Write this down before you decide

Hi Everyone,

Think about a decision you're weighing right now – a hire, a product bet, a pricing change. How confident are you that it will work? Now, give it a number. 70%? 85%? 95%?

Whatever number you picked, it's probably too high.

A decade-long study from Duke and Ohio State found that when CFOs made predictions they felt 80% sure about, they were only right 36% of the time. And the most confident ones? They spent more and borrowed more.

Here's a brief five-question checklist you can use before any major decision. It's quick, and it gets more useful every time you do it.

Answer these 5 questions before you commit

Before your next big call, take two minutes and write your answers down. Writing matters here because putting your reasoning on paper forces specificity in ways that thinking alone doesn't.

1. What are my real alternatives?

Write down at least two other options you considered, including "do nothing" or "wait three months." If you can't name alternatives, you haven't finished thinking.

A lot of decisions feel binary when they're actually not. Naming your alternatives often reveals an alternative path or a sequenced approach that lowers the stakes considerably.

2. How confident am I, as a percentage?

Put a number on it. Not "pretty sure" or "I feel good about this" but an actual percentage.

This changes how you think. A vague sense of confidence gives you nothing to check later, while a number creates a claim you can actually test. And research from Philip Tetlock at Wharton shows that people who practice thinking in probabilities become measurably better at predicting outcomes over time.

3. What would have to be true for this to fail?

List two or three specific, plausible conditions.

Be specific and write down actual conditions – a timeline that slips, a cost that doubles, a key assumption that turns out to be wrong.

Vague failure ("it doesn't work out") gives you nothing to watch for.

Concrete failure ("we can't hire fast enough to hit the Q3 deadline") tells you exactly what to monitor.

If you can't describe what failure looks like, you haven't stress-tested the decision.

4. What information am I missing?

Write down what you wish you knew but don't. Then ask yourself whether you're comfortable making the decision without it, or whether you could get that information within a week.

Jeff Bezos has talked about making decisions with around 70% of the information he'd ideally want, because waiting for 90% usually means deciding too late.

The useful question isn't whether you have all the information, but whether you know which pieces you're missing and whether any of them could change your answer.

5. If this fails, will I feel surprised or relieved?

This question catches decisions you're making for the wrong reasons.

If a small part of you would feel relieved by failure, you're not fully committed, and that ambivalence is worth understanding before you move forward.

You'll misremember your own reasoning

Annie Duke, a former professional poker player who now advises on decision strategy, points to a common problem she calls "resulting" – judging the quality of a decision by its outcome rather than by the reasoning behind it.

A good decision can lead to a bad outcome because of factors you couldn't control, and a bad decision can work out because you got lucky. If you only look at results, you'll learn the wrong lessons from both.

Writing down your reasoning, your confidence level, and your assumptions before you know the outcome creates a record you can actually revisit. Three months later, you can check whether your confidence matched reality (not whether things worked out, but whether you were well-calibrated going in)

Over time, that builds something you can't get any other way: an honest picture of where your judgment is sharp and where it tends to drift.

Try this today

Pick a decision you're working through right now. Open your journal or a note on your phone, and write your answers to the questions above. Set a calendar reminder to revisit it in 90 days.

If you want to build this into a regular habit, keep a running log with one entry per decision. After six months, you'll have plenty of data to see your own patterns and learn from them.

Go deeper

👉 Farnam Street: Decision Journal: How and Why to Start One – a practical template for tracking your reasoning before you know the outcome

👉 Harvard Business Review: Noise: How to Overcome the Hidden Cost of Inconsistent Decision Making – Kahneman, Sibony, and Sunstein on why even expert judgments vary wildly

👉 McKinsey: Bias Busters: Premortems — being smart at the start – a team-level approach that complements this personal check

👉 Good Judgment Project: goodjudgment.com – Tetlock's platform for improving your forecasting accuracy through practice and feedback

Coming up tomorrow

Tomorrow we're covering five AI risk questions your leadership team should be ready to answer.

Until next time!

P.S. Have you ever talked yourself into a decision you knew was shaky? What made you go ahead anyway?