Most organizations allocate deliberation by size. Big decisions get committees, papers, and three meetings. Small ones get handled by whoever is nearest.
The axis is wrong, and the error runs in both directions at once. Large reversible decisions are strangled by process they do not need, while small irreversible ones are made in a corridor by someone who did not know what they were closing.
The question that should govern
If this turns out to be wrong, how fast will we know, and what will it cost to undo?
A decision that reveals its error within a quarter and can be unwound for a modest sum needs a fraction of the deliberation it usually receives. You will learn more from six weeks of running it than from six weeks of discussing it, and the discussion has its own cost, which is that nothing is learned during it.
A decision that will not reveal its error for four years, or that cannot be undone at any price, deserves every hour you can give it regardless of how small the dollar figure looks on the page.
What is actually irreversible
Contracts and capital are the obvious cases and the ones people handle correctly. A fifteen-year lease on a building gets the attention it warrants. The irreversibility that gets missed is of a different kind.
People. Ending someone’s employment cannot be undone, and neither, in practice, can hiring the wrong person into a senior role in a small market where everyone will remember. The decision looks reversible because a person can be let go. What cannot be recovered is the eighteen months, the departures they caused, and the fact that the community watched.
Anything said publicly. A position stated to a customer, a council, or a reporter is fixed the moment it lands. The organization can change course afterward, at a price paid in credibility rather than money, and that price is never estimated in advance because it does not appear on any schedule.
Precedent. This is the one that does the most damage, because the decision itself is usually trivial. One exception granted to one person on one occasion establishes what the organization does in that situation. The next request cannot be refused without an explanation of why this person is different, and that explanation is nearly always worse than the original concession. A decision worth four hundred dollars can permanently reset a policy, and the person making it in a corridor has no idea they are doing anything of consequence.
Trust, spent in a specific direction. Backing a particular faction, overruling a particular person publicly, or committing your credibility to a forecast draws down a balance that refills slowly if at all.
The two misjudgments
Leaders get this wrong in a patterned way rather than randomly.
They treat reversible things as irreversible when they are anxious. A pilot programme, a trial of a new scheduling system, a temporary change to a shift pattern: these get studied for months as though the organization were signing something in blood. What is actually being avoided is not risk but the discomfort of visible change, and the study is a way of not doing it while appearing diligent.
They treat irreversible things as reversible when they are moving fast and the item is small. Nearly every precedent an organization comes to regret was set by somebody who believed they were handling a one-off.
What the sorting produces
Run against reversibility rather than size, a normal week’s decisions divide unevenly, and the division is useful.
The large majority are reversible, and the correct instruction for those is to decide today, at the lowest level with the relevant information, and set a date to look at the result. The gain here is not speed for its own sake. It is that you spend the intervening weeks holding evidence instead of opinions.
A small number are genuinely irreversible, and those warrant the full apparatus: independent estimates gathered before discussion, someone assigned to argue the other side, and a deliberate pause before commitment. Most organizations have enough process for perhaps five of these a year and spend it on thirty things that did not need it.
The habit
Ask the reversibility question out loud, at the start, before anyone has argued the merits. It takes about twenty seconds and it sets the correct weight for everything that follows.
The answer also changes who should decide. Reversible decisions belong as far down as the information goes, because the cost of a wrong one is a lesson. Irreversible ones belong with whoever will still be there to answer for them, which is frequently not the person currently holding the file.
Edited by Patrick J. Wolf, PhD