Message tags: the four doors out of the 24 hour window
Message tags let a Page message somebody outside the 24 hour window. Each one names a specific, narrow situation, and using one outside that situation is the fastest way to lose a Page's messaging permission entirely. For most businesses only human agent is relevant.
The tags, and what each is actually for
| Tag | Permits | The trap |
|---|---|---|
| HUMAN_AGENT | A person answering an inquiry, up to 7 days | Needs App Review, and must genuinely be a person |
| CONFIRMED_EVENT_UPDATE | Reminders for an event they registered for | Not an invitation to an event they did not |
| POST_PURCHASE_UPDATE | Updates on a transaction they made | Not an upsell attached to a receipt |
| ACCOUNT_UPDATE | A change to their account with you | Not news about your company |
The one that is actually useful
HUMAN_AGENT exists because a real support team cannot always answer within a day, and Meta accepts that. Seven days instead of one is the difference between a channel a small team can staff and one they cannot.
It requires the human_agent permission through App Review, and the condition is the name: a person is responding to something the customer asked. An automation using it is not a grey area.
This is the correct answer to "our team cannot always reply within 24 hours". The wrong answers are picking a different tag that sounds close, or removing the window from your own software - which changes nothing, because the window is Meta's and the send is refused at their end.
Njiwa does this for you, and does nothing else with tags. A reply somebody types in the shared inbox is sent normally first; if Meta answers that the window has closed, it goes again under HUMAN_AGENT, on a Page approved for the feature. Nothing an automation produces - an auto-reply, a campaign, a send from an API key - is ever tagged, because none of those is a human agent and the tag is the permission, not a retry button.
What misuse costs
Meta does not fine you or bounce the message. It restricts the Page's ability to message at all. Everything stops - support, order updates, the conversations already open - and the appeal is slow. The asymmetry is the point: the tags are trusted, and trust is withdrawn rather than metered.
How to stay on the right side
- Tag at the point of sending, from the actual reason. Not a default in a config file that somebody set once.
- Never let an automation send a human agent tag. If a bot composed it, it is not a human agent message.
- Log which tag was used and why. When Meta asks, the answer needs to be evidence rather than a recollection.
- Prefer answering inside the window. Every tag is a workaround for a queue that moved too slowly, and the queue is the fixable part.
Questions people actually ask
What is the human agent tag?
A tag that lets a person - not an automation - respond to a customer's inquiry up to seven days after their last message. It requires the human_agent permission, granted through App Review.
What happens if I misuse a message tag?
Meta restricts or removes the Page's messaging permission. The loss is the whole channel, not just the tagged messages, and it is not a warning system - the first consequence is the consequence.
Njiwa
One inbox for WhatsApp, Facebook and Instagram, and an API that sends over either WhatsApp rail. 7 days free, no card.