How to hide spam and scam comments on YouTube

The three kinds of comment spam

Link spam and scams. Crypto pitches, screenshots of imaginary earnings, fake investment advice, "message me on WhatsApp". Sometimes from a real-looking account, often from a fresh one.

Impersonation. An account using your name and photo, replying to your viewers to tell them they won something. This one is worse than ordinary spam, because it targets your audience using your credibility.

Low-effort promo. Sub4sub requests, "check my channel", follow-for-follow chains. Not malicious, but it arrives in volume and it buries the comments you want.

They look different and they need slightly different handling, but they share one property. They are pattern-based, which is exactly what makes them the right thing to automate.

What YouTube already does for you

YouTube runs its own spam filter and holds comments it thinks are inappropriate for review. You can add your own blocked words and phrases in Studio, and choose whether held comments are reviewed before they appear.

That covers the obvious cases and it is worth leaving on. Where it struggles is anything specific to your channel. Your audience has its own scam, the one that copies your giveaway, quotes your product, or borrows your catchphrase to sound legitimate. YouTube's filter has never seen that pattern. You have.

The manual version

By hand, the actions are:

1. Hide the comment. It disappears for everyone except the person who wrote it. They are not told and they can keep commenting, so hiding is a filter rather than a ban.
2. Block the author when it is a repeat offender. This stops that channel commenting on your videos at all.
3. Report impersonation to YouTube. Hiding an impersonation comment does not remove the fake account, and the account is the actual problem.

Hiding is the right default for spam. It is reversible, it is invisible to your audience, and it keeps the argument out of your comment section. Replying to spam, even to mock it, feeds the thing you are trying to remove.

The keyword version

The fastest durable fix is a keyword rule: a list of phrases that, when they appear in a comment, cause it to be hidden on arrival.

Two things make keyword rules work better than they sound.

The matching happens before the comment reaches your review queue, and it happens locally. Nothing is sent to the YouTube API to make the decision, so the rule costs no quota and adds no delay. The expensive API calls stay reserved for comments that need a written reply.

The rule is yours to write. You can add the specific scam that targets your channel, the phrase your impersonator uses, and the misspellings real spam uses to slip past a generic filter.

Writing keywords that do not over-block

The failure mode of keyword moderation is the false positive. A rule for "free" hides a viewer asking whether your course is free. A rule for "link" hides someone offering you a source you actually want.

A few habits reduce that:

• Prefer phrases over single words. "free crypto" is a much safer rule than "free".
• Test the rule against real comments before it goes live. A rule editor should show you what it would have matched.
• Start with contains-matching on phrases, and only widen the match once you have watched it behave.
• Add the specific rather than the general. "dm me for" catches the pitch without catching a viewer who wants to DM you.
• Check the first few days. A rule you never review is a rule whose cost you do not know.

Hiding is reversible, which is the reason this is a safe place to start automating. A wrong hide can be undone.

Handle repeat offenders by channel

Keywords catch a message. Some spam is a person, and that person changes the message.

For those, a blocklist keyed to the channel is the better tool. A rule that hides comments from a specific channel ID keeps working when the account changes its display name, which is exactly what a determined spammer does. You do not have to write a new keyword every time they reword the pitch.

What not to do

• Do not reply to spam. A reply is engagement, and engagement is what the spammer came for.
• Do not hide criticism. A hidden critical comment is a visible act of censorship, and it turns a viewer into an ex-viewer.
• Do not hide on a single weak signal. If a model is only mildly confident that a comment is hostile, leaving it up and flagging it for review costs you nothing.
• Do not hide first and check later. Hiding is reversible, so the cost of a wrong hide is small, but it is not zero.

A starting set

Write down the last ten spam comments you can remember and look for the phrases they share. That list is your first rule. Add the impersonation pattern if your channel has been targeted, and add a blocklist entry for anyone who keeps coming back.

RippleFlow applies keyword rules locally and hides on match, and the features page covers how the rules fit with the rest of the pipeline. Pricing shows the plan limits, and the moderation guide has the triage order that decides what happens to everything the rules do not catch.