Reversible and Irreversible Decisions

Most people spend their decision effort backwards. They agonise over choices that could be reversed in an afternoon, then rush the few that will bind them for years. The result is a life that feels both exhausting and oddly fragile: drained by small deliberations, exposed on the big ones.

The fix is a single question asked before any other: can this be undone, and at what cost? Reversibility is not a detail of a decision — it is the property that should set how much care and how much time the decision deserves. Get that match right and you move faster where speed is cheap and slower where caution actually pays.

Two-way doors and one-way doors

Some decisions are doors you can walk back through. Try a new tool, change a meeting time, run a small experiment — if it disappoints, you reverse it and little is lost. These are two-way doors. Other decisions are one-way doors: selling a house, shutting a business line, publishing something public, ending a relationship. Walking back through costs a great deal, or is impossible. The label is not about size or importance; it is about the cost of undoing.

Reversibility sets the speed limit

Once you know which door you face, the right tempo follows. Reversible decisions should be made quickly, by whoever is closest to the work, and treated as data-gathering: the cheapest way to learn is often to try and see. Irreversible decisions deserve the opposite — more information, more perspectives, a slower hand, and a genuine pause before commitment. The error is applying one tempo to both.

The two symmetric costs

Mismatch cuts both ways. Deciding too slowly on a two-way door has a real price — the cost of delay: opportunities that expire, learning postponed, teams stalled waiting for a verdict that barely matters. Deciding too fast on a one-way door risks a commitment you cannot escape. Good judgement is not "be careful" or "be fast" but knowing which the moment calls for.

A team debates two decisions in the same meeting. First: which project-management app to adopt. They spend forty minutes. But this is a two-way door — migrating away later costs a weekend. The right move is to pick a reasonable option in five minutes, try it for a month, and switch if it disappoints. Second: whether to sign a three-year office lease. This is a one-way door; breaking it costs six figures. Here the forty minutes — and more — is well spent. Same meeting, opposite tempos, because the doors differ.

Reversibility is not a licence to be reckless with anything that can technically be undone. Some decisions look like two-way doors but are not. A hasty public post can be deleted, yet the screenshot is forever. Cutting a corner on safety is "reversible" only until the accident. When undoing depends on nobody having noticed or nothing having gone wrong in the meantime, treat the door as one-way even though it appears to swing both ways.

The clearest articulation of this idea comes from Jeff Bezos. In his 2015 letter to Amazon shareholders, Bezos distinguished two types of decisions. Type 1 decisions are consequential and nearly irreversible — one-way doors — and must be made slowly and deliberately. Type 2 decisions are changeable and reversible — two-way doors — and "can and should be made quickly by high judgment individuals or small groups."

Bezos's warning was that large organisations tend to apply the heavyweight, slow Type 1 process to Type 2 decisions as they grow. The result, he wrote, is "slowness, unthoughtful risk aversion, failure to experiment sufficiently, and consequently diminished invention." His prescription was to sort decisions by their door before choosing a process for them — a framing he returned to in the 2016 letter's discussion of high-velocity decision-making. The wording here follows those published letters.

• "Big decisions are the irreversible ones." Reversibility is about the cost of undoing, not the stakes; some large decisions are easily reversed.
• "Careful is always safer." Slowness on a two-way door has a real cost in delay and lost learning.
• "If it can technically be undone, it's reversible." If reversing depends on nobody noticing or nothing going wrong, treat it as one-way.
• "Fast decisions are low quality." On reversible calls, speed plus a willingness to correct beats slow perfection.

You will be shown a set of real decisions and asked to classify each as a one-way or two-way door, then set the speed and level of care it deserves. The interesting cases are the ones that look reversible but hide a permanent cost.

Think Like a Maester: Before deciding, ask not "what is the right answer" but "can I walk this back" — the door you are facing sets how fast and how carefully you should move.

1. What single question distinguishes a one-way door from a two-way door?
2. Why should reversible decisions generally be made quickly and by those closest to them?
3. Name the two symmetric costs of mismatching tempo to reversibility.
4. Give an example of a decision that looks reversible but should be treated as one-way.
5. According to Bezos's 2015 letter, what happens when organisations apply a Type 1 process to Type 2 decisions?

Before weighing a choice, ask whether it can be undone and at what cost. Two-way doors — reversible decisions — should be made fast, close to the work, and treated as experiments. One-way doors — irreversible ones — deserve more information, more caution, and a slower hand. The costs are symmetric: dithering on the small and rushing the large both do damage. Jeff Bezos's Amazon letters give this its sharpest form, warning that treating reversible decisions as if they were permanent breeds slowness and timidity. Match the tempo to the door.

• Problem-Solving Under Constraints
• Ethical Decision-Making