Most descriptions of decision fatigue focus on the point where decisions start going wrong. That framing is why so many people who have it do not recognise it: their decisions are not going wrong. They are arriving late, expensively, and with none of the ease that used to come with them.
Degraded output is the late stage. Long before it, there is a stage where the quality holds and the cost quietly rises — and because organisations measure quality rather than cost, nothing in the environment flags it.
What Actually Degrades First
Judgement is not one capacity. It is several, and they fail in a predictable order.
Speed goes first. A call that used to resolve in a few seconds now takes a few minutes of deliberate reasoning. The answer is the same. The route to it has changed from recognition to calculation, which is slower and far more effortful.
Then confidence. You arrive at the right answer and find yourself checking it. Not because there is a new reason to doubt, but because the click of certainty that used to accompany a good decision is no longer arriving with it.
Then reversibility tolerance. You start preferring decisions that can be undone. Optionality feels prudent, and sometimes it is, but a steady drift toward the reversible option is usually a sign of reduced confidence rather than improved strategy.
Only then, quality. By the time actual errors appear, the state has been present for a long time and has been silently compensated for.
Most people do not notice until stage four, because stages one through three are all absorbed by working harder. Which is exactly the input that accelerates the process.
Why High Performers Are Particularly Exposed
Decision fatigue is a function of decision volume, and decision volume tracks seniority almost perfectly. But there are two aggravating factors specific to capable people:
Consequential decisions cost more than trivial ones. Volume is only half the equation. Twenty low-stakes calls are cheaper than three that materially matter. Seniority raises both terms at once.
Nobody else is available to decide. The decisions that reach you are precisely the ones nobody below you could resolve. You cannot reduce the load by delegating the hard ones, because being handed the hard ones is the role.
Together those produce a load that does not respond to the usual efficiency advice. Better prioritisation helps at the margins. It does not change the fact that the decisions that remain are the expensive kind.
The Thing That Stops Arriving
There is a specific experience that people in this state describe, and it is more diagnostic than any measure of hours or output.
Good judgement, when conditions support it, tends to be preceded by a particular quiet. A settling. The problem is held, nothing is forced, and the answer surfaces. Most people who make consequential decisions for a living know this state well enough to rely on it, without ever having named it.
Under depleted conditions that quiet takes longer to arrive. Then it takes much longer. Then there are days it does not arrive at all, and the decision gets made anyway — from analysis rather than from clarity, which produces a defensible answer and none of the confidence that used to come with it.
This is the clarity pattern: judgement running on conditions that no longer support it. The calls are still sound, which is precisely why it goes unaddressed for so long.
Why The Standard Advice Underperforms
The usual recommendations — decide important things in the morning, reduce trivial choices, batch similar decisions — are not wrong. They are just aimed at the volume side of the equation, and volume is rarely the part you control.
Nobody in a senior role gets to meaningfully reduce how many consequential decisions reach them. What can change is the state those decisions are made from. That is a different intervention with a different target, and it is available even when the calendar is not.
It also matters that this be identified specifically rather than generally. Depleted clarity and the urgency pattern both produce a sense that thinking has got harder, and they call for close to opposite corrections. Someone whose judgement is running on empty needs conditions restored. Someone moving faster than the situation requires needs the pace decompressed. Apply the wrong one and you will conclude, wrongly, that nothing works.
The Signal Worth Watching
Do not wait for a bad decision. By then the state has been running for a long time.
Watch the cost instead. Are the calls that used to be easy still easy? Does certainty still arrive with the answer, or some time after it? Is the quiet before a hard decision still showing up on schedule?
Those questions have answers well before anything shows up in your results — which is the entire window in which this is cheap to correct.