How User-Designed Social Media Could Create Healthier Online Communities

Eryk Salvaggio is a Research and Communications Consultant at the Siegel Family Endowment.

Social media, despite its flaws, may have heralded a golden era of participation. Yet participation hasn’t become the norm at the level of designing and developing social networks. While social media platforms may claim to be responsive to popular demands, these adaptations are driven chiefly to raise revenue, largely from advertisers, and not simply to serve users. Invitations to build the tools we want and need – such as privacy settings on Facebook or an edit button on Twitter– have been underwhelming.

For years, people have asked: where is the alternative? In a competitive marketplace, sites that don’t serve users would see those users go elsewhere. One of the restraints on movement was that the choices were limited; users could switch from one top-down, revenue-driven business model to another top-down, revenue-driven business model. The same issues would arise – and the challenge of migrating individually was compounded by the challenge of moving entire communities.

A Community-Driven Alternative

Enter the federated network. Federated networks, or the ‘Fediverse,’ are a loose organization of mostly small, open source, volunteer-run digital infrastructures. This includes Mastodon, but also versions of blogging software like MicroBlog (an alternative to Wordpress) and WriteFreely. European governments have embraced federated tools such as Matrix, a cross between WhatsApp and Slack.

Below are some of the key features that this approach offers over commercial platforms, and how they affect what we want our social and digital infrastructure to look like.

1. Unbound by Profit Motives

Many federated alternatives are run by nonprofits. That means the incentives of traditional social media companies are absent. Crucially, nonprofit status allows these tools to be built in ways that don’t rely on the collection or sale of data; grant greater agency over the ecosystems users participate in, and avoid the advertisement-driven incentives that amplify attention-seeking content. The people hosting these servers are not as interested in increasing the time you spend on site, nor do they typically have any interest in amplifying highly engaging, but controversial posts - an affordance known to drive the sharing of misinformation and polarizing content on ad-driven platforms.

2. Volunteer-Run

Ultimately, the infrastructure – both the hardware and the underlying code – for these platforms is installed, developed and maintained by volunteers – those who care to support a community using the platform, often at their own expense of time or money. It’s a direct contrast with the tech giants, whose warehouses of servers are subsidized by ad revenue shown to millions, if not billions, of users. In a federated network, users can communicate across instances, but are part of largely small, locally run servers, some of which are run out of people’s homes.

This open, interchangeable and decentralized approach also poses challenges to coordination between hosts, users, and the overarching network. When interest in these networks is high, many may be inclined to volunteer on development projects. Yet, without a dedicated team of developers or designers, new software updates, which can range from quality of life improvements up to much-needed community safety features, can remain on the backlog for years. Even if some were willing to volunteer, what to do with that interest, and how to sustain and direct it, is only one of the challenges facing this space.

“Right now the situation is extremely asymmetric,” said Nathan Schneider, director of the Media Economies Design Lab at UC Boulder. For example, “Twitter was bought for $44 billion, while Mastodon has been one person since 2017 with community help. Things are really just beginning to be tried.”

3. User Designed and Governed

Algorithms can be a source of discovery and delight. Yet, the complexities of developing and implementing such algorithms may be technically challenging, especially when relying on volunteers. Developing these algorithms in a way that is community-driven, rather than advertiser-driven, requires design research and user testing, which can be an expensive investment even if those driving the process did it for free. And while some servers may be more nuanced and thoughtful than others, many are either unprepared or outright unconcerned with the kinds of content moderation policies that sites like Twitter, for commercial purposes, have had to enforce with more rigor.

What has this looked like in practice so far? While Mastodon is just one decentralized platform that’s risen to fame (Bluesky promises to follow), it’s worth examining as a specific case study.

Mastodon: A Case Study

Mastodon is a free and open-source software and social network. Most of the development is run by a German non-profit that is just over a year old. This organization is responsible for developing and releasing the source code, and runs a sizable default server, which hosts many new users. No advertising money goes to Mastodon, allowing it to avoid algorithms with commercial intentions. Instead, the site relies on crowdfunding and sponsorships, and collects subscription fees from the volunteers who run their own server. Mastodon is extremely small. It had 1 million active users, before a recent movement away from Twitter brought it 1.5 million more – only to see active users decline, again, to settle at 1.5 million users altogether. But it is worth paying attention to if only to test the proposition that, if users can build and control the technologies they use, they will build more responsive alternatives to technologies they don’t like.

