Fighting spam in the Help Center

A public help centre attracts spam wherever anybody can post. Four defences, and one decision that removes most of it.

A public help centre attracts spam wherever anybody can post. Four defences, and one decision that removes most of it.

Where it arrives

• Article comments.
• Community posts, if you have one.
• The request form, which becomes suspended tickets.

The decision that removes most of it

Requiring sign-in to post. Not to read, which would cost you your reach, but to comment or post.

Automated spam does not create accounts, so this alone removes the great majority with almost no cost to genuine users.

The other three

• CAPTCHA on the request form, which stops automated submissions while leaving people through.
• Zendesk's own spam detection, which holds suspected content for review.
• Moderation, where posts are checked before appearing. Effective and it needs somebody.

Marking as spam teaches the filter

Deleting removes it. Marking it as spam removes it and improves detection, and it takes the same click.

Comments you do not need

Article comments are optional, per article or across the help centre. On a knowledge base where comments generate spam and little else, turning them off is a legitimate answer rather than an admission of defeat.

Feedback through voting and through the ticket channel usually tells you more anyway.

See also

• Turning off comments on an article
• Moderating end-user content
• Recognising and fighting spam