How to Fix a Shadowban on Twitter (X)
How do I fix a shadowban on X?
Check which parts of your account are restricted. Read any notice from X, complete the steps it requests, stop unwanted automation and repeated promotional messages, then rerun the same visibility checks. Only X can lift restrictions it applies.
Start with the check that fails and your account notice from X. The steps below explain what to do and how to verify the result. A visibility check measures what is shown; it does not reveal the internal reason for a restriction.
Not sure which restriction you have? Test your account first.
Fixing a search suggestion ban
A search suggestion ban means your account stops appearing in the dropdown when someone types your handle, even though your profile still works fine.
- Sign in to X. Complete any account verification or other step requested in its notice.
- Check your privacy settings. Keep posts protected if you want them available only to approved followers.
- Revoke access for apps you do not recognize or no longer want. Keep the legitimate tools you chose to use.
- Type your exact handle in search from an account that does not follow you. Record the date and result, and repeat under the same viewing conditions.
Keep in mind:
- Suggestions vary by viewer.
- A missing suggestion does not reveal its cause or duration.
- If X shows a decision you believe is wrong, use the review route in that notice.
Fixing a search ban
A search ban describes posts missing from the search results checked. The result depends on the query, period and viewing conditions; it does not establish what every person sees in every search.
- Read your account notice in X and complete its instructions. A requested step may have to be completed before a stated waiting period begins.
- Open the direct links to the missing posts. Confirm that the posts still exist and are public.
- Stop repeating the same promotional text across posts. Remove unwanted posting access and make future messages specific to their context.
- Search from:yourhandle in the Latest tab from a second account. Use the same settings and date range, and record which posts appear.
Keep in mind:
- Save the missing posts and their dates before making changes.
- Recheck old and new posts separately.
- A missing result does not establish whether the cause concerns one post or the account.
Fixing a ghost ban
A ghost ban describes a reply that is absent from a conversation in the viewing context checked, even though its direct link still works.
- Save the direct link to a missing reply. Open its parent post and confirm that the conversation still exists.
- Open the conversation from an account that does not follow you. Expand hidden reply sections and keep the viewing settings consistent.
- Stop sending the same promotional reply across conversations. Write replies that address the post, and stop apps from sending replies you did not intend.
- Record the reply link, date and result. Repeat the same check to see whether that reply returns to the conversation.
Keep in mind:
- Keep the reply and its conversation for comparison.
- A missing reply does not expose the internal cause.
- Deleting unrelated replies is not a demonstrated way to restore visibility.
Fixing reply deboosting
Reply deboosting describes a reply appearing lower in a conversation or behind an expanded section. Record the placement and viewing context before treating it as evidence of a persistent restriction.
- Find a specific reply from an account that does not follow you, including any expanded reply sections.
- Check more than one conversation. Record both the reply position and the viewing settings.
- Replace repeated promotional replies with messages that address each conversation. Posting more copies does not make them more relevant.
- Repeat the checks under the same viewing conditions. Distinguish a reply that is visible but gets few views from a reply that cannot be found.
Keep in mind:
- Reply order depends on the viewer and conversation.
- Volume alone does not identify a spam label.
- A low position in one thread does not establish an account-wide restriction.
What causes a shadowban: 13 patterns and conditions to review
X's published recommendation code documents parts of its For You ranking, visibility filtering and labeling systems. The categories below explain those mechanisms and what remains unknown.
A rule in that code does not prove it caused your search or reply result, and an absent expiry does not prove a permanent restriction. Use this list to investigate the observations, not to assign yourself a label.
- Reply patterns to review. X's published systems examine reply context and activity. Repeated or generic replies are worth reviewing, but our sample does not reveal a spam label.
- Repeated wording. Our sample can reveal repeated wording. X also labels clusters of similar posts, but a match in your sample does not prove that X applied that label.
- Links to review. X's published rules use internal link-reputation verdicts. We can show links in your posts, but cannot see whether X considers a destination unsafe or low quality.
- Mentions and following to review. Repeated promotional tags and unwanted follow activity deserve a review. X's link-and-mention rules also require signals that our public sample cannot establish.
