The New Employee Kept Correcting My Reports in Private—Until I Saw Who Had Trained Her
The first correction arrived on her third day.
Hi Julia—small note: the retention figure in section two is based on active accounts, not total accounts. I updated the footnote in my copy.
The email was polite.
That almost made it worse.
I had written monthly performance reports for four years. Maya Chen had been at the company for seventy-two hours.
I checked the number.
She was right.
I fixed it and replied, Thanks.
The next week she found another issue.
Then another.
A mislabeled quarter.
A source note that referenced an outdated dashboard.
A chart axis that made two segments look more different than they really were.
None of the mistakes were catastrophic.
All of them were mine.
And Maya never corrected me in meetings.
She sent private messages after.
At first I appreciated that.
By the sixth correction, I started resenting it.
Maya joined our analytics team after working two years at a regional bank.
She was quiet in meetings, quick with spreadsheets, and uncomfortable with small talk.
She also had the annoying habit of asking where numbers came from.
Not what they meant.
Where they came from.
“Which table feeds this?”
“Is this filtered for closed accounts?”
“Does this include refunds?”
Our team knew the answers most of the time.
When we did not, we usually said, “That’s how the dashboard is set up.”
Maya never seemed satisfied by that sentence.
One afternoon, after she sent me a correction about customer counts, I walked over to her desk.
“Can I ask you something?”
She looked worried immediately.
“Sure.”
“Are you reviewing all my reports line by line?”
Her face turned red.
“Not exactly.”
“What does that mean?”
“I read them because I’m trying to learn the business.”
“And you keep finding mistakes.”
“Sometimes.”
I heard defensiveness in my own voice and disliked it.
“You can tell me if something is wrong,” I said.
“I have been.”
That was fair enough to be irritating.
I went back to my desk.
That evening, I opened six months of my old reports.
I found three more small inconsistencies Maya had not mentioned.
That scared me more than her emails.
My reports went to senior leadership.
People made staffing and budget decisions using them.
We had never had a major error because our data pipeline was mostly automated.
But “mostly automated” had become an excuse for casual review.
The next morning I asked my manager, Peter, whether we had a formal report-checking process.
“You are the process,” he said, laughing.
I did not laugh.
“That might be a problem.”
Peter looked at me.
“Did something happen?”
“Nothing major.”
I explained the corrections.
He shrugged.
“Maya is intense. Her old manager was famous for that.”
“Who was her manager?”
Peter searched his memory.
“Ellen Ward, I think.”
The name meant something to me.
I opened LinkedIn.
Ellen Ward had worked at our company eight years earlier.
Before my time.
Her title had been Director of Analytics Governance.
I had never heard of that role.
I searched our shared drive.
There was an archived folder with her name.
Inside were documents no one on my current team had touched in years.
Validation checklists.
Metric definitions.
Source maps.
Change logs.
One file was titled: Monthly Executive Report QA Procedure.
I opened it.
There were twelve review steps.
Source verification.
Population definition.
Date-range consistency.
Visual scale review.
Footnote review.
Peer sign-off.
My current process had one step.
I looked at it before sending.
I felt my stomach drop.
Institutional knowledge often disappears quietly—not because anyone rejects it, but because the person carrying it leaves and no one realizes the process left too.
I called Maya into a small conference room.
“Did Ellen Ward train you?”
She looked surprised.
“Yes.”
“At the bank?”
“She consulted there for about a year.”
I turned my laptop around and showed her the archived checklist.
Maya smiled for the first time in the conversation.
“That looks like her.”
“Why didn’t you tell me we used to have this process?”
“I didn’t know you used to have it.”
Of course.
She was new.
I was blaming her for failing to explain our own history to me.
“At your last job, did you use something similar?”
“Yes. Every recurring report had a reviewer.”
“And here?”
“I thought maybe the review happened somewhere else.”
“It doesn’t.”
She hesitated.
“I figured that out.”
I laughed despite myself.
“Clearly.”
We spent an hour comparing Ellen’s old procedure with our current workflow.
Some of it was obsolete.
The old source systems no longer existed.
Some steps were unnecessarily manual.
But the principles were good.
Every metric needed a defined population.
Every chart needed a scale check.
Every recurring executive report needed another person to review it.
I asked Maya why she had corrected me privately instead of raising the issue with the team.
She looked embarrassed.
“I didn’t want to seem like the new person who thinks everyone is doing things wrong.”
That was exactly how I had been interpreting her.
“You were trying not to do that?”
“Very hard.”
I leaned back.
“You should know it wasn’t working.”
She laughed.
Then I apologized.
Not for asking questions.
For assuming motive instead of examining the pattern.
Maya apologized too.
She admitted some of her messages were overly detailed because anxiety made her over-document.
We agreed on a better system.
At our next team meeting, I showed everyone the archived checklist.
Peter raised his eyebrows.
“Where did you find that?”
“Under eight years of digital dust.”
We reviewed it together.
No one wanted twelve manual steps.
So we built six.
We automated source checks where we could.
We created a short metric-definition page.
We added a peer-review rotation.
And we stopped treating report ownership as report immunity.
The first month was annoying.
Review took longer.
People found inconsistencies in each other’s work.
Some corrections felt trivial.
Then we found a major issue before publication.
A dashboard filter had changed during a software update and excluded one customer segment.
The report would have understated churn by almost nine percent.
Maya caught it.
This time she did not send me a private message.
She flagged it in the review channel.
I replied, Good catch.
And meant it without resentment.
Three months later, Peter asked me to lead a broader documentation project.
“You’ve become obsessed with governance,” he said.
“Apparently.”
Maya overheard.
“Ellen would be proud.”
I emailed Ellen that afternoon.
I had never met her.
I told her we had found her old documentation and adapted part of it.
She replied the next morning.
I’m delighted someone found that folder. I always wondered whether it vanished with me.
That sentence stayed with me.
Organizations like to talk about knowledge as though it lives in systems.
Often it lives in habits.
In one person who remembers why a field matters.
In a checklist no one knows exists.
In a new employee who asks inconvenient questions because someone once taught her to.
Maya still corrects my work.
I correct hers too.
Now the corrections are part of the job rather than evidence of a relationship problem.
And every month, before I send the executive report, I open our six-step checklist.
The final line says:
Someone other than the author has reviewed this.
That line has saved us more than once.
It has also saved me from becoming the kind of experienced employee who confuses familiarity with accuracy.
