I’ve been looking into some social protocols and the one that I would bet on winning is Farcaster. I think they're doing multiple things that other protocols aren’t doing, building a product-driven protocol, the protocol is following the sufficient decentralization concept, a desktop app, and focusing on creating an early community.
For a recap, social protocols are open data graphs where users own the relationship with their audience, and developers can read and build on top of the graph without any limitation. My hypothesis for why this is important is simply, permissionless innovation. We don’t know what kind of products developers and founders will build if the social graph is truly open but we know that open systems where you don’t need anyone’s permission to build are one of the greatest things that ever happened. Brad Burnham says it perfectly:
The real reason the independent software industry emerged is that operating systems and APIs made it possible for independent software vendors to develop applications independently. They no longer had to ask permission of the hardware vendors. This same characteristic of permissionless innovation led to the explosion of independently created services on the internet.
#A Product-driven protocol
The first thing that you'll notice about how Farcaster is getting built is the product-driven protocol approach. This seems very simple but nearly all web3 protocols are not doing it. If you’re building a protocol and have some hype, it doesn’t mean that you don’t need product-market fit and that protocols aren’t boring. Protocols are boring and most of them died before and are going to die but what’s exciting are cool products. The Farcaster team is shipping every week, and I feel they're having as much focus on the product as they have on the protocol. They built token-gated posts on the product-level in the beginning but they weren’t used very much so they removed it and planining to reimplement it in a new way. This would’ve been much harder and slower to do if it was on a protocol level.
It also seems that on-chain activities are going to be a part of the Farcater product which I think is going to be an important hook. People now share their on-chain activities as screenshots on groups, and subreddits but I think there’s going to be a product where people can follow each other on-chain activities as social data become a bigger part of web3.
Another very product-focused thing in Farcaster that I am excited about is signing accounts which they implemented from day one.
Another important part of Farcaster is that it’s a sufficiently decentralized social network. It’s not fully on-chain and not fully off-chain. It leverages the decentralization of the on-chain record and the scalability and experience of off-chain systems. I was reading about some early social protocol experiments and a clear pain point was where do you implement the main user's directory which Farcaster implemented on top of Ethereum.
There’s also a huge upside of putting the data structure off-chain, it’s permissionlessly extendable which makes it much easier to build on top of and extend. An example of why this is better are NFTs likes. As the data structure was fully implemented in on-chain, every platform had to implement the likes data in their database. But if it was a JSON file off-chain mapped on-chain in the same way Farcaster works, it would’ve been only added by the first platform and other platforms would’ve been able to just use it.
I believe the Ethereum identity problem will also not be solved by a protocol-first approach rather it’s going to be one of those things where it just happens. This is why I think there’s a huge opportunity for Farcaster to become the social identity for web3 by introducing something sufficiently decentralized for other projects to build on top of and extend.
I also feel that the sufficient decentralization concept is using the properties and lessons that we could learn from web1 protocols and this is why I am excited about it! You can read the full concept of sufficient decentralization here.
#A Desktop App
A question that always gets asked about decentralized social apps is how can I verify what I am seeing, and how can I make sure to not get banned from the UI. The answer is why not decentralize the frontend? and you would do so by building open-source desktop apps. Farcaster is focusing on building a user-controlled product by building a desktop app first which I think is very important.
When Dan was asked about the biggest risk for Farcaster he said: “We lose the sense of community we have now too early. Need to keep for as long as possible!”. Kevin Systrom also talks about how important the early community and meetups were to Instagram and I think Farcaster is doing the same. I met multiple interesting people from the community, multiple other people are casting (posting) how they enjoyed meeting other people and many meetups happened. And I really feel there’s a sense of community when using the app.
I think (1) is the most likely way a decentralized social app is going to exist. It's going to acquire a large number of users by collective migration and what Balaji calls a proof-of-exist and the only way this is going to happen is by a philosophical or political based movement which I think is a part where web3 got very well and most people in web3 have so it makes sense to focus on web3 people.
About the third point we have a proof that this is going to happen if the product works and they’re the early days of Twitter:
"Yeah. The API has been arguably the most important, or maybe even inarguably, the most important thing we’ve done with Twitter.....So, the API which has easily 10 times more traffic than the website” - Biz Stone, Sep 2007