include user name resolver, storage registry, key registry and Id registry. Majority of Farcaster’s activity is handled by the offchain components.

ENS name resolver is used for user names on farcaster, farcaster username can be linked with ENS permissionlessly, or use warpcast to generate one. Warpcast usernames availability is centralised.

Users rent storage on farcaster for storing their casts, likes, followers onchain. And key registry is used to manage keys for various apps.

Farcaster’s offchain components is a peer-to-peer network of servers called hubs, hubs are similar to nodes on blockchain. They store user data and gossip recent messages to peers in the network. Hubs can join and leave farcaster network permissionlessly, without any stake. Hubs monitor peers and score peer behaviour. If a hub violates norms by gossiping excessively, does not accept valid messages or falls back from sync, they are ignored by its peers.

Gossip is based on libp2p, and is lossy. Hubs periodically selects a random peer and performs a diff sync and adds the lost messages.

Farcaster hubs do not receive any incentives to be run. All the fees collected for storage units goes to the protocol. Farcaster Hubs can be viewed as nodes of a distributed system following a leaderless model, collaborating to state changes occurring on all nodes at an instant via gossiping and diff sync, along with being permissionless so anyone can join and leave the Hub network.

With the mix of offchain components for performance critical tasks and onchain components for identity critical data, farcaster stands as a base layer for sufficiently decentralised social consumer application(s), especially social media.

## Warp Cast:

Apps on farcaster are categorised into Wallet apps, and Connected apps.

Wallet apps are used to create signing up, adding connected apps, posting and browsing messages. Connected apps are only used for posting and browsing messages.

Warp Cast is the major Wallet app on farcaster built by the farcaster team, with features similar to twitter and Reddit.

## Notable innovations on Farcaster/Warp Cast:

- Farcaster has a unique value proposition of data ownership for users. Users rent storage units for their casts, reactions, links, profile data, verifications.
- Farcaster frames were introduced early this year, allowing users to perform actions straight from warpcast or any farcaster app, such as registering for an event, buying an NFT and … This is similar to iFrames concepts in Browsers.
- Direct Casts on farcaster allow direct messages between individual accounts.
- Channels are an experimental feature on farcaster, similar to subreddits.
- Sign in with Farcaster, lets users sign in with their own farcaster handles and lets apps get necessary data about users and decide on the app flow. This is similar to SSO by Google, Facebook, Apple. But Sign in with Farcaster, being permissionless and having a stable API, stands out from other SSO offerings.

## Anticipated Growth and Challenges ahead:

With 600K+ accounts registered on farcaster, and avg DAU of \~50K and 2M+ USD revenue, it looks like farcaster got a good start in the market. It will be very interesting to watch them scale to masses and capture a decent share in global social media users, as the sufficiently decentralised architecture of them gets tested. As mentioned in [this section](#heading=h.c3two8fta7lm), apart from offering account ownership to users as a unique media proposition, farcaster apps need to improve their recommender algorithms a lot, to keep the user hooked to the application, as most users bother only about engaging recommendations(unconsciously). Huge amounts of publicly available user activity will enable many farcaster applications to build their recommender algorithms effectively and farcaster applications can be tuned into showing in app advertisements.

The current hub requirements for farcaster are minimal, hubs with the below specs handled 8M+ messages on a peak activity day. This is farless than 500M+ tweets posted everyday on twitter. For farcaster to handle messages of this volume a day, the hub requirements will increase along with hub handling costs.

- 16GB of RAM
- 4 CPU cores or 16 vCPUs
- 200GB of Free Storage

Farcaster messages are handled in a different way than most of the social media data is handled at present. Every farcaster message that reaches a node is verified for signature validity, deterministic collision rules are applied and much more. Increased volumes also necessitates software optimisations for signature verification, merkle tree, diff gossip, disk writes and network propagation. This given the centralised nature of current social media apps, is very easy to handle.

Being a permissionless platform for hosting social consumer apps, Farcaster can become a centre of anti-social groups to communicate with the public. Although no such activity has been identified till date. One benefit of being a centralised social consumer application is the ability to prevent anti social groups from their platforms.

