Empty Station

Pressure: The resignation takes a day. What follows belongs to the people who stayed.

Ray resigned on a Tuesday.

He did it the way you would want someone to do it. Weeks of notice. A written handoff for every account he owned. He sat down with the two people picking up his territory and walked them through which customers needed a call in the first month and which ones would hold until the quarter turned. He was in sales, he had been with us for years, and he left the way he had worked.

A few days later, in a room with a lot of other people in it, a boss of mine said that Ray had never really been a committed member of the team.

I want to be accurate about what that was. It was not rage. It was said evenly, almost in passing, folded into a longer conversation about pipeline. And it was not true. I had been in the trenches with Ray. He had carried quarters that would have flattened other people, and the person saying it had been in the room for most of them.

I said so. Not well, and not at length. Something close to: that is not right, Ray worked hard here and he left this place in better shape than he found it. The conversation moved on inside of a minute.

What I remember is the room. The room did not argue with me and it did not agree with me. People looked at their laptops. It took me a long time to understand that my correction had not mattered much, because the sentence had already done its work. Every person in that room had just learned something, and it was not about Ray.

What a team learns in the days after someone leaves is not about the person who left. It is about what happens to your record here when you go.

I believed at the time that I was watching a bad moment. A leader under pressure, saying something he would not say again. I filed it that way for years.

Then I left, and he said the same kind of thing about me.

That is when I understood it had not been about Ray, or about me. It was a reflex, and the reflex had a logic underneath it. If someone leaves and their work still counts, their leaving says something about the place they left. If their work did not count, it says nothing. Rewriting the record is cheaper than asking why they went.

Meanwhile the actual work Ray had been doing did not go anywhere. It sat there, unowned, for about a week, and then it went somewhere. Not by decision. By gravity.

Pattern: A load left unassigned still gets assigned. It settles on whoever is nearest and least likely to say no.

Identify the Pattern

Two things happen when a key person leaves, and organizations usually only manage one of them.

The first is the staffing problem, which is the visible one. A seat is open. A requisition gets written. Somebody starts interviewing. It takes months, the months are understood as a given, and so it gets treated as the hard part.

The second is the load problem, and it resolves itself in about two days whether anyone attends to it or not. The work does not wait for the req. It moves. And absent a decision, it moves along the path of least resistance: to the person physically closest to it, to the person who already knows the system, to the person who has historically said yes. Conscientiousness is a selection criterion here, which is why the people who absorb an empty station are frequently the people least able to afford it.

The load also travels below the level of anything written down. It shows up as three extra approvals in someone’s week, a standing meeting that person now runs, a customer who now calls them. It sits in no plan and carries no end date. And because the output holds for a while, the arrangement looks like it worked.

Then there is the part that has been measured. Coworkers watching someone leave are not neutral observers. Felps and colleagues studied this directly and found that a person’s likelihood of quitting is shaped by the job search behavior of the people around them, over and above their own satisfaction with the job.^[1] Departures are informative, and people update on them. The first person to leave changes the arithmetic for the people who watched them do it.

Across a large body of studies, turnover’s drag on an organization tends to surface first in operational measures such as quality, safety, and customer service, and only later in financial results.^[2] Which means the empty station is usually visible somewhere before it is visible in the numbers a leadership team reviews. Somebody is already seeing it. It is rarely the person reading the dashboard.

In the days after a departure, a team is doing something more careful than gossip. They are collecting evidence about what this place does with your history once you are gone. They will get that evidence from how you describe the person, from whether you say anything at all, and from whether the version you give matches the one they lived through. This is not a soft concern. It determines how much anyone tells you the next time they are thinking about leaving, which determines whether you find out early enough to do anything about it.

What It Costs

None of this stays in the register of how people feel. Four places it lands, none of them soft:

Decision quality and speed. When a team concludes that leaving gets your record revised, they stop telling you early. Not out of spite, out of arithmetic: there is no upside in signaling that you are looking. So the next departure arrives as a resignation rather than as a conversation three months out. You lose the lead time, and every retention decision after that gets made later and with worse information.

Cross-functional throughput. The load does not distribute evenly across a job. It moves to whoever is nearest, and nearest usually means inside a function. The parts of the role that crossed boundaries are the parts least likely to get picked up: the standing relationship with the other team, the escalation this person handled because they knew who to call, the translation between two groups that did not quite speak the same language. Those belonged to no job description and were visible to few people. Work at the seams degrades first and reports last.

Organizational prioritization. Interim coverage with no end date commits real hours against nothing on paper. Next quarter you plan against a model that assumes those hours are available, and they are not. You will prioritize with confidence against a number that is wrong, and the gap will surface later as slipped commitments that get attributed to estimation.

Trust, which compounds the other three. Trust here is not a mood. It is the mechanism that governs how much people give beyond what was asked, and most of what an organization calls excellence lives in that margin: the catch that was not assigned to anyone, the second look at something that already passed review, the hard conversation someone starts because they care about the outcome more than the awkwardness. When trust drops, that margin closes first, and it closes without an argument. Compliance holds. Discretion goes. Output looks stable for a quarter or two, which is exactly why this is easy to miss until it is expensive.

