I had not spoken privately with Grant in years.
Then one afternoon, an email arrived.
Daniel,
I’m writing a piece for an industry leadership program about incentive design and operational failures. I’m not asking you to endorse me or Hartwell. I would value one conversation about what I missed when we built the accountability system.
If you’d rather not, I understand.
Grant
I read it twice.
Laura read it once.
“You want to go?”
“I don’t know.”
“Then don’t answer tonight.”
Good advice.
Two days later, I said yes.
Public café.
One hour.
Grant looked older.
So did I.
He had left Hartwell and now advised industrial companies on acquisitions and operations.
Apparently, part of his work involved helping firms avoid exactly the kind of governance failure he once helped create.
Life has humor.
We ordered coffee.
He opened a notebook.
I laughed.
“What?”
“You really came prepared.”
“I’m trying not to rely on memory.”
Good.
He asked:
“When did the system first feel wrong to you?”
“The two-hundred-dollar sensor photo.”
“Why didn’t you escalate?”
“Because Derek made it sound like policy.”
Grant wrote.
“And later?”
“Customer survey.”
“Why not then?”
“Same reason.”
“Did you ever see the full policy?”
“No.”
He stopped writing.
“That’s one of the biggest failures.”
“Yes.”
“We thought managers had explained it.”
“Managers thought HR had validated it.”
“HR thought operations owned the substance.”
“Finance processed.”
We both smiled grimly.
Exactly.
Responsibility diluted until no one owned the whole mechanism.
Grant asked:
“What should I have measured instead?”
I thought.
“You were measuring project leakage.”
“Yes.”
“You should also have measured reversals, disputes, employee variance, and manager-level usage.”
He nodded.
“If Derek used adjustments ten times more than another manager, that should have triggered review.”
“Yes.”
“At the time, we treated that as evidence his team was more accountable.”
There it was.
Metrics can reward the wrong story if nobody asks what creates the number.
Grant continued.
“I thought workers would become more careful if they felt cost.”
“Did you share savings when projects performed unusually well?”
“Through bonuses.”
“Were those bonuses directly linked to the same cost categories?”
He paused.
“No.”
“So losses were immediate and personalized. Gains were broad and conditional.”
“Yes.”
“That’s not symmetrical ownership.”
He wrote that down.
Then he closed the notebook.
“I’m sorry.”
I nodded.
Not dismissing.
Not dramatizing.
He said:
“I used to think your resignation embarrassed me.”
“Why?”
“Because Evelyn saw your pay stub and immediately saw something I had defended for three years.”
“That must have hurt.”
“It did.”
“Not as much as the paycheck.”
He laughed.
“Fair.”
Then:
“I blamed Derek for a long time.”
“He deserved some.”
“Yes. But he was using a system I created.”
“That matters.”
“I also blamed Evelyn for overruling me publicly.”
“She was CEO.”
“I know.”
He looked out the window.
“I didn’t understand then that limiting my authority was part of keeping me, not simply punishing me.”
That sounded mature.
I said so.
He smiled.
“Expensive lesson.”
We talked about leadership more broadly.
I told him about Priya’s forty-thousand-dollar error.
He asked:
“What did you do?”
I explained.
Root cause.
Coaching.
System changes.
Written bonus policy.
No invented repayment.
Grant listened.
“That’s what I wanted the culture to become.”
“Then you built the wrong mechanism.”
“Yes.”
Simple.
That conversation gave me something I did not know I wanted.
Not vindication.
Completion.
For years, Grant existed in my memory as the man adjusting his cuff while defending my $312.64 paycheck.
Now he was also a man willing to examine the design failure without asking me to absolve him.
Both.
Afterward, he sent the draft section he intended to publish.
I refused to edit.
“Why?”
“Because it’s your lesson.”
“Fair.”
The final article did not name me.
Good.
It described a failed compensation incentive at a former company, the dangers of combining cost assignment with manager performance metrics, and the need for independent review.
One line stood out:
“When employees bear downside they cannot predict, leadership should not call uncertainty ownership.”
That was good.
Grant learned.
So did I.
I never became his friend.
We exchanged two messages afterward.
That was enough.
Accountability does not require permanent closeness between the person harmed and the person who learned from causing harm.
Sometimes a truthful conversation completes the useful part.
Then both lives move on.
Grant also told me something I never expected.
After the policy crisis, he had spoken with several former employees who refused to meet him.
“At first I was angry.”
“Why?”
“I thought if I was willing to apologize, they should at least hear it.”
I said nothing.
He smiled grimly.
“I know.”
“What changed?”
“My therapist asked who the apology was for.”
Good therapist.
“If hearing me creates more work for them, they don’t owe me access.”
Exactly.
That was a lesson beyond management.
Accountability does not create entitlement to reconciliation.
Grant had learned it.
I had too.
I could accept his apology without becoming available for an ongoing relationship.
Workers who wanted nothing from Hartwell after repayment could leave the story closed.
Evelyn could reform the company without getting everyone back.
Repair respects the other person’s choice about participation.
That may be the most overlooked part of institutional apology.
Laura’s payroll dispute also helped me see how quickly expertise can become anxiety disguised as preparedness.
For days after her differential was corrected, I still asked:
“Did they update the year-to-date total?”
“Yes.”
“Tax withholding?”
“Yes.”
“Retirement contribution impact?”
“Daniel.”
I stopped.
She smiled.
“It’s fixed.”
I laughed at myself.
Hartwell had trained me to expect a hidden second problem.
Sometimes there is one.
Sometimes a correction is just complete.
Learning when to stop checking became its own skill.
I practiced it at work too.
If an employee dispute was reviewed, corrected, documented, and closed, I did not keep reopening it because I feared missing something.
Trust requires endpoints.
Without endpoints, vigilance becomes another form of dysfunction.
That insight helped me understand Evelyn too.
After the scandal, she could have micromanaged every payroll line forever.
Instead, she built controls and eventually stepped down.
Good leaders create systems they do not personally have to watch every minute.
The lesson works at home.
At work.
Everywhere.
Grant’s industry article eventually became required reading in one leadership course, according to a colleague who sent me a screenshot. I felt a strange mix of amusement and respect. Years earlier I would have resented him receiving professional credit from a failure that hurt workers. Later I understood the value of leaders documenting their own mistakes accurately. The article did not erase who paid the cost. It made the lesson portable. If another executive avoided a similar design because Grant wrote honestly about it, that was useful. Learning from failure is not the same as being rewarded for causing it.
Click here to continue reading: PART 15: When I prepared for retirement, I realized Hartwell had changed how I thought about money forever — but not in the fearful way I once expected
I expected my resignation meeting to last ten minutes, but one pay stub forced the CEO to confront a system her own brother had built
Part 14 of 16
