Should we turn on auto-replies for our Instagram and Facebook DMs?

The hesitation is earned: everyone has DMed a business and hit the chatbot wall — canned cheer, useless menu buttons, no path to a human. That experience is why "should we automate replies" feels like asking "should we hang up on people politely." But the wall and the auto-reply are different machines, and the difference is one design rule.

The rule: automate the acknowledgment, never the conversation

A good auto-reply does one job — it converts dead air into a held place in line: "Thanks for the message! A real person (usually Dana) replies within the hour during the day. Booking? Fastest is right here: [link]." It promises only what's true, names the human, and offers a real path for the people who don't want to wait. Bot theater fails because it pretends to BE the conversation; the honest version just holds the door until the conversation starts.

Why the acknowledgment matters that much

The DM clock runs in minutes — the sender is watching for the "seen" that isn't coming, with alternatives one swipe away. In a live business we measure, inquiries that aged before engagement converted at a third the rate of fresh ones. The auto-reply can't close anyone, but it reliably stops the leak between "message sent" and "human free" — which is where the measured losses live.

The wording that works (and the words that don't)

Keep it: short, honest about timing, signed with a human name, carrying one real link. Kill it: fake typing indicators, "our team is thrilled you reached out," menu mazes, and any promised response time you can't keep — an auto-reply that lies about "we'll respond in minutes" burns more trust than silence. Off-hours gets its own honest version: "We're closed until 9am — book directly here, or we'll reply first thing."

Whether the auto-reply is even your gap

The free score scores the whole chain in three minutes — where DMs land, who owns them, off-hours coverage, follow-through — and names your weakest link, with the monthly cost estimated from your own numbers, assumptions labeled.