A daily log for decisions, not for feelings
Search for a daily journal app and you'll find roughly the same product everywhere: a prompt of the day, a mood to tap, a streak not to break, sometimes a gratitude line. These apps are well made. They answer a real need.
It's probably just not yours.
Because the question you'll ask in six months isn't how did I feel on 20 August. It's: why did I decide that?
You always remember your decisions. You almost never remember what you knew when you made them. And that is precisely the information you need in order to tell whether you decide well.
The thing an introspective journal can't do
A decision can be judged two ways.
By its outcome: it worked, or it didn't. That's how everyone judges, because it's the only information still available later.
By its process: given what you knew at the time, was it reasonable? That's the only useful way, and it becomes impossible the moment the context fades.
An ordinary example. In March you turn down a project because the brief looks vague and the client has no approved budget. In September you hear it went well with someone else. You conclude you got it wrong.
Maybe. Or maybe the brief really was poor in March and somebody fixed it in April. With no record you can't tell the two apart — so you adjust your judgement based on an outcome you didn't control. It's the most efficient way to un-learn.
A mood journal doesn't help here. It records what you felt, not what you knew.
What a decision log actually records
Four things, and they fit in three lines.
What was decided. Phrased as an action, not an intention. "Turned down the Duval project," not "thought about the Duval project."
What you knew. The two or three facts you leaned on. This is the most valuable line and the fastest to evaporate.
What you expected. A prediction, however loose. "I think they'll come back in September with a budget." Without a written prediction your memory will always rewrite the story in your favour — you'll remember having anticipated whatever happened.
What would change your mind. The line almost nobody writes, and the one that turns an opinion into a revisable position.
In practice:
20 August — Turned down Duval. Brief unclear (3 contacts, no identified decision-maker), budget not approved. Expect them back in autumn with a written scope. If the budget clears sooner, I'll reopen it.
Thirty seconds to write. In September that note is worth more than three pages of reflection.
Why the date does most of the work
An undated note loses its value within weeks. "Just changed jobs" is useful in March and misleading in November — and nothing in the sentence tells you which one you're reading.
A decision log isn't a collection of texts. It's a chronology. Its value doesn't come from any single entry but from being able to read them in order. That's what lets you see, six months on, that you've deferred the same decision four times — something no individual entry says.
It's also why the commonly recommended folder of markdown files holds up badly over time: every note needs a name and a location, and you pay that decision on every single write.
What to look for in an app
The shortlist is simpler than it looks. Four criteria.
Entries are dated without you doing anything. If dating is an action, you won't date.
Writing takes under ten seconds. The cost of writing determines what gets written — and a decision gets logged in the minutes after it's made, never in the evening. It's the same mechanism as quick capture on mobile: past a few seconds of friction you only record what already feels important, which means filtering before you've had time to think.
An entry can attach to a person and to a project. Most decisions involve someone. If you can't pull up every decision tied to a client before you get them on the phone, the chronology can only be read whole — which means never.
Search runs on the text, not on tags. You'll search "Duval" or "budget," not a category you picked eight months ago.
What you're not looking for: daily prompts, streaks, mood trackers. Those mechanics exist to build a habit of writing every day. But a decision log doesn't need to be daily — some days contain no decisions, and nothing should push you to write on those. A streak would do exactly that: make you fill the gap.
Re-reading, which is half the system
Writing without ever re-reading produces an archive, not a tool. Two moments are enough.
Before a conversation. Thirty seconds on that person's chronology, before the call. It's the highest-return use, and the most immediate.
Once a week, ten minutes. Go through the week's decisions looking for one thing: predictions that have come due. Did what you expected actually happen? The point isn't to score yourself, it's to calibrate — after a few months you know which areas you read accurately and which ones you're reliably optimistic about.
It's the same mechanism as the write-to-reopen ratio: a system nobody re-reads stops working without anything announcing it.
When the introspective journal is the right tool
In fairness: if what you want is to put the day down before sleeping, track a mood across months, or keep a writing practice for its own sake, then conventional journaling apps are exactly right and this article isn't about you.
The two uses don't conflict. They simply don't call for the same tool, and trying to hold both in one always ends up sacrificing the quieter of the two — the one with no streak, no notification, and whose usefulness only shows up six months later.
Where to start
Don't open a fresh notebook to record future decisions. Do the opposite.
Write down today the last three decisions you made and still remember. For each: what you knew, what you expected. You'll probably find the second line is already hard to reconstruct for the oldest of the three — and that is the demonstration.
Then log as you go. Three lines, in the minutes that follow. The value isn't visible in the first week. It shows up the day someone asks why you decided that, and you have the exact answer.
Keepsake dates every entry automatically, attaches it to the people and projects involved, and shows the full chronology on a contact's page. Try it free for 30 days.