Account-level enforcement against a business account is documented by every major platform, and the documents disagree with each other about almost everything except the existence of a strike. This page reads the current policy and help pages for seven platforms, records what each one publishes about restricting an account, and separates that from the beliefs that circulate among people who run those accounts. Every citation below carries the date shown on the page and the date it was read, which was 2026-08-24 for all of them.
Two things emerged from the reading that are worth saying up front. Six of the seven platforms publish a different vocabulary for the same broad idea, so a manager who says “we got restricted” has said almost nothing precise. And the single most common fear about business accounts, the silent account-wide penalty, has no published source anywhere in the set.
What a restriction actually is, in each platform’s own words
Seven platforms publish six vocabularies for account-level enforcement. Meta uses one set of names across Facebook and Instagram. LinkedIn, X, TikTok, YouTube and Pinterest each publish their own.
A solo social media manager running a restaurant group’s Instagram professional account and Facebook Page sees a notice and cannot tell from it whether posting, advertising or recommendation eligibility is affected. That confusion is not a reading failure. The three effects sit in three different documents.
| Platform | The platform’s own published terms | What the documentation says the term limits | Source and date |
|---|---|---|---|
| strike, feature restriction, restriction from creating content, disabled account; for Pages, unpublished and reduced distribution | Meta publishes a graduated ladder: one strike is a warning, two to six strikes restrict “specific features, like posting in groups”, seven strikes bring a one-day restriction on creating content, eight strikes three days, nine strikes seven days, ten or more strikes 30 days | Restricting accounts, Meta Transparency Center, page states updated Nov 12 2024, read 2026-08-24 | |
| strike, disabled account | Meta counts strikes on Instagram but states that the graduated Facebook feature restrictions “generally only apply to Facebook accounts” | Restricting accounts and Counting strikes, page states updated Apr 23 2025, read 2026-08-24 | |
| account restriction, permanent account restriction, limited visibility | LinkedIn publishes that it “may limit the visibility of certain content, label it, or remove it entirely”, that “repeated or egregious offenses can result in account restriction”, and that egregious violations may bring permanent restriction after a single violation | Professional Community Policies (no date published on the page) and How we enforce our Professional Community Policies (page shows “Last updated: 1 year ago”), read 2026-08-24 | |
| X | suspend, terminate, limiting visibility, discontinuing access | The X Terms of Service name “removing your Content, limiting visibility, discontinuing your access to X, or taking legal action” as enforcement actions. The X Rules and the suspended-accounts help page could not be read in this session (see the source ledger) | X Terms of Service, page states Effective April 10 2026, read 2026-08-24 |
| TikTok | warning, strike, permanent ban, ineligible for recommendation, and five named account check states | TikTok’s account check publishes five separate restriction states: restricted from logging in, from posting, from commenting, from editing your profile, and from sending messages | Your account status, TikTok Support (no date published on the page), rendered and read 2026-08-24 |
| YouTube | warning, Community Guidelines strike, termination | YouTube publishes that a first strike blocks uploads, live streams, thumbnails, posts and playlist edits for one week, a second strike inside the same 90 days blocks posting for two weeks, and three strikes in one 90-day period “may result in your channel being permanently removed” | Community Guidelines strike basics on YouTube (no date published on the page), read 2026-08-24 |
| deactivation (also called removal or suspension), limiting distribution | Pinterest publishes exactly two enforcement actions. Deactivated content “is no longer available to anyone on the platform”. Distribution-limited content stays reachable by the poster and by direct link but “won’t be featured in recommendation or discovery surfaces” | Enforcement, Pinterest Policy, page states Last updated: April 2025, read 2026-08-24 |
One more piece of vocabulary worth recording, because it removes a document practitioners still cite. The URL that used to serve the Instagram Community Guidelines, https://help.instagram.com/477434105621119, resolved on 2026-08-24 with a 301 to Meta’s Community Standards at https://transparency.meta.com/policies/community-standards/. Instagram no longer publishes a separate account-level policy document. Meta’s Community Standards cover Facebook, Instagram, Messenger and Threads together.
These definitions apply to business or professional accounts on the seven platforms named. They do not describe personal-account-only enforcement paths, developer or app-level enforcement, or advertising account enforcement, and none of the documents read claims to cover those in the same place.
What the policies name as triggers
Six grounds for account-level action appear across the seven platforms’ policy documents. Grouping them by ground rather than by platform is what makes the overlap visible.
