Skip to main content

Comment automation

Comment automation (sidebar → Engage → Auto Reply) lets you define rules that act on incoming comments without a human in the loop. Each rule watches one post on one account, matches comments against keywords, and responds — with a public reply or a DM.

How a rule works​

A rule has:

  • Scope — one account and one post (picked from the account's 25 most recent posts, with thumbnail and comment count). Account and post are fixed at creation; clone the rule to cover another post.
  • Trigger — one or more keywords, matched case-insensitively against the comment text. Enable whole word match only to require word boundaries (so "price" doesn't match "pricewise").
  • Action — Reply with comment or Send DM, subject to the platform's capabilities (options that a channel can't do are disabled with a note).
  • Message — the response text. Use {username} to address the commenter by name (falls back to a generic greeting) and {keyword} to echo the matched keyword.

Rules can be paused and resumed; every fire is logged.

The activity log​

Each rule has an Activity tab listing every action it took — user, comment, action, status (sent / failed / skipped), error, and time. Skipped rows carry the reason, which is how you'll notice a quota or rate cap firing.

When rules fire​

  • Real time — a comment arriving via webhook that identifies its parent post triggers matching rules within seconds (Meta platforms).
  • Every 5 minutes — a sweep catches everything else (polling platforms, events without a parent post id).

Each comment is processed once per rule (deduplicated), and keyword matching plus send happen inside your plan's quotas:

  • Monthly auto-reply quotas — autoDmPerMonth and autoCommentsPerMonth per plan (see plans). Over-limit actions are logged skipped with the limit in the error.
  • Per-platform hourly rate caps — automation paces itself per account: e.g. Instagram ~190, X ~300, YouTube ~100, LinkedIn ~80, Facebook ~100, Threads ~50 replies per hour. Over-cap comments are skipped, not queued.

The full listen → match → reply loop works where the platform exposes comments: Instagram, Facebook, Threads, X, YouTube, LinkedIn, Reddit, Mastodon, and Bluesky. DM delivery is narrower — Instagram, Facebook, X, Mastodon, Reddit, Slack, Discord, Telegram — see the capability matrix for sendDm.

warning

TikTok, Discord, and Telegram can receive replies but don't surface incoming comments to automation (no comment access or webhook), so rules can't trigger there.

Common patterns​

  • "Comment 'LINK' to get the link" — match the keyword, DM the resource. The classic lead-gen pattern.
  • FAQ deflection — auto-reply to "price"/"pricing"/"how much" with a pointer to your pricing page.
  • Personal thanks — reply to a keyword with a {username}-addressed thank you to boost engagement.

Limits and safety​

  • Rules run inside the platform's rate and spam limits — pace aggressive campaigns or the platform will block the action.
  • Automation never edits or deletes the commenter's original message.
  • A lapsed subscription pauses rule execution cleanly (rules survive; they resume when billing is active).
  • There is no AI in auto-reply — it's pure keyword matching on your template, and consumes no AI-action allowance.