Why Knowledge Management Initiatives Fail

The Second Attempt

We have written before about why knowledge management gets overlooked, setting out ten reasons it slides down the priority list. Culture was one of those ten, sitting alongside ownership, measurement and cost.

This article argues that one of those ten does more damage than the other nine combined, and that it is the reason funded initiatives fail. Not the ones that never get off the ground, but the ones with a budget, an owner, a tool and a launch, that are quietly dead eighteen months later.

When those fail, the post mortem usually blames the tooling, the templates, or the fact that people were too busy. Occasionally it blames the process. In our experience it is almost never any of those things.

Culture, Not Process

There is a direct relationship between organisational culture and the willingness to share what you know.

Many people hold a belief, often without examining it, that knowledge is power. Nobody who holds that belief is keen to surrender it. If what you know is what makes you valuable, writing it down and handing it to the organisation looks a lot like reducing your own worth.

That is an ego problem, but not the loud kind. Not the chest beating and the trumpet blowing, though that exists too. The more damaging version is quieter, and it is built on fear.

What People Are Actually Afraid Of

In a lot of organisations, fear runs quietly through the culture. Fear of losing your job. Fear that the organisation stops depending on you. Fear that the thing you are uniquely good at becomes a document anyone can follow.

Those fears do not have to be well founded to be effective. If the organisation cannot demonstrably disprove them, the fear behaves exactly as though it were justified, and the outcome is the same. People do not share.

You can see it in the artefacts. Articles that describe the happy path and omit the difficult parts. Documentation written at a level of detail that still requires the author present. Contributions that stop the moment the launch enthusiasm fades. None of that is laziness. It is rational behaviour in an environment where sharing feels risky.

Which Is a Leadership Problem

This is uncomfortable, because it means knowledge management is not primarily a service management problem. It is a leadership one.

Culture cannot be fixed from the bottom up. A knowledge manager, however capable, has no mandate to change what people are afraid of. Only leaders can do that, and only by doing it openly.

The behaviours that build the right conditions are the ordinary, unglamorous ones:

• Is failure treated as something to learn from, or something to attribute?
• Are the people who document their work visibly valued for it, or just given more work?
• When someone shares knowledge that makes them replaceable in a task, what actually happens to them?
• Is there any evidence, that employees can point to, that sharing has been rewarded rather than punished?

Trust is the operative word. Where trust exists, sharing follows without much prompting. Where it does not, no amount of templating, tooling or process design will produce it.

Why This Matters More Now

There is a practical reason this has become urgent rather than merely important.

A great deal of what organisations want to do with AI in service management depends on the knowledge base. Answer deflection, agent assistance, automated resolution: all of it is built on the assumption that the organisation has written down what it knows, accurately and completely.

Most have not. And teams often sense this before they can articulate it, which is why conversations about AI readiness so frequently stall at the point someone asks what the assistant will actually be trained on.

An AI capability built on a thin knowledge base does not fail quietly. It fails visibly, in front of users, at scale.

Where to Start

If you are considering another knowledge management initiative, the first question is not which tool or which template standard.

It is whether the conditions that killed the last one have changed. If they have not, a second initiative will consume real resources and reach the same place.

That is a harder conversation than a tooling decision, and it needs to happen with the people who can actually change the answer.