Social and Community Content

Comment Moderation Decision Log

Document a proportionate moderation decision with the applicable policy, minimum evidence, action, notice, appeal route, reviewer, and follow-up.

Free editable Markdown · Community moderators, trust and safety teams, and publishers · Updated July 30, 2026

Download Markdown

Accessible HTML preview

Blank template

The downloaded file contains the same fields in editable Markdown.

Case intake

Case ID: [Non-identifying reference]
Channel and content location: [Controlled link or platform reference]
Reported at: [Date, time, and zone]
Report source: [User, moderator, system, or another]
Content author reference: [Minimum necessary identifier]
Affected party: [Role or group, if relevant]
Immediate safety concern: [Yes or no, with action]
Assigned moderator: [Name or role]

Policy assessment

Policy version: [Date or identifier]
Applicable clause: [Exact rule]
Observed behavior: [Neutral description]
Context reviewed: [Thread, history, media, quotation, or another]
Severity factors: [Reach, repetition, target, credibility, or sensitivity]
Mitigating context: [If relevant]
Uncertainty: [Missing evidence or interpretation question]
Comparable precedent: [Case or none]
Assessment: [Violation, no violation, or escalate]

Decision and notice

Options considered: [List]
Selected action: [No action, warning, hide, remove, restrict, ban, or another]
Scope and duration: [Exact effect]
Proportionality rationale: [Why this option]
Decision maker: [Name or role]
Second review required: [Trigger and reviewer]
User notice: [Message or reason notice is unsafe]
Reporter update: [Permitted status]

Criticism or viewpoint was not substituted for a policy violation.
Reporter identity and private evidence are protected.
Related copies or replies were handled consistently.

Evidence, appeal, and follow-up

Evidence retained: [Minimum record]
Access restriction: [Who may view]
Retention or preservation date: [Rule]
Appeal route and deadline: [User-facing process]
Appeal reviewer: [Independent role where practical]
Appeal result: [Uphold, modify, reverse, or pending]
Restoration action: [If reversed]
Precedent or policy issue: [What needs review]
Case closed: [Date and owner]

How to use this template

Assign a case identifier and preserve only the content, report, thread context, timestamps, and platform details necessary for review.
Identify the policy version and exact clause, then describe the observable behavior, affected party, severity, and uncertainty.
Compare available actions for effectiveness, proportionality, urgency, reversibility, and impact on evidence or conversation context.
Record the decision, reviewer, user notice, escalation, duration, enforcement scope, retention rule, and appeal route.
Complete any appeal or follow-up, note reversals and precedents, and feed recurring ambiguity into policy and training review.

Evaluate behavior against a specific rule

Capture the exact content and context needed to understand the report, then identify the precise policy clause in effect at the time. Describe observable behavior rather than labeling the person. Distinguish disagreement, criticism, sarcasm, and unpopular opinions from threats, targeted harassment, doxxing, spam, impersonation, or another defined violation. Check thread context, quoted material, target, repetition, power dynamics, and credible real-world risk. A keyword may signal review but should not decide the outcome without context.

Choose a proportionate and reversible action

Record the plausible options: no action, reply with guidance, add a warning, limit reach, hide, remove, temporarily restrict, permanently ban, preserve and escalate, or another platform-specific measure. Select the least restrictive response that reasonably addresses the policy harm, while acting quickly when exposure itself creates danger. Consider whether the action affects replies, quoted copies, or evidence needed for an appeal. Do not remove critical feedback to improve brand appearance. If the policy is unclear, escalate the interpretation and mark the decision as precedent pending formal review.

Explain, review, and learn

Where safety and system capabilities allow, tell the affected user what rule applied, what action occurred, its duration, and how to appeal. Avoid revealing reporter identity, investigation methods, or private information. Separate the original moderator from an appeal reviewer when practical. Track reversals, inconsistent treatment, repeated policy confusion, disproportionate effects, and evasion. Use aggregated findings to improve rules and training without republishing harmful material. Delete evidence on the approved schedule unless a lawful preservation requirement applies.

See the fields in context

Fictional example: repeated promotional replies

Cedar Square Makers and all accounts, posts, and rules in this record are invented.

Observed behavior: One fictional account posts the same unrelated sales link under twelve workshop discussions within twenty minutes.
Policy basis: The current community rule prohibits repeated unsolicited commercial promotion; the content contains no threat or personal data.
Decision: Remove the duplicate replies and apply a 24-hour posting restriction after a warning, while leaving one relevant product discussion untouched.
Notice: The account receives the rule, duration, affected content references, and appeal route without identifying reporters.
Appeal result: A separate reviewer upholds the restriction but corrects one content reference that had been included by mistake.

Frequently asked questions

Should every removed comment receive a full log?

Match detail to risk, scale, policy, and review needs. Routine spam may use a compact record, while threats, appeals, permanent restrictions, precedent, or high-impact accounts require fuller evidence and reasoning.

Can screenshots be stored indefinitely?

No. Define a lawful purpose, access limit, and retention period. Screenshots can contain usernames, faces, health information, or unrelated conversation. Preserve longer only when a documented legal or safety basis requires it.

What if the policy does not clearly cover the behavior?

Do not invent a retrospective rule. Escalate, consider a limited reversible action if immediate harm requires it, document uncertainty, and ask the policy owner to clarify future treatment transparently.

How should a reversed decision be handled?

Restore content or access where possible, notify the affected user, correct downstream records, and explain the operational lesson. A reversal is useful quality evidence, not a reason to conceal the case.