We saved the most important decision of the quarter for the last fifteen minutes of a three-hour meeting, then spent the next two quarters wondering why it was the one that went wrong.
You know the meeting. It opens sharp. There is coffee, there is a sense of occasion, and the first items on the agenda get argued the way decisions are supposed to be argued. Someone pushes back. Someone revises. A number gets challenged and sent back for rework. The discussion has teeth. Then the clock begins its quiet work. The agenda is long, because the calendars of the people in that room are the scarcest resource the company owns, so everything that needs a senior decision is stacked into the same afternoon. By the sixth item the pushback has thinned. By the eighth, people are agreeing to things mostly to keep the room moving. And the decision that actually mattered, the capability bet, the restructure, the commitment that will shape the next three years, sits at the bottom of the page in the slot marked "if time permits," where it receives the depleted, get-me-to-my-next-call judgment of a room that spent its best thinking ratifying a travel policy.
The agenda cast a vote.
We treat the order of an agenda as an administrative detail, a matter of flow and courtesy, the kind of thing an assistant arranges so the meeting runs smoothly. It is nothing of the sort. The order of an agenda is a rationing plan for the single most expensive and least renewable thing in the building, which is not money and not time but the capacity of a senior group to think hard about something. That capacity is finite, it depletes across an afternoon, and it does not come back with a biscuit. Whatever you put first gets the full force of a room that is fresh, sceptical and willing to fight. Whatever you put last gets whatever is left, which on most afternoons is close to nothing. The sequence is not neutral. It is a vote, cast in silence, before a word of the actual discussion is spoken.
Watch what a depleted room actually does, because it does not announce its exhaustion. It does something far more dangerous. It keeps performing competence while quietly lowering the bar. It stops generating alternatives and starts reaching for the nearest acceptable option. It defers to whoever sounds most certain, because evaluating certainty is tiring and accepting it is free. It defaults to the status quo, because change requires imagining consequences and imagination is the first faculty to go. And above all it approves, because approval ends the meeting and rejection reopens it. A tired room carries a structural bias toward yes that has nothing to do with the merits of what it is saying yes to. The proposal that would have been taken apart at ten in the morning sails through at a quarter to five, not because it got better, but because the room got weaker.
The same proposal, the same people, the same room. Only the hour changed. And the hour decided.
The organization, naturally, tells itself the opposite story. It believes it gave the big decision pride of place. We put it last because it was the most important, so we would have the context of everything else in front of us. We cleared the deck for it. This is the flattering inversion that lets the practice survive: the most consequential item is placed exactly where the judgment is thinnest, and everyone agrees to call that placement respect. The company has convinced itself that the slot at the end of a long meeting is the slot of honour, when it is in fact the slot of least resistance, the place where things go to be approved rather than decided. A board that spends ninety minutes on an audit sub-item and eleven on a strategy that commits the firm for a decade is not prioritising. It is running out of road and calling the destination a choice.
And so, with real reliability, the decisions that deserve the most judgment receive the least. The pattern is not random. It is close to a law. The bigger and more ambiguous the decision, the more time it is assumed to need, so the later in the meeting it gets scheduled, so the more depleted the room is by the time it finally arrives. The importance of the item and the quality of attention it receives run in opposite directions, by design, inside the most disciplined companies in the world. Nobody chose this. It emerged from a hundred reasonable calendar decisions, each sensible on its own, that together guarantee the firm brings its worst thinking to its most important choices.
There is a second failure folded inside the first, and it is quieter. It is not only that the last decision is made badly. It is that the room cannot tell. Decision fatigue is invisible from the inside. A depleted mind does not feel depleted, it feels efficient. The room at a quarter to five experiences its own lowered standards as decisiveness. We are aligned. We moved fast. We did not overthink it. Everything that is actually the fingerprint of exhaustion gets relabelled as a virtue, written into the minutes as consensus, and carried out of the room as a decision the group believes it reasoned its way to. The fatigue does not merely damage the call. It erases the evidence that the call was ever in doubt.
Notice, too, which decisions get reopened. The small item argued hard at the start, the one that got the room's real attention, is the one somebody circles back to next week with a second thought. The enormous item waved through at the end is the one nobody ever revisits, because reopening it would mean conceding that the company committed three years of itself in the fifteen minutes it had the least capacity to judge, and no room volunteers for that admission. So the trivial decision stays under review and the consequential one hardens, untouched, precisely because it was made worst. The organization ends up protecting its biggest mistakes behind the embarrassment of having made them tired.
Here is the part that should unsettle anyone who runs a leadership team. This is not a discipline problem, and it will not yield to the usual remedies. You can send the room to decision-making training, hand out the bias cards, appoint a devil's advocate, and none of it will survive contact with the ninth item on a Thursday afternoon. The sharpest, best-intentioned executive you have makes a measurably different decision at nine than at five, and she cannot feel the difference while it is happening. The quality of your most important decisions is therefore not primarily a function of how good your leaders are. It is a function of when, in the arc of their depletion, you ask them to decide. You have been treating as a test of character something that is mostly a test of scheduling.
What rebuilding looks like.
So what would you actually change. Not the training. The sequence. You would stop hiding the most important decision at the bottom of the agenda and start giving it the one thing it has never had, which is a fresh room. You would decide the biggest thing first, while scepticism is still cheap and the appetite to fight is still intact, and let the travel policy have the tired fifteen minutes it can safely survive. You would cap how many real decisions a single meeting is allowed to carry, on the understanding that judgment is a budget and not a tap. And you would learn to read the tells of a depleted room the way a clinician reads vitals, because the moment you can see the fatigue is the moment you stop mistaking it for agreement.
That last skill is the one almost no organization can build from the inside, because its own instruments are all calibrated to the fresh room and the confident deck. The appraisal, the strategy review, the planning offsite, each of them catches leaders at their rested best and asks what they think. None of them catches leaders at hour three, under load, when the decisions are stacking up and the quality is quietly falling, and shows them what they actually do. That is the precise condition a crisis simulation is built to create. It applies pressure and fatigue on purpose, runs decisions on top of decisions until the room is operating exactly as it operates at the bottom of a real agenda, and makes the degradation visible while it is happening. SSUNDAR built that instrument because the gap it exposes is invisible everywhere else: the distance between how a leader decides when the stakes are abstract and the mind is fresh, and how the same leader decides when the stakes are real and the judgment is nearly spent. On paper, in the rested room, every leader looks deliberate. Under load, at the ninth item, you find out which of them actually is.
Look at the last three decisions your leadership team got wrong, and check where on the agenda they sat.