The Decision Journal: Compound Interest for Judgement

Four months ago you made a hiring call you weren’t sure about. The experience was thinner than you’d have liked and two references came back lukewarm, but the candidate showed real initiative and brought something the team was missing. You weighed it up, sat with the doubt, and went ahead.

It worked out. And now, looking back, you can barely find the doubt you felt. The lukewarm references have faded out of the story. What you remember is the initiative, the gap they filled, and a version of you who saw what everyone else missed and trusted your gut. The uncertain call you actually made has been tidied into a clean story about foresight.

That tidying feels like remembering, but it’s closer to editing. And unless you wrote down what you actually thought at the time, you’ll never catch it happening. The one decision you could have learned the most from ends up teaching you nothing.

The evidence

Start with what your memory does to you the moment an outcome lands. This is hindsight bias, and it’s been pinned down again and again: once you know how something turned out, your memory of what you predicted slides toward what actually happened. It usually isn’t you flattering yourself on purpose. Your mind takes the new information, the outcome, and writes it back over the original memory of what you thought. The old version doesn’t survive the edit, and the rewritten one feels every bit as true. So when you try to learn from a past decision by thinking back, you aren’t consulting a record. You’re consulting a story that already knows the ending.

That’s why honest reflection can’t run on memory alone. There’s a long line of work on how experience turns into expertise, and the part that matters here is blunt: experience by itself doesn’t make you better at deciding. What does is experience plus honest feedback on how your calls actually turned out. This is deliberate practice, and the load-bearing word is honest. Without a true account of what you expected, every outcome gets absorbed into “I knew that,” and you collect years of practice that teach you next to nothing. What you build instead is confidence, which feels like competence and isn’t the same thing.

How it works

So the fix is almost embarrassingly simple: write the decision down before you know how it ends. At the moment you decide, you record the call, the options you weighed, what you were trying to get, and what you expect to happen. Now there’s a fixed point the editing can’t reach. When the outcome arrives, you compare it to what you actually wrote, not to the polished version your memory would have served up. Every gap between the two is a lesson hindsight would have erased.

The bigger payoff is in what you write about your reasoning. It isn’t enough to note “I hired her.” You note that you hired her because you judged the initiative to matter more than the thin experience, and you expected her to be up to speed inside three months. When she takes six, you have something to work with. You can go back and ask which belief was off: was the initiative worth less than you thought, or was the ramp-up always going to be slower? That’s the difference between adjusting your next move and upgrading the judgement that produced it, which is the shift from single-loop to double-loop learning.

This matters more the more successful you are, which is the part most people have backwards. A strong track record teaches you to chalk up the wins to your skill and the losses to bad luck or other people. It’s a comfortable habit, and it’s why capable, experienced professionals are often the slowest to sharpen their judgement: nothing in the working day forces them to check the story they tell about their own calls. The journal forces it. And it only works because of how it’s built, written down rather than carried in your head, read back after enough time has passed that you’ve lost the thread, and kept in a consistent shape so you’re analysing the decision instead of reliving it.

Without a written record, you don’t learn from the decision you actually made; you learn from the one your memory rewrote once it knew the ending.

How to use it

So here’s the practice, and it costs about five minutes. When you make a decision worth learning from, open a note and write the date, the call, the options you looked at, what you chose, why, and what you expect to happen. Set a reminder to come back in thirty days. That’s the whole investment, and it pays out a month later when you put your prediction next to what actually happened and see where your reasoning held and where it bent.

Once a week, spend a few minutes re-reading recent entries. One entry on its own tells you about one decision. A dozen of them together start showing you yourself: the kind of risk you reliably underrate, the situations where you run too hot or too cautious, the sort of call where your gut is gold and the sort where it leads you astray. None of that shows up in the moment. It only appears across the pile.

And pay closest attention to the entries where you got it wrong. The pull is to enjoy the ones where you nailed it, but those just feel good. The ones where your reasoning seemed solid and the outcome went the other way are where the real information sits, because that’s where an assumption you trusted got exposed. Treat a wrong prediction as data rather than a black mark, and the journal turns from a diary into the thing that actually trains your judgement.

Why it matters

Step back and look at how most experienced people learn. They’ve been deciding for years, and they trust that the experience has sharpened them. Some of it has. But a lot of what they “know” was assembled by the same memory that edits the hard calls into clean wins, so the lessons they carry can sit a fair distance from what the raw events would have taught. They’ve got plenty of experience; what they haven’t got is an honest record of it.

The journal is the cheapest fix there is. No special tools, no training, no real time: five minutes when you decide, five minutes when you review. What you get back is a true and growing account of how your judgement actually performs, which is the one input that turns ordinary experience into the kind that compounds. Almost everything else you do improves as a by-product of a hundred other things. Your judgement only improves if you feed it the truth about its own track record, and the journal is how you keep telling it.