Rule 10 of 33 · Chapter II — How Decisions Get Made
Make reversible decisions fast
Why this rule exists
We waste an astonishing amount of energy deliberating over choices that we could undo in an afternoon. Not every decision deserves the same rigor, and treating a one-way door and a revolving door with equal caution is its own kind of mistake, because it makes us slow where we could be fast. If a choice is cheap to reverse, the fastest way to learn which option is better is usually to just pick one and see, since the information you get from doing beats the information you get from more arguing. Reserve the careful, deliberate, write-a-long-doc process for the decisions that are genuinely hard to walk back: the database, the core data model, a hire, a public API. For everything else, speed is a feature, and the cost of being wrong is small and recoverable. A team that decides quickly on small things has far more energy left for the big ones.
In practice
Before deliberating, ask how expensive it would be to reverse this if we're wrong. If the answer is 'not very,' pick the most reasonable option quickly and move on, treating it as an experiment rather than a verdict. Set a low bar for reversible calls: a Slack thread and a decider, not a meeting and a doc. Save your real scrutiny for one-way doors, and be honest about which is which, because we tend to inflate small decisions into big ones to justify agonizing over them. When you pick fast and it turns out wrong, reverse it fast too, without shame, since that was the whole deal. Speed on reversible things is how you buy patience for the irreversible ones.
When it doesn't apply
Some decisions look reversible but quietly aren't, because they set a precedent, leak into a public interface, or become load-bearing before you notice. If a cheap-to-reverse choice will be hard to reverse once other things depend on it, treat it as one-way. And 'fast' never means skipping the people who'll have to live with it.