Recovery

Quiet Leavers

The bill for sustained operational pressure rarely arrives during the incident. It arrives three months later, in a resignation letter from the person who held the line.

This piece discusses pressure, burnout and recovery. If any of it lands close to home, Resources lists support lines that answer at any hour.

Three months after the incident closes, the load-bearing engineer hands in notice.

They were not the one who complained during the response. They were not the one who pushed for the review. They were the one who kept the bridge running on day three when no one else could remember what they were supposed to be doing. They held the team together while it was on the line, then quietly went home, slept four hours, and came back. The org has been running on what they did for weeks afterwards without quite noticing.

The resignation letter does not mention the incident.


The interval

The bill for sustained operational pressure does not arrive when you expect. The technical incident has a clear ending. Containment is declared. Recovery begins. The Slack channel goes quiet. The team gets back to their normal work.

The person who held the line goes back to their normal work too. They are tired. They sleep. They take a long weekend. They are surprised to find themselves still tired after it. They notice that work which used to feel meaningful now feels like work. They notice the gap between what they put in and what they get back has changed shape.

They keep going. The org praises them. They are functional. Functional is not the same as recovered.

Somewhere in the quiet weeks that follow, something gives. It is rarely dramatic. They start having a conversation in their own head about whether this is a job they want to keep doing. The conversation is patient. It comes back. It refines itself. Eventually they are editing a resignation letter in their head, sleeping with it next to them.

Then they hand it in.

The exit interview does not mention the incident. The reason on the form is something like “pursuing new opportunities”.

Who it is

It is almost always the load-bearing one. The person who absorbed the most without making noise. The one with the highest signal-to-complaint ratio during the response. The one who knew the system well enough to keep working when the rest of the team was waiting for direction.

This is not random. The person with the most capacity is the person the system asks the most of. Over a long enough horizon, they are also the person with the most reason to quietly decide the job cost more than it returned.

It is rarely the loudest person who leaves. It is rarely the most stressed-looking person who leaves. It is the steady one. The one who, if asked at the time, would have said they were fine.

They were fine for the duration of the work. They were not fine afterwards. They did not say so because afterwards is not a time the organisation has formally allocated for being not fine.

What the review misses

The review accounts for the incident. It does not account for the person who carried it.

The review is built to capture what happened. It is good at the technical sequence. It is good at the decisions and the times. It is sometimes good at the comms gaps and the detection misses. It is not built to capture what happened to the people who were in the room.

The review closes. The action items get owners. The lessons get added to the wiki. The team congratulates itself, gently, on having handled it.

Months later, the person who held the bridge through the longest stretch sends the resignation letter. The organisation loses institutional memory it did not know was load-bearing until it was gone.

The next incident, when it comes, lands on a thinner team. The team is tired in ways the review did not record. The organisation is also less aware than it thinks it is, because the most aware person has left.


Swap the incident for anything with a hard middle and a clean ending. The launch that ran on one team’s nights. The school term that ran on one teacher. The year of care that ran on one member of the family. Wherever the work ends cleanly and a person carried it there, the interval starts.

Quiet leavers are not a culture problem. They are a structural one. They are what happens when the response gets a design and the recovery gets none.

Nothing in the interval is fixed. It is the one stretch where design still works: load actually moved rather than praised, a check-in that is scheduled rather than hoped for, recovery treated as part of the response. Organisations that build for the interval keep the person.

The response had a design. Give the recovery one.

Come back