Account health, strikes, suspected shadowbans and appeals

Name what you observed before naming a cause Start with the observable event: a post was removed, a warning appeared, publishing failed, login access changed or reach declined. Record the account alias, platform, time, affected post and exact visible message. If the platform does not state a reason, write unknown. An unsupported diagnosis can send the team toward the wrong remedy.

Use separate categories for content enforcement, recommendation eligibility, technical delivery and account access. They may overlap, but they are not interchangeable. A scheduler connection expiring does not establish a content strike. A post remaining visible does not prove it is eligible for every recommendation surface. Check the available native information before changing the queue.

For a fictional example, three recent posts receive less reach while the account shows no visible notice. Record the change, compare equivalent observation windows and check whether formats or posting volume changed. Escalate the uncertainty with evidence. Do not report a confirmed shadowban merely because a third-party checker or a comment thread uses that word.

Build a timeline that can support a decision Include the last known normal publication, the first affected item, relevant edits, access changes and notices. Save the minimal screenshots or exports needed in the authorized incident record. Remove private messages and unrelated personal information. A chronology helps the owner separate what occurred before the problem from actions taken afterward.

Record who observed each fact and the source. A native notice, scheduler error and teammate’s recollection carry different information. Keep them labeled. If a timestamp is local, include the timezone. Without that detail, a team spread across countries can accidentally reverse the order of a publishing attempt, a restriction and a later edit.

Keep a decision log alongside the evidence: what was paused, what remained permitted, who approved the response and when the next review is due. Do not overwrite the initial observation when new facts arrive. Add an update. A preserved timeline is more useful for an appeal and later prevention review than a rewritten story with no uncertainty.

Contain the issue without making it worse Pause affected scheduled work if its approval or account eligibility is uncertain. Do not automatically stop unrelated accounts unless the incident owner determines a shared risk. Identify the actual scope: one asset version, one platform connection, one account or a common permissions problem. The scope determines which teammates need to act.

Avoid repeated retries when the public state is unknown. Check the native account and the scheduler’s delivery record. If access appears compromised, use the approved security escalation and authorized recovery process. Do not ask a broad team channel to share passwords, recovery codes or identity documents as a quick fix.

Tell the marketing manager what is affected and which decision is needed. A useful update says which posts are held, why, what evidence exists and when you will check again. It does not promise a recovery deadline or speculate publicly about the platform. Community messaging, if needed, should follow an approved communication decision.

Prepare an appeal from the relevant evidence Read the notice and current official instructions for the specific platform and action. Identify whether an appeal is available and what it asks for. Keep the submission factual: the affected item, the decision being reviewed, the relevant context and the reason you believe review is justified. Do not invent permissions or omit a material part of the content.

Assign one appeal owner and record the submission, reference and next check. Multiple teammates filing contradictory explanations can reduce clarity and waste time. Keep a copy in the authorized incident record. Follow the actual process and any stated limits rather than assuming repeated submissions will accelerate the response.

An appeal is a request for review, not a guarantee of reinstatement or distribution. If the platform upholds the decision, document the result and ask the responsible owner how future publishing should change. Do not re-upload removed content under a different caption or use another account to bypass the action.

Separate prevention from enforcement evasion A prevention review asks what the team can control: asset permissions, approval precision, cover framing, captions, destinations, access and publication checks. If the cause remains uncertain, say so. You can improve a weak handover without claiming it caused the original restriction. Corrective actions should fit the evidence rather than a convenient story.

Back up approved assets, version histories and recovery ownership. Maintain only legitimate accounts and permitted access arrangements. A backup plan must respect the platform’s restrictions. An account created or repurposed to continue prohibited activity is not ordinary business continuity simply because somebody labels it a backup.

Review any shared template that contributed to the issue. If a caption or destination was used across a batch, identify the affected records and seek the appropriate decision. Do not quietly change every queued item without preserving which version was approved. The prevention process needs both a corrected workflow and an understandable history.

Report the incident without inflating certainty A concise report includes status, affected scope, evidence, actions, owner, next check and open questions. Keep confirmed facts distinct from hypotheses. For performance changes, include the measurement period, post age and comparison basis. Do not use a single weak post as proof that an entire account is restricted.

After resolution, record the outcome even if it is inconclusive. A connection restored after reauthorization is a different result from a content decision overturned on appeal. A later reach recovery does not automatically establish which action caused it. Preserve that distinction so the team does not turn an anecdote into an unreliable playbook.

Close with practical learning. Perhaps every held item now needs an owner, appeals have one coordinator or destination approval refers to a specific version. Those changes make future work easier to investigate. The social manager’s value is a reliable response under uncertainty, not a promise to defeat every moderation decision or restore a particular audience size.

Official references and scope

The workflows and fictional examples above are editorial guidance. Consult the current official sources for platform and product decisions; access to some support pages may require the native app or a signed-in account.

• X: authenticity, multiple accounts and enforcement evasion
• TikTok: Community Guidelines (check current promotion and destination rules)
• TikTok: official content violations and appeal guidance (may require native access)

Related execution guides

• Community management: comments and inbound DMs
• Captions, hooks, covers and hashtags
• A day in creator social media operations