How Kyle Draper Scaled Training for 40 Agents Without Burning Out
The "Softball" Trap
Why peer-to-peer role playing wasn't working.
Kyle Draper runs a team of 40 agents. Quality control was always the thing keeping him up at night. He tried peer-to-peer role-playing at first. It seemed logical enough—get everyone practicing together to build team chemistry. However, it did not work out the way he expected.
The problem with agents practicing together is they are too nice to each other. Nobody wants to be the person throwing curveballs at their teammate. Kyle watched it happen over and over.
As a result, his agents were walking into calls unprepared. They were essentially learning by burning through actual leads, which meant lost deals and revenue walking out the door. You cannot afford that when every lead costs money to generate.
Sharpening the Sword
That inefficiency pushed Kyle to look for something that would actually prepare his team. MaverickRE changed how Kyle's team practiced.
The AI does not pull punches. It throws objections that actually sound like objections, not the polite pushback agents give each other during role-play sessions. Agents could finally screw up without consequences—making mistakes with the AI, not with someone who is deciding whether to list their $800,000 home with you.
The 70% Rule & Instant Feedback
Once Kyle saw his agents actually improving from these practice calls, he needed a way to track what was working. Kyle's team leans hard into the feedback loop.
Diagnosing the "Pitch"
Kyle found something useful in the metrics that surprised him. The talk-time ratio became his primary diagnostic tool. If an agent is talking more than 70% of the call, something is wrong with their approach.
"It shows their talk time versus the lead's talk time... for us, it's a big red flag if the agent is talking more than 70% of the time."
There is a whole philosophy behind that number. Realtors who dominate conversations are not listening; they are pitching instead of understanding what the prospect actually needs. While this sounds obvious, the data made it impossible to ignore. You could see exactly who was selling versus who was serving.
Scaling Management for 40 Agents
Kyle used to be the training department himself. Every agent would ping him with call recordings asking, "Hey, when you have a minute, can you listen to this?"
With forty agents doing that regularly, the math does not work. You would need three of him.
MaverickRE took that entire burden off his plate.
The system delivers coaching at scale without Kyle needing to clone himself. The feedback is consistent, detailed, and honestly more thorough than what he could provide while juggling everything else.
Summary of Results
The shift from manual review to AI-driven coaching created immediate operational improvements for the Serene Team.
| Metric | Impact |
|---|---|
| Feedback Loop | Shifted from "when manager has time" to instant automated scorecards immediately after calls. |
| Conversation Quality | Agents stopped "pitching" and started listening, utilizing the < 70% talk-time rule. |
| Managerial Capacity | Allowed Kyle to manage 40 agents without needing to manually review every training call. |
| Training Realism | Eliminated "polite" peer role-play in favor of challenging, realistic AI objections. |