Knowledge mapping is presented as a large exercise with a methodology attached, and the useful version is small enough to do in an afternoon. It answers three questions about what a company knows, and the answer to the third one is the only reason to do it at all, because it is the one that names a risk rather than describing a situation.
What the company knows
The topics somebody would have to be told about to do a job here. Not documents, topics: how pricing exceptions are approved, how the deploy works, why the second warehouse is different. Thirty to eighty lines for a small company.
Where each one lives
Article, drive, chat thread, or nowhere. The nowhere rows are the point of the exercise and there are always more of them than anybody expects, because the topics nobody has written down are exactly the ones everybody assumes somebody else has.
Who holds it, and how many people
One name against a topic is a risk. This is the column that turns a mapping exercise into an argument for doing something, because a topic held by one person who is also the only person who could write it down is a specific, nameable exposure rather than a general worry.
The afternoon version
A spreadsheet with those three columns, filled in by two or three people who have been here longest. It is not rigorous and it does not need to be: the single-holder rows are visible immediately and they are the whole output.
What to do with it
Write the single-holder rows first. That is your article list, ranked by risk rather than by volume, and it is a genuinely different list from the one the support inbox gives you: the inbox ranks by what is asked, and this ranks by what would be lost.
Questions people ask about knowledge map
Is this the same as a knowledge audit?
A knowledge audit is the formal version and is usually much larger. The three-column spreadsheet gets most of the value in an afternoon, which is why it is the version that actually happens.
How often should we redo it?
When somebody leaves or joins in a role that holds topics, which is a better trigger than a calendar. Between those, once a year is plenty.
Does the map become the base structure?
It informs it but should not become it. The map is organised by what the company knows; the base has to be organised by what readers ask, and those are different orderings.