X's enforcement and search help pages never use the word shadowban, but they describe the same effects: X can keep posts out of search, trends and recommended notifications, drop them from the For You and Following timelines, restrict them to your profile and push them down in replies. You can check most of these from a second account that doesn't follow you, and when X labels a post as visibility-limited, its enforcement page says you can appeal the label.
This guide turns those pages into a self-check table, a testing method that doesn't fool you, and an audit of the media settings that quietly hide posts. Posting habits that keep promotional accounts out of trouble are in our X growth strategy guide. If your account is suspended or locked rather than merely quieter, switch to the X suspension appeal guide.
Five reasons X gives for reduced reach
X's page When a post's reach may be limited lists five situations. The first is a rules violation, where X makes a post less discoverable. The second is other people's controls: blocks, mutes and notification filters, which let users skip posts from accounts without a profile photo or without a confirmed email address or phone number. The third is quality and safety ranking, including safe search, which keeps potentially sensitive posts and posts from blocked or muted accounts out of search results, and the collapsing of replies X judges abusive or low quality. The fourth is experiments. The fifth covers legal, quality and technical problems, including posts wrongly identified as spam and service issues X announces through its @Support account.
The same page explains that ranking leans on behaviour signals. If many of the accounts you post at block or mute you, X reads that as a strong sign they don't want your posts; if people reply when you mention them, the signal runs the other way. X says no single signal is used in isolation, which is why one bad week rarely has one neat cause.
Account-level limits can look similar from the outside. X's locked and limited accounts page says an account in a temporary limited state may be filtered out of search results and notifications, and some limits leave only your followers able to see what you do.
Labels, silent filters and what each tells you
Some limits are announced and others are not. The enforcement options page lists the post-level actions: excluding a post from search, trends and recommended notifications, removing it from the For You and Following timelines, restricting it to the author's profile, downranking it in replies, and switching off likes, replies, reposts and other engagement. It says that from April 2023, posts identified as breaking the rules began receiving labels that tell both author and viewers their visibility was limited, and that authors can appeal the label.
Older guidance is weaker on that last point. X's April 2023 blog post Freedom of Speech, Not Reach described authors submitting feedback on labels with no promise of a reply, and said X was still working on appeals. The help page is the more recent and more specific source, so use its appeal route, but don't expect a quick or detailed answer.
Search filtering, safe search, reply collapsing and other people's mutes produce no label at all. That is why the self-check below tests each surface separately instead of hunting for a single shadowban switch.
Visibility self-check table
Run each test from a second account that doesn't follow you and has never interacted with you. Causes and fixes come from the X help pages linked in this guide.
| What you notice | How to test it from outside | Cause X documents | Fix or appeal route |
|---|---|---|---|
| An exact phrase from your recent post returns nothing in search | Search the phrase from the test account with safe search off, checking both Top and Latest | Search quality filtering for duplicate posts, hashtag abuse, automated replies or aggressive following | Stop the listed behaviour; X says your followers still receive your posts in the meantime |
| Searching your name or handle doesn't bring up your account | Use the People results in search on the test account | An incomplete name, username or bio, or little recent activity | Complete the profile; X says profile edits can take a few days to reach search |
| Posts with images drop out of search while text posts still show | Toggle safe search on the test account and compare the results | Media marked or labelled as potentially sensitive, which safe search excludes | Expected for labelled media; run the settings audit below |
| Your replies land at the bottom of threads or behind a show-more prompt | Open the same conversation from the test account and scroll through it | Conversation ranking, plus collapsing of replies judged low quality | Reply on topic; X's spam rules count promotional replies unrelated to a post |
| A post carries a notice that its visibility has been limited | View the post on your own account and again on the test account | A post-level action for a policy violation | Appeal from the label, as the enforcement options page describes |
| Visitors meet a sensitive-content warning before your profile loads | Open your profile from the test account with sensitive content hidden | Your sensitive media setting is on, or X switched it on after reports | Work through the audit below; X may lock the setting after repeated mislabelling |
| Some viewers can't open your media at all | Check from a test account that has no birth date on its profile | Age restriction of adult content for under-18 viewers and those without a birth date | Nothing to fix; this is how X restricts labelled adult media |
| Only followers see your activity, or you can only message followers | Log in and look for a message about limited features | A temporary limited state or read-only mode | Follow the in-app steps; our suspension guide covers each prompt |
| A post or the whole account shows as withheld in one country | Look for a withheld notice when viewed from that country | A legal request from an authorised entity in that country | X's country withheld content page explains the notice |
| Impressions fall with none of the signs above | Compare a week of analytics against any change in how you post | Ranking experiments, spam misclassification or technical faults | Check @Support for known issues and report persistent problems through X's help forms |
The search rows draw on X's search rules, which list repeated duplicate posts, trending-topic abuse, automated posts or replies, keyword bots, similar messages across several accounts and aggressive following, and its search troubleshooting page, which adds protected posts, an unconfirmed email address and sensitive content. That troubleshooting page also says X does not inject specific posts into search.
