Revised 2023-10-02.
Twitter is dying, everyone knows this. There’s a bunch of ways this is true, but this article isn’t meant to be about that. This article is meant to be about the many projects that are trying to be a successor to the niche is holds in the social media space. I will assume readers already have some working understanding of how the Fediverse works, like what an instance is. If you want some more background knowlege, here’s some resources:
See also reddit: What is to be done?.
Off the top of my head there’s a handful of projects that are in some way trying to be successors to Twitter.
(please suggest others so I can include them for completeness)
The key lesson for why “mainstream social media platforms” have existed and stayed dominant for so long is that they heavily leverage network effects and low interest rates letting them offer unfairly-good deals for a long time. But as others have noticed this can’t really last forever, eventually it comes time to pay the piper. Author and technologist Cory Doctorow named this phenomenon enshittification.
The point I would like to make here is that Mastodon (and the Fediverse broadly) has already won the race to replace Twitter, most of the users (including current Fedi users) just haven’t realized it yet. Not in the sense that, somehow, there’s a hundred million Fedi users that nobody noticed. But in the sense that, assuming the current trends continue, Fedi will eventually displace all existing competitors. There’s two prongs to my argument here, the first point is based on development and funding models of these protocols constraining the ways they can grow, and the second point is based on the nature of network effects in social media and the nature of the users that have already migrated.
The common theme with the issues with the other supposed competitors is over the legitimacy of their claim to be the successor. I’d like to link Vitalik’s article on this, in which he articulates what I mean by this better than I could. He then goes and talks about NFTs which I don’t really like, but that’s more of a end thought that I don’t think takes away from the core argument.
I want to discuss some more focused points particular to some of the competitors that are coming.
The big ones trying to pick up the reigns are Bluesky and Threads.
Bluesky is a vanity project that was funded personally by Jack Dorsey, based around a more generic “AT Protocol”. There’s a lot of intersting design decisions they made, decisions that people have already critiqued elsewhere. But Bluesky is kinda directionless at the moment. It does have some active users, reporting 1.2M, but it still has a waitlist for registrations and is otherwise invite-only. It’s not really clear what the timeline for opening it up further is. And it’s reported that Jack even deleted his own Bluesky account inexplicably.
Threads is Facebook’s answer to Fedi. It bootstrapped itself by essentially being built out of Instagram infrastructure like a chestburster from Alien, borrowing all its userbase and making it easy onboards a Threads profile from Instagram. This gave it a huuuuge initial bonus, claiming 200M users. Of course this led people to say that “Meta did it! They’re going to replace Twitter!”, whether that’s a good or a bad thing in their worldview.
But it seems that this also dropped precipitously, with some estimates suggesting it’s sitting at somewhere around 1M daily active users. That sounds like a shitload, but remember, this is a Facebook product. They’re the company that told their investors that they had burned through the set of possible new users and that there isn’t much of anyone left to turn into a new active user. So it’s a pretty big flop if you put it into perspective of the power they theoretically have.
As an aside, a large goal for Threads was ActivityPub support, allowing interop with Fediverse instances. This hasn’t happened as of the time of writing (as far as I’m aware), but it wouldn’t be likely to change much in the eyes of people trying to pick between it and Fedi more broadly, since everyone that would be considering it knows it’s still just Facebook. Fedi users aren’t Facebook’s target audience.
They’re mostly still just curiosities at this point. Even the people I’ve talked to that are jumping onto Bluesky aren’t wholly sold on it. Threads carries a very astroturfed and advertizer-friendly feel to it. I have yet to meet anyone that actually likes using Threads. If you do, then I’d love to talk to you.
But like, a more important question we should be asking is “why should people trust these platforms anyways?”. People can kinda tell when they’re being taken for a ride, even if they can’t articulate it, and these platforms don’t really do anything to address the systemic problems that people have with Twitter. I’m confident that if we swapped places and had a feature-mature Fediverse with as many users that Twitter had at its height going against even a mature Twitter with all of the features it has now (but starting from 0 users), it would struggle to ever get off the ground. There’s no feature a centralized social platform can have that a decentralized social network couldn’t possibly provide in some way, and a platform that has to make money will always have to extract it from its users once the VC money runs out.
