Amidst all this talk about the future of Twitter pending a still-theoretical Elon Musk acquisition, some people have been asking whether there’s a viable alternative they could move to immediately.

Inevitably, some of the obvious decentralized projects have been suggested. The most notable is Mastodon, a federated social network that might as well be a full Twitter clone, albeit based on the ActivityPub standard.

I have nothing against Mastodon. I’ve been using it for years, alongside my other social networks. The community there is a little nerdier, and certainly quieter. Twitter is where the action is; Mastodon has so far been for the handful of enthusiasts who want to experiment with federation. Even among them, a hefty percentage simply syndicate content from their Twitter accounts.

Since the Elon Musk news broke, Mastodon founder Eugen Rochko reports that the network has grown by 84,579 users. That’s great, but hardly a drop in the bucket when you consider Twitter’s 330 million monthly active users. It’s possible we’re at the beginning of a larger move, but it’s more likely that these users represent a spike in new sign-ups that will settle down to something closer to their usual level.

Although it’s a decentralized network with no corporate owner, Mastodon is fundamentally no better than Twitter from an end user perspective. In some ways, it’s a little bit worse: the username scheme is necessarily more complicated and harder to understand, there aren’t any anti-harassment protections for vulnerable communities, and the news feed is less likely to immediately show you content you’re interested in. That’s if you even get that far: you’ve got to pick a home server and one of multiple client apps.

Again: I’m not knocking Mastodon (or any decentralized project). It’s an important step towards a web that is not under the corporate control of a handful of companies.

What I am knocking is the design approach of emulating Twitter. While Twitter has tweets, Mastodon has toots; while Twitter has a 280 character limit, Mastodon has 500. The two networks have fundamentally the same content and interaction models, with what amount to slightly different settings. I say tweet, you say toot: let’s call the whole thing off.

“Twitter, but decentralized” is an example of a solution to a problem that has been defined in technical or ideological terms, but doesn’t come from a direct user need. As ideological proponents of decentralization, we might want the user to need federated Twitter, we might think they need it, but without a deep understanding of the users, all we’re doing is projecting our hopes and dreams onto them. Is their need decentralized Twitter, is it a network where they can connect to breaking news but also feel safe from abuse, or is it something else entirely?

The only true approach is to go back to a well-defined, core group of users, and learn from them holistically. Instead of making a problem to solve from whole cloth, we should start with the real-life points of view of a number of real people. (Not market segments; not invented personas; real-life humans who are representative of who you want the users to be.)

[Name], a [description], needs a way to [verb] in order to [surprising insight].

These POVs can only be arrived at through getting to know those people - and are the first step in a long human-centered design process that must encompass not just the product being made, but the structure of the organization that makes it. You can make a decentralized tool, but if the underlying organization is a C-Corp that could be bought by a billionaire, the effectiveness of your solution to a problem created by another C-Corp that was bought by a billionaire is limited. And if you want to build a platform where diverse, vulnerable communities feel safe, you’d better give them a say in running it.

We are never absolved from doing the hard work of deeply working with real people in order to serve them. A technology-first approach never wins. When 86% of Americans get their news via the internet, and when the platforms providing that news are owned by a very small handful of commercial companies and an even smaller gaggle of rich men, this isn’t a problem we get to half-ass.

Whether he ultimately does or not, the idea of Musk owning Twitter is a problem. The solution is not “Twitter but decentralized”, or a protocol, or an open standard, although it might potentially incorporate any of those things. The solution is something new that more deeply serves its target users better than they have been served before. The technology is secondary to the need, always.