Hey there,
We have been discussing Web3 social networks internally. A few applications are beginning to see traction. In today’s piece, we won’t go into the specifics of why it matters. We covered the key differences and a thesis for a new internet in the past here. What I want to focus on today is how Web3 native social networks can go from zero to one.
On October 11, Farcaster opened the platform to everyone. Developers could use the platform to build applications, and users could sign up without anyone’s approval. I wanted to try out what an application built on Farcaster would feel like. I downloaded Warpcast, a microblogging application like X (formerly Twitter).
The first thing is it asks you for an upfront $1 per month payment, which is used for gas. My first reaction was that this step creates friction for an average user. After all, how many paid users does X have? ~640,000 out of 330 million active users is ~0.2%. But the fee is 1/8th of X and can be considered a feature, as it helps fight bots and keep things real. The good thing about this payment is you don’t need crypto. You can pay in your local currency through the app store.
Once you get past the fee barrier, the experience is similar to Web2 counterparts (more on this later). I wanted to see how the metrics changed after Farcaster went permissionless, so I looked at the number of ID registrations and casts.
The number of users registering on Farcaster shot up after October 11, and they contributed to ~20% of the new casts (posts) on the network. Farcaster has ~191k IDs registered compared to ~125k profiles on Lens. Note that Lens is not yet permissionless, although they are soon to be releasing a new version which will be. You can see how opening access to the app has impacted daily casts on the protocol in the chart below.
A few questions stood out to me while using the app:
Why do users keep going back to X?
Why did Threads fail to retain users in the initial weeks?
Why would anyone move from X to Warpcast?
Or What can Web3 social networks do to onboard more users?
The following is a breakdown of how I believe social networks in Web3 will first imitate and then redesign incentive mechanisms on the Web.
Replicating Incentives
X’s moats are network effects and incentives for creators. Network effects are straightforward. I go to X every day because the content is interesting to me. The content is interesting because the people I want to hear from are active there. The people I want to hear from are there because their incentives are aligned.
Two kinds of incentives exist for creators:
Direct, where X recently started sharing revenues with creators. In September, X paid out $20 million to its creator community.
Indirectly, as a creator, your audience is on X. When you try to move to a different application, it’s unlikely that all your audience will move there unless there’s a strong reason. It is because creators to audiences is a many-to-many relationship, not one-to-many. That is, you are not the only creator for your audience. So, your move doesn’t guarantee that your audience will move.
When Meta launched Threads, users were excited about it for one weekend. After a few days of headlines around the demise of X (then Twitter), eventually, the migration tapered off. According to Similar Web, after its launch, the DAUs plummeted from ~49 million on July 7 to 10 million on August 7. A new application fights for users’ attention with the incumbents.
It offers novelty as a hook for users. As Chris Dixon once said, come for the tool, stay for the network. Threads could not bring out one character or feature over X (and Instagram) that resonated with users and, as a result, failed to take over X in terms of the number of users.
Threads wasn’t the only competition for X. Mastodon, a decentralised social network built in 2016, started gaining traction in 2022 when Elon Musk bought Twitter. But the Mastodon UX wasn’t intuitive. The growth spurt had been a flash in the pan. When a user joins the network, they have to join a server, which is an interest-based community. Managing and growing a user’s social graph on the server is cumbersome.
When something feels like work without incentives, it is unlikely to attract a critical mass. The cost of switching from Twitter to Mastodon was too high for little gain.
Besides UX, Mastodon had an incentive problem. Servers usually have volunteer moderators who struggle to scale server capabilities due to the growth spurt. The non-profit nature of Mastodon meant that moderators had to ask for user donations. Although users foot the maintenance bill, server maintenance and moderation require time and energy, resources that are difficult to obtain for free.
Although user growth on Farcaster and other Web3 social ecosystems has been sluggish, Friend.tech (FT) offers clues for user onboarding. In over three months, FT has gained more than 800K unique subjects (users). Granted, the category differs from what platforms like Warpcast and Lenster try to achieve, but the drivers of success overlap.
Users are flocking to FT, hoping for an airdrop. But airdrops are anticipated for almost every protocol and application. So, airdrops cannot be the only reason FT is so far ahead of others. Another obvious answer is they have the backing of leading VCs like Paradigm. But VC backing can only take you so far. In my view, there are two reasons for FT’s success.
Using existing Web2 social graphs.
FT used X for onboarding users. It is safe to say that out of all the Web2 social media applications, X has the most active participation from crypto natives. Using X to let users borrow their social graph has immensely helped FT to reduce user onboarding friction. It allows FT to piggyback on X’s network effects.Gamification.
Call it speculation/gamification/incentivisation, FT hooked users. When there’s money involved, attention typically follows. What percentage of FT users are actually there to be connected to people they follow is up for debate, but it’s clear that FT has managed to grab and retain attention so far. Users may go to FT to trade shares of influencers but keep using it to chat with them occasionally.
Returning to the third question of what Warpcast can do to attract more users, there are a few learnings from the hits and misses discussed above.
All About The Hooks
Every successful social media application has had its unique feature or character that helped with the cold start problem – no users, so no good content; no good content, so no users. It can be considered a “cyclical loop of boring”, as anybody who has spent time on emergent Web3 social apps would have seen.
With the benefit of hindsight, most social networks tapped into user behaviour that was emergent but not captured at scale yet. X, for instance, was initially a text-based social network. You could send texts (through a mobile network) to have them show on a feed, which is why the platform originally had a 140-character limit.
