What to write in a decision journal entry (and what makes it worth rereading)

The format of a decision journal entry is short enough to learn in thirty seconds: what you are deciding, what you expect, why, what would change your mind, how you are feeling. Knowing the fields is not what makes an entry useful. Two people can fill in the same five lines and end up with completely different amounts of value six months later, because one of them wrote something checkable and the other wrote something that only sounds like it. This is about the difference — what actually belongs in each line, and the small rewrites that turn a forgettable entry into one worth opening again.

The test for every line you write

Before you write a sentence, ask whether it could turn out to be wrong. A sentence that cannot be wrong cannot teach you anything when you reread it, because there is nothing to compare it against. "I think this will probably be fine" cannot be wrong — almost anything technically counts as fine. "I think we lose two of the four clients by March" can be wrong, and finding out which way it was wrong is the entire point of keeping the journal in the first place. If a line you just wrote would still sound true no matter what happened, delete it and try again.

Name what you turned down, not just what you chose

Vague: "Decided to take the new role." That sentence tells future-you nothing about what the decision actually cost, because it does not say what you gave up to get it.

Specific: "Took the new role over staying another year to finish the project I started — the deciding factor was that I don't think the project will still exist in its current form by the time it's done either way." Now the entry records a trade-off, not just an outcome, and the trade-off is usually where the real reasoning lives. Most of what you will want to know later is not what you picked, it is what you were choosing it over.

Write a prediction specific enough to be graded

Vague: "I think this will work out." There is no version of the future where this sentence gets marked wrong, so rereading it teaches you nothing about your judgment.

Specific: "I think she'll come around within a week once she's had time to think about it, maybe 65%. If it's still tense after two weeks, I was wrong about how she'd take it, not just about the timing." A number attached to a date does something a feeling cannot: it gives you a moment, later, where you can honestly say "that was right" or "that was wrong" instead of quietly reinterpreting the sentence to match whatever happened.

Write the reason you would not say out loud

Vague: "Felt like the right time to have the conversation." This is the version you would tell a colleague. It is not false, exactly, but it is polished, and a polished reason teaches you nothing about what was actually driving the decision.

Specific: "Honestly, I picked tonight because I was still angry from the thing yesterday and wanted to get it over with, not because tonight was actually a good time for her." This is the line that stings a little to write, which is usually the sign it is the right one. A decision journal is not a place you are presenting to anyone, including your own better self. If you are writing the version you would defend in front of someone, you are keeping a different, less useful document.

Name a tripwire you would actually notice

Vague: "I'll reconsider if it doesn't work." This sets no trigger at all — "doesn't work" is defined only in hindsight, which means it cannot warn you of anything in advance.

Specific: "If I'm still avoiding opening the shared doc by next Friday, that means I don't actually believe this plan, whatever I'm telling people." A real tripwire names a concrete, noticeable behavior — not a feeling, not a vague sense that things are off — so that when it happens, you catch it instead of explaining it away in the moment.

Three ways an entry quietly stops being useful

The defense brief. If the entry reads like something you would show someone to justify the decision, it has already stopped being a record and become an argument. A useful entry can include real doubt sitting right next to the choice you made — in fact it usually should.

The conclusion dressed up as reasoning. Sometimes you have already decided before you sit down to write, and the entry becomes a search for reasons that support what you were going to do anyway. One tell: if every consideration you list points the same direction, you were probably not actually weighing anything.

The wish written as a prediction. "I think this will make things better" is usually hope wearing a prediction's clothes. A real prediction names what "better" would concretely look like and by when, so it can fail to arrive on schedule instead of just quietly not happening.

A full entry, rewritten

Vague version: "Told my sister I can't help with the move this weekend. Felt bad about it but I think it's the right call. Will see how she takes it." Every sentence here is true and none of it will mean anything in six months — there is no rejected alternative named, no checkable prediction, no real reason.

Specific version: "Told my sister no to helping with the move, choosing a weekend to actually rest over the third weekend in a row of saying yes to something. Real reason: I'm tired and a little resentful that I'm usually the one who says yes, not just genuinely busy this particular weekend. Expect some coolness for a few days, then back to normal by the following week — if it's still tense after two weeks, this was about more than one move and I should actually bring that up with her instead of assuming it'll pass." Same decision, same two minutes to write, but this version gives future-you an actual claim to check against what happened.

None of this requires a special app or a particular notebook — a decision journal works on paper as well as anywhere else. CreedOS's daily check-in leaves room for a short note next to whatever you are already tracking, which is enough space for an entry like the specific version above, and the AI coach in the review is there if you want to talk through what a pattern across several entries is actually telling you. It's free, with no in-app purchases.