Universal Knowledge Management Rules

Davide Wiest
4 min readMay 5, 2024


This article is about using a Knowledge Management System (KMS) effectively. The rules I’ll cover are universal in the sense that anyone using a knowledge management system (KMS) can benefit greatly from them. No matter whether your goal is to write a great research paper, learn, or simply be more productive at work, these tips will help you get the most out of your knowledge and your system.

Why only 6 rules? Because all other possible candidates weren’t generally relevant enough. I might write one or two more articles about more specific use cases, e.g. knowledge management for education or projects.

Create a positive (learning) environment

For information and ideas to flow, your PKM has to be YOUR space. This is the undisputed number one goal of your knowledge base. Make your brain feel at home. Are you more … or …? Do you NEED order, or is chaos your natural habitat? What’s the atmosphere in real life where you function the best? How can you replicate it in your application? Know these answers and then choose an application that feels right, and organize your system as a natural, more organized extension of your mind. There are two important aspects to the last sentence: 1) If you choose from a purely utilitarian perspective, you’re making a mistake from the start on. You can probably compensate utilitarian aspects, but not a lack of motivation to commit to using and expanding your knowledge base. 2) Why more organized? Your mind is for having idea, not storing them. Your knowledge base is for storing ideas, not having them. If possible, customize the design, font, font size. If you’re a minimalist, hide everything that you don’t use more than once a week or so.

Here‘ s a helpful article about positive learning environments.

Break rules fast

Breaking rules fast is about leaving yourself space for efficiency. I am a very orderly person. When I break a rule, I damage the structural integrity of my PKMS, and it feels like it instantly got a little more chaotic. But this isn’t necessarily bad. The order you’re creating is just for you. The chance that you have programs reading your files based on your rules is astronomically small, so you can stop predenting it were the case. However, what is bad is holding yourself accountable to persisting on a rule that makes no sense. The first few times when it was useful might have just been random. You can break the rule whenever you think that’s the best decision. My personal solution to this problem is to leave empty the space where information concerning the rule would be. Maybe implement an alternative. Or being descriptive about the information that doesn’t fit into the established pattern. Otherwise, I just have to get used to it.

Keep discoverability high

A body of knowledge that you don’t have access to is not a body of knowledge. It has a value of 0 to you. The worst thing besides knowledge being wrong is that it’ll never be used. I hope that it’s clear enough that discoverability is one of the most important metrics of a knowledge management system. I’d rather sacrifice quality and other aspects partly to improve discoverability. Measures to improve discoverability are your standard organizing measures (e.g. tagging, moving to a folder that’s relevant to you), adding metadata (creation date, the area of knowledge it belongs to, what kind of note it is), adding related phrases, creating a reference to a parent-node of your knowledge base, mastering your application’s search functionality, or even using a vector/embedding search tool.

Don’t do BDUF

BDUF (Big design upfront) is a software engineering term that describes establishing the architecture (how all parts of the application work together) before actually implementing anything. It’s counterproductive because unexpected circumstances are basically guaranteed to reveal themselves during implementation. Not during designing. The same is true for your knowledge management system. When you do BDUF on your system, you might break some workflows you didn’t know were essential, or add overhead (e.g. the necessity of filling out a field of a template) that’ll slow you down or even make you resent your system in the long term. Implement architectural changes incrementally. And focus on the changes you know your system needs. A sizable restructuring can take place that way, but it isn’t upfront.

Don’t delete

I thought my old notes on neural networks were of no use to me. Even though I didn’t touch them for 2 years straight, they helped me write my research paper that uses a neural network based model for a prediction. And even if you won’t need your knowledge again-you can never be sure of that-other’s might. And sharing knowledge is a zero-cost way of building rapport. If a part of your stored knowledge does not seem to fit in anymore-maybe it’s outdated or partly wrong-then archive it. The cost of storage is virtually zero, and the time spent on organization can also be virtually zero. Just move it into a folder called “Archived”.

Make backups or use a sync service

Your knowledge is far too important to be dependent on the health/operability of a single device. I use my KMS every day and losing access to it for a week would be terrible. To mitigate that risk, you should have at least 1 independently stored, relatively up-to-date version of your knowledge base. This can be achieved with a syncing service, a backup service, or, as the last option, manual backups (at fixed time intervals, e.g. Sundays).



Davide Wiest

Programmer, Data/AI/QS Enthusiast, Student | Writing technical, knowledge management, PKM, productivity, abstract