Blog

What is knowledge management

Knowledge management is the set of practices, systems, and habits that ensure valuable knowledge is captured, organised, findable, and used rather than lost, forgotten, or locked in someone's head. It sounds dry until you consider the alternative: a team where critical information lives only in the memories of specific people, where important context disappears when someone leaves, where the same research gets done three times because nobody knew it had already been done, and where finding what you need takes longer than doing the work itself.

Research consistently finds that knowledge workers spend between 20% and 30% of their working time searching for information. That's roughly a day and a half per week spent not on productive work but on the retrieval step that should take seconds. The gap between having knowledge and being able to access it when you need it is what knowledge management exists to close.

A brief history

The roots go back further than most people realise. Peter Drucker coined the term "knowledge worker" in 1959, recognising that the economy was shifting from manual labour to work where the primary input and output was information rather than physical goods. His observation was that knowledge workers need to manage their own productivity, because unlike factory workers, nobody can stand over their shoulder and measure their output by the hour.

The discipline of knowledge management as a formal practice emerged in the late 1980s and early 1990s, driven by two overlapping concerns. Organisations were realising that their most valuable asset (what their people knew) was the one they managed least systematically. And the rise of networked computers made it technically possible to capture, store, and share knowledge at a scale that wasn't feasible with physical filing systems.

Ikujiro Nonaka and Hirotaka Takeuchi's 1995 book The Knowledge-Creating Company was the watershed. Their SECI model (Socialisation, Externalisation, Combination, Internalisation) described how tacit and explicit knowledge convert into each other through four processes, and it gave organisations a framework for thinking about how knowledge moves, transforms, and creates value. The model remains the most cited framework in knowledge management research.

In the 2000s and 2010s, personal knowledge management (PKM) emerged as a parallel discipline, driven by people like Tiago Forte (Building a Second Brain), the Zettelkasten revival through Sönke Ahrens, and the rise of tools like Evernote, Roam Research, and Obsidian. PKM applied the same core ideas to the individual: your brain is for having ideas, not holding them. You need an external system. The system should make knowledge more useful over time, not just store it.

Personal knowledge management

PKM is the individual practice. It's about how you capture what you learn, organise it so you can find it later, and connect ideas across different contexts and time periods.

The foundational insight is the same one David Allen built GTD around: your brain is for having ideas, not holding them. Every piece of knowledge you're holding in your head, every insight you haven't captured, every connection you're trying to maintain mentally, is consuming cognitive resources that could be used for actual thinking.

A personal knowledge management system typically involves:

Capture. Getting information into the system as you encounter it. Web clipping for articles and pages. Voice notes for ideas on the go. Annotations on PDFs and documents. Email forwarding for messages worth keeping. Screenshot sync for visual captures. The lower the friction, the more consistently you'll capture, and consistency is what makes the system trustworthy.

Organisation. Structuring what you've captured so it's findable and connected. The PARA method provides one framework (Projects, Areas, Resources, Archives). Zettelkasten provides another (atomic notes linked by concept). Smart organisation and AI-powered tagging reduce the manual overhead of filing, which is where most PKM systems fail, because the filing becomes a job in itself.

Retrieval. Finding things when you need them. This is where the gap between traditional folder-based organisation and semantic search is largest. Folders require you to remember where you filed something. Semantic search lets you describe what you're looking for and finds it based on meaning, regardless of where it lives or what you called it.

Connection. Linking related ideas across sources, time periods, and domains. This is what transforms a collection of notes into something that generates new understanding. Evergreen notes and Zettelkasten are built around this practice. The value of a PKM system increases non-linearly with the density of connections between notes, because connections produce insights that individual notes alone cannot.

Expression. Using your accumulated knowledge to produce something. Writing, research, creative work, decision-making. Forte's CODE framework (Capture, Organise, Distil, Express) makes this explicit: the point of the system is output, not accumulation. A commonplace book that's never consulted is just a box.

Organisational knowledge management

At the team and company level, knowledge management addresses a different but related set of problems.

Tribal knowledge. The informal, undocumented understanding that exists only in people's heads: why things are done a certain way, who to ask about specific systems, the context behind decisions made years ago. When the person who holds this knowledge leaves, the knowledge leaves with them. Capturing tribal knowledge in shared documentation, onboarding materials, and searchable team libraries is typically the highest-leverage knowledge management activity for any organisation.

Documentation. Policies, procedures, reference materials, templates, and checklists. The challenge isn't creating documentation (most organisations have plenty); it's keeping it current, findable, and actually used. Documentation that nobody reads is indistinguishable from no documentation.

Knowledge sharing. Making what one person or team knows available to others. Meeting notes that are shared and searchable. Project retrospectives that capture lessons learned. Collaborative workspaces where work and its context are visible to everyone who needs them. The cultural challenge is often larger than the technical one: in many organisations, knowledge hoarding (keeping what you know to yourself) is rewarded implicitly even when knowledge sharing is encouraged explicitly.

