How to Make Better Business Decisions

Judgment improves through process, not through gathering more information. Here is what actually changes outcomes.

Most owners try to improve decisions by gathering more information. Past a fairly low threshold this stops helping and starts hurting, because additional data mostly gets recruited to support the position you already hold. What actually improves outcomes is structural.

1. Separate reversible decisions from irreversible ones

Most decisions can be undone at modest cost, and should be made fast. A small number cannot, and deserve real deliberation. Treating both alike is the most common process failure: owners agonise over reversible choices and rush the permanent ones because those arrive with urgency.

Ask first: if this is wrong, what does it cost to reverse? The answer sets how much process the decision earns.

2. Write the decision down before you make it

Record what you are deciding, what you expect to happen, and what would tell you that you were wrong. This takes ten minutes and does two things. It exposes reasoning that felt solid until it had to be written in sentences. And it creates a record that defeats hindsight bias later, since memory reliably rewrites what you expected to match what occurred.

3. Name what would change your mind, in advance

Before committing, state the evidence that would prove the decision wrong. If you cannot name any, you are not making a decision, you are defending a conclusion. Doing this beforehand is what lets you actually change course later without it feeling like a personal defeat.

4. Get disagreement from someone with no stake in your comfort

This is the practice most owners skip, and it is the one that matters most. Your team has incentives. Your friends want to be supportive. Your customers see one facet. What you need is someone competent who has no reason to protect your preferred answer.

The value is not that they know more. It is that they have no reason to agree.

5. Set a decision date and hold it

Not deciding is a decision, made by default and usually badly. Indefinite deliberation feels like diligence while the option quietly expires. Set the date when you frame the question, then decide on it with what you have.

6. Review outcomes against what you wrote

Once a quarter, reread your decision records and compare them to what actually happened. Look for patterns rather than individual errors. Most owners have two or three recurring biases, and seeing yours named in your own handwriting changes behaviour in a way that general advice does not.

The trap underneath all of it

Almost every owner who describes a decision as difficult has already chosen. What they are looking for is not analysis, it is permission, or the reassurance of someone confirming it. That is a human and understandable thing to want. It is also why so much advice seeking produces no change in behaviour.

A useful test: notice how you feel when someone argues against your leaning. If the reaction is relief at finding a flaw, you were genuinely deciding. If it is irritation and a search for counterarguments, you had already decided and are shopping for agreement.