Decision fatigue: why your best calls never happen after six in the evening
Decision fatigue sabotages your evening calls. What the brain actually does, and how to protect your bandwidth.
It is a quarter past seven in the evening. You are alone in your office with the door closed. One last email is waiting for an answer: a quote to approve, a hire to settle, a discount to grant. You read the same line three times. And you click “yes”, simply to be done. That is what decision fatigue looks like.
The next morning you reopen that email. And you cannot understand how you said yes.
I see this scene with almost every leader I work with. It has nothing to do with intelligence or experience. And it may be the worst-managed hidden cost in your company.
The hidden cost nobody measures
Let us start with an image that stayed with me. In 2011, three researchers published a now-famous study in PNAS. They analysed more than 1,000 parole decisions handed down by eight experienced judges in Israel.
The result is dizzying. Just after a meal break, judges granted parole in nearly 65 % of cases. Then, file after file, the rate collapsed, sometimes to zero. One break, and it climbed back to 65 %.
The facts of the files had not changed. The judge had. His brain was simply exhausted from deciding. A caveat: the study is debated, and the order in which files were heard probably plays a part. But it illustrates a truth neuroscience confirms elsewhere.
What happens in the prefrontal cortex
Here is the mechanism. Every decision consumes resources in your prefrontal cortex, the area that arbitrates, compares and projects. That cortex is not an infinite reserve. As it tires, it hands over to faster and more primitive circuits: the amygdala, the reflex, avoidance.
So by the end of the day you are no longer really deciding. You choose the option that costs the least effort: postpone, say yes, follow the habit. The worst of it is that you do not feel it coming.
The multitasking myth: your brain is only juggling
I often hear the same pride: “I handle ten things at once.” I say this with affection: no. Nobody does. Our brain does not multitask.
What you call multitasking is in fact a succession of very fast switches from one task to another. And every switch has a price, paid in accuracy more than in speed.
The clearest experiment on this dates from 1999. Cheri Speier, Joseph Valacich and Iris Vessey, in Decision Sciences, distinguish two kinds of task. On a simple task, interruption saves time at no cost. On a complex task, decision accuracy falls from 0.13 to 0.08, and the effect follows the dose: at four interruptions it still holds at 0.22, at twelve it collapses to 0.05.
Translated for a leader: a fragmented day does not degrade the handling of routine subjects. It degrades exactly the category of decisions you are paid for.
Attention residue
There is worse, and it is more insidious. The researcher Sophie Leroy described in 2009 a phenomenon she called “attention residue”. When you leave an unfinished task to start another, part of your attention stays stuck to the first. You are in the meeting, but your brain is still circling the email from earlier.
Add the cost of interruptions. You may have read that it takes “23 minutes and 15 seconds” to refocus. That figure circulates everywhere and has no published primary source: it appears in no scientific paper, it comes from interviews. Here is the one that was actually measured.
In 2008, Gloria Mark, Daniela Gudith and Ulrich Klocke published an experiment with 48 participants at the CHI conference. Interrupted work is finished faster than uninterrupted work, roughly 20.4 minutes against 22.8, at equal quality: participants compensate by speeding up. But measured stress rises from 6.92 to 9.13, frustration from 4.73 to 6.48, effort from 9.50 to 11.04.
So you absorb the acceleration. You pay for it in load, not in delay. And since delay is the only thing your dashboard measures, you never see it coming. Your mental bandwidth does not wear out because you work too much. It wears out because it is fragmented.
Mental load: when the brain saturates in silence
There is a third factor, and this one is invisible: mental load. It is the sum of everything your brain keeps open in the background. The supplier dispute. Month-end cash. The employee who is wavering. The meeting you forgot to prepare.
Why working memory saturates so fast
Your working memory, that mental workbench where you handle problems, is tiny. It holds only a few items at a time. When you saturate it with open files, there is no room left for the essential thing: thinking clearly.
On the surface, you feel hyperactive, indispensable, everywhere at once. Underneath, you decide less and less well, because every micro-decision eats into the reserve you should be keeping for the real ones. That is the heart of decision fatigue.
This is where I often ask a leader a simple question: how many decisions did you take today that should never have reached you?
AI does not save you time, it gives you attention back
You may be expecting me on the ground of artificial intelligence. Fairly: it is half my work. But I am going to surprise you.
AI will not save you time. Not the way it is promised, at least. A study by the METR organisation, published in 2025, followed experienced developers equipped with AI. They believed they were roughly 20 % faster. In reality they were 19 % slower. The gap between felt speed and real speed is startling.
The lesson is not “avoid AI”. The lesson is finer. Well deployed, AI does not give you time back: it gives you attention back. And for a leader, attention is worth far more than time.
What properly framed AI actually absorbs
In practice, well-framed applied AI absorbs the micro-decisions that saturate your day. Sorting requests. The first version of a report. Routine client follow-up. Filing information before it reaches you. It is a direct weapon against decision fatigue.
In other words, it protects your prefrontal cortex from the calls that do not deserve your brain, so that you keep reserve where nobody can replace you: vision, judgement, relationships. That is exactly the pairing I argue for: neuroscience to understand your brain, AI to unload the trivial from it.
But beware the trap: badly framed AI adds noise instead of removing it. It generates, you have to review; it proposes, you have to decide. Without method, you move the load, you do not reduce it.
A working manual: protecting your decision bandwidth
So what can be done, concretely? Here is the manual I give the leaders I work with.
Four rules that hold under pressure
First, put your heavy decisions in the morning. Your cortex is fresh, your reserve is full. A strategic decision taken at nine and the same one taken at seven in the evening do not come out of the same brain.
Second, sequence instead of juggling. Block slots for a single task, notifications off. You do not lose responsiveness: you gain depth.
Third, let as few decisions as possible climb to you. Every call you delegate or automate is bandwidth returned to what matters. Ask yourself, for each one: is this really mine?
Fourth, make recovery a ritual. Sleep, movement and real breaks are not comfort. They are the only moments when your brain recharges its capacity to decide. You do not run a company on a dry cortex.
Deciding is a muscle, not a character trait
The quality of your decisions is not a question of character. It is a question of the state of your brain at the moment you decide. Composure, clarity, the accuracy of a call under pressure: all of it can be trained, protected and managed, exactly like the mental preparation I borrow from elite sport.
The good news is that decision fatigue is not a fate. It can be mapped, like a risk. It can be reduced, with method on one side and the right tools on the other. And the day you take back control of your attention, you do not only decide better: you lead differently.
If you recognised yourself in the quarter-past-seven scene, that is already a signal. A leader’s first performance lever is not working more. It is protecting the brain that decides.
Want to put figures on your own decision load and see where your bandwidth goes? Let us spend thirty minutes on it, with no commitment.
Frequently asked questions
What is decision fatigue in a company leader?
It is the gradual degradation in the quality of your judgement calls as the day goes on, independently of your intelligence and your experience. It does not feel like tiredness: it feels like wanting to be done with it.
When in the day should important decisions be taken?
In the morning, with a fresh cortex. A strategic decision taken at nine and the same one taken at seven in the evening do not come out of the same brain.
How do I reduce the number of decisions that climb up to me?
Ask yourself, for each one: is this really mine? Most leaders discover that one decision in two could be delegated or automated. Every call handed over is bandwidth recovered where nobody can replace you.
Does artificial intelligence actually help?
A study by the METR organisation published in 2025 followed experienced developers equipped with AI: they believed they were 20 % faster, they were 19 % slower. Well framed, AI absorbs the micro-decisions that saturate the day; badly framed, it moves the load rather than reducing it.