It’s also possible that Facebook doesn’t really care that much about Threads and that it was more of a corporate show for investors rather than a serious strategy. They have a shitload of money they can throw into a project in the event it works out, without the expectation that it does. But if they don’t make a move then investors will complain that they’re missing the opportunity on gobbling up users in the exodus from Twitter imploding.
For other reasons that I’ve already articulated in the reddit article, I and others think that it’s pretty important that whatever successor to Twitter there is is decentralized. But people have lots of different visions for this.
Most of the crypto-based social media protocols are accepting VC funding. So many of them don’t even have organic user behavior, with Lens literally just using bots to be able to say to investors “this is what it will look like when it takes off”. The core development team will not be likely to give up control over the project to turn it over to community development any time soon, and so there won’t be much community development, outside of potentially some grants coming from the VC money to astroturf an ecosystem into existence.
Since they’re backed by VCs, there will always be a desire to extract a revenue stream to pay back the investors. The enshittification cycle will manifest in some way, sooner or later.
On the other hand, Mastodon development is driven entirely through donations. Currently, the nonprofit that manages development and the flagship instance receives $27k/mo just throgh Patreon. Mastodon doesn’t need to have exponential growth. It doesn’t need to find a way to monetize its users through the core protocol. It doesn’t need to find ways to lock in its users like corporate-owned platforms do. It doesn’t need to be advertiser-friendly and get corporate customers onboard. The costs of operating the network are borne by its own (power) users directly, and development model is already self-sustaining.
Additionally, in most cases, these crypto social media projects associated with some scammy crypto coin that is going to push away normal users that don’t have the crypto brain poison, which means they’re also standing at a disadvantage.
There’s also the simple UX issues of these up-and-comers too. Normal people do not want to deal with private keys, especially for something as trivial as social media. You can convince them to care if it’s something more serious like keeping their crypto funds safe, but usually that’s best done through the abstraction of a hardware wallet. A completely unacceptable UX burden for something like social media. And all the negative issues that come out of it (like usually losing access to your account if you lose your private key, as it is in Nostr) also are unacceptable for most users. Bluesky has a much better approach in this area.
While network effects can manifest in many different contexts, in social media they take an interesting form. The 1% rule or 1-9-90 rule is the idea that 1% of users in a mature social media space actively produce new content, 9% engage with it passively, and 90% are lurkers that only consume content. In reality, it’s more like a power law distribution rather than 3 clean buckets, but it’s a rule of thumb after all.
A related idea in network theory is the idea of a small-world network, which is characterized by most nodes only having a few connections but a few nodes having many connections. These highly connected nodes are the people that tend to be the ones that produce the majority of the content that all of the rest of the people (nodes) interact with. You may be more familiar with the observation that everyone in the world only has “6 degrees of separation” from anyone else in the world, which relies on that most social networks take this form.
These creators that make a platform worth it for these lurkers are also the real power users that are especially sensitive to changes with the platform, making the risks of pissing them off that much greater. This is why the response to the reddit situation was so organized and kicked off a healthy community in the “threadiverse” (the subset of the wider Fediverse based on reddit-like discussion boards). I’m using the term “creator” as a catchall term for very active users that many other users pay attention to.
A conclusion we can make from these observations is that if you win over the 1% (or the 9%), the 90% of lurkers are likely to follow.
This does appear to be happening. New groups of users are joining, and they’re producing lots of content and activity, especially since Musk’s takeover of Twitter in November 2022 giving people a strong reason. Journalists post a lot of content and are very active about their topics of interest. The instance newsie.social was created specifically focusing on supporting journalists, and currently has >2k MAU (a lot for a single instance focused around one topic) and has >796k posts. Another group of content producers is game developers on mastodon.gamedev.place, with >8k MAU. I’ve also seen a number of webcomic artists around in various places.
