Bots and follow bait
Follow bait, mentions and the accounts that only exist to post them.
Mentions and follow baitOff: your customers see all of these under your ad.
One inbox for every page, Facebook and Instagram together, each comment tagged by what it is.
See it workInstagram ads draw the same spam as Facebook ads, with more bots behind it. Hidden Words catches some words. CommentWatchDog catches the meaning, on every ad and boosted post, the moment a comment lands.
Hidden Words hides comments by word and emoji and reaches your ads when the account is connected to your Page. Ads Manager shows one ad's comments at a time, and only with that connection. Nothing reads meaning, and nothing touches the backlog.
| On your Instagram ads | Instagram and Ads Manager | CommentWatchDog |
|---|---|---|
| Hide or delete one comment | One at a time, with a connected account | Yes, and by rule |
| Hide by word or emoji | Hidden Words, set on the account | Yes, with spelling variations |
| Hide by meaning: scams, negativity, off-topic | No | Yes, AI reads the comment |
| Delete spam automatically | No | Yes |
| Every ad and boosted post at once | One ad at a time | All of them, one setup |
| The backlog on an older ad | By hand | Post cleanup, in bulk |
| Turn comments off | Per post; for ads, some accounts and formats | Not the goal: the harmful ones go, the questions stay |
Sources: Meta Business Help Center, "Manage comments on your ads in Meta Ads Manager"; Instagram Help Center, "Hide comments or message requests you don't want to see on Instagram", read on 5 September 2026.
Meta's own systems remove scam ads at scale and still publish their error rate. Under a legitimate Instagram ad, the comments are the advertiser's to handle, and the ones that hurt rarely contain a word you would think to list.
Six comments from under one Instagram ad, in the order they landed
Four comments arrive under an Instagram ad and pass two gates. Hidden Words stops the one with a listed word. The rules read the other three for meaning: two are bait, one is a customer.
Turn on the rules you want. Each runs on every new comment under every ad. These are real comments the watchdog removed from Instagram ads.
Follow bait, mentions and the accounts that only exist to post them.
Mentions and follow baitOff: your customers see all of these under your ad.
Accusations, insults and "reported", in any language, read for meaning.
Read by AI, in any languageOff: your customers see all of these under your ad.
Links, phone numbers and "send pic to", the road to a scam.
Links, numbers and emails, any formatOff: your customers see all of these under your ad.
The Instagram ad's comments sit next to the Facebook ones. Same rules, same undo, one place to look.


Ads first, posts too. The same rules run on your Instagram posts and reels, so one setup covers everything the account publishes. See what the rules take off Instagram posts.
You connect and pick your rules. The watchdog runs them on every new comment under every ad, day and night.
Sign in with Facebook and pick the Page the Instagram account is linked to.
No code, no passwords shared.
Bots, links, scams, numbers, profanity, negativity. Choose hide or delete for each.
Add your own topics in plain words.
Every ad's comments in one inbox, tagged. What was hidden or deleted carries the rule that did it. Undo is one click.
Reads every language, day and night.
Free for 7 days, no credit card required
Short answers, from Meta's own pages where the answer is Meta's.
You can turn off comments on your own posts. For ads, Meta treats comments as something to moderate: for most advertisers anyone who sees the ad can comment, and Ads Manager lets you hide or delete them one at a time. Some accounts see a comment control for certain formats.
In Ads Manager, open the Ads tab, click Edit under the ad, choose More options, then Manage Instagram comments. It works only when an Instagram account is connected to the Page; ads run under the Page identity cannot be managed there. CommentWatchDog shows them in one inbox with the ad marker on each row.
Yes. Filters set at the account level extend to ads from that account. They act on the words, phrases, numbers and emojis you list, not on meaning.
Delete spam, scams and bots; nobody needs them back. Hide negativity you may want to answer. Both are one rule in CommentWatchDog, and both can be reviewed in the inbox.
Meta documents nothing either way, and neither do we. The ad keeps running; the comment is gone.