1. A single severe violation. Every platform in the set publishes that one violation can be enough. Meta’s Account Integrity standard states that violations posing severe safety risks “will lead to an account being disabled” (Account Integrity, change log’s most recent dated revision May 28 2026). TikTok lists its permanent-ban grounds explicitly, including child sexual abuse material, promoting or threatening violence, and content depicting real-world torture (Content violations and bans, rendered 2026-08-24). YouTube names “a single case of severe abuse” as a termination reason (Channel or account terminations). LinkedIn publishes that for egregious violations “we may permanently restrict your account after a single violation” (How we enforce). Pinterest publishes that it may remove an account “after a single instance of a severe policy violation” (Enforcement).
2. Accumulated violations. Each platform publishes its own accounting, and the counters differ more than the grounds do. Meta states that all strikes on Facebook or Instagram expire after one year, and that strikes are not counted for violating content posted more than 90 days ago for most violations or more than four years ago for severe ones (Counting strikes, updated Apr 23 2025). TikTok counts strikes “by policy area as listed in our Community Guidelines … or by feature”, and publishes that strikes expire after 90 days. YouTube runs a 90-day window in which three strikes may end the channel. LinkedIn publishes no counter at all.
3. Frequency and automated activity. Meta’s Spam standard is the only document in the set that names activity rate as a ground, and it names it without a number: “Posting, sharing, engaging with content or creating accounts, Groups, Pages, Events or other assets, either manually or automatically, at very high frequencies.” The same standard adds that Meta “may place restrictions on accounts that are acting at lower frequencies when other indicators of Spam (e.g., posting repetitive content) or signals of inauthenticity are present” (Spam, change log’s most recent dated revision Jun 26 2024, read 2026-08-24). LinkedIn’s spam policy names the behavior rather than the rate, prohibiting “gratuitously repetitive messages or similar content” and instructing members not to “do things to artificially increase engagement with your content” (Professional Community Policies). Pinterest lists Spam as one of four violation areas that can deactivate an account (Account deactivation).
4. Identity and authenticity. Meta’s Inauthentic Behavior standard defines its target as “a network of inauthentic assets controlled by the same individual or individuals” (Inauthentic Behavior, change log’s most recent dated revision Dec 11 2025). LinkedIn’s requirement is blunter and reaches further into ordinary business practice: “do not share your LinkedIn account with anyone else”, and do not “associate yourself on LinkedIn with a business or organization that you are not actually professionally associated with”. TikTok names deceptive impersonation of another person or entity as a permanent-ban ground. Pinterest lists account security, “including impersonation and third-party logins”, as a deactivation ground.
5. Evading an existing restriction. TikTok publishes that creating or using another account to avoid a restriction or ban on an account you own is itself a permanent-ban ground, and that this applies “for as long as the restriction remains active on your other account”. YouTube publishes the same rule for channels, extending it to “any channels in which you are repeatedly or prominently featured”.
6. Reasons that have nothing to do with content. The X Terms of Service are the clearest published example. X states that it may suspend or terminate an account where it reasonably believes the account created “risk or possible legal exposure”, where the account should be removed “due to prolonged inactivity”, or where “our provision of the Services to you is no longer commercially viable” (X Terms of Service, Effective April 10 2026).
An agency lead responsible for 18 client accounts across Meta, TikTok and LinkedIn who inherits an account already carrying a prior violation is reading ground 2, and the answer differs by platform: a Meta strike on that account is live for a year, a TikTok strike for 90 days, a YouTube strike for 90 days, and a LinkedIn violation for an unpublished period.
These are the grounds the documents name as of 2026-08-24. None of the documents claims to be exhaustive, and every platform in the set revises enforcement policy without advance notice, which is why the read date sits on each citation rather than at the foot of the page.
What is observed: the published record
The published record of individual enforcement decisions against Meta accounts sits with the Oversight Board, and two decisions from 2026 show enforcement behaving differently from the way the policy documents describe it.
In 2026-032-FB-UA and 2026-033-IG-UA, published August 19 2026, the Oversight Board overturned Meta’s removal of two posts about sexual harassment on public transport. The first post ran on the Facebook Page of a news organization in India with more than 1.1 million followers. Human reviewers and policy subject-matter experts removed it under the Adult Sexual Exploitation standard. The Board records that the content was viewed more than 60 times before removal, and that “the posting users failed in their appeals against the respective decisions”. Meta reversed itself only after the Board selected the cases, restoring both posts with age-gating and a warning screen. The Board also found that making the posts non-recommendable “was not a legitimate restriction on freedom of expression”, and recommended that Meta decouple the sensitive-content warning screen from the non-recommendable enforcement action (Posts Condemning Bus Harassment).