Testing without fooling yourself
- Use an account that doesn't follow you, hasn't muted or blocked you and has no history with you. Your own feed and your friends' feeds are shaped by the ranking signals you are trying to test.
- Give the test account a birth date and let it see sensitive media, or it will hide your labelled posts and make things look worse than they are. X's notices page says viewers under 18 and those without a birth date on their profile can't view adult content.
- Run each search with safe search on and then off. X's search guide says safe search excludes potentially sensitive content and muted or blocked accounts, and that the setting carries across the web and the apps.
- Check both the Top and Latest tabs, because they can return different results for the same query.
- Don't judge from a sign-in screen. If X asks a logged-out browser to sign in before showing a profile, post or search result, use the test account rather than reading anything into the prompt.
- Skip third-party shadowban checkers that ask for your login. X's third-party apps page says anyone given your username and password can take complete control of the account, and recommends connecting apps only through X's OAuth screen.
- Write down what you tested, when, and what you saw. A support request or label appeal with specifics gives a reviewer something to check.
Sensitive-media settings audit
On accounts that share mature or adult material, media settings are the most common silent reason posts go missing from other people's views. What must be labelled is set out in our X adult content policy guide; this audit only checks how the settings change who sees you. The steps follow X's media settings page.
- Open Settings and privacy, then Privacy and safety, then Your posts, and see whether Mark media you post as containing material that may be sensitive is switched on.
- If it is on, expect visitors who haven't opted in to see a warning on your profile; X says people who chose to see sensitive content see your account without it.
- If you share adult media, keep the setting on. Turning it off for reach invites reports, and X says it may then label the media and switch the setting back on for you.
- Look for a label on your account saying X changed the setting. After repeated mislabelling, X says it may label the whole account and permanently adjust the setting, so this can't always be undone.
- For an occasional mature post on an otherwise general account, use a one-time warning instead: the flag icon on an attached photo or video lets you mark it as nudity, violence or sensitive.
- Check your profile photo, header, List banners and Community cover photos. X's adult content policy bars adult content in those highly visible places and in live video, with read-only mode or suspension for repeat breaches.
- Strip unrelated trending hashtags from promotional posts, since hashtag abuse is one of the reasons X's search rules give for filtering accounts.
- Re-run the media rows of the self-check table a few days after any change, and note the date you changed it.
Appeal routes by type of limit
Each kind of limit has its own route, and some have none. A labelled post is appealed from the label itself. For an account in a limited state or read-only mode, complete the in-app steps first; the locked and limited accounts page says that if you believe the limit was a mistake you can contact X support, and account-level decisions go through the locked or suspended account appeal form. Search filtering and ranking have no appeal form at all: the fix is to stop whatever the search rules describe and wait for the filtering to lift. For apparent bugs, report through X's help forms and watch @Support.
Whichever route applies, quote what you saw: the test account, the search phrase, the date and a screenshot. A visibility complaint without specifics gives a reviewer nothing to look at.
After the limit lifts
Because ranking responds to blocks and mutes, posting promotions at strangers tends to bring limits back. X's conversations page says replies rank higher when the original author has replied, when the replier is someone the reader follows, or when the replier subscribes to Premium; our guide to getting verified on X covers what that subscription does and doesn't change. Messaging limits are set out in our X DM spam rules guide. If Reddit is also part of your promotion, the equivalent checks there are in our guide to Reddit shadowbans and bans.
Limitations of this self-check
X doesn't publish its ranking thresholds, how long a label or search filter lasts, or which experiments an account is part of, so this self-check can show where a post isn't appearing but can't prove why. A single test account brings its own settings and history, which can skew results. The help pages cited were read on 1 October 2026 and can change, and where the 2023 blog post and the current enforcement page disagree, this guide follows the current page. Be wary of any claim about secret reach scores that X itself doesn't document.