Bluesky is Twitter 2.0
Ever since Elon perverted Twitter into what Truth Social wished it was, many of us have been trying to find an adequate replacement, and it looks like Bluesky, which is currently #1 in Apple’s app store, is it.
How did we get here? After Elon took over, many Twitter users left and ended up at either Mastodon, Threads, or Bluesky.
But Mastodon was never going to be mass consumer viable. It looks like it was made by linux hobbyists, and it feels that way too. Every minor decision and user flow is mired in power user choices; you can’t even use the app without finding and joining a server, and don’t get me started about trying to change them. It’s an obtuse maze of user friction, created by well intentioned people who do not understand consumer product.
Threads, Meta’s Twitter replacement, came out of the gate strongest, but in typical Meta fashion, reflects the interests of Meta over the interests of users. Many table stakes features on Twitter — DMs, detailed search, ironically the ability to actually see threads (its namesake) in the home timeline, and even mapping the in-app identity of users to their actual name rather than whatever obscure handle they grabbed on sign up — were missing, shifting the experience in a different direction from the Twitter users knew and wanted.
But even worse was Meta’s heavy hand on the algo and content preferences, which downranked any posts with links and prevented algo boosting of any content it deemed political. While old Twitter was many things, it was centrally where global news broke and where the global political discussion happened. Meta’s fundamental misunderstanding of what made Twitter valuable, why people used it, and what its untapped potential was, combined with being really bad at 0 to 1 product creation, led to what I think will ultimately be the squandering of the deca-billion dollar opportunity Elon handed them, on a platter, for free. It’s an ironic blunder coming from the company whose CEO once (and accurately) said that “Twitter was a clown car that fell into a goldmine.’
Initially Bluesky didn’t look promising either. And its origins are… odd. In 2019, Twitter CEO Jack Dorsey allocated several million dollars from corporate funds to create an open protocol version of the social interest graph that was independent from Twitter. He also hoped it would one day replace Twitter. A weird move from the CEO and board of directors of a publicly traded company who have a fiduciary duty to their shareholders. But, bizarrely it’s possibly one of the only good things Jack Dorsey has done in the last 10 years. He is, unsurprisingly, no longer part of the project and has fucked off to tinker with Nostr while cheerleading bitcoin and doing everything he can to hangout with Jay-Z (a continuing trend from when he had Square, his financial payments company, buy Tidal, a not widely used music streaming platform owned by Jay-Z, for $300M for no discernable reason; the only explanation provided was Jack posting a picture of him and Jay-Z drinking wine with the word “artists” circled).
At first Bluesky was invite only, missing many key features, and looking extremely unpolished, similar to Mastadon. The team was (and still is) small and mostly unheard of in the bustling tech community that formerly existed on Twitter. As Threads picked up steam, Bluesky wasn’t yet ready for prime time; they were making sure their tech and product could handle scaling. And back then, other small details made Bluesky and Mastodon seem unserious contenders for replacing Twitter on the global stage, like having their posts be called ‘skeets’ and ‘toots’ respectively.
But America’s 2024 presidential election results changed everything. In the final stretch, Musk came out in strong support of Trump. He donated $200M, took over ground operations, and ran scams in battleground states claiming to give away $1M a day to random voters to drive up turnout — implicitly for Trump (never mind that the winners were pre-selected and not a random lottery as he implied). Perhaps most importantly, he turned X into a megaphone for pro-Trump sentiment, normalizing DJT to the masses of dense dudes who still held Elon in high regard, despite the hundreds of very good reasons he gave them not to.
The shock of Trump’s resounding victory, and the role Xitter’s owner played it in, seemed to have the effect on many left-leaning Twitter users that years of his terrible behavior had not: *finally*, they started leaving. Maybe it took people this long because their vanity, ego, and sense of self importance wouldn’t allow them to part with their precious followers, even if many had already left or gone dormant. The moral imperative should have been for us to migrate the graph the instant Elon and his goons showed up, but at least we seem to have one final chance now.
Since the election Bluesky has grown to over 17 million users, at times averaging 1 million daily signs ups, and having user follow operations spike 37x from pre-election levels. Major news figures, politicians, celebrities, and popular Twitter personalities are all starting to show up, from Dril to Chris Hayes to AOC, who as of today has 575k+ followers. For the first time since Elon killed the bird, it’s starting to feel like the old Twitter graph is coalescing around a new platform to replace it.
And while Threads boasts 200M users with 1M sign ups a day, there are some really important differences afoot. First of all, despite Threads’ large numbers, popular users have been A/B testing posts and claim to get better engagement on Bluesky. But more importantly, I think much of Threads “growth” and “strong user numbers” are actually just clever, low-converting ad campaigns it runs against its own pre-existing 4B users, resulting in misleading and inflated user counts. In the screenshot below we discuss the user number to activity discrepancy.
