Reversible decisions
one-way doors · two-way doors · reversibility
The distinction between decisions that can be undone cheaply and those that cannot. Reversible ones should be made quickly by whoever is closest; irreversible ones deserve the slow, careful process most organizations apply to everything.
In practice
Most decisions that consume a week of deliberation could be reversed in an afternoon. Treating them as though they could not is how an organization becomes slow without becoming careful.
The common mistake
Assuming reversibility is a property of the decision. It is a property of the decision plus the situation — a hire is reversible in month one and much less so once they hold the only knowledge of a system.
Decisions differ in one respect that matters more than their size: whether you can take them back.
A decision you can reverse cheaply is a decision you can make on partial information, because the cost of being wrong is the cost of reversing it. A decision you cannot reverse has to carry the whole weight of the uncertainty, because there is no second attempt.
The failure in most organizations is not that they decide badly. It is that they apply one process to both — usually the slow one, because the slow process feels responsible — and end up deliberating at length over things they could have tried, while irreversible commitments go through on the same approval as everything else.
The two processes
Reversible: decide fast, decide low, expect to be wrong sometimes. The correct amount of analysis is small, because analysis is expensive and the experiment is cheap. Push the decision to whoever is closest to it. If it goes wrong, reverse it; that is what reversible means.
Irreversible: decide slowly, involve more people, look hard for the case against. These are the decisions worth a week. Long leases, senior hires, a public position, an architecture everything else will be built on, anything with a notice period or a contract.
The test is not importance. Plenty of important decisions are reversible and plenty of trivial ones are not.
Reversibility is situational
This is where the distinction is most often applied badly. Reversibility belongs to the decision in its circumstances rather than to the decision type, and it decays.
A hire is reversible in the first month and much less so in the sixth, once they hold knowledge nobody else has. A pricing change is reversible until customers have built expectations around it. A technology choice is reversible until three other things depend on it.
Which means the question is not "is this reversible" but "for how long, and what makes it stop being so?" — and often the useful move is to extend the window rather than to decide better: a trial period, a pilot, a contract with a break clause, a decision made in a way that leaves the alternative available.
The two failure modes
Treating reversible decisions as irreversible is the common one and it is expensive in a way that never appears in any account. The cost is the delay, the meetings, the opportunity not taken, and the signal sent to everyone that they cannot decide anything. It also degrades judgment: a team that never gets to be wrong cheaply never calibrates.
Treating irreversible decisions as reversible is rarer and worse. Signing a five-year lease with the reasoning that it can be sublet. Taking an investor on the assumption that a disagreement can be resolved later. Publishing something on the assumption it can be taken down.
The second is rarer because the commitments are visible. It is worse because there is no recovery, and it tends to happen in exactly the situations where speed feels like the virtue.
The related trap
Once a reversible decision has been made, reversing it runs into sunk cost — the effort already spent argues for continuing, and it should not. The whole value of a reversible decision is the reversal, and a business that makes fast reversible decisions and then never reverses any of them has taken the risk without buying the option.
Which makes this the practical partner of optionality: the point of keeping the door open is that you are willing to walk back through it.
Concept web
Open the full webQuestions
What is a reversible decision?
One that can be undone cheaply, so the cost of being wrong is the cost of reversing it. Reversible decisions should be made quickly and by whoever is closest; irreversible ones deserve a slower, more careful process.
How do you tell if a decision is reversible?
Not by its type or importance. Reversibility is a property of the decision in its circumstances and it decays — a hire is reversible in month one and much less so once they hold knowledge nobody else has.
What is the cost of treating every decision as irreversible?
Delay, meetings, opportunities not taken, and a team that learns it cannot decide anything. It also degrades judgment: people who never get to be wrong cheaply never become well calibrated.