Facebook Moderation Assist is the free comment filter built into Facebook Pages. It automatically hides comments that match rules you choose: comments with links, images or videos, custom keywords, or comments from accounts with no profile picture, no friends or followers, or a very new account. Hidden comments stay visible to their author and the author’s friends, so nobody is notified and threads rarely escalate.
This guide shows how to set it up, what each rule catches, how it behaves under ads, and where a rule-based filter stops being enough.
What is Facebook Moderation Assist?
Moderation Assist is a set of automatic rules attached to your Facebook Page. Each rule is a criterion (for example “comment contains a link”) paired with an action. On Pages, the action is to hide the comment. Meta documents the feature in its help center under Manage comments with Moderation Assist.
A hidden comment is different from a deleted one. The public no longer sees it, but the person who wrote it and their friends still do. You can review everything the tool has hidden, then unhide, delete or ban the author manually.
That design is the main reason to use it: spammers and trolls keep posting into what looks like a normal thread, while your real audience sees a clean comment section.
What Moderation Assist can hide: the criteria
You combine several criteria. A comment that matches any active rule is hidden.
Criterion | What it catches | Watch out for |
|---|---|---|
Link in comment (any link or a specific site) | Scam links, phishing, “DM me for a deal” sites, competitor URLs | Customers sharing a legitimate link, such as a tracking page |
Image or video in comment | Off-topic memes, shocking images, spam banners | Customers posting a photo of a faulty product or of their order |
Keywords | Insults, scam vocabulary, competitor names, words you never want under an ad | Literal matching: “free” also catches “free delivery?” |
No profile picture | Throwaway and fake accounts | Real people who simply never added a photo |
No friends or followers | Accounts created only to spam | Private users with hidden friend lists |
Very recently created account | Bot waves and ban-evasion accounts | New users who genuinely just joined |
Two older Page settings sit next to Moderation Assist: the profanity filter (Meta’s own list of offensive terms) and keyword blocking for Page comments, described in the Facebook Help Center. They work on the same principle: matching comments are hidden automatically.
How to set up Moderation Assist step by step
You need admin access (or a Page role with moderation rights) on the Page. Meta moves the menu from time to time, so if a label differs, search “Moderation Assist” in the Page settings.
Switch into your Page from your personal profile, then open the Page’s Professional dashboard or Settings.
Open Moderation Assist. On many Pages it is listed in the Professional dashboard; on others it sits under Settings, then Privacy.
Add a criterion. Start with links and profile-based rules (no profile picture, no friends or followers, new account). Test each rule and review hidden comments for mistakes.
Add a keyword list. Separate entries with commas and include plurals, common misspellings and abbreviations (scam, scams, scm). Add the words scammers use under your ads, such as “investment”, “crypto” or “WhatsApp me” if they never apply to your business.
Turn on the profanity filter in the Page settings if your audience is general public.
Review the hidden comments after 48 hours. Unhide anything legitimate and remove the keyword that caught it.
Repeat weekly. Look at what slipped through and add new terms. Spam vocabulary changes fast.
Does Moderation Assist work on Facebook ad comments?
Ads run from your Page are Page posts, so Page-level comment rules are designed to apply to them too. In practice, check the hidden queue after every campaign launch, because ad comments are where volume and scams concentrate. For the full workflow, see our guide to monitoring comments on Facebook ads.
Meta has also been adding comment controls directly in Ads Manager, which let some advertisers limit or turn off comments on certain ad formats. Availability depends on the account and format, as explained on Meta’s page on managing comments on your ads. Turning comments off removes spam, but it also removes the questions that signal buying intent.
Moderation Assist vs keyword blocking vs profanity filter
Native tool | What it filters | Best use |
|---|---|---|
Moderation Assist | Links, media, keywords, suspicious profiles | Your main automatic filter for spam and fake accounts |
Keyword blocking | A custom word list | Brand-specific terms, competitor names |
Profanity filter | Meta’s list of offensive words | A quick baseline for public-facing Pages |
Manual hide, delete, ban | Anything you review yourself | Edge cases, repeat offenders, legal issues |
For a wider view of Facebook’s native options and how to organise a team around them, read our Facebook comment moderation guide and the step-by-step on filtering comments on Facebook.
Limits of Moderation Assist
Moderation Assist is useful and free. It is also literal, and that shows quickly on busy Pages.
No understanding of context. “This problem was solved fast, thanks” and “your product is a problem” look the same to a keyword rule.
Easy to bypass. Spammers write “c r y p t o”, swap letters for numbers or use look-alike characters. Every variant needs its own keyword.
Hide only. It does not delete, ban or reply. Repeat offenders and questions still need a person.
One Page at a time. Rules do not carry over to your other Pages or to Instagram, which has its own Hidden Words settings (see Instagram moderation).
No prioritisation. It cannot tell you which of 300 visible comments are buying questions that deserve an answer first.
For a small Page with a few dozen comments a week, these limits are acceptable: Moderation Assist plus a daily ten-minute review is enough. They start to cost money when ads bring hundreds of comments and scams reply to your prospects faster than you do.
Common mistakes to avoid
Hiding legitimate criticism. A complaint answered politely builds trust. A thread with only praise looks curated. Keep hiding for spam, abuse and scams, and reply to the rest (see how to answer a negative comment without feeding it).
Keywords that are too broad. “Price”, “problem” or “refund” catch your best leads and your most urgent customers.
Never checking the hidden queue. A false positive hidden for weeks is an unanswered customer.
Setting it once. Spam patterns shift every few weeks; a list that never changes slowly stops working.
A simple rule set for Page and ad comments
Comment type | Automatic action | Human action |
|---|---|---|
Spam or scam link | Hide (link rule) | Ban repeat accounts |
Insults or hate | Hide (profanity and keyword rules) | Ban and report if it breaks Meta’s rules |
Question about price, stock or delivery | None | Reply publicly, move personal details to DM |
Fair criticism | None | Acknowledge, offer a fix, follow up |
Misinformation about your product | None, or hide if abusive | Correct once with facts, do not argue |
Related guides
Where Blabla fits
When keyword rules stop keeping up, Blabla’s comment moderation connects to your Facebook Page and Instagram account through Meta’s official API and handles spam, scams, hate and competitor links day and night, on paid ads and dark posts as well; unlike Moderation Assist, it can delete as well as hide. Besides keyword lists, you describe what to remove in plain language (“hide any comment offering investment returns”), and an AI filter applies it with context, so “this problem was fixed fast” stays visible. Questions land in one inbox with your DMs, where a teammate answers, or an AI reply drafted from your saved answers does. If Moderation Assist covers your volume, keep it; Blabla is for Pages where ads and scale make manual review too slow.
Conclusion
Turn on Moderation Assist today: link and profile rules first, then a short keyword list, then a weekly review of what it hid. Review its effects on your own comments. Treat it as a first filter, not a moderation strategy, and add context-aware tools once ad comments grow beyond what your team can read.