In 2026-036-IG-UA, 2026-037-IG-UA, 2026-038-FB-UA, 2026-039-IG-UA and 2026-040-FB-UA, published June 16 2026, the Board recorded five removals from Greece, Germany, Ecuador, the United States and India, all under the Hateful Conduct standard, all for slurs used self-referentially or in a positive context. Meta restored all five posts after the Board brought the appeals to its attention. The same decision notes that as of May 2026, 73 recommendations issued by the Board remain “in progress” (Slurs in Positive Context).
Two facts sit in those decisions that the policy documents do not carry. Meta’s own reviewers applied the same standard incorrectly across multiple reviews. And an in-product appeal failed in every one of those cases, with reversal arriving only through a body most account holders never reach.
For scale, the European Commission’s DSA Transparency Database is the only public record that spans platforms. Read on 2026-08-24, its home page reports the summary statistics it defines as covering “statements of reasons submitted by providers of online platforms to the Commission in the last six months (180 days)”: 3,703,748,843 statements of reasons in total, 365 active platforms, and 42% of decisions fully automated. The three restriction types the database lists as most frequently reported are disabling access to content, removal of content, and “other restriction (please specify)” (DSA Transparency Database).
Those figures describe the 180-day window the database itself defines, and nothing else. They support no claim about the base rate for any individual account, and the platform-by-platform comparisons need their metric definitions checked before they mean anything, which is a separate job.
What has no published source
Six beliefs about business account restrictions were tested against the 24 platform policy and help documents read for this page. Each row records what the documentation says on the point and what the search found. A finding of “no published source found” is a statement about the source record on a date, not a statement about what platforms do.
| The belief, as practitioners state it | What the documentation says on the point | Finding |
|---|---|---|
| A shadowban is a silent account-wide penalty that platforms never announce | TikTok publishes a named state, ineligible for recommendation, and states it notifies the account in inbox notifications and on the profile. Meta publishes reduced distribution in Feed and surfaces “content shared by your Page that isn’t eligible for suggestions” inside Page Status. Pinterest publishes limiting distribution as one of its two enforcement actions | No published source found. The word shadowban appears in none of the 24 platform policy and help documents read on 2026-08-24. Every reach-reduction mechanism found in the set is named and notified |
| There is a safe number of posts per day that keeps a business account out of trouble | Meta’s Spam standard names “very high frequencies” and adds that lower frequencies may draw restrictions when other spam indicators are present. It publishes no figure | No published source found. No numeric posting-frequency or daily-action threshold appears anywhere in the 23 documents read on 2026-08-24 |
| Low views mean the account has been penalized | TikTok’s For You feed Eligibility Standards state the opposite: “If a video isn’t getting many views, it also doesn’t necessarily mean it broke a rule” (FYF Eligibility Standards, Released August 14 2025, Effective September 13 2025) | Contradicted by published text on TikTok. No document in the set connects low views on their own to an enforcement action |
| Deleting the post clears the strike | TikTok publishes that “deleting your content does not remove strikes, and we may still issue strikes after you delete violative content”. YouTube publishes that “deleting your content doesn’t remove a strike” | Contradicted by published text on two of the seven platforms. The other five publish nothing on the point, so the belief is unsourced there rather than refuted |
| A restriction on one client account spreads to the other accounts the same person manages | TikTok and YouTube both publish circumvention rules, but both attach to accounts used to evade an existing restriction. Meta publishes that a strike may also count against the Page or group the content was posted to | No published source found. No document read states that enforcement against one account propagates to unrelated accounts under the same manager |
| An appeal is read by a human | Pinterest publishes that “we may also use automation to handle appeals more efficiently, for example by expanding a decision made on one Pin to other similar Pins”. The DSA Transparency Database reports 42% of decisions fully automated for its 180-day window | No published source found. No document read guarantees human review of an appeal |
The ledger above is accurate as of the search date recorded in each row. It does not rule out undocumented enforcement behavior, and it is not a compliance guarantee. Absence of a published source means the platforms have not written it down, not that it never happens.
For the Instagram-specific half of the reach and recommendation question, the sourced baseline sits in what Instagram has published about reach and recommendations. This page stays on account-level enforcement.
What to do the day it happens
Five steps exist because a platform document says the action is available. Nothing here is a prevention checklist, and no step is invented.
- Read the notice and name the state it uses. The account holder needs the platform’s own term before anything else, because the appeal path differs by term. Meta publishes that an account holder can see their violation history and current restrictions in Account Status on Facebook and Instagram (Restricting accounts). TikTok publishes five distinct account check states (Your account status).
