Article
Followers Still Matter on X. The Number Mostly Doesn’t.
9 Oct 2026
X no longer shows your posts to most of the people who follow you. Reach now depends on whether the system predicts someone will stop, reply, share, or follow—not on how many people once clicked Follow.
X no longer shows your posts to most of the people who follow you. Reach now depends on whether the system predicts someone will stop, reply, share, or follow—not on how many people once clicked Follow. That is why accounts with hundreds of thousands of followers routinely reach only a fraction of a percent of them.
The follower relationship has not been deleted. It has been demoted from a delivery guarantee to a seed.
What actually broke
The old model was simple: people followed you, so they saw you. That model is gone. Candidate selection pulls from accounts a user follows and from accounts they do not. Ranking then scores predicted actions. In the current open-sourced weights, a copied-link share dwarfs a like. Replies and quotes sit far above passive approval. A report can bury a post. Early engagement velocity matters more than the size of the list attached to the author.
Users describe the result in plain terms. Large accounts call their follower lists souvenirs. Creators report that people who opted in still say they never see the posts. Product comments match the complaints. Nikita Bier has noted that people overestimate how many of their followers are active in any short window and get frustrated when a post fails to reach all of them. X added active-follower counts for that reason. Elon Musk has described the goal as maximizing time people do not regret spending, and has said strong posts from small accounts should surface more often.
What did not get deleted
In-network posts are still retrieved. Models still estimate whether you and an author actually interact. A 2026 change restored a mutual-follow boost in replies because that signal had been missing and friends were losing to strangers. A predicted follow of the author remains one of the stronger positive weights in the ranker. Small-account lifts exist in the code, but they are narrow and temporary. They do not erase the advantage of an account whose readers reliably reply and share.
Off the pure math, the number still does work. Visible scale is social proof. It affects who follows next, who answers, and who pays for attention. The accounts that dominate daily conversation are not small. Elon’s earlier point—that verified followers would be the durable metric—was a warning about fake totals, not a claim that audience size stopped mattering.
What to do with this
Treat the displayed count as a lagging scoreboard. The working asset is the active neighborhood underneath it: mutuals who reply, readers who dwell, posts that earn follows rather than passive likes.
Chasing a higher raw total does little for distribution if those followers are inactive. An account with fewer followers but real early engagement will usually outperform a large dormant list. Ratio hygiene and posting volume still matter at the margins—flooding the single author slot in followers’ feeds can suppress your own earlier posts.
The useful distinction
Followers matter as the seed of distribution, the source of the highest-confidence engagement predictions, and the external signal of credibility. They do not matter as a guaranteed impression quota.
Blizzard! $35. Buy