THE MAN THEY PRICED BY HIS SHOES

FairDrive Access arrived wrapped in the language everyone wanted to believe.
Its website showed smiling families, diverse employees, and a promise that every customer would receive equal attention. Its representatives called the old system crude and biased. The new platform, they said, removed human prejudice by assigning leads automatically.

No clothing field appeared on the screen.

No employee could select Exit.

The operating council suspended deployment and ordered a technical review. The vendor protested that its source code was proprietary. A lender warned that refusing modern lead management could reduce sales enough to threaten payroll. After weeks of crisis, even reformers were tired. Efficiency began to sound less like greed and more like survival.

Ethan voted for a limited pilot.

“We need evidence,” he said. “We can’t reject every tool connected to the old company.”

Lena voted no. Maya abstained because reception would be responsible for running the pilot. The measure passed with strict conditions: no credit data before consent, random auditing, and equal maximum wait times.

For the first ten days, FairDrive appeared to work.

Customers were assigned in rotation. Complaints fell. Sales rose modestly. The council began discussing a public launch.

Then Jamal Carter, a twenty-three-year-old service scheduler, noticed that some appointments disappeared from the main dashboard during lunch. They were not canceled. They moved to a secondary queue labeled Nurture.

The customers in that queue shared no obvious race, age, or gender. What they shared were older phones, prepaid numbers, apartment addresses, long public-transit travel times, and ZIP codes with lower median incomes. FairDrive had removed the appearance questions and replaced them with data gathered before the customer reached the door.

The software did not ask what someone’s shoes looked like.

It estimated what those shoes were likely to look like.

Jamal reported the pattern to his supervisor. The supervisor told him not to create panic until the pilot ended and drafted a warning accusing him of unauthorized analytics access.

Maya found the warning before it was issued.

The old dealership would have protected the supervisor and reputation first. The council had forty-eight hours to prove it was different.

They stopped the pilot, preserved the logs, and announced the problem to employees before reporters discovered it. Jamal received independent advice and continued pay. No one demanded that he appear at a press conference or turn his experience into marketing.

Ethan disclosed that he had supported the pilot.

“I thought a machine would be fairer because it could not see a coat,” he told the staff. “I did not ask what else it could see.”

Some workers wanted his resignation from the technology committee. Ethan offered it. The committee accepted, then invited him to testify during the review as the person whose bad judgment had helped expose a better policy.

Accountability did not require permanent exile. It required surrendering the power connected to the mistake.

The audit found that FairDrive calculated a hidden value score using dozens of proxies: device age, browsing time, distance traveled, insurance-history estimates, neighborhood property values, and whether a visitor opened financing pages before viewing safety information. Customers placed in the Nurture queue waited an average of twenty-six minutes longer.

That delay mattered. Sales reports later labeled those customers “low engagement” because many left before speaking to anyone. The system created the behavior it claimed to predict.

Arthur attended the public council meeting but sat in the second row.

Janice Webb, the nursing assistant from Ethan’s first warning, spoke from the front. She described finishing a night shift, changing buses twice, and waiting forty minutes while later arrivals received keys for test drives.

“Your old manager thought my uniform looked poor,” she said. “Your new computer thought my bus ride looked poor. Which version was supposed to feel fair?”

No one offered her a discount in exchange for forgiveness.

The council refunded disputed deposits, paid for independent reviews of affected finance contracts, and created a customer compensation fund financed first by canceled executive retention payments. People could submit evidence privately. Public storytelling was not a condition of repair.

Sales fell during the disclosure. Some employees feared the council had destroyed the dealership by telling the truth.

Then something unexpected happened. Customers who had avoided Cole Meridian began booking appointments because the failure had been made public. Local unions and community organizations agreed to observe the new policies. Trust did not return as applause. It returned as people willing to verify.

The council replaced automated prioritization with a visible rotation system and permitted exceptions only for language access, disability accommodation, scheduled appointments, or customer request. Every exception required a reason employees could audit.

FairDrive threatened legal action over its proprietary information. Naomi responded with the vendor contract, which guaranteed compliance and permitted review for discriminatory effects. For once, language drafted to protect a company protected the people outside it.

During the forensic examination, technicians found an encrypted directory left by the prior vendor administrator. It contained installation packages for twenty-three dealerships across six states. Each used a different public product name. All sent scoring data to the same analytics company.

One folder was labeled Reputation Recovery.

Inside were draft advertisements, synthetic voice samples, and a video model trained on Arthur’s public interviews.

The final file was scheduled for release the following morning.

It showed a digitally generated Arthur Bennett praising FairDrive as “the future of equal service.”