- Posting activity to review. X's behavior models read internal action sequences. Posting times alone can show regularity or volume, but cannot establish whether an account is automated or penalized.
- Reports only X can assess. Some published models use spam reports and engagement signals from outside the follower network. We cannot measure those internal inputs from follower count or public posts.
- Sensitive content to review. X can limit recommendations or show warnings for sensitive content. Public text and profile markers do not reveal the labels applied to your media or account.
- Engagement requests to review. X's published policy pipeline can limit engagement bait and farming. Our text patterns identify requests to review, not an internal policy decision about your post.
- X notices and review options. X can apply post restrictions, account labels or suspension after a rules decision. A failed visibility check alone does not identify the decision or policy involved.
- Account security to check. A compromised or read-only account status can affect recommendations. Only the account owner can see the notice and complete the required security or review steps.
- Audience context, a built-in limit rather than a sanction. Account size and history affect some recommendation inputs, but a small audience is not itself a restriction and no follower threshold unlocks all distribution.
- Reach trend to review, a built-in limit rather than a sanction. Views can change while visibility checks keep passing. X personalizes recommendations, and our public view counts cannot reveal the audience predictions behind that ranking.
- Visibility settings and context, a built-in limit rather than a sanction. Privacy, reader relationships and the surface being checked can explain different views of a post. These conditions are separate from a measured restriction.
What is a reply spam flag on X?
X's published systems examine reply context and activity. Repeated or generic replies are worth reviewing, but our sample does not reveal a spam label.
What to review:
- Repeating the same response across unrelated conversations.
- Publishing many replies close together without adding context.
- Using reply automation or a reusable message that does not answer the conversation.
- Answering your own reply inside another account's thread, when that thread's root author has more than 1,000 followers.
- Sending a reply that a published flow scores at zero: that score is never raised back for the message it was written on.
What the published code says about duration: Some published labeling paths use 14 days or 30 days; another does not expose its timer. We cannot identify your label or its start time, so these are not recovery dates.
What does not help: Deleting unrelated replies or repeatedly reposting them to reset a score. Treating pasted text, reply volume or a published timer as proof of the cause.
What is the duplicate content flag on X?
Our sample can reveal repeated wording. X also labels clusters of similar posts, but a match in your sample does not prove that X applied that label.
What to review:
- Publishing near-identical messages repeatedly.
- Reusing the same closing paragraph across several posts or replies.
- Changing only a ticker, handle or link in an otherwise repeated message.
What the published code says about duration: The published duplicate-label calls show no expiration argument. Other removal or review paths may exist outside the published code; no recovery date can be inferred.
What does not help: Deleting posts you merely suspect are flagged. The code does not establish that this clears a restriction. Reposting the same message with tiny variations as a reset.
Do links to flagged domains cause a shadowban?
X's published rules use internal link-reputation verdicts. We can show links in your posts, but cannot see whether X considers a destination unsafe or low quality.
What to review:
- A destination or redirect chain marked unsafe by X's internal systems.
- A pinned post linking to a destination with a bad reputation verdict.
- Repeatedly sharing a destination that is misleading, broken or unsafe.
What the published code says about duration: One published pinned-link pathway writes a 7 day account label. Other link-label calls expose no timer. We do not know the verdict, label or time of application for your account.
What does not help: Deleting unrelated posts or removing every legitimate link as a supposed reset. Treating a shortened URL or the number of links as proof that X flagged a domain.
Can mass mentions or follow churn cause a shadowban?
Repeated promotional tags and unwanted follow activity deserve a review. X's link-and-mention rules also require signals that our public sample cannot establish.
What to review:
- Repeatedly tagging people who are not part of the conversation.
- Combining unsolicited mentions with a destination X classifies as low quality.
- Tools sending unwanted follows or repeated promotional interactions.
What the published code says about duration: Some published mention-and-link calls expose no timer, while one pinned-link account pathway uses 7 days. Neither provides a recovery date for a pattern seen in our sample.
What does not help: Following a magic mentions-per-post limit. A review threshold is not a safe quota. Deleting suspected posts or mass unfollowing as a way to clear a label.
Does automation trigger a shadowban on X?
X's behavior models read internal action sequences. Posting times alone can show regularity or volume, but cannot establish whether an account is automated or penalized.
