Content moderation and comments

Ref: C10 · Track: Knowledge Base Admin · Time to complete: ~4 min

Why this matters

If your help center allows customer comments on articles, you need a plan for moderation. Without one, you eventually get one of: spam, rants, or something NSFW that stays on your public site until someone notices.

Turn comments on — or off

Per article, per section, or globally. Options:

• Enabled — customers can comment. Comments appear immediately or after moderation (see below).
• Disabled — no comment field shown.

Most B2B orgs disable comments globally, because: - The signal-to-noise ratio is poor (mostly spam or support requests that should be tickets). - Moderation is work. - There's often no clear benefit.

Turn on comments deliberately, for articles where community input is actually useful (e.g. a tutorial where customers share tips).

Moderation modes

If comments are enabled:

Immediate publication

Comment appears on the article instantly. Moderators can remove after.

Require moderator approval

Comment enters a queue. A moderator approves or rejects. Only approved comments go live.

Default to require-approval unless your comment volume is high and trusted. It's the safer starting point.

The built-in swear-word list

Zendesk ships with a default profanity filter. If a comment contains a listed word, it's held for moderation regardless of your default mode.

⚠️ The default list is extensive — some would say aggressive. Review it before enabling in any customer-facing setup. Your admin can edit it (add or remove words) in Guide settings → Moderation.

Depending on your customer demographic, the default list may be too loose or too strict. Common edits: - Add brand-specific reserved terms (your competitors, your CEO's name if embarrassing). - Remove clinical terms if you're in a medical help center.

Moderating the queue

In Guide admin → Comments → Pending: 1. Each pending comment is listed with the article, the commenter, and the text. 2. Click to approve or reject. 3. Optionally ban the commenter if they're a persistent spammer.

Rejected comments don't appear on the article. The commenter isn't notified (they just see their comment hasn't appeared).

Reporting on comments

• Comment count per article — in the book view, sort by comments.
• Active commenters — user profiles show comment history.
• Flagged comments — other users can flag; flagged ones re-enter the moderation queue.

Setting a moderation SOP

If you're leaving comments on, put an SOP in place: 1. Moderator checks queue at least daily. 2. Approve genuine questions/feedback; convert any actual support requests into tickets. 3. Reject spam silently. 4. Ban repeat offenders. 5. Weekly review of approved comments for tone/patterns.

Assign someone specific. "Everyone moderates" = nobody moderates.

Gotchas

• Comments are hard to migrate. If you ever move help-center platforms, historical comments may not come with you.
• SEO. Comments do contribute to page content and can help (or hurt) SEO. Low-quality comments can tank rankings.
• Data retention. Comments are user-generated content and may fall under your privacy policy — plan retention rules.

Quick check — have you got this?

• You're a B2B org with ~50 articles. Comments on or off by default?
• A customer comments with profanity. What happens by default?
• You want comments on tutorial articles but not on billing ones. Possible?

Related articles

• C05 — The book — managing all your content
• C20 — Global KB settings: require sign-in, unsafe content, GA