How do you tell a reversible decision from an irreversible one?

What is the undo test?

Most people size a decision by how big it looks: money, headcount, visibility. That is the less useful half of the picture. The better question is concrete: if I do this and it’s wrong, what exactly does it cost to undo, in time, money, and trust?

Two generic examples show why size misleads. Changing how the morning stand-up runs feels significant because it touches the whole team. If the new format flops, you change it back next Monday at the cost of one awkward meeting. Big-feeling, easily reversed. Promising an angry client “a one-hour response on all tickets” in an email feels like a small concession. Once it is in writing it works like a contract term, and walking it back costs trust with that account. Small-feeling, hard to reverse.

When the feeling and the undo test disagree, go with the test.

Where does the one-way door idea come from?

From Jeff Bezos’s 2015 letter to Amazon shareholders. He split decisions into two types. Type 1: “Some decisions are consequential and irreversible or nearly irreversible – one-way doors – and these decisions must be made methodically, carefully, slowly, with great deliberation and consultation.” Type 2: “But most decisions aren’t like that – they are changeable, reversible – they’re two-way doors.”

His warning is the part most managers need: “As organizations get larger, there seems to be a tendency to use the heavy-weight Type 1 decision-making process on most decisions, including many Type 2 decisions.” The same thing happens to one person. Treat every call like a one-way door and you end up both exhausted and slow.

How do stakes and reversibility work together?

Reversibility alone doesn’t separate a big bet you can pull out of from a trivial one. Add stakes as a second question (if this goes wrong, is the damage minor or serious?) and you get four boxes, each with its own level of effort.

What if a decision is partly reversible?

When a call seems to belong in two boxes, it is usually two decisions stuck together. Split them. “Renew or switch our software platform” can hold a reversible piece (keep the current platform for now) and an irreversible one (commit to it for three years). Decide the first quickly; let the second wait for better information.

Hiring works the same way. “Do we need more capacity this quarter?” is often reversible through contractors, overtime, or a temp. “Do we add a permanent person?” is much harder to undo. Two questions help:

• Which part would I regret most if it turned out wrong, and can I defer just that part?
• Is there a smaller version I could do now that keeps the big version open?

Can you make an irreversible decision reversible?

Often, partly. A pilot, a shorter contract term, a smaller first version, or a trial period moves part of a decision from the hard column to the easy one. A six-week trial of a new on-call rotation tells you more than a month of spreadsheets guessing at the same thing.

But reversibility only counts if you use it. Bezos’s 2016 letter puts the condition plainly: “If you’re good at course correcting, being wrong may be less costly than you think, whereas being slow is going to be expensive for sure.” So write down the signal that would make you reverse and put the check date on your calendar. More on choosing under doubt in how to decide when you can’t be sure.

Why do managers over-analyze reversible decisions?

Because analysis used to be expensive and isn’t anymore. When running the numbers took an analyst’s week, the cost forced you to save deep work for big calls. Now a tool will produce a forty-row comparison for a call that deserved a coin flip, and nothing in it asks whether the decision deserves that. The restraint has to come from you. Extra analysis also feels more convincing than it is: in three studies, Tsai, Klayman, and Hastie (2008) found that “when judges receive more information, their confidence increases more than their accuracy.” Before a reversible call, check that all five are true, then decide within five minutes:

• If I’m wrong, I can walk it back without major cost.
• Every option is acceptable; the worst case is “slightly worse than ideal.”
• A deeper look wouldn’t flip my call.
• Waiting has a price, even a small one.
• More time would buy me reassurance, not a better decision.

How do you protect the few irreversible calls?

Their main threat is momentum. After a week of snapping small decisions, the big one arrives and you snap it too, especially with a “discount expires Friday” attached. A useful signal: it will still matter in a year and you can’t easily picture undoing it. When both are true, slow down even if you have been moving fast all week. One good use of that time is Gary Klein’s premortem: assume the decision has already failed and have everyone write down why, a method Klein built on 1989 research finding that imagining an event has already occurred increases the ability to correctly identify reasons for future outcomes by 30%.

Urgent requests often bundle both kinds. Answer the part that needs speed now, and the part that is hard to take back by a set date: “You’ll have [X] today and [Y] by Thursday.” Then record the big calls, with your expectation and confidence, in a decision log so you can grade them later.