Reversible and irreversible choices deserve different amounts of thought

One variable that reorganises everything

Decisions are usually sorted by how important they feel, which is a poor sorting key because importance is vague and mostly reflects how much attention the decision is currently demanding. A better key is reversibility: if this turns out badly, what does it cost to undo?

Sort by that and the appropriate amount of deliberation falls out almost automatically. Where reversal is cheap, extended analysis is usually a waste, because you can learn more by trying than by thinking and the cost of a wrong turn is one wasted attempt. Where reversal is expensive or impossible, the analysis is the only protection available, and speed is not a virtue.

What makes a decision hard to reverse

Several things, and they are worth naming separately. Financial irreversibility, where the money is gone or the asset cannot be sold for what it cost. Contractual irreversibility, where exit is possible but priced. Physical irreversibility, where the thing has been demolished, cut, mixed or published. And what might be called social irreversibility, where the information is now known to other people and cannot be unknown.

The distinction is not always where people assume, either. A large purchase with a resale market may be far more reversible than a small commitment that quietly reorganises how a household or a team operates. Amount of money is a poor proxy for reversibility, and using it as one is how organisations end up approving genuinely irreversible small decisions through a process designed for trivia, while subjecting recoverable large ones to months of review.

There is also a quieter form: path dependence. Some choices remain nominally reversible while steadily raising the cost of reversal, because systems get built around them and people organise their expectations accordingly. A technology choice at the start of a project is like this. Nothing prevents changing it on day one, and by month nine the same change is a different decision entirely.

Speed is the right answer more often than people think

For genuinely reversible decisions, the dominant cost is usually delay rather than error. Time spent deliberating has an opportunity cost, the information available before acting is often much worse than the information available after acting once, and a wrong choice you can undo generates exactly the evidence you were trying to obtain by reasoning.

This is the argument for deciding quickly on things that can be unwound, and it holds up well in practice. It is also frequently misapplied, because people take the general endorsement of speed and use it on decisions that are not reversible at all. The recommendation is entirely conditional on the classification, and the classification is the part that requires thought.

Buying reversibility on purpose

Reversibility is not fixed. It is often possible to convert an irreversible decision into a reversible one, and doing so has a price that is worth paying explicitly rather than accidentally. Renting rather than buying, piloting in one location before rolling out, running a fixed-term arrangement before a permanent one, keeping the old system running alongside the new one for a period.

Each of these costs something real: rent exceeds ownership costs over time, a pilot delays the benefit, parallel running is expensive and tedious. The right way to think about the extra cost is as the price of an option, purchased because the outcome is uncertain. When you are confident, the option is not worth much and buying it is waste. When you are genuinely unsure, it is often the cheapest part of the whole exercise.

The two failure modes

The first is treating everything as reversible and moving fast through decisions that cannot be unwound. This tends to happen in environments that have been rewarded for speed and have not yet encountered a decision where speed was the wrong instinct. The damage arrives late and is attributed to something else.

The second is treating everything as irreversible, which produces long deliberation over choices that could simply have been tried. This is the more common failure in careful people and in large organisations, and its cost is invisible in exactly the way opportunity costs always are. Nobody is ever criticised for the six weeks spent evaluating something that could have been tested in an afternoon.

The remedy for both is to ask the classification question first, out loud, before any discussion of the merits. It takes a minute and it determines what kind of conversation the rest of the meeting should be.