Seriously, go check out JoinMastodon’s server list. There’s a ton of new instances that weren’t here a year ago that are growing and active. There’s even pretty great active instances focused on people’s own cities, like better.boston for my own city and sfba.social for the SF Bay Area, among many others in the US and Europe.
While this is mostly conjecture and I haven’t empirically studied the flows of users, if you’re an active Fedi user you might agree with the reasoning above.
Another observation that I made (all the way back in 2019) was that what sets Fedi apart from most of the rest of the alternative social media projects was that its users don’t talk about Fedi itself that much. They talk about whatever other random things interest them a lot more than they talk about Fedi. If you look at a lot of other projects, their users (when they’re actually organic users) mainly just talk about how cool they think the project is. Nostr is pretty bad about this in my experience with this, as its users mainly just talk about how cool Nostr is.
Since Fedi is still experiencing an influx of new creators we can’t extrapolate easily to compare the sizes of the engaged userbase there is. Since social networks are small world networks we can’t just compare plain MAU numbers to compare the strengths of the network effects it has against Twitter. And since Fedi isn’t mature yet, it’s still experiencing a large influx of the most active kinds of users, we can’t just use the 90-9-1 heuristic.
Another interesting phenomenon that has not happened anywhere else is governments running their own Mastodon instances, like EU Voice. As it turns out, they don’t really like being dependent on the social media platforms based in another country. Having a free and open network where they can maintain an official feed for the many different institutions within them is pretty desirable, without the meddling that comes from privately owned foreign actors.
There’s still a lot of issues with the user experience of Mastodon, we know this. These will improve and the avenues for getting user feedback are much more transparent than any corporate social media site. A big one people want is a better recommendation algorithm to improve the new user experience.
People forget how functionally shitty Twitter was in its early days, it’s unfair to compare the UX of a massive mature social platform with millions of person-hours of engineering time against a foss project with like 2 full time developers.
We can complain all day about little features Mastodon needs and why it not having those means it will never be successful. But, somehow, people don’t consider that those features can (and will, if they’re actually desirable) be added.
Some people, in particular the pro-crypto-social-media ones, argue that because Mastodon isn’t fully distributed that it will trend towards centralization over time. This is a really funny critique because some of the crypto social media protocols are themselves centralizing around for-profit-owned indexers and other central providers. Bluesky itself even implicitly depends on an indexer that maintains a large economy of scale.
But if you look at FediDb, the opposite phenomenon seems to be happening with Fedi. Since each instance is entirely independent and there isn’t any need to rely on “supplementary” third party infrastructure, and both users and administrators prefer diffusing out into separate instances to avoid the moderation concerns of many different users with different expectations in the same space. The total population of the small/medium-sized instances is growing proportionally more quickly than that of the larger ones.
This isn’t meant to be a thorough analysis. It’s really hard to concretely study social phenomenon, so most of this is conjecture. But I hope that I provided some interesting propositions that you may not have have considered before if you aren’t actively paying attention to these projects.
To be fair to Nostr, it wasn’t really meant to be used as a social media platform at first. It was a design consideration, but it was more meant to be used as a building block in other cryptographic protocols. But most of the investment around it now is mainly on its use as a social media protocol, since that’s what sells to VCs. I think it could be useful for some future thing that doesn’t exist yet, but I don’t think it’ll be social media. Nostr developers and users tend to hate Fedi, with some valid critisisms, but these criticisms are not existential to the idea of the Fediverse and mostly I don’t think are as severe as presented there.
Like most people, I don’t think Twitter will have a specific death date. It’ll just be a slow rot as all the people that make it worth being there leave. Some people think that we’re simply reaching the end of the era of megaplatforms and that many smaller platforms will flourish. But for reasons I outlined in the reddit article (linked at top of page), whatever comes after this has to be decentralized and not vulnerable to capture otherwise we’ll just repeat the cycle in ~15 years. Protocols over platforms.