Twitter’s New Communities Must Avoid The Same Toxic Mistakes As Facebook Groups

At 19, Maryaam Lewis-Herbert felt a little adrift in life. To right herself, she started to learn about astrology, the study of celestial bodies and how they, purportedly, overlap with our lives. (Venus, for instance, is said to have a powerful influence over love, while Jupiter governs luck.) She boned up enough on the subject—chiefly through YouTube videos—to become a sought-after astrologist online. Lewis-Herbert today has a 112,000-person Twitter following as @shawtyastrology; offers readings through Etsy for as much as $149 and Zoom classes for $99; and writes a column for the website of Paper magazine, a glossy pop-culture quarterly.

Most recently, she’s become one of the founding moderators of AstroTwitter, one Twitter’s new so-called Communities. That’s a fancy term for groups, a novel concept on Twitter: For the first time, the 15-year-old site is allowing collections of people to band together on its site. “Astrology has brought me so much clarity and a sense of purpose and a sense of belonging in this world,” says Lewis-Herbert, who’s now 24 and lives in Toronto. Through AstroTwitter, she says, “we wanted to create a safer space for people just to talk about astrology without like, backlash from other people on Twitter”—astrology tending to draw believers and non-believers with equal force—“because that does happen.”

Groups on social media are a powerful force and powerfully alluring to the sites hosting them. As often in real life, when people get together someplace online, they’re more engaged and spend more time there than they might solo: More engagement, more time—it’s what social networks like Twitter crave. (The longer users stay on a site, the greater number of ads they’re shown; if they’re particularly tuned in, as they might be through interacting with a group, they’re increasingly likely to notice those ads.) It’s the same for Facebook. In 2019, it announced a redesign putting groups at the center of its core app. Facebook Groups would “end up creating a more trustworthy platform,” CEO Mark Zuckerberg said at the time. “Everywhere you can see and connect with friends, you’ll be able to see and connect with groups; it’s going to be woven into the fabric of Facebook.” Other platforms prospered through similar focuses, including Reddit and Discord, both valued at around $10 billion and for their ability to virtually convene people en masse.

But allowing people to form groups online can be unquestionably dangerous, and they’ve played a role in countless violent events and misinformation campaigns during the past few years: the 2017 Charlottesville protests, the 2020 Stop the Steal efforts, the Jan. 6 insurrection—any list of these incidents could go on and on and still remain unexhauastive. Angelo Carusone, president of Media Matters for America, a liberal tech watchdog group, puts the potential outcome of such social media activity bluntly: “Group functions,” he says, “tend to produce extremism and harassment and become gravitational centers of misinformation.”

So why is Twitter adding groups? There’s a business imperative. Twitter has set deeply ambitious goals for itself in the next few years after years of diminished expectations and lackluster new features. By September 2023, it hopes to have at least 315 million daily users, an approximately 40% increase from current levels. Under the same projections, sales would rise to $7.5 billion, from $3.7 billion last year, while it doubles the number of features it rolls out. Those features need to increase engagement and users, so groups could seem like a natural addition. With this mindset, Twitter, which wouldn’t comment for this story, launched a small test bunch of Communities on Wednesday. They’re an important test for Twitter: Can they produce the desired user and revenue growth without the company committing the same mistakes that’ve tripped up Facebook and other competitors?

One feature of Twitter Communities immediately separates them from groups on other platforms. They’re all public. While public groups can be misused—PizzaGate largely happened in open forums, and it’s where the right-wing Proud Boys first organized—misinformation researchers generally say private ones are the most problematic, operating in secret, away from other users who might report them.

The Communities have, for now, a couple other added safety features. They’re invite only and small, most less than 1,000 people. Twitter HQ must approve any requests to form new Communities as well as the moderators, the users who control the group and its content. Twitter handpicked the initial moderators, selecting what to be veteran Twitter influencers with clean reputations: Folks who the company can trust to marshal a community without worrying they’ll lead it to eventually broadcasting Qanon theories. And put bluntly, the first few Communities revolve around extremely benign topics, ones unlikely to court controversy or attract misinformation. Along with Lewis-Herbert’s astrology crew, there are others devoted to beauty, weather and dogs. The most possibly contentious Community topic? One for NFTs, the burgeoning asset class of digital artwork and collectibles.

But small, invite-only groups built on a few pleasant topics aren’t likely going to produce the type of growth Twitter wants to see. In a series of tweets announcing Communities, the company made it pretty clear that the groups will eventually open their gates to the public:

“That's very enticing for us,” says Matthew Nelson, 24, of Los Angeles. He’s the founder of @WeRateDogs (followers: 9 million) and a moderator of the new DogTwitter Community. “It’ll spread like wildfire. I think that dogs are incredibly popular, and once people get wind that their entire Twitter experience could just be dogs by staying within that Community, it'll grow like crazy.”

The influencers-turned-moderators could be well incentivized for the groups to expand. It’s unclear whether Twitter will directly add features to the groups to allow moderators to directly monetize their roles. But the site is already hastily adding many new revenue streams, such as Super Follows, where influencers can add a paywall to their tweets, and Spaces, which are ticketed audio events. It’s not hard to imagine a moderator looking at a Community as a means to funnel new follows to their paywalled content, their Spaces events or their main Twitter feed, where they can market themselves even more fully.

In all, the existing model for Communities is fairly safe and fairly limited. Expanding is what could trip up Twitter. Communities could certainly launch faster if Twitter switched to allowing anyone to create one about anything while letting anyone be a moderator. But the high-touch effort of approving each group and moderator is what’s Twitter on solid footing right now.

The thorniest issue will be whether Communities will become algorithmically recommended to users. Such algorithmic-based feeds are what has gotten Facebook and YouTube in trouble, those streams sending users down increasingly dark holes. Facebook knows this is a problem: After the Capitol riots in January, it decided to permanently stop recommending political and civic groups, something it had done leading up to the presidential election. Perhaps Twitter will try to strike a similar balance: recommending some Communitites—the ones about safe topics like weather, dogs, astrology—but not every Community.

“It’s just a matter of waiting and seeing,” says Katie Paul, director of the Tech Transparency Project, a think tank partly devoted to tracking online misinformation. “Only time will tell us if Twitter is willing to take the careful measures to ensure it's not used for primarily nefarious purposes, like what we've seen with Facebook Groups.”