- Open the platform’s published record, not the notification. Facebook publishes that anyone with Facebook access to a Page can open Page Status through Settings, then Page setup, then Page Status, and see recent Community Standards violations, fact-checked content, intellectual property violations, clickbait, events take-downs, violations from other people who manage the Page, and content that “isn’t eligible for suggestions” (About Facebook Page Status). TikTok publishes two paths to account check, through TikTok Studio and through the in-app Safety Center. YouTube publishes that strike details arrive by email and appear in channel settings.
- Check whether the notice is policy enforcement or a rights complaint. YouTube publishes that it removes content for reasons other than Community Guidelines violations, naming a first-party privacy complaint and a court order, and that those removals carry no strike (Community Guidelines strike basics). YouTube also runs copyright strikes as a separate system with its own counter-notification process. A copyright or rights notice is a different problem with a different resolution path, and this page does not answer it.
- Assemble the appeal before submitting it, because the attempts are countable. LinkedIn publishes that “you may appeal each decision only once” (Redress rights, page shows “Last updated: 11 months ago”). YouTube publishes that “there is also a limit on how many times you can appeal a single channel termination”. Pinterest publishes that it “may also limit the number of times that a particular decision can be appealed”.
- Submit through the published mechanism and note the window where one exists. YouTube is the only platform in the set that publishes appeal deadlines. The other six publish a mechanism without a window, which is covered in the table below.
An agency lead restricted on one client’s TikTok account on a Friday can work steps 1 through 5 in order and reach a filed appeal, and the documentation says nothing about how long TikTok takes to answer. The absence is the honest answer to the client asking for a timeline.
These steps apply where the platform has issued a notice the account holder can see. They do not apply to an account with no notice and no visible enforcement state, and no document in the set describes a way to diagnose an enforcement action the platform has not named.
The documented appeal path, platform by platform
| Platform | Published mechanism | Published window | What the documentation says the outcome is | Source |
|---|---|---|---|---|
| Request review from Profile Status or Page Status, then escalation to the Oversight Board for eligible decisions | No window published, read 2026-08-24 | Content restored and the strike removed if Meta finds the content followed the standards. Meta states “not all decisions are eligible for appeal to the Oversight Board” | I don’t think Facebook should have taken down my post and Restricting accounts | |
| In-app request review, and Meta states the strike and restriction are removed on a successful review | No window published, read 2026-08-24 | Meta publishes that it will “put the content back on Facebook or Instagram” and remove the strike and restriction | Restricting accounts | |
| Appeal submitted after the notice. EU members may also select a certified out-of-court dispute settlement body under Article 21 of the Digital Services Act | No window published. LinkedIn publishes a cap instead: one appeal per decision | LinkedIn publishes that it reviews and may reverse, and that a judicial remedy may also be available | Redress rights | |
| X | The X Terms of Service point to an appeals form in the X Help Center | Not readable in this session | Not readable in this session. See the evidence gap below | X Terms of Service |
| TikTok | In-app appeal from the ban banner at login, from the post analytics screen for a recommendation-ineligible post, or from the inbox notice for an ineligible account. Status is viewable in the in-app Safety Center | No window published, read 2026-08-24 | TikTok publishes that on a successful appeal the content or account is reinstated and “the strike will be removed from your account” | Content violations and bans, Why is my account not being recommended?, Enforcement |
| YouTube | Appeal from YouTube Studio for a strike or removal, and Begin Review from the termination screen for a terminated channel | Six months for warnings and strikes from the date issued. One year for content removals from the date of removal. One year for channel terminations from the date of termination | YouTube publishes that a successful termination appeal returns the creator to the Studio dashboard, and that a rejected appeal signs them out on acceptance | Appeal a Community Guidelines strike or video removal and Channel or account terminations |
| In-product appeal, a Help Center request, or a one-click appeal link in the enforcement notice email. For a deactivated account, Pinterest directs the account holder to contact support | No window published, read 2026-08-24 | Pinterest publishes that it updates the decision if it finds it made a mistake, “or in some cases to give people another chance to abide by our rules” | Enforcement and Account deactivation |
The appeal paths above are as published on 2026-08-24. This page states no expected success rate for any individual appeal, because the only outcome data published is aggregate and each platform defines its metrics differently.