As a result of removing the profit incentive, Mastodon has already made design decisions that set it apart from commercial platforms. However, design is not always easy to do through democratic participation alone, and doesn’t always have the same impact for all users. New features can also introduce new limits or tools for abuse on a platform. One of the earliest groups of people involved in shaping Mastodon were queer and trans communities, who created servers to operate as “safe spaces'' from online harassment. A number of cultural norms developed on Mastodon as a result, as did some design priorities. One of those is a content warning button that allows posters to blur out content that they think some people might not want to see. There have been controversies about this, too, when a server’s admins encouraged Black users to use content warnings when discussing racism.

For all the potential for healthier online communities that such a model might make possible, questions about sustainability and scale linger. How might philanthropy have a role in cultivating a sustainable social infrastructure that plugs directly into digital infrastructure – specifically, the design and improvement of open source social networks?

Nurturing the Fediverse

Building digital infrastructures for and with social infrastructures is a challenge, but it isn’t one philanthropy needs to shy away from. To understand a possible role, we might look to a system that has a similar support structure: Wikipedia’s relationship with the Wikimedia Foundation. The Wikimedia Foundation is a steward of Wikipedia. It runs the servers, develops the infrastructure, and manages the millions of dollars that flow in from donors every year. It directs those dollars to community groups, such as local Wikimedia chapters who organize events. They also invest in cultivating communities of contributors abroad, especially in language groups where Wikipedia faces wide content gaps. (Every language edition of Wikipedia is unique, rather than simply translating the English Wikipedia). Meanwhile, external organizations designed to support specific Wikipedia infrastructure or initiatives exist to support programs, such as designing tools for educators.

The Wikimedia Foundation’s donations and grants support community driven initiatives. Mastodon’s user base is considerably smaller than Wikipedia – though Wikipedia’s user base, the core contributors, are just 1% of its registered users. Mastodon may be small, but a dedicated number of active contributors with a clear roadmap might benefit from the kind of grant structure for development and design that Wikimedia has used for years.

One example is Social.coop, a Mastodon instance created to experiment with platform cooperatives, where users can coordinate decisions and costs. Every member pays in, sharing the burdens of that infrastructure. And each member has a say in governance decisions around how the server is run and maintained.

“Observing Mastodon instances, I felt that a very important movement was at risk of failing because we were simply moving from oligarchy to feudalism. There was no accountability for node admins. I wanted to create an alternative model for organizing nodes within federated networks — participatory, democratic, cooperative, transparent — and to then scale it horizontally rather than vertically.”

— Mayel De Borniol, Social .coop

UC Boulder’s Schneider suggests that the space still has risks: "The growing adoption and the relative lack of investment in the fediverse means there is still a danger that it could become captured by corporate actors,” he noted.

A Better Way Forward

“There is space for mission-oriented public organizations to build public-option funded infrastructure,” Schneider says. For example, creating platforms for community participation around decision-making for what to invest in. He points to models like Quadratic Funding, which has been adopted in cryptocurrency projects like Gitcoin. In this model, large organizations donate to matching pools. Then, community projects are funded by people in the community – with weights assigned not only by how much money is raised, but also by how many donate.

This could be a more sustainable, more social way of developing a social network. At the moment, power is concentrated in the hands of those with the skills and resources to host a server. Many may have good intentions, but systems for broader governance simply aren’t in place. Developing that social infrastructure may be a first step to creating a truly participatory social media network.

“Any investment at this point has ripple effects,” Schneider said.

While Mastodon and the Fediverse are not maintaining their rate of growth, this may point to a strength, not a failure of these ecosystems. Simply put, they are there if users need them. A non-profit social media ecosystem such as Mastodon provides a space not only to imagine new arrangements for social media and communications technologies, but also offer a testing ground from which to build new ecosystems and tools. This type of “sandbox” space is one that both the public sector and purpose-driven investors, like foundations, should watch, and may play a role in growing.

Preserving these alternative ecosystems remains a challenge, and funding them from voluntary contributions is likely to hamper the exponential growth we are used to seeing in commercial projects. The mere existence of a viable alternative may create incentives for larger social media platforms to build more carefully, in ways that are less likely to alienate their core users. But the spirit of experimentation in design and policy by these communities may prove more useful than the size of their core user base. These are the stepping stones that may lead to the types of future social networks that we want and deserve - ones that are built with the public interest at the center.