Decisions.

The framework

1. 01

   Test the door before you choose the process — cost, time, and scar.

   Before deciding anything, decide what kind of decision it is, and use three questions rather than instinct. What would it cost to walk back? How long would the walk take? And what residue would remain even after you had reversed it — trust spent, data handed over, a supplier who now knows your numbers? Expensive and irreversible are not the same test, and most people only ever run the first. A six-figure software purchase you can abandon within a quarter is a two-way door with a price tag; a quiet promise to your largest customer costs nothing today and cannot be taken back. Size is not the test. The walk back is.

2. 02

   Two-way doors: one owner, a fast answer, and a date to look again.

   Reversible decisions should be delegated hard and answered quickly, because on a two-way door the cost of being wrong is smaller than the cost of the meeting you held to avoid it. Name one owner rather than a group, give them the call outright, and set a date when the decision gets reviewed against what you have learned by then. The failure mode is consensus theatre: three senior people, two weeks, and a choice that would have taken an afternoon to undo. A mistake has someone's name on it; the fortnight spent avoiding it has nobody's. Put the review date in writing and the speed stops feeling reckless.

3. 03

   One-way doors: slow down, and buy information rather than simply wait.

   On an irreversible decision, delay has genuine value, but only when the waiting is doing work. The point Dixit and Pindyck formalised is that the option to wait pays when you can learn something in the interval that would change the answer. So make the waiting explicit: name what you do not know, name what would change your mind, and buy that knowledge deliberately with a pilot, a trial contract, a week of customer conversations, a call with someone who has already walked through this door. Waiting without a purchase is not prudence; it is procrastination, and it carries a risk of its own — the door can close from the other side while you deliberate.

4. 04

   Fit a hinge — and stop welding shut the doors that still swing.

   The category is partly a design choice, and the cheapest move available is converting a one-way door into a two-way one before you walk through it. Take the irreversible version and ask what a smaller, shorter, switch-off-able version of the same commitment would look like — Day 80 puts a price on that flexibility, so today it is enough to notice the hinge is available at all. The inverse deserves equal attention, because you weld doors shut without noticing: announcing a decision publicly, restructuring around it, or telling a team it is permanent turns a reversible call irreversible for no benefit at all.