## Monetization opportunities:

Hubs/Nodes are not incentivised for running farcaster. They can be run by any individual/entity that wants to be part of the network, either to have better performance for their apps or when other nodes started censoring their content.

All the storage rent goes to the farcaster protocol. Currently with 600K+ registered accounts, 50K+ DAUs farcaster has earned 2M+ USD as storage rent. Apps on farcaster can earn by including advertisements on users feed(it is always upto who offers the engaging experience along with data ownership). With every new user, farcaster protocol earns at least 5 USD as storage rent. Users on farcaster apps are incentivised to stay active on platform in certain intervals, as not doing so will result in loss of their account activity.

Farcaster YAU’s may reach 1.5 Million in next 5 years. At $5 per storage unit, farcaster protocols projected revenue.

| Yearly Active Users | Yearly protocol revenue(at 1 storage unit per user) | Yearly protocol revenue(at 1.15\*\* storage units per user) |
| :---- | :---- | :---- |
| 300K | 1.5 M USD | 1.64 M USD |
| 500K | 2.5 M USD | 2.88 M USD |
| 1 Million | 5 M USD | 5.8 M USD |
| 1.5 Million | 7.5 M USD | 8.6 M USD |

\*\* Current storage units to total users ratio is 1.13.

If storage unit costs are $7 per unit with storage units to YAU ratio being 1.4, the revenue looks as.

| Yearly Active Users | Yearly protocol revenue(at 1 storage unit per user) | Yearly protocol revenue(at 1.4 storage units per user) |
| :---- | :---- | :---- |
| 300K | 2.1 M USD | 2.9 M USD |
| 500K | 3.5 M USD | 4.9 M USD |
| 1 Million | 7 M USD | 9.8 M USD |
| 1.5 Million | 10.5 M USD | 14.7 M USD |
| 5 Million | 35 M USD | 49 M USD |
| 10 Million | 70 M USD | 98M USD |

## Investment:

Social consumer applications on blockchains are a new space with only a few products operational, one of them being farcaster. it's highly unlikely to predict how a decentralised social consumer application will succeed in the future, given the nascent market. Farcaster is architecturally different from lens protocol. Lens protocol is transitioning [into an evm compatible ZK Stack volition](https://blockworks.co/news/lens-network-migrating-to-zksync), with every user transaction(messages and financial activity) getting posted to Ethereum as part of a state root. This way lens protocol aims to verify all user activity on ethereum. Lens is aiming to build multiple interoperable volitions with private and public state to coordinate and settle on Ethereum L1.

Major difference between lens and farcaster is in the way the state is handled. Being an EVM compatible app chain gives lens a lot of disadvantages from the start, as EVM is known for inefficiencies when higher throughput is needed. Higher throughput is very much needed in social consumer applications. Another notable difference between farcaster and lens protocol is that users are charged on the go for DA usage, L1 Settlement by lens protocol, wherein farcaster users buy in storage units at a nominal cost in a subscription based model. WarpCast by farcaster has been a great client providing the expected UX for social consumer applications.

The Farcaster team has no plans to launch a token in the near future.

Farcaster is valued at $1B after its recent series A round led by paradigm. As a bull case if the farcaster platform manages to gain 10 M YAU, in the next five years with 1.4x storage units to YAUs, its platform revenue would be \~100 M per year, excluding its potential revenue that can be generated with advertisements made by its apps. With advertisement revenue from apps taken into consideration, farcaster may be valued in between 8-10 Billion USD, without any token launch. Generating a return of 8-10x over the initial investments.

For a case with 1 Million YAU in next five years with 1.4x storage units to YAUs, platform revenue would be \~10M per year, excluding its revenue that could be generated with advertisements. With advertisement revenue, Farcaster may be valued at 3-4 Billion USD, without any token launch. Generating a return of 3-4x over the initial investment.

If the farcaster team decides to launch a token, within the next few years, the valuation of the farcaster may be entirely different. Farcaster may be valued very highly as an initial mover in the decentralised social consumer applications market. Valuation of farcaster could not be predicted accurately, if they go for a token launch.