Post

A decade of walled gardens trained us to be consumers not creators. So telling someone they can “own their data” assumes they see themselves as a data-producing participant in a network. Most don’t. Their mental model is much simpler: people I follow; things I care about; where to find them.

Web participation patterns makes this clear: ~90% of people consume 9% interact/react 1% create original content. People aren’t looking for infrastructure. They’re looking for relevance, relationships, info, entertainment, belonging (and increasingly ways to make or save money).

That creates a tension for Atmosphere builders. We naturally notice tech capabilities cos we’re close to it. We see what becomes possible with decentralised identity, portable data, composability and an open social graph.

But capability is not the same as market demands!!! Elegant solutions can still be highly irrelevant to the masses of people we’re trying to attract.

And there’s a much bigger acquisition problem hiding underneath all of this: social networks have two-sided lock-in. Creators won’t move without an audience. Audiences won’t move without the creators **they already follow**. “We have lots of creators” doesn’t solve either side’s problem.

Imagine someone decides to leave Instagram tomorrow. They don’t ask: “Does Bluesky have a better decentralised architecture?” They ask: “Are the people I care about here?” Is my favourite journalist here? My niche community? The creators I follow? My friends? Can I find them? Are they active?

Bluesky’s biggest growth moments have largely been driven by push, not pull: people becoming dissatisfied with the other site, and start looking for somewhere else. The trigger is external. The opportunity is converting that moment of dissatisfaction into a successful migration.

That changes what “growth” means. Adding 100,000 creators is less useful than making it possible for one person leaving X to instantly reconstruct the social world they care about. Growth is not just increasing the population. It’s reducing the cost of moving between networks.

discovery and graph portability deserve much more attention across Bluesky and the Atmosphere. If someone arrives with a list of people they care about, finding those people must be almost frictionless. If they can’t, the network can have every feature imaginable and still feel very empty.

(Which is feedback we’ve all heard uttered often: “where is everbody?”)

There’s a strategic distinction here: building a bigger network vs making my existing network portable. The 1st grows slowly and abstractly. 2nd unblocks migration at exactly the moment someone is ready to leave a walled garden. Thats a more consequential growth lever than incremental feature.

The same applies to discovery. “There are millions of users here” is **meaningless** to a new arrival if they can’t find the 50 people, 5 communities or 3 topics that make their network useful to them. Network effects are experienced personally, not statistically.

This is where I think bsky optimises for the wrong users. Features that help existing users have value, obviously. But if our ambition is the next 100m, we need to spend much more time designing for people yet to come with absolutely no reason to migrate.

They don’t need another explanation of why decentralised social media is interesting. They need to know the people they follow are here, the conversations they value happen here, that their communities thrive here, and that joining doesn’t mean abandoning communities they’ve built elsewhere

This is why I think we need a shift from a (primarily) developer-relations mindset to a creator/community-relations mindset. Devs are essential to building out the network. But creators and communities are what make a social network worth joining, and the needs of those groups are not the same.

Creators need reach, discovery, identity, reputation, sustainable audiences and eventually ways to earn. Communities need continuity, moderation, norms, belonging and reliable discovery.

Audiences need relevance, people, information and connection. “Data sovereignty” can underpin all of this without necessarily being the reason anyone shows up.

(None of this is an argument against decentralisation btw. Quite the opposite. It’s an argument for treating decentralisation as infrastructure for better human outcomes, rather than assuming the infrastructure itself is the consumer value proposition)

The Atmosphere has a genuinely unusual opportunity: build an open social network where people can move, carry relationships, discover communities and participate without being trapped by one company’s walls. But we need to sell human-centric outcomes, not plumbing.

So I’d frame the next 100m-user challenge much more simply: Who are the people they already care about, where are they now, what would make them move, and how do we remove every bit of friction between wanting to move and actually arriving?