Nobody Checks On the People Who Are Fine

Attention goes to the fire.

It goes to the new hire who needs onboarding, the struggling report who needs a plan, the loud problem in the channel, and the escalation that lands at four o’clock on a Friday. All of that is triage, and triage is mostly correct.

Then one of your best people resigns and everyone in the leadership meeting uses the same word. Sudden.

The apparent contradiction: managing by exception starves your best people

Ask a leader what their job is and you will hear some version of managing by exception. Handle what is off track. Do not hover over what is working. This is sound advice, it scales, and it is the only way anyone survives a span of twelve people.

Ask the same leader who they are investing in and they will name their strongest people without hesitating. They believe both things and both feel true.

They cannot both be true, because they draw from the same account. Exception handling is a rule that routes every available hour toward whoever is failing, new, or making noise. Your best people generate no exceptions. That is the definition of what makes them your best people. So the rule quietly guarantees that the person carrying the most load receives the least contact, and the leader never notices, because nothing ever appears on the calendar to be noticed.

I have watched this from both chairs. The leaders doing it are not negligent. Most of them are working harder than the people they are neglecting. It is a cousin of the failure in making someone untouchable, where the story you tell about a person replaces any actual contact with them.

The deeper truth: why good employees quit suddenly

Both commitments are real. The relationship between them is what nobody names.

Managing by exception is correct for systems and wrong for people, and the reason is narrow. A system reports its own status. A build fails, a queue backs up, a dashboard turns red, and the exception announces itself the moment it exists. A person does not work that way. A person’s output is the last thing about them to move, which means it reports the state they were in rather than the state they are in.

That delay is the whole problem. Someone can decide they are done and keep shipping excellent work well past the decision. The tickets close. The reviews are clean. The dashboard says green, and the dashboard is telling you the truth about last quarter while being completely wrong about next quarter. By the time output moves, the decision is finished. That is a notification arriving where a warning should have been.

Here is the part that stings. Silence reads as content to the person on the receiving end of it, and what it says to them is rarely what you would have said.

Aristotle has the right word for what is happening on the other side. In the Rhetoric he builds his account of anger on oligoria, slighting, treating someone as being of little account. Slighting does not require cruelty to land. Being passed over registers as a judgment about your worth, and the person doing the slighting is usually unaware they have done anything at all, because from where they stand nothing happened. Nothing happening is the entire event.

So the person who keeps the shop running spends eight months in a room where no one asks them anything. They are not resentful about the workload. They draw a conclusion about how much they matter, and the conclusion is reasonable given the evidence available to them.

Then they take a call from a recruiter who asks them three good questions about what they want, which is three more than anyone internally has asked all year.

A word to that person, since most of this is addressed to the people with the calendars. You can ask. Nobody is going to discover you, because the system your manager runs surfaces problems and you are not one, so you are invisible to it by design. What your leader fails to say is not in your control and never was, and reading their silence as a verdict on your worth is a conclusion you are choosing to draw from thin evidence. Put thirty minutes on the calendar and say what you want next. You may not get it. You will stop guessing, which is worth the thirty minutes on its own.

The integration: giving top performers attention without giving everyone equal time

The fix is not equal time. prosoche, the Stoic discipline of attention, is treated as a practice precisely because attention is finite. Rationing it is the job. A leader who spreads themselves evenly across twelve people ends up thin rather than fair.

What has to change is the order of operations.

Right now your best people are funded out of the remainder. Whatever is left after the fires, the escalations, and the person who needs a performance plan. There is never a remainder. The leftover hour does not arrive, has never arrived, and planning around it is a fantasy you keep re-funding.

So the contact gets scheduled first, and it survives the week. That is nearly the whole intervention. It sounds administrative and it is the difference between having a relationship with someone and having a reporting line to them.

Two old ideas make this concrete. epimeleia is care in the sense of stewardship, the ongoing tending of something you are responsible for, and it does not switch off when the thing you are tending is healthy. A vineyard that is producing still gets walked. And axia is worth, what is actually due to someone. Give each their due, in proportion to what they carry and what they are capable of becoming. Those two together, not seniority and not volume of noise. What is due to a person who has been carrying the load for two years is not silence, and calling that silence trust is a story leaders tell themselves after the fact.