What to review:
- Tools sending interactions you did not intend.
- Repetitive replies, likes or follow activity generated in bulk.
- Regular posting intervals, which can also come from legitimate scheduling and need context.
What the published code says about duration: The published behavior pipeline does not establish a recovery timer we can apply to your account. Its decision thresholds and some enforcement details are not public.
What does not help: Randomizing automated actions to look human instead of stopping unwanted activity. Assuming a scheduler, regular timing or high posting volume proves a bot label.
What is the report ratio flag on X?
Some published models use spam reports and engagement signals from outside the follower network. We cannot measure those internal inputs from follower count or public posts.
What to review:
- Spam reports recorded by X from people outside the follower network.
- Internal engagement counts used alongside those reports.
- Previous enforcement signals available to X but absent from our scan.
What the published code says about duration: Two published Agatha paths write a 7 day label. This does not show that your account has that label, when it began, or whether another restriction applies.
What does not help: Organizing likes, replies or reports to compensate for an assumed internal ratio. Blaming a small follower count or a suspected report wave without evidence.
Do adult content flags cause a shadowban?
X can limit recommendations or show warnings for sensitive content. Public text and profile markers do not reveal the labels applied to your media or account.
What to review:
- Media identified by X's review systems as sensitive content.
- Sensitive media included through a quoted post.
- An account or media setting, or a notice shown to the account owner.
What the published code says about duration: Two published aggregation paths each use 7 days and exclude accounts already carrying their label. Other paths differ. We cannot infer your label, its renewal or a recovery date from the scan.
What does not help: Disabling an accurate sensitive-content label to increase reach. Deleting suspected media or counting seven days from a scan as a reset.
What is the engagement bait flag on X?
X's published policy pipeline can limit engagement bait and farming. Our text patterns identify requests to review, not an internal policy decision about your post.
What to review:
- Making useful information conditional on a reply, tag or share.
- Repeating follow-to-see or share-to-unlock messages.
- Trading engagement or requiring it to receive a reward.
What the published code says about duration: The published label call for this category exposes no expiration. The scan cannot identify that label or an expiry date; absence of a timer here does not prove permanence.
What does not help: Treating every question or ordinary call to action as prohibited engagement bait. Buying or coordinating engagement to offset an assumed restriction.
Does a policy strike cause a shadowban on X?
X can apply post restrictions, account labels or suspension after a rules decision. A failed visibility check alone does not identify the decision or policy involved.
What to review:
- A policy notice attached by X to a specific post.
- An account restriction or suspension notice shown when you sign in.
- A security or verification challenge that X asks you to complete.
What the published code says about duration: Some published enforcement paths use a 30 day label; others expose no timer. These are separate paths, and the scan cannot establish which applies or when a decision ends.
What does not help: Deleting unrelated posts or submitting the same unsupported explanation as a reset. Treating a published label duration as the deadline for X to answer an appeal.
What is an account security hold on X?
A compromised or read-only account status can affect recommendations. Only the account owner can see the notice and complete the required security or review steps.
What to review:
- Sessions or applications the account owner does not recognize.
- A compromised-account or password-reset notice from X.
- A read-only or authenticity notice with a specific instruction.
What the published code says about duration: The published visibility rules read account status, but do not establish a recovery deadline. Completing the matching step and seeing a changed status are separate observations.
What does not help: Deleting unrelated posts or waiting for an arbitrary timer instead of following a notice. Sharing your password or session cookies with a service promising to restore visibility.
Is a small or new X account shadowbanned by default?
Account size and history affect some recommendation inputs, but a small audience is not itself a restriction and no follower threshold unlocks all distribution.
What to review:
- A small follower network, measured from the public profile.
- A recently created account with limited posting history.
- A small sample of views that makes changes difficult to interpret.
What the published code says about duration: There is no restriction timer to calculate from follower count or account age: neither is a recovery threshold in X's published rules.
What does not help: Buying followers or creating a new account to cross a supposed recovery threshold. Reading fewer than 400 followers as proof that X cannot recommend your posts.
What is predicted audience rejection on X?