What moves into the space discretionary effort leaves is cynicism, and cynicism is costly in a specific way.

A person who watched a good colleague get written out of the record, and drew a conclusion about the place, is not being cynical. They are reasoning correctly from evidence you supplied.

That is the part a message cannot reach. You do not argue someone out of an inference they made from your own behavior. You change the behavior, and then you wait, because they will want to see it more than once.

Reframe the Mindset

Two readings get in the way here, and they fail in opposite directions.

The first is that a departure is simply a resourcing event. Post the role, run the process, restore the headcount, and the problem closes. This reading is not wrong so much as slow. It attends to the thing that takes three months and ignores the thing that resolved on Thursday. By the time the new person arrives, the load has already settled into an arrangement, and the arrangement is now the baseline that the new hire gets measured against.

The second is that a departure is a verdict, on the company or on the person. Either they were disloyal, or the place is broken. Both versions are satisfying and both stop useful inquiry, because once you have a verdict you no longer need the facts. The facts are duller and more actionable: this person was carrying a set of things, some of which were on their job description and some of which were not, and those things are now somewhere. Finding out where is not an emotional exercise. It is an inventory.

There is a third reading available, and it is the one worth practicing. A departure is a short, unrepeatable window in which a team is paying close attention to what its leaders do. Most weeks you have to work hard to be heard. This is not one of those weeks.

Transform the Tension

Three moves. None of them require you to feel good about the person leaving, which matters, because sometimes you will not.

Say the true generous thing in the first two days, out loud, in front of the team. Not a tribute, and not a statement you have to walk back later. One accurate sentence about what the person contributed, said in a room, early. Losing someone good hurts, and the hurt does the editing before you notice it happening. The version that comes out is rarely a lie. It is a true account with the best parts left out. So the work is not being generous, which is easy to fake and easy to spot. It is being accurate while it still stings. The team is not evaluating your warmth. They are checking whether the record survives contact with your feelings.

Redistribute on purpose, with names and an end date. Write down what the person was actually holding, including the parts that were not in their job description, and then assign each item to a named person for a stated period. Say the period out loud: until the backfill starts, or for six weeks, whichever comes first. Absorbing for a defined stretch is a plan. Absorbing with no end date is a role change in disguise. Pay particular attention to whatever crossed a boundary, because that is the part your own function will not think to claim.

Ask the people who stayed a specific question, not a general one. “How are you holding up” gets you a polite answer. “What landed on your plate this week that was not there two weeks ago” gets you an inventory. Ask it individually, ask it within the first two weeks, and write down what you hear, because the sum across four people is usually larger than any of them realizes, and larger than any one of them can see from where they sit. That sum is also the correction to next quarter’s capacity model.

Shift: Who picked up what this person was carrying, and was it intentional, or was it gravity flowing in the path of least resistance?

Reflection (Questions to Ponder)

• When someone left recently, what did I say about them in the first two days, and would they recognize the version I gave?
• When the person leaving was someone I counted on, did the account I gave of them survive my own disappointment?
• What was this person carrying that was not written down anywhere, and do I actually know?
• Whose week got longer when that seat emptied, and did I decide that, or did it settle?

Connection (Questions to Discuss)

• What have we each picked up since the last departure, and did any of it come with a decision attached?
• If someone here were thinking about leaving, what would they expect us to say about them afterward?
• What are we currently treating as temporary coverage, and how long has it been temporary?
• Which of our commitments next quarter assume hours that are already spent covering someone who left?

Subtraction (Things to Stop)

• Stop treating the requisition as the response. The req addresses a seat. The load moved days ago and is sitting on someone specific right now.
• Stop revising a departed person’s contribution downward. It costs nothing in the moment and it teaches the room exactly what it will get.
• Stop reading steady output as evidence that the coverage is working. Output is the thing the coverage exists to protect, so it will be the last place strain appears.
• Stop treating cynicism on a team as an attitude problem. It is usually an accurate reading of something the organization did, and it responds to changed behavior rather than to messaging.

Action (Things to Try)

• This week, write down the last person who left and list what they carried, including three things that were not in their role description. Then write a name next to each item. If you cannot fill in a name, you have found something that is either not getting done or getting done by someone you have not thanked.
• Give every piece of interim coverage an end date, in writing, and tell the person the date. Then keep it or renegotiate it out loud when it arrives.
• Before the next planning cycle, subtract the hours currently going to uncovered work from the capacity you are about to commit. Plan against the number you actually have.
• Ask one person who stayed what changed for them, using the specific question rather than the general one, and resist the urge to solve it in the same conversation. Write it down first.

Years after that meeting, I left too, and I got the same treatment Ray had. I would like to say the symmetry taught me something noble. Mostly it taught me that the reflex is common and it is not personal, which is a smaller lesson than I wanted and a more useful one.

What I do differently now is narrow. When someone leaves, I write down what they were carrying before I write the requisition, because the second thing is easier and I will do it first if I let myself. And I try to say one true thing about them in the first forty-eight hours, in a room, while people are still listening. Not because it helps the person who left. They are gone, and it does very little for them.

It is for the people who stayed, who are working out, without saying so, what this place will say about them someday. They are going to reach a conclusion either way. The only question is whether you are the one who gives it to them.