Most of what was written about quiet quitting was nonsense, and the term itself is the worst part of it.
Doing your job and going home is not quitting. It is the arrangement. The reason it needed a name — and a slightly sinister one — is that a lot of employers had quietly come to depend on unpaid discretionary effort and were startled when some of it went away.
So let's set the framing aside and ask the question underneath it, which is real: is there a measurable pattern that predicts someone leaving, and can you see it early enough to do something?
Yes. But it is not what the discourse said it was.
What actually shows up in the data
Disengagement rarely looks like doing less work. That is the intuition, and it is wrong, which is why managers miss it.
What changes first is the shape of the effort, not the amount. Specifically, the discretionary parts go quiet:
The person who used to leave detailed review comments now approves with "LGTM". The one who used to wander into other teams' design discussions stops. Someone who reliably picked up the ambiguous, unowned problems now works strictly inside their ticket. Questions in the team channel go unanswered by the person who always used to answer them.
Core output holds. Tickets close. Standups sound normal. On every dashboard that counts throughput, this person is fine.
They have just stopped doing the parts nobody assigned them — which, in most knowledge teams, is where a disproportionate amount of the value was.
The second pattern: the boundary snap
There is a distinct version that looks different and is worth separating.
Someone who has been working late for months suddenly stops. Their after-hours activity, which had been creeping up steadily, drops to zero more or less overnight.
The naive read is that they have disengaged. The more common read, in our experience of what these curves look like, is that they hit a wall and set a boundary — and that the months of creep before the snap were the actual signal. That is the burnout pattern, and the moment to intervene was six weeks earlier when the after-hours line was still climbing.
Both patterns end with reduced discretionary effort. They have completely different causes and completely different fixes. If you cannot tell them apart, you will apply the wrong one about half the time.
Why the "crack down" response makes it worse
The instinct, when a manager notices withdrawal, is to increase visibility. More check-ins, tighter reporting, closer supervision.
This reliably accelerates the exit, and the mechanism is not mysterious. Someone who has pulled back their discretionary effort is usually already reconsidering the relationship. Responding with surveillance confirms the thing they were starting to suspect.
The same data, used to open a conversation instead — your review comments have gone quiet, has something changed? — sometimes recovers the person, and almost always tells you something true about the job. Occasionally what it tells you is that they are leaving regardless, which is still useful information three months earlier than you would otherwise have had it.
What this means for measurement
The tempting move is to build a disengagement score. Do not build a disengagement score.
The moment people believe a metric flags them as a flight risk, two things happen. The engaged ones start performing engagement, and the actually-disengaged ones — who are already halfway out — get very good at looking normal. You will have destroyed the signal to build the dashboard.
What works is duller. Watch collaboration and discretionary activity at team level over months, not weeks. Notice when an individual's pattern changes relative to their own baseline rather than against a team average, because people are legitimately different and ranking them is how you get the performance-theatre failure mode. And treat every change as a prompt for a conversation rather than a conclusion, because the software genuinely cannot tell you why.
The wider method is in how to track employee productivity, and the principle that keeps it honest is in ethical monitoring: if you would not be comfortable showing someone the data you hold about them, you are already doing it wrong.
The part managers do not want to hear
Discretionary effort is a gift, not an entitlement. When it stops, the question that produces useful answers is not "what changed in this person" but "what changed in this job".
Usually something did. A reorg that removed their scope. A promotion that went to someone else without explanation. A manager change. Eighteen months of being the person who absorbs every escalation without that ever appearing in a review.
Quiet quitting, in the cases we have seen described, is almost always a lagging indicator of a management decision. The data tells you when. It does not tell you what — but if you know when, you can usually work out what.
ProdView shows collaboration and working patterns at team level from activity metadata, with everyone seeing their own data, which is the only version of this that does not backfire. Free for three seats.