Chapter 3: The Pattern in the Schedules

The HR director returned before the day was over with two investigators from the outside review team and a stack of records thick enough to make my desk disappear.
The roster for Emily’s next week was still on hold. No one had confirmed who entered the change, when it happened, or why. I had told myself that waiting for the records was responsible. Looking at the stack, I understood that waiting could also become an excuse.
The HR director set a laptop beside the folders.
“We’ve started building a timeline,” she said. “Leave requests, posted schedules, draft schedules, and the complaints we’ve been able to locate.”
“Across every property?”
“Across the properties covered by the initial review. Some records are still being collected.”
One investigator opened a spreadsheet. Rows of dates, shift totals, and file references filled the screen.
“This is not a finding of retaliation,” she said. “It’s a comparison tool. We’re looking for changes that occurred near an emergency-leave request, then checking what else was happening at the property.”
“What else could explain them?”
“Room occupancy. Staffing levels. A role change. A schedule correction. Whether the employee requested fewer shifts. And whether the records are complete enough to tell.”
I pulled a chair closer.
“Show me what you can say with confidence.”
She turned the laptop toward me. A few entries had been highlighted—not as proof, she emphasized, but as cases where the timing warranted another look. An employee had requested leave for a family emergency; a later roster showed fewer hours. In another file, the employee’s hours had changed, but the earlier schedule was missing. Elsewhere, a complaint described lost shifts, while the surviving records showed only a revised roster, without a reason attached.
The sequence looked simple when reduced to a line on a screen.
Request. Schedule change.
But a line couldn’t tell me whether a manager had punished someone, whether a schedule had been adjusted for an unrelated reason, or whether a record had been entered after the decision was made.
“What can we establish about the dates?” I asked.
“Sometimes the date a schedule was saved,” the investigator said. “Not necessarily the date the change was decided. The system records a user account, but that may not tell us who initiated the request or who approved it.”
“And the complaint files?”
“Some have a clear record of follow-up. Some don’t. We’re matching them to schedules and communications where possible.”
I looked at the highlighted rows again. A company that could track the temperature of every wine cellar and the turn-down time of every suite could not always tell why an employee had lost a shift.
“Can you compare the hours before and after each emergency absence?” I asked.
“We can,” the investigator said. “But we need to be careful about what the comparison means. A drop after an absence is a reason to ask questions. By itself, it doesn’t answer them.”
I felt the old impulse to reach for a decisive sentence, something a board could vote on and an employee could understand. Instead, I looked at the signature on the labor-efficiency approval, reproduced in the board packet beside my name.
Six months earlier, I had let a number stand in for a question.
“How many cases are we talking about?” I asked.
“Several currently merit closer review,” the HR director said. “The complaint count is higher, but not every file has enough information to match it to a schedule.”
“Do not present the complaints as proof. And don’t dismiss them because the records are incomplete.”
The investigator nodded.
“That’s our approach.”
“I want every case separated into what is verified, what is reported, and what remains unknown. If the records are too thin to tell us, write that down. Don’t convert a blank field into a conclusion.”
The HR director began taking notes.
“And no one should be asked to prove retaliation just because the company failed to keep a usable record,” I added. “We still have to investigate fairly, but the gaps belong in the findings.”
For a moment, nobody spoke. The request sounded obvious. The fact that it had to be said did not.
The HR director changed the display to a new tab. “We also found inconsistencies in how properties record emergency leave. Some use a specific code. Some enter a general absence. Others appear to rely on manager notes outside the scheduling system.”
“So the same event could look different depending on the hotel.”
“Exactly. That makes a company-wide comparison harder. A missing code doesn’t mean no emergency happened, and a code doesn’t tell us what followed.”
The investigators began listing the records they would need next: schedule versions, user histories, messages, approval logs, and any relevant manager guidance. The task sounded less like a search for one bad actor than a search for the path a decision took—from policy, through a manager, into a worker’s paycheck.
That was uncomfortable. It was also what I had asked them to do.
A message from the COO appeared on my phone. The board wanted an update before its afternoon discussion of interim guidance. The COO asked whether we had enough information to recommend a temporary pause.
I read the message twice.
“We have enough information to say the current records are inadequate,” I said. “We don’t have enough to say every schedule change was retaliatory.”
The HR director watched me.
“I’ll tell the board the same thing,” I continued. “We need a temporary rule that protects emergency leave while the facts are gathered. But I won’t pretend the investigation has already established who did what.”
“That may leave directors asking what they’re supposed to decide,” she said.
“Then we give them the distinction. What we know, what we don’t, and what harm can continue while they debate.”
I sent the reply. It did not solve Emily’s schedule. Her individual paid-leave arrangement remained in place, and the draft roster stayed frozen for review. No company-wide protection had been approved.
The investigators returned to their files.
Later, I stood by the window while the HR director and one investigator checked the next batch of records. Atlanta spread below us in winter-gray blocks of glass and concrete. At street level, guests were arriving with suitcases, expecting the ease our company sold them. Somewhere beneath that promise, housekeepers were waiting to learn whether a family emergency might cost them hours.
My phone remained silent. Emily had not called, and I would not ask her to carry the investigation for us. She had already told me what she needed. The burden of proving our systems were fair belonged to the company.
The HR director broke the quiet.
“We found a few files where the complaint language and the schedule history don’t line up cleanly.”
“What does that mean?”
“In one case, the complaint says hours kept decreasing after an emergency absence. But the file was marked resolved, and the final note doesn’t explain what changed.”
“Did the employee’s hours recover?”
She studied the screen.
“That’s what I’m checking.”
I turned from the window. The spreadsheet had grown, the highlighted rows now marked with new questions rather than answers. One entry was open in a separate window. The complaint summary was brief. The resolution field held a single word.
Resolved.
Beneath it, the investigator scrolled through the schedule history.
The employee’s hours had continued to shrink.
May you like
The HR director leaned closer to the screen.
“This complaint was closed,” she said. “But I can’t find anything showing the employee’s schedule stopped declining.”