The proposal was small. Change a default setting, ship it to a tenth of the base, watch one number for two weeks. If it moved the wrong way, flip it back on a quiet afternoon and nobody outside the room would ever know it had happened. Total cost of being wrong: fourteen days and a shrug.
It took eleven weeks. Three steering reviews, a risk memo, a legal read, a stakeholder map with color-coded owners, and a final sign-off that got deferred once because a director was on leave. The same quarter, in the same building, a two-year platform commitment that could not be unwound without penalties and a migration went through in a single Thursday meeting, because, as someone said with a straight face, the team needed to move fast.
The organization had one deliberation budget and spent it on the wrong door.
Every organization has a finite amount of collective attention it can spend on any given decision. Meetings, escalations, memos, the hours senior people spend actually thinking rather than performing thought. That budget is real and it is small. The only question that matters is where it gets spent.
Here is what almost no governance process gets right: it spends the budget in inverse proportion to how much the decision deserves it. The reversible call, the one you could undo before lunch, attracts a committee. The irreversible one, the door that only swings in a single direction, gets waved through on instinct and calendar pressure. The cheap-to-be-wrong decision is deliberated like it is expensive. The expensive-to-be-wrong decision is rushed like it is cheap.
This is not stupidity. It is a rational response to the wrong signal. Deliberation in most companies is calibrated to visibility and blast radius of blame, not to the actual cost of being wrong. A default toggle is visible. If it goes sideways, a lot of people will be able to point at the person who shipped it. So it accretes reviewers, each one adding a layer of shared cover. The two-year lock-in is abstract, its damage lands eighteen months out, and by then the room that approved it will have reorganized twice. Nobody guards a door whose consequences arrive after they have moved on.
The time a decision takes in your organization correlates with how many people can see it, not with how hard it would be to unwind. That single misalignment is quietly more expensive than any bad call your leaders will ever make.
There is a tax nobody puts on a slide. Every reversible decision slowed to irreversible-decision speed is optionality left on the table, weeks of learning you could have bought for the price of a two-week test and chose not to. And the mirror image is worse. The irreversible decisions, the ones sprinted through because the room mistook urgency for importance, are exactly where the compounding damage lives. You are simultaneously over-insuring the decisions that carry their own insurance and under-insuring the ones that do not. Both errors come from the same blind spot.
Watch how the reversible decision actually accretes its weight. Someone reasonable asks to be looped in, because if it goes wrong they would rather have been consulted than surprised. That is not obstruction. It is self-protection, and it is contagious. Each new name on the thread makes the next name feel prudent, until a change you could have tested and unwound now carries a quorum of people who each own a sliver of the outcome and none of the decision. The deliberation was never really about the toggle. It was about assembling enough shared cover that no single person would be left holding a mistake that, by construction, was never going to be very costly in the first place.
You do not have a decision-making problem.
This is the part that lands badly in the room, so it is worth saying plainly. The failure is not that your leaders decide slowly, and it is not that they decide recklessly. They do both, on the same day, and they cannot see the contradiction because they are not sorting decisions by the one property that should govern everything downstream.
They are not asking the first question. Before a decision is deliberated, staffed, or escalated, exactly one thing needs to be established: if this is wrong, what does it cost to reverse? A decision you can walk back in a week is a different animal from one you can never take back, and it should be handled by a different process, at a different speed, by a different number of people. Your governance treats them as the same animal because it was built to distribute accountability, not to classify risk. Every layer you added to feel safer made the classification worse, because process optimizes for who is covered, never for which door you are standing in front of.
The tells are consistent once you look for them. Reversible decisions that need three approvals. Irreversible ones with a single owner and no dissent in the record. Steering committees that spend an hour on a pilot and ninety seconds on a commitment. A culture that celebrates moving fast on the things that should be slow, and calls it caution when it crawls through the things that should be fast.
The reason this survives every attempt to fix it is that both failures feel like virtues from the inside. Slowing down a reversible decision feels like rigor. Speeding through an irreversible one feels like decisiveness. Nobody in the room experiences either as a mistake, which is why no retrospective ever catches it. The pilot that took a quarter is filed under thorough. The commitment that took a Thursday is filed under bold. The organization congratulates itself on holding both values at once and never notices it has attached each one to precisely the wrong kind of decision.
What rebuilding actually looks like.
Imagine the discipline reversed. Every decision gets classified before it gets deliberated. Reversible, and cheap to reverse: push it down to the person closest to the work, decide today, and instrument the undo so the reversal is real and not theoretical. Irreversible, or expensive to reverse: this is where the deliberation budget belongs, so widen the room, slow the clock, invite the dissent, and spend the attention you just freed up from the toggles. The speed of a decision stops being a function of who is watching and becomes a function of what is actually at stake.
That reflex cannot be installed with a policy, because the misclassification happens in the moment, under pressure, when a real decision is on the table and the clock is running. It is a judgment pattern, and judgment patterns only surface where the pressure is real. This is precisely what the SSUNDAR simulation exposes. Put leaders inside cascading crises where reversible and irreversible calls arrive in the same minute, and you watch the default play out with uncomfortable clarity: they guard the two-way doors and sprint through the one-way ones, every time, until someone shows them the pattern in their own decisions. You cannot coach what you cannot see, and this is not visible on any competency framework or 360.
The organizations that decide well are not faster and they are not more careful. They have simply learned to tell the two doors apart, and to spend their scarce, expensive attention on the one that does not swing back.
A reversible decision made slowly is not caution. It is the most expensive thing an organization will ever mistake for it.