A working method for harder calls.

Start by writing the decision down.

The first step in any non-trivial decision is to write the decision down in a single sentence. 'Do I replace the roof now or attempt one more season of repairs?' 'Do I hire this candidate as a foreman or as a senior crew member?' 'Do I sign this contract on the carrier's terms or counter?' Forcing the decision into one sentence does two things at once: it surfaces what you are actually deciding, and it filters out the related-but-different decisions you are tempted to fold into the same conversation.

Most decisions feel paralyzed because they are actually three decisions tangled together. Writing the sentence forces a clean cut: which decision is on the table today, and which ones are downstream of this one and therefore can wait.

“Most decisions feel paralyzed because they are three decisions tangled together. Writing one sentence forces a clean cut.”

Is this a one-way door or a two-way door?

The single most useful question to ask about any decision is whether it is reversible. A reversible decision (a one-month trial of a new vendor, a soft hire into a 90-day evaluation period, a small initial purchase) can be made quickly, with imperfect information, and corrected cheaply. An irreversible decision (replacing the roof, signing a multi-year contract, terminating a long-tenured employee) deserves slower thought, more inputs, and a stronger threshold of confidence.

Most people invert this. They agonize over reversible decisions and rush through irreversible ones. The correction is to ask, before anything else: if this turns out wrong, what does it cost me to unwind it? A two-way door deserves boldness. A one-way door deserves patience.

Gather the small number of inputs that actually matter.

After the decision is framed and its reversibility assessed, gather inputs — but only the ones that would actually change the decision. Many decisions stall in the input-gathering stage because the operator is collecting information that, even if true, would not move the call.

For a roof replacement: the inputs that matter are the documented condition of the roof, the documented condition of the deck where visible, the age and warranty status of the existing system, the comparative quotes from two or three reputable contractors with verified credentials, and a clear-eyed read of the homeowner's planning horizon for the property. Almost everything else is noise.

For a hire: the inputs that matter are the candidate's recent on-the-job behavior (not their résumé), the testimony of two or three people who have actually worked with them in the role you are hiring for, and your direct observation of how they handle a small piece of real work. The interview, by itself, is among the least predictive inputs in the file.

What does this decision make easier or harder later?

A discipline that separates good decision-makers from mediocre ones is the habit of asking, before any meaningful call: what does this decision make easier later, and what does it make harder? Almost every decision sets up the next one. Almost every shortcut today is a constraint tomorrow.

Choosing a slightly cheaper underlayment on a roof installation, for example, is rarely a meaningful first-order decision. The second-order effect — a roof that fails three years earlier, a warranty that is harder to enforce, a homeowner who is harder to keep as a long-term customer — is where the actual cost lives. Good decisions are evaluated at the second order. Bad decisions are evaluated only at the first.

“Almost every shortcut today is a constraint tomorrow. Evaluate the second-order cost before the first-order savings.”

The right number of people to ask.

On any non-trivial decision, talk to the small number of people whose judgment in this specific domain you trust. Not the general advice-givers. Not the loudest opinion in your circle. The specific people who have actually faced the decision you are facing, more than once, and have lived with the consequences of their answer.

The right number is usually two or three. Fewer than that and you are over-weighting a single opinion. More than that and you are collecting noise and giving yourself permission to delay. After the second or third conversation, you usually have the answer. Additional opinions are a substitute for decision, not an input to it.

Decide cleanly, then move.

Once the decision is framed, the reversibility assessed, the inputs gathered, the second-order effects considered, and the small handful of trusted opinions taken — decide. Write down the decision and the reasoning behind it. Communicate it to the people it affects. Then move.

Most lingering damage in companies and households comes not from bad decisions but from undecided ones. Ambiguity is more expensive than wrong answers. A clear decision, even an imperfect one, lets the rest of the system organize around it. An undecided question keeps every adjacent system stalled.

After the decision is made, set a date to review it — usually thirty, ninety, or three hundred sixty-five days out, depending on the time horizon — and put it on the calendar. Reviewing your own past decisions, honestly, is the single most underused source of decision-making improvement available to anyone.

The method in six steps.

Write the decision in one sentence. Assess reversibility — one-way door or two-way? Gather only the inputs that would actually change the call. Evaluate the second-order effects, not just the first. Consult the small number of people whose judgment in this domain you trust. Decide, communicate, act — and set a date to review your own reasoning.

The method is not glamorous and is not original. Its value is in the consistency of its application. Decisions made this way, across a working lifetime, compound into a track record that the alternative methods cannot match.