Community Manager Agent:
AI Replies to Comments and Messages

The Community manager handles engagement on Zoe's team:

• Writes replies to comments on posts and Reels.
• Drafts answers to common messages in the Unified Inbox.
• Turns keyword comments into private messages.
• Drafts replies to Google reviews.
• Flags complaints and sensitive conversations for people.
• Hiding spam is coming soon for the agent. Inbox automations already hide abusive comments.

Reply to every comment

Thank-yous, quick answers and friendly follow-ups are drafted promptly and go out once you approve them, so your audience feels heard. Unhappy people come first, then unanswered questions.

Comment keyword reply

When someone comments "price," "details" or "link," the agent drafts a public reply and a private message with the information, and both are sent after you approve.

Message helper

For new messages, the agent writes replies from your facts: timings, prices, locations and policies, ready for you to copy and send.

Review replies

Thank happy customers and respond calmly to unhappy ones, with an offer to resolve the issue privately. Every review reply is a draft you approve.

Smart handoff

Complaints, refunds, legal, medical and angry messages are set aside for you with the reason. The agent never answers them from memory.

Spam control

The agent leaves spam and abusive comments without a reply. To hide abusive comments, turn on the hide abusive comments automation in your inbox, and it hides them on its own.

1. 1Set your rules: what the agent can answer, what needs a person and your tone.

1. 1Add your facts: prices, timings, links and policies.

1. 1Turn on recipes: reply to every comment, comment keyword reply, review replies.

1. 1Agent drafts: replies are written within your rules and wait for your approval. Comments you want answered instantly can go to inbox automations, which reply on their own once you turn them on.

1. 1You handle the rest: sensitive conversations are set aside in your inbox, with the reason.

Comment: "How much for keratin?"

Public reply: "Hi Priya, sending you the details in your DMs now!"

Private message: "Keratin treatments start from ₹3,500 depending on hair length. Would you like to book a slot this week?"

Comment: "Worst service last time."

Action: no public argument; the agent drafts a polite reply asking them to message, for you to approve, and alerts the salon manager.

People often message several businesses at once. The first to reply clearly usually wins the booking. Fast replies also signal to platforms that your account is active, which can support reach.

• Arguing publicly with unhappy customers.
• Copy-paste replies that feel robotic.
• Answering with outdated prices.
• Ignoring comments on older posts.

• Keep your facts updated every time prices or timings change.
• Review handed-off chats weekly and add common answers.
• Use a friendly, human tone; avoid corporate language.

Start in approval mode. The agent's replies, private messages and review answers wait for your yes on a card. For the first two weeks, use it for compliments and simple factual questions such as timings and location, and review its drafts daily. Once you trust the tone and accuracy, switch routine comments to automatic by turning on inbox automations, which then reply to comments and send commenters a DM on their own, instantly and 24/7. Automatic replies to incoming DMs and Google reviews are coming soon; until then, those stay as drafts you approve. Complaints, refunds and anything health-related, legal or money-related always stay in the "hand to a person" group. This step-by-step approach builds confidence and keeps your brand safe.

Busy businesses that receive more comments and messages than they can handle, brands running ads that generate many comments, multi-location groups with a separate page for each outlet, and agencies managing engagement for several clients. Pair it with WhatsApp agents to move conversations from comments to booked appointments.