Shadowbanned on X? How to fix it, step by step

So a check came back with something restricted. Before you change anything about how you use X, the useful thing to know is that recovery is mostly subtraction. There is no button, no form, and no growth hack. You stop the thing that tripped it, you keep being visibly human, and it lifts on its own, usually within days to a few weeks.

The part people get wrong is the middle. They panic, go silent, delete posts, and then wonder why nothing improved after a month. Going quiet does not speed anything up. Here is the order that actually works.

Step 0: make sure it is real

One empty search result is not a shadowban. Search is inconsistent, reach swings hard week to week, and a bad week looks identical to filtering from the inside.

Two things are worth doing before you accept the diagnosis. Open X's own account status page, under Settings and privacy → Your account → Account status, which reports formal enforcement actions in X's own words. And re-run a real check a day later, because filtering is persistent by nature. A signal that vanishes overnight was noise.

Our free shadowban checker runs the search, suggestion and reply tests for you and tells you when it cannot answer rather than guessing. The full explanation of what each test means is in am I shadowbanned on X.

If the account status page reports an actual enforcement action, stop reading here and appeal it. That is a different problem with a real appeal path, and waiting it out is the wrong move.

Step 1: find the trigger, today

Visibility filtering is almost never about one post. It is about a pattern that looked automated. Look honestly at the last week or two and find yours:

• A burst. Forty or fifty replies inside an hour reads as a script no matter how good the replies were.
• Repetition. The same reply, or near-identical replies, across many posts. This is the one that catches people using AI carelessly, and it is by far the most common cause we see.
• Follow and unfollow churn. Cycling follows to farm reciprocity is exactly the behaviour the limits exist to discourage. More on that in why can't I follow anyone on X.
• Link volume. The same domain, repeatedly, from a young account.
• Negative signal. Enough blocks, mutes and reports from real people is its own trigger, independent of anything automated.

You need one honest answer here, because everything that follows is just "stop doing that and prove it."

Step 2: the two weeks that follow

Recovery looks boring on purpose.

Stop the specific behaviour. Not all activity. The one thing you identified.

Keep posting, at a human pace. Space actions out. A handful of thoughtful replies across a day beats thirty in an hour, and the gaps are as much of the signal as the content.

Write things people reply to. Genuine engagement from real accounts is the strongest counter signal available to you. A conversation under your post does more for recovery than anything you can do alone.

Vary your output. If repetition was the trigger, the fix is not writing less, it is writing differently. Every reply should read as though a person considered the post in front of them, because that is the thing being measured.

Wait, and check weekly. Not hourly. You cannot see day-to-day movement in filtering, and checking constantly only makes you act on noise.

Step 3: what to expect per ban type

The four types lift at different speeds, which is why a single "am I clear yet" question is not that useful.

Search suggestion ban. The mildest and usually the first to go. Often clears in days once the pattern stops.

Reply deboosting. Your replies are there but ranked below a fold. This responds directly to reply quality, so it tends to improve as soon as your replies start earning responses.

Ghost ban. Replies missing entirely from the public view. Slower, typically a week or two of consistent normal behaviour.

Search ban. The most severe and the slowest, since it is the one most often attached to something bigger than pacing. If a search ban is still in place after three or four weeks of clean behaviour, that is when the account status page and an appeal become the right move rather than more waiting.

The four things that make it worse

Going completely silent. A dormant account does not recover faster. It just stops producing the engagement that would demonstrate you are a real person.

Deleting everything. Mass deletion does not remove whatever was recorded, and it costs you the posts that were working.

Starting a new account. You throw away every follower and every bit of account age, and you carry the same habits into the new one. If the trigger was behavioural, the new account meets the same limits faster than the old one did.

Paying for shadowban removal. Nobody outside X can lift visibility filtering. Anyone selling that is selling you the wait you would have done for free, and anyone who asks for your X password to do it is doing something worse. The same reasoning as in are X unfollow tools safe applies exactly.

How to tell it is lifting

Re-run the same tests weekly, and look for structure rather than numbers. Filtering is all-or-nothing per surface: your posts are absent from search entirely, or they are not. Recovery shows up as a surface coming back, not as a gentle rise in impressions.

If your posts appear in search logged out and your replies show without a fold, you are clear, and whatever is left is a reach problem rather than a visibility one. That is a different thing to fix, and how the X algorithm works is the more useful read.

The prevention side, as a set of rules to work inside rather than symptoms to diagnose, is in how to grow on X without getting banned.

Where SpacerrX fits

Two parts of this are hard to do by hand, and they are the two we built for.

The check itself runs on demand from the shadowban checker, and inside the SpacerrX extension it runs automatically against your own account and shows the result as a small seal on your X sidebar, so you find out that something changed without ever going looking for it.

The follower checker answers a different question: who you follow does not follow you back right now. It is free and runs in your browser, but it does not store follower history or prove why reach changed. Use the visibility seal as evidence, not as a guaranteed diagnosis.