One more thing about why, because the argument so far has been mostly about cost. Everything above is true if the person could never quit. You owe attention to someone doing hard work in front of you because it is owed, and the turnover is what it costs you when you do not pay it. If retention is the only reason you can find to talk to your best people, they will hear that in the first four minutes.

The mastery: what a one-on-one with a high performer is actually for

Scheduling the time is the easy half. The harder half is knowing what the conversation is for, because most leaders get the slot on the calendar and then waste it.

A status update goes nowhere. Their status is fine, that is the problem, and asking a high performer what is blocked produces a polite nothing and four minutes of dead air. Praise goes nowhere either. They have heard it, and the good ones can tell the difference between being valued and being complimented.

Most leaders finally show up once the resignation is already written, which turns the conversation into a counter-offer. A counter-offer works as a bid, and by then you are bidding against someone who asked better questions than you did.

The conversation is the one place where a person who is good at their job gets to talk about something other than their job going well. It is where they get asked what is boring them, what they would build if nobody stopped them, which problem they want next, and whether the thing they are best at is the thing they still want to be known for in three years. Those answers do not surface in a blocker check. They surface when someone asks and then stays quiet long enough to get a real answer.

What the conversation is finally for is their continued development, and everything else follows from serving that. Retention included. A person who is still becoming something does not spend much time reading recruiter email.

Seven things that make this operational.

Put your strongest people on the calendar before the fires, recurring, and do not cancel it. The cancellation is the message. Cancel it twice and you have said everything.

Expect the first few to produce nothing. This is the obvious objection and it deserves an answer: a person who does not raise problems unprompted will not suddenly raise them because it is Tuesday. The first conversation gets you a polite summary. The fourth one gets you something real, because by then the person believes the slot is going to keep existing. Repetition is the mechanism, which is the actual reason cancelling costs so much. Cancelling resets the count rather than postponing the conversation.

Change the question. “What’s blocked?” is built for someone who is stuck. For someone who is not stuck, ask what is dull, what they would change if it were theirs, and what they want to be doing a year from now.

Track contact the way you track output. If you cannot say when you last had a real conversation with your top three people, you have your answer, and the answer is not going to improve by itself.

Spend the currency that is actually scarce. Your best people mostly do not want more praise or more money. They want harder problems, access to the room where the decision gets made, and time thinking with someone who can keep up. Those cost you something, which is why they land. A deliberately harder rep is a gift to someone who has run out of edges.

Stop using output as your trigger for concern. It is the last thing to move. If you wait for the work to slip, you have chosen to be informed last.

Ask what they are becoming, not only how they are doing. The other six items get you accurate information about a person. None of them tell you whether the work is still making them better, and it is possible to run a perfectly informed team of people who stopped growing two years ago.

And when someone does resign, retire the word sudden. The signal was the absence of a signal, and it was available the whole time. Missing it is a fair mistake to make once.

I keep coming back to how ordinary this failure is. It does not require a bad leader. It requires a busy one with a sensible rule, and the rule does the damage on its own. It is the same arithmetic that makes nice leaders ruin people, running in the other direction.

Final Thoughts

Excellence is not self-sustaining, and that assumption is buried underneath all of this. It fails for the same reason fitness fails. Excellence in a person is a condition, and conditions need conditions. Take away the hard problems, the honest response, and the sense that someone is watching what you are turning into, and the capability does not hold steady at its current level. It quietly declines while the output stays flat, which is the part that makes it so easy to miss. The people who need the least supervision still need contact, and the fact that they will not ask for it usually says more about the kind of person they are than about what they want.

arete is something you practice, and it decays quietly when you stop. It runs on hard problems, honest feedback, and someone paying attention to what you are becoming. Take all three away and the work will look fine for a long while, which is exactly what makes the neglect so easy to sustain.

Go look at your calendar for the last month. Find the people who never appear on it. That list is where your risk actually lives, and you built it by accident, one reasonable Thursday at a time.

Paying attention to the right people is a discipline, and disciplines have to be practiced somewhere. That is what MasteryLab is for.