One reason I’m really excited is the ethos of the Bluesky team seems to be incredibly be good, which is definitely not true for the ethos of Meta’s leadership or the smoldering black hole where morality should be for the overlords of Xitter. The Bluesky team are very competent technologists who first and foremost care about what’s best for the user. They want the tools that create the platform to be open and accessible so any developers can tinker, and so the products and features users want most gain traction, even if that means Bluesky itself is eventually not the primary client atop the ATProtocol. It’s a truly unique experiment in tech, where historically, centralization and capitalization, while initial catalysts for compelling new products, invariably shackle them to late stage enshitification, beholdened to the greed and myopia of Wall Street. A process that seldom ends well for the user.
So what is Bluesky? It’s a Public Benefit Corporation: a for-profit corporation that includes a legally defined mission to produce a positive impact on society, the environment, or a specific public benefit, in addition to generating profit for shareholders; well known examples, I believe, include Patagonia and Etsy.
They do plan to generate revenue to sustain the product, possibly through subscriptions. It seems they may not currently have a willingness to explore ad revenue, and there is no precedent of subscription models generating enough revenue to sustain the operations of a scaled social media platform alone. But it’s clear they plan to try various approaches to find a sustainable path forward.
It’s unclear how the big problems of scale can be addressed by such a small team, when in the past, it’s taken $10s if not $100s of millions and many contractors and FTEs to deal with content moderation, abuse and bad actors, verification, and privacy and safety issues, but these are early days and at the very least, we know the team cares deeply about these issues (which arguably Meta does not, and Xitter definitely does not). Here’s one example from Bluesky’s CEO and Flavor Flav:
It’s been ~10 days since Bluesky started popping off and I think the next things we can expect to see is more user growth, especially from bigger name figures and celebrities, and as well as more features, both from the Bluesky team itself and from 3rd party developers. Here are some examples of how open the code base is from Bluesky CTO Paul Frazee:
While Bluesky has some important Twitter features Threads is lacking and refuses to add, many are still missing. The growing list of product requests includes: Edit, Drafts, Bookmarks, Group DMs, better notification controls (eg block notifs from accounts you don’t follow), and even lower level things like the ability to crop an image in the compose flow when attaching it to a post; I’m probably forgetting a bunch. It’s also not yet clear who owns or runs product at Bluesky, or what the company’s product approach or philosophy is.
One of the benefits of Bluesky being so open and welcoming to 3rd party developers is they can make tools that work inside the Bluesky client, or alongside it externally. For example, any developer can make a feed that any user can add to their core app experience. In fact, I was just chatting with an ex-Twitter home timeline engineer on how to get likes to show up as a first order feed insertion event. Feeds, I believe, can be generated based on things as simple as lists, or they can have technical complexity with variable rules discerning what shows up. One fun feed is “Dogs of Bluesky”, I highly recommend it :).
What is probably the most exciting, beyond finally having Twitter back, is the chance to explore all the untapped potential that pre-Elon executives left on the table. It was a near universal consensus among smart, accomplished technologists in Silicon Valley that Twitter never came close to reaching its full potential, from user growth, to revenue, to societal impact, even though it did always have a very outsized impact on news, culture, and politics.
Part of that untapped potential is solving onboarding, which was broken for the entirety of Twitter’s existence. Over 1 billion people signed up, tried Twitter, and left never to return. Even Twitter’s first PM of User Growth & Onboarding, Josh Elman, admitted one of their core learnings early on was not being able to connect the granular interests of users to specific accounts to follow. There are potential solutions for this, but I don’t want to go down that tangent right now; suffice to say there’s a huge opportunity to help people quickly find the people they would most enjoy following, which can yield a good first home feed experience. And not having one means a user will bounce and likely never come back.
I think the Bluesky team should consider reaching out to some pre-Elon ex-Twitter vibe-matched experts who thought about these product spaces deeply for potential advice. In addition to Josh Elman for onboarding and Christine Chen for the home timeline, many other well intentioned experts exist like David Gasca and Tom Hauburger, former PMs who ran notifications.
Onboarding aside, the social interest graph can be harnessed to do many things no app does well today. Like solve hard, open-ended questions such as what podcasts should I listen to, what books or articles should I read, or what events should I go to? Using people you respect and trust the most as a proxy for your interests, and tapping into their recommendations, can unlock so many more powerful use cases than just reading and composing posts.
In addition to hopefully being the death knell of the app formerly known as Twitter, and the solidification of X being reduced to nothing more than Truth Social with crypto bros, I hope we are in the early stages of witnessing the true potential of the social interest graph finally being explored. If nothing else, we could all use one good thing to be excited about in these bleak and truly insane times.
If you’re on one of these apps and not on Bluesky, you are missing out. Come join the party, it’s fun!
Alex is junior software developer, product manager, and co-founder of the first social podcast app on iOS. He nearly convinced pre-Elon Twitter to make a podcast app and has just begun work to create a social podcast app using Bluesky and the ATProtocol.
Related reading on the potential of the social interest graph:
Twitter Should Build a Podcast App (2017) by Alex Carter
The Future of Personalized Discovery (2020) by Alex Carter
Recommended context for Elon x Twitter:
Battle for the Bird by Kurt Wagner
Extremely Hardcore by Zoe Schiffer
Character Limit by Ryan Mac and Kate Conger
