I got serious about organising notes because my own notes failed me once.
I was writing a proposal and I was certain I had recorded a key set of numbers somewhere. I searched for half an hour: bookmarks, chat history, several documents. Eventually I found it in a file called “New Document (3)”.
That was the moment I understood: having written it down is not the same as owning it.
Decide where things go first
Most note chaos starts with “where should this go”. So the first step is not writing, it is fixing the folders.
I settled on eight:
- Diary: what happens each day, the raw data.
- Note: structured knowledge points.
- Essay: experience, methods, reviews.
- Project: documents for work in progress and finished.
- Index: gathers related items scattered elsewhere.
- Principle: criteria and rules for decisions.
- Resource: external material and tools.
- Archive: finished things, moved out of the way.
One rule: the same kind of content does not live in two folders. Before creating a file, spend one second deciding where it goes. It saves half an hour of reorganising later.
The filename is the index
My filenames look like this: [prefix][date][number] Title.md.
The prefix marks the type: ND for a diary, N for a knowledge point, E for experience, PJ for a project, P for a principle. The date is six digits; the number is the order of same-type files that day.
For example, the third essay of 13 April 2026 is E2604133 Personal Knowledge Management Methodology.
The diary is simpler: one file per day, ND260413, no number. Messages in group chats, meeting notes, random thoughts — all go into that day’s diary first, and get sorted at the weekend.
The benefits are practical. Sorting by filename is sorting by time. You can tell what a file is without opening it. And backlinks do not break when a title changes.
A diary at five-minute resolution
My diary is not a mood journal. It is more like a log.
Each event is one line: place, time, what I did, an energy score, and one objective sentence. Times round to five minutes — 12:21 becomes 12:20. No “around” or “approximately”.
It sounds obsessive. But when you fill it in later, the details are gone and what you write is interpretation. Filling it in the same day at least keeps the facts.
The log is the raw material for everything else. A review, a principle, an essay — each can be traced back to the afternoon it happened.
An example. The diary entry says: “📌 customer visit 📍 factory ⏱ 14:00–16:30 🌊 6; customer asked about lead time three times, did not move on price.” If I later write an essay about sales reviews, I start from there instead of inventing from memory.
Four kinds of knowledge
When distilling, I tag each item with one of four letters.
C is concept: what it is, where its boundary is. P is principle: when this kind of choice appears, what to decide by. M is method: how to do it — steps, commands, templates. E is evidence: the concrete case that proves it works.
Splitting them has an immediate benefit: within one topic I can see which kind is missing. Lots of concepts and no principles means I have been learning without using. Lots of methods and no evidence means the methods may never have been verified.
Now, whenever I finish a note, I try to tag it. If I cannot tag it, I have not thought it through yet.
Having written it down is not the same as owning it.
Why not a fancier tool
I use Obsidian, plain text, plus backlinks.
The reason is simple: I want this to survive ten years. Plain text opens a decade from now, any editor can read it, and I can feed it straight to an AI. Screenshots and PDFs cannot do that.
Backlinks solve another problem: folders are trees, knowledge is a network. A link from a diary to a principle, from a principle to a project, stops folders from chopping content apart.
When I write an essay about quotes, for example, I link it to the relevant principles, a few concrete customer cases, and the process notes I wrote before. Three months later, two clicks get me back to that context instead of relying on memory.
There is a migration cost to this, so I basically stopped changing tools. I used to look for the perfect app. Now I know a familiar broken tool beats an unfamiliar perfect one.
Maintenance
A system is not finished when it is built. My rhythm: daily entries, a weekly tidy, a monthly archive, a quarterly look at the principles.
I do the weekly tidy on Wednesday afternoons. Go through the week’s diary, turn what can be distilled into notes, leave the rest. It takes about half an hour a week.
The hard part is not the operation. It is “I’ll organise it when I have time.” That time never comes.
Mistakes I made
The first mistake was folders that were too fine-grained. For a while I made subfolders for every field; each ended up with two or three items and finding things got slower. I learned: keep folders coarse, put the detail in tags and indexes.
The second was changing tools too often. Each migration took a week, and in the end the content grew very little while the time went into moving and redecorating.
The third was collecting without distilling. Saving good articles and never opening them again. Now I have a rule: anything saved needs one sentence within a week explaining why it is worth keeping. If I cannot write that sentence, I delete it.
Last
My knowledge base still has plenty of unfiled cards. It is not perfect, but at least finding things does not depend on luck anymore.
Knowledge management sounds abstract. In the end it is one sentence: next time you need it, you can find it.
If I ever stop using this system, that will probably mean I found something simpler that suits who I am at that point.