Views can change while visibility checks keep passing. X personalizes recommendations, and our public view counts cannot reveal the audience predictions behind that ranking.
What to review:
- A decline in sampled views across several dated observations.
- Different post formats or topics with different audiences.
- Recent posts with less time to accumulate views than older examples.
What the published code says about duration: A reach trend is not a timed label. The published ranker scores candidates for a viewer, and public counts do not tell us when or whether distribution changes.
What does not help: Deleting a low-view post or coordinating engagement to reset an assumed ranking penalty. Using model weights as a formula that converts one report into lost likes.
What X design limits look like a shadowban?
Privacy, reader relationships and the surface being checked can explain different views of a post. These conditions are separate from a measured restriction.
What to review:
- A Protected account setting that limits access for non-followers.
- A reply or repost being considered by a For You recommendation pipeline for a non-follower.
- A changed post version, unavailable parent post or country-specific notice.
What the published code says about duration: These conditions do not provide a recovery timer. Their effects depend on the surface, reader and current state; a public check cannot describe every reader's personalized feed.
What does not help: Treating every low-view post as a shadowban or reposting the same content to reset it. Applying an internal candidate-pool limit as a safe daily posting quota.
Common mistakes
- Deleting your posts or your whole account. You may lose useful history and evidence without resolving the actual issue. First read any notice from X and identify the specific change you intend to make.
- Creating several backup accounts "just in case". Creating another account does not resolve a decision on the original one. Keep the focus on the account notice and the visibility problem you can reproduce.
- Paying for a "shadowban removal" service. No third-party payment guarantees removal of an X restriction. Only X can lift restrictions it applies.
- Assuming every result needs the same appeal. Read the exact notice first. X offers review routes for decisions you believe are wrong. Relevant URLs, dates and an existing case reference are more useful than an unsupported claim about an internal label.
- Tweeting "am I shadowbanned" and asking your followers to check. Their observations can help, but followers may see a different view from strangers. Record who can see which content, and compare with a second account that does not follow you.
How long it takes
There is no universal recovery timer. Some paths in X's published code write 7, 14 or 30 day expiries; other paths leave the duration outside the published code.
A public check does not reveal which path applies or when its timer began. An account notice may also require you to complete a step before its stated period starts. See what is known about shadowban duration for the limits of these estimates.
How you'll know it's over
Check whether the missing posts return to search and the missing replies return to their conversations. Compare the same checks, dates and measurement coverage. A passing result confirms the surface tested at that time; low views alone do not mean the restriction persists.
For a personal audit, practical actions and ongoing checks, see pricing.
Frequently asked
Which restriction do I actually have?
Run the four visibility checks: profile suggestions, post search, reply visibility and reply placement. Test from an account that does not follow you, or use the free checker. Read the result of each check and its measurement coverage before choosing an action.
Does deleting my account and starting a new one fix it faster?
No proven shortcut comes from deleting your account. Starting over does not repair the original account and discards its history and connections. Keep the account, read the notice from X and address the specific problem.
Can I pay someone to remove a shadowban?
No third-party payment guarantees removal of an X restriction. Only X can lift restrictions it applies.
Will X notify me when it's fixed?
Not always. X shows notices for some account restrictions, but a change in search or reply visibility may arrive without a message. Rerun the same checks and compare their dates and results to see what changed.
I stopped the behavior that caused it. Why is it still there?
Stopping a pattern does not confirm that X has lifted a restriction. Complete any outstanding step in the account notice and keep checking the affected surface. A public scan does not reveal which internal label applies or when it started, so a code timer is not your recovery countdown.
What causes a shadowban on X?
Spam-like activity, rule violations, account security conditions and visibility settings can limit distribution on X. The 13 categories below identify patterns and conditions to check. X's published code covers For You recommendations; search and public reply threads are different surfaces. These mechanisms do not establish the cause for a particular account.
Does deleting posts fix a shadowban?
Deleting posts at random is not an established way to restore visibility. Follow the process if X names a specific post in a notice. Otherwise, save the relevant links and dates, stop the repeated pattern in future posts and compare the next checks. Removing a post does not prove that an account restriction has been lifted.
Test it yourself
Run the 4 checks on any handle
More on each restriction
Last reviewed