My note after a round has three lines and takes about forty seconds. The first line is what I heard: one instrument, one texture, one production detail, in whatever words come out. The second is where the guess came from: the voice, the drums, a lyric, a feeling that turned out to be a different artist's feeling. The third is a single word for the round, like "close", "blind" or "lucky".
The template is deliberately small because a note that takes five minutes does not happen on a busy morning. Forty seconds is under the threshold where my brain starts negotiating about whether the note is worth it. And three lines is enough to show a pattern over a month: I can read twenty of these in five minutes and see which of my two guesses was the real one.
The third line is the one I expected to be useless and the one that turned out to matter. "Lucky" rounds teach nothing and should be recognised as such, while "close" rounds are where the learning is. After a few months I noticed a steady shift from lucky to close, which is exactly the shape of progress in a guessing game with a small information budget.
If you only write one line, write the hearing line. Start guessing once tomorrow with a scrap of paper next to the keyboard and describe the first clip before you type an answer. Even on a terrible round it leaves something useful behind, and the song list in your head grows faster when the descriptions are written down rather than remembered.
The template has survived one revision, and the revision was subtracting. For a few months I added a line about confidence, scored from one to five. The scores were almost always wrong in the same direction and they made the note slower without making it more useful. The lesson was that the useful part of a note is the description, and descriptions do not need ratings attached to them. Three lines, one of them about hearing. If you are starting a note habit, start with fewer lines than you think you need and let the useful ones grow back on their own.