As a medium, it allowed more users to share relevant news in shorter spurts than traditional media. Reddit, on the other hand, was a reinvention of Google in a loose sense. Where Google indexed pages by rank (hence the term PageRank), networks like Reddit and Hacker News used human curation (upvotes) as a measure of how user submissions should be ranked. This transfer of the filtering mechanism from Google’s algorithm to human upvotes was powerful at a time when content was just exploding on the internet.
Similarly, WhatsApp launched when many still communicated through text messages or the BlackBerry Messenger. Using the internet for text messaging through a mobile interface was not as common then. For a consumer to switch over, the value proposition was simple: You save on how much you would pay for texts.
Every social network scales with a value proposition that feeds into a certain user behaviour. To a certain degree, Friend.tech managed to replicate this model for success. The team enabled vanity (like Instagram) by giving value to a person’s share and allowing users to trade it. They combined a person’s need for validation with an industry’s desire to speculate.
In our view, the next generation of social networks (in Web3) will need to have a strong hook that taps into an existing human behaviour or emergent user need.
When I was using Warpcast, most of it looked like an X clone. But the bar on the left side is cleaner and has channels. These channels are pre-curated communities of sorts. Users can tweak channels as per their interests. Instead of ‘onboarding creators’, if Warpcast can help create niche channels unique to the platform, the likelihood of some users getting hooked increases.
Once an initial base of loyal users is created, scaling becomes easier. Wield and Farquest, similar to Rabitthole and Layer3, can make the discovery easier for users.
Users typically use applications when people they know also use the same application. It is difficult to judge which Web3 social application will be the one with critical mass. Aggregators can encourage users to be application-independent. Firefly and Yup are creating a platform interface that allows users to interact with their social graphs across platforms like X, Threads, Farcaster, Lens, Bluesky, etc.
The internet experimented with RSS (Really Simple Syndication) feeds as a means of distribution. RSS feeds allow content to be distributed without users going to individual websites. If a user has an RSS reader installed, the reader checks all the feeds the user has subscribed to and automatically brings updates to the user. Although it is efficient, it is limited to text.
RSS feeds are less user-friendly than applications like Facebook. They are also decentralised, so a platform or company has little control over the content broadcast via its sites. So, the adoption of RSS feeds has been limited. New aggregation models are significantly more user-friendly compared to RSS feeds. They are more collaborative.
Allowing users to sign up with X or Facebook helps with two things: the friction of switching is reduced, and it helps with network effects. FT is an example of how using existing network effects can help. While FT leverages X, Song.Tech is using Spotify to bootstrap its user base. We will likely see multiple social apps trying to replicate these models in the coming months.
The reason why these apps make use of existing social graphs and borrow elements of UX that already exist is to obfuscate the complexity of the new tech stack. When paradigms change in technology, we take elements from the previous era to make the transition easier for the user.
Redesigning Incentives
We have been internally debating whether Web3 is about ownership or changing incentives. Most users do not care about custodying their assets or identities. It is a burden they would instead outsource. What they care about are incentive mechanisms that directly impact them. Web3 design principles offer a way for users to simultaneously own their identity/data and change the incentives on social networks.
For instance, social networks have historically leaned towards aggressive bickering because it is a good way to keep users hooked longer. The headlines in our newspapers have become more negative in the past few decades because fear keeps us returning and buying more of the newspapers. The incentives for media in the last century have been to evoke emotions that can be bad for the consumer.
Quite recently, a documentary on Juul dropped on Netflix. For quite a while, people looked towards it as an alternative to cigarettes. Then, the regulator clarified that it can release considerably higher amounts of nicotine into the human body. Social networks today have a similar impact on the human mind. And the regulators might be sleeping at the wheel.
We need them for the sake of our democracies and free speech. The world is better off with more people voicing their opinions in the great town square that is the internet. But if the incentive mechanisms are designed to leave users feeling worse off, we must consider how to fix them.
Blockchains being payment rails means Web3 social networks can reward users globally without needing platforms to tap into user data or run advertisements. A social network that focuses solely on allowing users to be compensated (in stablecoins) for the content they post has value.
It is not here yet, but it will emerge sooner or later as individuals realise that many social networks leave them worse off mentally. Much like the food we consume, the ideas we entertain and source through social networks have a bearing on our minds. For us, the question is — does changing the incentive mechanisms for social networks pave the way for a better internet? Do Web3 social products have a shot at doing that?
It is a big vision to build towards. But a worthy one that goes beyond speculating on tokens.
Signing out,
Saurabh
Acknowledgment: Conversations with Jakub and Mitch helped me crystallise ideas for this piece. I’m grateful to both for spending time with me on calls to discuss all things Web3 Social
While I was reading the article, I liked the humility of the writer that he has no maximalist position despite being the one dedicated to crypto writing. I will second his thoughts by the use of a principle from behavioral economics. This is called Nudge as popularized by its writers Richard Thaler and Cass Sunstein. There are lots of spaces in social networks where people can be gently nudged to try out something that may solve a part of their current troubles. Being a founder myself, I have located this problem in dating networks where finding a person with commitment is hell of a problem. People don't prefer either of the two choices "jumping into marriage" and "jumping into bed" but dating networks don't offer what they need. This is what I call a pre-marital commitment space where commitment can co-exist with the option of an exit. Smart contracts can be used to convert this space into floating bonds of a huge variety. This is just to highlight the potential of this article for other readers. However, I would expect more from the writer in the next article around how to try out a bigger range of user number with dApps.