X Is Killing Communities. IShowSpeed's 100K Group Averaged 287 Active Users.

X is shutting down Communities on May 30, 2026, after admitting the feature reached less than 0.4% of users while producing 80% of platform spam, scams, and malware. Head of product Nikita Bier disclosed that IShowSpeed's 100,000-member community averaged just 287 active users in April. Brand admins have five weeks to migrate to XChat group chats, capped at 500.

The 0.4% That Made the Decision

Communities launched in 2021, back when the company was still Twitter, and never cracked half a percent of the user base in the five years since. That is an extraordinary number for a feature with that much surface area inside the app. Premium is also a minority product at X, and it reportedly clears somewhere in the 0.6 to 0.8 percent range depending on whose estimate you trust. Communities was smaller than the paid tier, and it was the primary spam vector.

Bier put the number bluntly: 80 percent of spam reports, financial scams, and malware on X came out of Communities. A feature that reaches 0.4 percent of people but produces 80 percent of the incidents is a ratio no product team can defend for long.

This part surprised me. I had assumed Communities was neglected because X was busy shipping XChat, AI-powered Custom Timelines, and Grok integrations. It was actually the opposite. Bier said Communities was consuming "half the product team's time some weeks." Half, for 0.4 percent.

Why "100,000 Members" Meant 287 Active People

The sharpest moment in the whole announcement was Bier's reply to IShowSpeed. Speed pushed back publicly, saying he had 100,000 members in his community and it was valuable for reaching them. Bier disclosed that Speed's community averaged 287 active users per day in April.

287.

If you have ever built an audience on any platform, this is the number that should stop you. Member counts are sign-up metrics. They are not audience metrics. Speed's community was 0.287 percent active, and that is for a tier-1 creator with tens of millions of followers. Brand-built communities usually perform worse than creator-built ones, so whatever your member count looked like in the admin dashboard, the honest reading is probably south of 1 percent.

This is the same math that quietly kills most Facebook Groups, most Discord servers you got invited to, and most Slack communities that brands launched in 2022. The sign-up is cheap. The recurring engagement is the product, and most community features do not produce it on their own.

The Brands That Actually Built on Communities Lose Twice

Bier said the quiet part openly: most of the successful Communities on X were user-acquisition channels for Kick or compensated clipper communities paying out Creator Rev Share. The legitimate brand use case, the thing every community-led growth post of the last three years promised, was not what Communities mostly produced.

If you were a marketer who bought the community-led growth thesis and picked X Communities as the surface, you lose the platform and the build time. The compensation is zero. Your members will not port to the next tool. They are just X users again, sitting inside an algorithmic feed the same as everyone else.

The lesson is not "do not build community." Notion still runs community as one of its largest acquisition channels, and the math there still works. The lesson, I think, is to not build community inside a feature you do not own. Every marketer who built on Twitter Lists, Facebook Groups, Google+, Vine, or Clubhouse already learned this. X Communities is just the current entry on the list.

XChat Group Chats Are Not the Replacement X Is Pretending They Are

X's migration pitch is simple: create an XChat group link, pin it to your Community, done. The cap started at 350 members, got raised to 500 after backlash from KSI, xQc, and IShowSpeed, and X says it will reach 1,000 in coming weeks.

This is not a Community replacement. It is a different product with a different engagement model. Communities had a public-feed metaphor. XChat groups have a group-chat metaphor. If your Community was a place where a few hundred people engaged daily against a feed of posts, a 500-person group chat will likely feel worse, not better. Group chats tend to collapse into a small core of high-posters and a long tail of lurkers who never unmute, and the higher the cap climbs, the faster that collapse happens.

xQc argued publicly that XChat will fail and that removing Communities takes the human element out of the platform. He is not entirely wrong about the engagement model mismatch. A group chat is not a community feed, and the behavior you were building against was feed behavior.

On paper, raising the cap from 350 to 500 to 1,000 sounds like an upgrade. In practice, it probably just pushes more users into a product that most of them will mute within a week.

The 10-Minute Migration (If Your Community Is Actually Active)

If you run a Community with any real activity on it, here is the playbook:

1. Open XChat. Create a group chat. Toggle the joinable link option.
2. Pin that link to your existing Community page before May 30.
3. Post the link across your own X account, your email list, and any other channel where you actually own the audience relationship.
4. Export nothing, because X does not give you a member export. The only tool you have is publicizing the new destination.
5. Assume 20 to 40 percent of "members" follow you across. That matches the porting rates most brands saw on Google+ to YouTube, Vine to TikTok, and Clubhouse to Twitter Spaces.

If your Community was inactive, do not migrate it. A dead Community turning into a dead group chat does nothing for you and publicly signals deadness on a more active surface. The right move is to write it off and invest the migration time somewhere else.

A useful comparison is what Dove, Netflix, and Nike are doing on Reddit. They are treating subreddits as places to listen more than post, which is a very different discipline from the feed-within-a-feed behavior X Communities encouraged. If your brand budget for "community" was anchored on X, it might be worth reallocating that time to a surface where restraint is still rewarded.

Audience Math, Not Platform Loyalty

The thing worth keeping from this announcement is the number 287. For a long time, X creators and brand managers have cited Community member counts in decks as if they meant audience. X just made it public that they do not. From what I have seen in most internal reports, "members" and "subscribers" still get used as if they are equivalent to reach, and this week's news is a cheap, high-visibility way to kill that assumption.

The LinkedIn 360Brew change from a few days ago pointed at the same thing from a different angle: the platform decides what "audience" means, and the definition can shift overnight. The platform graveyard is long, and X Communities will not be the last feature shut down this year. What protects you, I think, is building your audience inside something you can actually move. An email list, a phone list, a logged-in app. Everything else is a short-term lease on attention, and the landlord has always reserved the right to take the unit back on five weeks' notice.

Notice Me Senpai Editorial