Your Best People Didn’t Leave

I went to an event a while back — one of those rooms that’s more or less your whole career standing in one place. People I worked with fifteen years ago. People I worked with last year. People who retired and came anyway, because these are still their people.

I expected the usual. War stories, the good kind of complaining, somebody’s boat photos.

It was quieter than that.

Not angry. Angry would have been a better sign.

What I heard, over and over, in slightly different words, was the same sentence: I don’t recognize the place anymore. And underneath it, the part that was harder to hear: so I’ve stopped trying to.

These weren’t people on the edges. I want to be clear about that, because it’s the whole point of what comes next. These were the engines. The ones who carried rollouts that had no business working. The ones who knew where every body was buried and used it to help instead of to block. If you’d asked me ten years ago to name the people who actually got things done, I’d have named most of that room.

Now they’re keeping their heads down, waiting out a clock, or gone in every way except the paperwork.

Disengaged is the word we’d normally reach for, and it isn’t quite right. Disengaged sounds like they stopped caring. That isn’t what I saw. They still care. They’ve just stopped spending it here.

We have a word for this and it’s the wrong word

We say burnout, and everybody pictures a collapse. The person who can’t come in. The dramatic version.

The real one is much quieter, and it comes in stages. In the Burnout Spectrum I use with clients it runs: Off Week, Sustained Drag, Visible Strain, Withdrawal, Gone.

That room was sitting at Withdrawal.

Withdrawal is the dangerous one, for one reason: it still shows up. Withdrawal hits its numbers. Withdrawal answers the email inside the deadline. Withdrawal sits through the town hall and doesn’t ask a thing.

Withdrawal will do the job exactly, and not one thing more — and the one thing more is where every good idea your organization has ever had came from.

You can’t see Withdrawal on a dashboard. That’s not a flaw in your dashboard. That’s what the stage is.

The thing that matters most is the thing we’re worst at

Here’s the number that made me rethink this.

Perceptyx’s 2026 benchmark — 23 million employee responses, 490 organizations, 113 countries — ranks what actually drives engagement. Change management is number one. Not pay. Not belonging. Not your manager. How the place handles change.

Four of the top five are about how change gets handled and what the future looks like. Not the day-to-day.

Now the other half of it. Change management scores 55% favourable — dead last inside that same top five. Belonging scores 75%.

Put those two together and it’s hard to look away from: the biggest reason your people are still switched on is the thing your organization is worst at.

That’s not a communications problem. That’s a design problem, and it’s sitting at the top of the list.

The part I’ll probably get email about

In my experience, every organization says it believes in change management. And where the whole organization doesn’t, there’s a group inside it that does — a division, a program team, one leader who means every word of it.

They say it in the job posting. They say it when they bring you in. They say it warmly, in the hallway, to somebody else: you should talk to them, they’re good at this.

And then push comes to shove, and they don’t choose it.

I want to be fair about why, because the cynical read is the wrong one. It’s almost never malice. It’s almost always politics and cost. The decision is already made. The date is already out there. Somebody has put their name on it. And the consultation would turn up something inconvenient at the exact moment inconvenient gets expensive.

So what’s left standing is the furniture. The comms plan, the training deck, the launch email, the cascade — all of it visible, all of it defensible, none of it the thing. What gets quietly dropped is the expensive half: going and finding out something you’d rather not know, early enough that knowing it would change the plan.

But that is the discipline. The rest is scheduling.

And people can tell the difference. That’s the finding I’d put on the wall of every executive floor I’ve ever worked on. In Perceptyx’s May 2025 survey of more than 3,000 employees, 77% of executives thought senior leaders were guiding change effectively. 49% of the people doing the work agreed.

Twenty-eight points. That gap isn’t a perception problem you fix with better messaging. That’s two groups accurately describing two different versions of the same event.

Why it piles up, and why your most experienced people go first

Between 2016 and 2022, the number of major changes the average employee absorbed went from two a year to ten. Over the same stretch, willingness to support those changes fell from 74% to 38%.

Five times the volume. Half the willingness.

This is the thing I believe most strongly after twenty years of this: every badly led change makes the next one harder. Not as a metaphor. Mechanically. You’re not just spending this quarter’s goodwill, you’re borrowing against the next transformation’s. It adds up, and it comes due.

And it comes due on your most experienced people first.

Not because they’re resistant. Because they’re calibrated. They’ve seen it done well and they’ve seen it done badly, and they can tell you which one is happening by about week two. They know the difference between being consulted and being told after the fact by somebody using the word consultation. They’re not cynical. They’re experienced, which from a distance looks identical and is the opposite.

So they do the sensible thing. They stop spending effort nobody asked for on a process that’s shown them repeatedly it doesn’t want their input, and they put it somewhere it pays off. Their team. Their craft. Their life outside the building.

That’s not a character problem. That’s a fair price, and your organization set it.

The signal nobody reads properly

Silence is not agreement. I’ll say that before anything else, because it’s the mistake I see most and it’s the expensive one. A room goes quiet, somebody upstairs reads it as they’re fine with it, and the programme carries on into a wall.

A quiet room usually isn’t agreement. It’s a conclusion. People have worked out that talking doesn’t change anything, and they’ve adjusted accordingly.

But there is a silence that means it worked. And two rooms can sound identical from the doorway and mean opposite things, which is why silence gets misread so often and so confidently.

Here’s how you tell them apart — and it isn’t by listening harder to the silence. It’s by checking whether anything else is still moving.

I watch five channels to know whether a change has taken root:

That last row is the whole point. Silence only counts as a good sign when the other four are still alive.

Full silence is a change nobody talks about anymore because it stopped being an event and became the way the work sounds. Boring. Absorbed. Underneath it, the other four are all still moving — people are teaching it, refining it, flagging their exceptions.

Emptied silence looks exactly the same from the doorway. Attendance holds. Compliance holds. Everything else has stopped. No questions, no pushback, nobody teaching it, workarounds nobody bothers to flag anymore because flagging them never did anything.

One is the sound of a change that’s finished. The other is the sound of a room that’s finished with you.

The room I described at the top of this was the second kind, scaled up to a whole career’s worth of people.

What I’m not saying

I’m not arguing for less change. Places that don’t change get overtaken, and the people inside them lose a lot more than one hard year. Nobody in that room wanted it frozen in 2011.

I’m also not saying everyone who has checked out is somebody’s fault. People check out for reasons that have nothing to do with you.

What I am saying is narrower, and harder to dodge: the cost of doing change badly doesn’t get paid at go-live. It gets paid years later, by the next initiative, in people who used to put their hand up and now wait to be told. And because none of them ever show up in an attrition number, nobody gets a bill with their name on it.

What I’d do about it

I said recently that most of change management is furniture, and there are only three things I’d keep. The diagnostic, the sponsor, and the close. Here’s what each one looks like when the thing you’re trying to head off is this.

And then one more, which isn’t a practice and doesn’t belong in the three.

Go back to the people who’ve seen it before. The person who’s been through three of these is the most under-used asset in most transformations. They’re not blocking you. They’re holding twenty years of pattern recognition that your project plan doesn’t have, and most of them will hand the whole lot over the first time somebody asks a real question and then does something with the answer.

It’s the cheapest thing on this list, and I’ve watched it work in rooms where nothing else did. It costs one conversation and a willingness to hear something you didn’t want to hear.

And most of those people would still say yes. That’s the part I keep coming back to. They haven’t gone anywhere. They’re three metres from a manager who’s decided they’re disengaged, waiting to be asked one decent question.