X Killed Communities. Acorn Wants $150/Month to Run Yours on Your Domain.

X is shutting down Communities on May 30, 2026, after the feature reached less than 0.4% of users while generating 80% of the platform's spam reports, financial scams, and malware, according to Head of Product Nikita Bier. Acorn, a community infrastructure product from Blacksky, debuted on May 4 with a $100 to $150 monthly price and tooling that runs on a brand's own domain. The signal for marketers: the second platform in 18 months has yanked a community feature you were told to invest in.

0.4% of users, 80% of the spam, all of the brand managers

That ratio is what made the kill inevitable. Bier's framing, calling Communities a "Temu version of subreddits," landed because the math says so. A feature that touches a fraction of a percent of users but produces most of the unwanted activity is a moderation tax, not a product. According to Bier, the few Communities that actually worked were "user-acquisition channels for Kick or compensated clipper communities," which is a polite way of saying clip farms and stream funnels.

Communities launched in 2021 under the old Twitter brand, sold to brands as the answer to building owned audiences inside the platform. Five years on, the feature is gone, and the migration path is XChat group chats capped at 500 members (1,000 coming soon, per Bier). For anyone who built a content calendar around posting into a Community, the calendar evaporates on May 30.

This is the second time since late 2024 a major platform has pulled a community surface brand managers were told to lean into. Meta deprecated Groups inside business pages with limited fanfare, Reddit kept changing the rules on subreddit promotion, and now X is calling the whole experiment dead. From what I have seen, the pattern is consistent: community features ship, brand teams build playbooks around them, the feature gets a usage problem the platform cannot or will not solve, and the surface gets sunset two years later.

Acorn shipped 11 days after the kill notice. That timing is the tell.

Acorn's launch post on Blacksky's site is dated the same week as TechCrunch's coverage of the launch, which is May 4. X announced the Communities shutdown on April 23. Eleven days between "your community is being deleted" and "here is a paid replacement that runs on your domain" is not a coincidence. Blacksky has been building this for months. The shutdown gave them a launch window.

Lead engineer Rishi Balakrishnan and the Blacksky team built Acorn on the AT Protocol, the same decentralized infrastructure that powers Bluesky. The product is named after the seed-bank community in Octavia Butler's Parable of the Sower, which is a slightly more dignified literary reference than most Series A pitch decks I have read this year. The tooling is concrete: custom community homepages on your own domain, starter packs for onboarding, customizable feed tabs for announcements and resources, a reputation system with badges, moderation built on Bluesky's open-source Ozone interface, and the option to run your own Personal Data Server.

Three communities are already paying for it: Latinsky, Medsky, and a filmmaker community called The Invite. Blacksky says it is in active discussions with media companies and nonprofits.

What $150 a month actually buys you (and what it does not)

The honest version. $100 to $150 monthly is the average customer price right now, scaling later through a tiered SaaS model based on community size and tooling depth, TechCrunch reported. At that price you get infrastructure: hosting on your domain, moderation tools, custom feeds, governance pipelines for things like community polls. What you do not get is reach. AT Protocol communities sit on a much smaller base than X or Reddit, and Bluesky itself only crossed 30 million accounts earlier this year.

This is the trade. Reach in exchange for control. The reach number is going to look bad in a pitch deck against the 600 million monthly active users X claims, and brand managers chasing top-of-funnel will run the math and walk away. From what I have seen, that is the wrong frame for community in 2026.

The brands that got real value from Reddit, Discord, or X Communities were never running them as awareness channels. They were running them as retention channels, support channels, or feedback channels. For those use cases, owning the moderation queue, owning the member graph, and owning the URL is worth more than having access to 600 million users who will never see your post.

I think most teams will overcomplicate the decision. The simpler question: do you have at least 1,000 people who actively want to talk to you and to each other? If yes, you can probably justify $150 a month. If no, you do not have a community problem, you have an audience problem, and Acorn is not the fix.

The case studies are quiet because nobody has run this play yet

Here is the part that is genuinely uncertain. Acorn is days old. There is no "we moved our 50,000-member community off X and ROAS went up" case study because nobody has done that yet. The three current customers (Latinsky, Medsky, The Invite) are AT Protocol-native communities, not migrants. The first wave of brand migrations from X Communities is going to happen in the next four weeks, and most of those teams will pick the easy option, which is XChat group chats inside X. That keeps them on the same platform that just told them their last community surface was 80% spam.

For a related read on owned versus rented audience, see our breakdown of viral campaigns that worked on zero paid spend. The pattern there is the same one Acorn is selling: own the asset, run it on your terms, accept that distribution costs more.

How to triage your community footprint before May 30

Three actions, in order, this week.

Audit what you actually have on X Communities. Pull the member count, the post-per-week rate, and the engagement rate. If you have under 500 active members and under 20 posts a week, the answer is to let it die. Most brand-run Communities will not clear that bar. Migrating 200 ghosts to XChat is busywork.

Run a one-month test on Acorn or a self-hosted alternative. $150 a month for one month is cheap diligence. Spin up a community for your top 100 customers, give it a custom domain like community.yourbrand.com, and see whether they show up. A 30-day pilot tells you more than six months of strategy decks. The pilot answers one question: do these people want to talk to each other when you are not paying X to put them in the same room?

Stop greenlighting any community work that lives entirely inside one platform's walls. The lesson from X Communities, Meta Groups, and Reddit's API price hike is the same. Anything you build on rented community land has a five-year half-life, on a good day. If a roadmap item only works while a single platform decides to keep the feature alive, push it down the priority list.

The Acorn pricing might be wrong. The AT Protocol might lose its current momentum. Bluesky might never matter for B2C reach. None of that changes the underlying math: the cost of building on a platform that can pull the rug is now visible in three different post-mortems, and the next one is on the calendar for May 30.

I do not think Acorn wins this category. There will be five products like it within a year, two of them from existing martech vendors with bigger sales teams. What Acorn does is mark the moment when "your own domain, your own moderation, your own community" stopped being a developer's side project and started being a $150 line item a marketing director can actually approve.

Notice Me Senpai Editorial