"Then validate it," said Iván from the back. "Run a real test. One week. Human-only protocols. No AI scoring." Lerma agreed—not out of conviction, but because Iván threatened to leak the "Solucionario Lie" to a trade magazine.

She said, "I’m very sorry. Tell me about him."

"You’re showing us the wrong numbers," she said. She clicked a USB drive. Iván’s graph appeared on the main screen.

"See this gap?" Clara pointed. "The higher our Solucionario score, the lower our customer loyalty. We’re teaching agents to solve problems but ignore people. McGraw’s own book—the full book—says that communication isn't about being right. It's about being present."

The system was ruthless. It penalized a pause over 1.2 seconds ("Hesitation = Insecurity"). It flagged any deviation from the script ("Creativity = Deviation"). It gave Clara’s star agents a 68% score because they laughed naturally with a client ("Informal register detected—professionalism compromised").

For one week, Clara’s team turned off the live grading. They used the forbidden techniques. They let pauses breathe. They admitted mistakes. They once spent 14 minutes on a call with an elderly man who just wanted to tell someone his internet was the only thing working in his life.