Lean Personal Knowledge Management

Most personal knowledge management advice starts with what a system can hold. Mine starts with what I can keep using when I am busy, interrupted, or unsure whether the work will matter later.

I learned that from freelancing. A meeting would happen once, with a person I might never speak to again. We would make a decision, change the shape of a project, or name a problem that was not obvious before the call. If I did not write it down soon after, the useful part of the meeting disappeared.

Employment records are not much better. They preserve job titles and dates. They do not reliably preserve the work between them, especially informal work, failed attempts, or the decisions that prevented a problem. A résumé compresses even more.

I wanted a record I could maintain myself without turning note-taking into a second job. The system that survived has five parts.

Write in time order

Start with a daily note. Put events, decisions, observations, and work under the date when they happened.

Chronology gives a note useful context before you organize anything else. You know what came before it, what followed it, and what else competed for attention that day. That is often enough to reconstruct why a decision made sense at the time.

The daily note does not need to be a diary. A few sentences can preserve the event:

Revised the career-continuity essay. Removed an unsupported placeholder instead of inventing an example. Kept the friend’s identity and family photographs private.

That entry records an action, a decision, and a boundary. I can add links if the work recurs. I do not need to classify every sentence before I write it.

Roll small periods into larger ones

Daily notes are good capture surfaces and poor summaries. At the end of a week, I look for movement: what changed, what repeated, what remains unfinished. A monthly note keeps only the outcomes and decisions that still matter at that scale.

The rollup is not a copy of every daily entry. It is a new account written with more distance. On Tuesday I may record that I changed an essay title. At the end of the month, the useful fact may be that I published the first piece in a four-part series and learned that the source record was strong enough to remove an unsupported anecdote.

Each level answers a different question. What happened today? What moved this week? What changed this month?

Use a small action vocabulary

I describe recurring work with plain action words such as planning, writing, prospecting, presenting, and preparing.

These words tell me what I was doing. Broad subjects often do not. A note tagged software might contain almost anything. A note marked writing tells me that I was producing or revising words, whether the subject was software, work, or parenting.

The vocabulary should stay small. I add a word after I have repeatedly wanted to find that kind of activity and could not. A classification system built in advance usually describes the person I expect to become, not the work I actually do.

Link what comes back

A person, project, or document set deserves its own note when it recurs. The link lets the daily note stay chronological while the linked note gathers a relationship over time.

This is especially useful when a name changes, a project moves, or a document ends up in several contexts. The link carries the relationship without forcing one file to live in three folders.

Person notes need restraint. I keep enough context to remember who someone is, what we worked on, and what decisions involved them. I do not need a private dossier on everyone I meet.

Keep folders shallow

Folders are good at answering one question: where does this file live? They are worse at representing all the reasons I may want it later.

A shallow set of folders can separate daily notes, references, projects, and source material. Links can express the messier relationships among them. This avoids the recurring chore of deciding whether one note belongs under a person, a client, a subject, or a year.

The point is not to eliminate organization. It is to stop asking the folder tree to do work it cannot do well.

The test

I consider the system useful if I can return months later and answer a few ordinary questions:

• What happened?
• Why did it matter then?
• Who or what was involved?
• What happened next?

If the system answers those questions, I can reconstruct a project, a difficult year, or a change in direction. If maintaining it takes more time than the work it records, I have made it too large.

Lean PKM is not a promise that I will remember everything. It is a way to keep important periods from becoming blank just because nobody else was responsible for recording them.