“Break by customer ZIP code.”
Cassandra frowned.
“Why?”
Simone ignored her.
Then:
“Also by credit tier.”
Kevin typed.
“And?”
Simone thought.
“Salesperson.”
“Finance manager.”
“Trade-in present or not.”
“Optional products.”
“Cancellation attempt.”
Counsel shifted.
“What are you looking for?”
Simone said, “I’ll know when distribution looks wrong.”
It did.
Thirty minutes later.
Customers from three predominantly Black neighborhoods showed higher average dealer markup.
Customers from two predominantly Latino ZIP codes showed significantly higher optional product penetration.
Not slightly higher.
Thirty-eight percent higher.
Credit scores did not explain difference.
Income did not fully explain it.
Trade-ins did not.
Sales mix did not.
Simone leaned toward screen.
“Filter deals handled by Cassandra’s department.”
Kevin did.
Numbers worsened.
Martin whispered:
“Oh, God.”
Cassandra stood.
“This is statistical nonsense.”
Simone looked up.
“Sit.”
Cassandra stared.
Then sat.
“This is correlation.”
“Yes.”
“Not proof.”
“Correct.”
Cassandra looked almost relieved.
Simone continued.
“That is why we investigate.”
She turned to Kevin.
“Can you export deal-level data?”
“Yes.”
“Send to outside counsel and Carter audit.”
Cassandra said, “You can’t accuse me based on ZIP codes.”
Simone’s voice remained level.
“I haven’t accused you of anything.”
“You keep acting like—”
“Like numbers deserve questions?”