Folders, tags, or neither: I have reorganised four times and none of it made any difference

The rule that ends the reorganising: pick one organising idea and do not run two.

Folders and tags together is where people lose weekends, because every new note poses the question of which system it belongs to, and the answer is arbitrary. Two systems means two decisions per note and no consistency in either.

So:

Folders only, if you think in projects and places.

Tags only, if you think in themes, and keep the vocabulary small - under twenty. A tag used once is not a tag, it is a word, and search already handles words.

Neither, with a flat folder and good titles, which works better than people expect if you also write good titles.

Whichever you choose, the thing that does most of the work is the title. A note called meeting is unfindable by any means. One called client onboarding call, decision on pricing tiers is findable by search, scanning and memory, in any structure or none.

And on your notes you know exist and cannot find: that is nearly always a titling problem rather than a filing one. Try searching for what you would have written rather than what it was about: your own phrasing is usually in there.

Ended up with one note per day, no folders, no tags, and search. Four years in and I find things faster than I did under any of the systems, because the date is a fact and a category is a judgement I have to reproduce later.

The only structure I kept is a link when one note genuinely follows from another.

Your suspicion is correct and it is worth stating plainly: for retrieval, search beats structure, and it has for years.

If you can remember any distinctive word in a note, search finds it faster than any hierarchy. Filing exists partly as a habit from paper, where search was impossible.

Which means the four reorganisations changed nothing because they were solving a problem you did not have.

What structure is actually for, and it is a much smaller job than people assume:

Finding things you cannot describe. Search needs a word. Structure helps when you are looking for the thing about that project, whose contents you have forgotten.

Seeing what you have. A list of projects tells you what is active. A flat pile of four thousand notes does not.

Separating states. Draft against finished, active against archived. This is genuinely useful and it is about status, not subject.

So the useful minimum: a small number of folders reflecting what you are doing, not what things are about, and search for everything else.

The reason subject folders never work is that a note is usually about three things and you have to pick one. Status has one answer, which is why it is the axis that survives.

One reframe that stopped me reorganising: organise around what you are producing, not around what things are.

Subject categories are infinite and arbitrary: that is why every system based on them eventually collapses and gets replaced by another. Outputs are finite and obvious: the article, the project, the course, the decision. A note either serves one of those or it does not.

In practice that means a small number of folders that are things you are working on, an inbox for anything uncategorised, and an archive for anything finished. Three states and a handful of active projects. When a project ends, its folder goes to the archive whole, and it never needs to be reclassified.

What this fixes: notes stop needing a permanent home, which is what makes subject filing so hard. A note is temporarily attached to work, and then it is archived, and search finds it afterwards.

And a practical note on the reorganising urge itself: do not migrate the old four years. Leave them where they are, apply the new arrangement to new notes only, and let the archive be an archive. Every weekend you have spent so far went on moving old notes you were never going to read, and that is the part that produced no benefit.