Institutional memory. The accumulated knowledge about what the organisation has tried, what worked, what didn't, and why. Without systematic capture, institutional memory disappears with staff turnover, which means organisations keep making the same mistakes and rediscovering the same solutions.

Why most knowledge management fails

Knowledge management has a reputation problem, particularly at the organisational level. Many people have experienced KM initiatives that produced elaborate systems nobody used, mandatory documentation processes that consumed time without producing value, or knowledge bases that were impressive at launch and abandoned within months.

The failures tend to share common traits:

Too much friction. If contributing to the knowledge system requires significant extra effort on top of doing the actual work, people won't do it. The capture and documentation need to be integrated into the work itself rather than being an additional task layered on top.

No retrieval payoff. If people contribute to a knowledge base but can't find what they need when they need it, the system loses credibility. Poor search, confusing organisation, and stale content all erode trust in the system, and once trust is lost, adoption follows.

Maintained by nobody. Knowledge management systems need ongoing attention: outdated content needs updating, gaps need filling, the structure needs adjusting as the organisation changes. Without clear ownership, the system rots. The content that was accurate two years ago becomes misleading, and people learn to ignore it rather than trust it.

Treating KM as a technology problem. The tool is rarely the issue. The issue is the practice: who captures what, when, how, and why. A culture where people share knowledge willingly and consistently will succeed with a simple tool. A culture where knowledge is hoarded will fail with the most sophisticated system money can buy.

What AI changes

AI, particularly large language models and semantic search, is shifting what's possible in knowledge management in a few specific ways.

Search by meaning, not keyword. Traditional search requires you to guess the right keywords. Semantic search understands what you're looking for and finds relevant content regardless of exact wording. This is a fundamental improvement in retrieval, which is where most KM systems fail in practice.

Automatic organisation. AI-powered tagging and categorisation reduce the filing burden that kills most personal knowledge systems. Instead of deciding where something belongs at capture time, the system handles initial categorisation and you refine as needed.

Synthesis across sources. An AI assistant that can answer questions about your own collected material transforms retrieval from "find the document" to "answer the question." Asking "what have I read about the relationship between attention and working memory?" and getting a synthesised answer from your own library is a qualitative shift from browsing folders and hoping you find the right note.

Knowledge that travels with you. Through protocols like MCP (Model Context Protocol), your personal knowledge library can serve as the memory layer for any AI agent you use. The AI doesn't start from zero; it starts from everything you've captured, organised, and refined. This is what turns a knowledge management system from a passive archive into an active thinking partner.

Getting started

If you're an individual looking to build a personal knowledge management practice, start simple:

Choose one capture tool and use it consistently. A notes app with a web clipper and mobile capture covers most situations. The tool matters less than the habit.

Adopt a simple organisational framework. PARA is a good default: four categories based on actionability. Don't build an elaborate taxonomy before you have content to organise.

Review regularly. A weekly review keeps the system current. Without review, the system accumulates without improving.

Focus on retrieval, not filing. A system where you can find things by meaning is more resilient to imperfect filing than one that depends on you always putting things in the right folder.

If you're building a team knowledge management practice, start by making tribal knowledge explicit. Identify the information that currently lives only in specific people's heads and document it in a shared workspace. The rest, the templates, the processes, the elaborate knowledge base, can develop over time. The tribal knowledge is the urgent part, because it's the knowledge you lose when someone leaves.

Frequently asked questions

What's the difference between knowledge management and information management? Information management is about storing and organising data and documents. Knowledge management includes information management but goes further: it's about making that information usable, connecting it to context, converting tacit knowledge into explicit knowledge, and creating conditions where knowledge is shared and applied rather than just stored.

Do I need a special tool for knowledge management? No. The practice matters more than the tool. That said, tools with semantic search, automatic organisation, and multi-source capture make the practice significantly easier to sustain because they reduce the manual overhead that causes most systems to fail.

What is personal knowledge management (PKM)? PKM is knowledge management applied to the individual rather than the organisation. It's about how you capture what you learn, organise it, find it when you need it, and use it to produce work. The main frameworks include Building a Second Brain (CODE: Capture, Organise, Distil, Express), Zettelkasten (atomic linked notes), and PARA (Projects, Areas, Resources, Archives).

How is knowledge management different from note-taking? Note-taking is one component of knowledge management: the capture step. Knowledge management also includes organisation (how you structure what you've captured), retrieval (how you find it later), connection (how you link related ideas), maintenance (how you keep the system current), and expression (how you use your knowledge to produce output).

What are the different types of knowledge that need managing? The main types are explicit knowledge (can be written down and shared), tacit knowledge (known but difficult to articulate), implicit knowledge (could be articulated but hasn't been), procedural knowledge (knowing how), and tribal knowledge (informal organisational know-how). Each type requires a different approach to capture and management. The types of knowledge guide covers these in detail.