Comment moderation
Handle spam. Keep useful conversations.
Build moderation workflows for Facebook and Instagram comments. Choose which rules can hide or delete content, and which conversations need your team.
Automatic moderation uses Advanced or Elite workflows. Free includes AI drafts you send yourself.
Different comments, different decisions
Illustrative review, not a live account
Repeated, unrelated promotional link
Check against your spam rule“My order arrived damaged.”
Reply and investigate“Someone sent me this suspicious link.”
Read the context first
A keyword match identifies text. It does not establish intent.
Hide, delete or reply?
Hide
Change a comment's visibility without deleting it. Visibility depends on the platform and type of moderation; do not treat hidden content as confidential or erased.
Delete
Remove the comment through the platform's supported action. Treat this as a removal decision, not a temporary filter. Do not assume OutReply can restore it.
Keep and respond
Leave legitimate questions and complaints for your team to address. A critical tone alone is not evidence of spam or abuse.
Review Meta's Facebook comment controls for the platform's visibility rules. A private reply does not hide the public comment or resolve the underlying complaint.
Test the rule against the context
These fictional examples show decisions a team might make. They are not AI classifications or actions taken on real comments.
Unrelated promotion: “Visit bit.ly/example for an offer.”Inspect against the spam policy
Rule check: Contains bit.ly matches this exact text. It does not inspect the destination or prove the link is harmful.
Team decision: If the comment is unsolicited promotion that violates your policy, a supported hide or delete action may fit. A shortened link alone is not enough to make that decision.
“Is the bit.ly/example link you sent me genuine?”The same keyword can catch a real question
Rule check: Contains bit.ly matches here too, although the person is asking for help.
Team decision: Read the context and verify the link before replying. If your rule catches questions like this, narrow it or leave these cases for a person instead of attaching an automatic removal step.
“My order arrived damaged and nobody has replied.”A complaint needs an answer
Rule check: A negative-sentiment label would not establish spam or abuse. This comment describes a service problem.
Team decision: Acknowledge the issue and direct the customer to an appropriate support channel. Do not ask for order details in public. Review mixed cases individually if a complaint also contains abuse or personal information.
“Visit BIT.LY/example for an offer.”Capitalization changes a literal match
Rule check: Contains bit.ly does not match uppercase BIT.LY in OutReply. The rule is case-sensitive and also preserves spaces.
Team decision: Test variants that matter to your policy. OutReply's literal keyword rules should not be assumed to behave like a platform's native word filter or an AI condition.
Try literal matching in the workflow rule preview, or draft a response with the free comment reply generator.
Set up a rule you can verify
Choose the account and post
Start with a connected Facebook Page or Instagram professional account and one post. Select the new-comment trigger and confirm the scope.
Define the rule and exceptions
Use case-sensitive equals/contains conditions, combine checks with AND/OR, or ask an AI condition a specific question. Include legitimate complaints and quoted keywords in your test set.
Connect the intended action
Choose a supported hide, delete or reply step for the matching branch. Leave uncertain cases without automatic removal. The preview on this page does not configure a workflow.
Enable, test and verify
On a plan with automatic triggers, test fresh matching and non-matching events. Inspect execution details and the platform outcome, then expand the scope only when the rule behaves as intended.
Give AI a specific question
Ask about the behavior you want to identify, such as “Is this an unrelated promotional solicitation?” Avoid treating “Is this negative?” as a reason to remove a comment. A dissatisfied customer may be describing a real problem.
AI conditions can misclassify quoted text, sarcasm and mixed messages. Test both unwanted comments and legitimate ones before connecting a condition to a moderation action. Use clear business instructions for any reply you generate, and leave uncertain cases without an automatic removal action.
What if someone posts personal information?
Review the comment promptly and use the platform's available controls as appropriate. Do not repeat private details in a public reply. Hiding is not a guarantee that nobody can see the information, and an automated rule cannot promise complete detection.
Supported accounts, plans and credits
OutReply's automatic comment moderation uses new-comment workflows for connected Facebook Pages and Instagram professional accounts. It depends on the selected account and post, permissions, incoming events and the platform accepting the action.
These live comment workflows are not available for X or YouTube; their comments can be reviewed with AI-drafted replies in the social media inbox. The current TikTok integration covers publishing, not comment moderation.
Automatic triggers are on Advanced ($30/month) and Elite ($45/month). Keyword conditions and hide/delete steps do not use AI credits. Each AI condition uses one credit; additional AI reply steps have their own costs. See plans and credit details.
Check the outcome, then refine the rule
Review execution details and check the comment on the connected platform. A run count or log entry alone is not proof of the final visibility. If an unexpected comment is affected, pause the relevant workflow and inspect its condition and account/post scope before enabling it again.
Record mistaken matches as well as correctly handled spam. A high number of filtered comments is not automatically a good result. Add those mistakes to your test examples and review whether a narrower rule or a person should handle the case.
Comment moderation FAQ
Can OutReply automatically hide spam comments?
Yes. Advanced and Elite workflows can attempt hide or delete actions on eligible new Facebook and Instagram comments. The selected rule, permissions, incoming event and platform response determine what happens; delivery is not instantaneous or guaranteed.
Should I hide every negative comment?
No. Critical feedback may identify a genuine problem. Separate unsolicited promotion or abuse from questions and complaints. A sentiment label or keyword match alone does not establish a policy violation.
Does hiding a comment delete it?
No. Hiding changes visibility under the platform's rules; deletion removes the comment through a separate action. Do not treat hidden content as erased or confidential, or assume a deleted comment can be restored in OutReply.
Do keyword rules ignore capitalization?
No. OutReply's contains and equals conditions are case-sensitive. Contains bit.ly matches a comment quoting that exact text, but not uppercase BIT.LY. Test legitimate examples as well as spam before enabling an action.
Does comment moderation use credits?
Keyword conditions and hide/delete steps do not use AI credits. Each AI condition costs one credit. Additional AI generation steps have their own costs, and automatic workflow triggers require a supported paid plan.
Will enabling a workflow clean up all old comments?
Do not assume so. A new-comment trigger acts on incoming events after activation. Review existing comments separately, and test a fresh event on the account and post selected in the workflow.
Explore the next step
Start with one post and a clear policy
Explore OutReply on Free; enable moderation workflows with Advanced or Elite.