Three things this page does not cover
Rights-based takedowns work differently. A copyright or trademark notice is a rights complaint, not a policy enforcement action, and the platforms route it separately. YouTube publishes that copyright strikes are a distinct system with a counter-notification process that begins a legal process. Pinterest points copyright and intellectual property appeals to its Copyright policy rather than its account appeal path. A companion post on reposting rights by platform is commissioned but not yet live, so no link to it appears here; the link goes in at publication if the post is live by then.
Enforcement numbers need their definitions read first. Aggregate figures like the DSA Transparency Database totals above are only usable once the reporting period and the metric definition are pinned down, and platforms change definitions between periods. The method for doing that is set out in how to read platform transparency reporting.
Policy revision will date this page. The read date on every citation above is 2026-08-24. The internal review date is six months out, because enforcement policy on all seven platforms changes without notice. Updates are made in place, dateModified is bumped, and the change is recorded at the foot of the page. The slug does not change.
What is documented and what is not
The documentation supports a narrow set of claims. Every platform in the set publishes grounds for account-level action, publishes that one severe violation can be enough, publishes some form of notice, and publishes an appeal mechanism. Only YouTube publishes an appeal window. Only Meta publishes a numbered escalation ladder. No platform publishes a numeric posting-frequency threshold, and no platform document read on 2026-08-24 uses the word shadowban or describes an unannounced account-wide penalty.
The documentation does not support the rest of it. It does not tell an account holder how long an appeal takes, whether a person will read it, or how often appeals succeed. The Oversight Board decisions from June and August 2026 show that reviewers apply Meta’s own standards inconsistently and that in-product appeals failed in all seven of those cases. This is a source-record finding about seven platforms on one date, and it is not legal advice.
For platform policy changes read the same way, with the primary sources and the dates attached, the archive is at platform policy changes, read properly.
Questions people actually ask
What does it mean when a platform says my business account is restricted?
The platform’s own term decides what it means, and the seven platforms use six different vocabularies. On Facebook a restriction is a numbered strike attached to a feature block that Meta publishes as a ladder from one day to 30 days. On TikTok it is one of five named account check states covering login, posting, commenting, profile editing and direct messages. On Pinterest it is either deactivation, meaning the content is gone for everyone, or limiting distribution, meaning the content stays reachable by direct link but leaves search and the home feed. Read the term in the notice before doing anything else.
Which behaviors do the platforms’ policies actually name as grounds for restricting an account?
Six grounds recur across the seven platforms: one severe violation, accumulated violations, high-frequency or automated activity, identity and authenticity failures, evasion of an existing restriction, and reasons unrelated to content such as legal exposure or prolonged inactivity. None of the documents read on 2026-08-24 claims its list is exhaustive, so treat these as the published grounds rather than the complete set.
Can I appeal a restriction on a business account, and how long do I have?
You can appeal on all seven platforms, and only YouTube publishes how long you have. YouTube gives you six months from the date a warning or strike was issued, one year from the date content was removed, and one year from the date a channel was terminated. Facebook, Instagram, TikTok and Pinterest publish an appeal mechanism with no window. LinkedIn publishes no window but caps you at one appeal per decision. X’s published appeal documentation could not be read for this page, so check it directly before you rely on a deadline.
Is there any published evidence that shadowbans work the way people describe?
No platform document read on 2026-08-24 uses the word shadowban, and none describes an unannounced account-wide penalty. What the platforms do publish is narrower and noisier: TikTok publishes recommendation ineligibility with an inbox notice and an appeal button, Meta publishes reduced distribution in Feed and surfaces suggestion-ineligible Page content inside Page Status, and Pinterest publishes limiting distribution as one of its two enforcement actions. TikTok also publishes that low view counts do not by themselves mean a rule was broken. The gap between those published mechanisms and the folklore is the finding.
What information do I need before I submit an appeal?
You need the platform’s own term for the enforcement state, the specific content the notice names, and the policy the notice cites, because each platform’s appeal form is anchored to its own decision record. Facebook and Instagram expose that record in Profile Status and Page Status, TikTok in account check and in post analytics, and YouTube in Studio and the strike email. Assemble it before you file, because LinkedIn allows one appeal per decision, and YouTube and Pinterest both publish that they limit repeat appeals.
Does a restriction on one account affect the other accounts I manage?
No document read on 2026-08-24 states that a restriction on one account propagates to unrelated accounts under the same manager. What TikTok and YouTube both publish is narrower: using or creating another account to get around an active restriction is itself a ground for a ban, and YouTube extends that to channels where the same person is repeatedly or prominently featured. Meta publishes one related rule, that a strike may also count against the Page or group the content was posted to. On the general question the documentation is silent, and silence is not permission.
#
