Almost every business is sitting on the same buried asset: thousands of leads inside the CRM that nobody ever calls.
So we ran a simple test. We let an AI voice agent call 2,800 of those dormant leads, qualify them, and book calls, with zero manual dialing. The only question that matters: did it make money?
It did. Here's the playbook, and the result.
The dormant-list problem
In our coaching business we ran an ascension model: a low-ticket course, then an upgrade to high-ticket. Over the years that left us with roughly 15,000 clients and a long tail of leads that went idle, some for months, some for years.
We'd worked lists like this before with human setters. And if you've ever managed setters, you know the truth: it's a soul-crushing job with high turnover. Nobody loves dialing hundreds of unresponsive numbers a day. So even good teams don't work the list systematically, at scale, forever.
The reactivation script
The agent opened by re-engaging, not pitching: "Hi, this is Emma from the team, you accessed [the thing] a while back, does that ring a bell?" From there it ran the same qualification our humans used, BANT, to find the people still trying to solve the problem, and offered to book a call with a human.
Non-negotiable: the script was already proven by humans. That's the first question I ask any client. If you don't have a script that converts, this isn't your system yet.
How the agent works a list
- Multi-touch cadence. Call, and if there's no pickup, retry after an hour, a day, three days, whatever you set. On this run it averaged about four dials per lead. A human won't do that on lead number 1,900. The agent doesn't argue.
- Book or tag. If the lead qualified, it booked the call itself. If they were interested but couldn't book right then, it tagged the interest and pinged a human rep for a warm follow-up. Those tagged follow-ups booked extra calls on their own.
What came back
| Metric | Result |
|---|---|
| Window | 15 days |
| Spend | ~$450 |
| Leads called | 2,800 |
| Total dials | 11,400 |
| Pickups | 542 |
| Booked by the agent | 22 |
| Closed by a human | 2 ($6,800) |
The response rate from dials was 4.7%, versus 3.5% with our human team, again, not because the AI is more charming, but because it actually calls every lead the number of times it should.
The point about consistency
Here's how I think about human setters versus an agent. A human might start at a higher skill level, but the line is jagged: good days, bad days, coaching, sick days, turnover. An agent might start lower, but every fix you make only moves it up. It can't have a bad Tuesday. Over months, that difference compounds hard.
And it happened on a cold, ignored list. If it books and closes from leads nobody had touched in two years, imagine what it does the second someone opts in and is actually expecting the call.
-Ruben
Grow your revenue or cut your costs in 90 days. Or you don't pay.
Done-for-you AI systems, the same ones I run in my own agency and build for 7 and 8-figure coaching and consulting businesses. Want them installed for you? Let's talk.
