Just of late I've encountered a number of capability maturity models aimed at knowledge management. Capability maturity models have been around for a while in other disciplines, most notably in software development projects. Almost all of the models owe their origins to the collaboration between the US Department of Defense and the Software Engineering Institute of Carnegie Mellon University. The Capability Maturity Model was originally a tool to assess processes - in particular the processes of a contracted third party. In that sense its intent was to reduce risk.

Capability models now abound and even have been internationally standardised as part of ISO 15504 . ISO 15504 sets six levels of capability maturity as follows:

• Level 5 - Optimising Process,
• Level 4 - Predictable Process, 
• Level 3 - Established Process,
• Level 2 - Managed Process, 
• Level 1 - Performed Process, and
• Level 0 - Incomplete Process. 

Notice anything here? It's all about process. Putting aside definitional issues, last time I looked knowledge management was about people, process, technology, and content. Capability maturity models are about process, so it begs the question "Does knowledge management really need a maturity model"?

Well perhaps I put the definitional issue to one side too early. Capability models measure, or at least attempt to measure, maturity. But how can we measure something if we can't agree on what that something is! And what do we actually want to measure?

I've blogged about this dilemma before . Coming up with measures for a knowledge management initiative is particularly difficult. It's too easy to report activity rates - how many children had a bath (to use an analogy) - because these are tangible and relatively easy to measure. Measuring and reporting the true impact of the initiative on the organisation - had a bath and came out clean - is much more difficult; if only because the impact will be variable, and not everyone will agree the strength of the outcome. Will a formal knowledge management maturity model help us?

Six months into the development of TARDIS we were subjected to Manfred Langen's 2004 Knowledge Management Maturity Model™ . This model has two parts - an analysis model and a development model. In 2004 Langen's maturity levels had the following definitions:

• ‘Initial means that information and its processes are not consciously controlled. Success in information management activities are a stroke of luck.
• Repeatable means that the importance of information management is recognised and pilot projects on information exist.
• Defined means there are stable and practised activities, which effectively support the information management of the organisation. These activities are integrated into day to day work and the technical systems enabling this are maintained.
• Managed means there are common organisation wide strategies and standards, with regularly measured indicators of effectiveness and efficiency of information management. The activities are secured in the long term by organisation wide roles and compatible socio-technical systems.
• Optimising means the organisation has developed the ability to adapt flexibly in order to meet new requirements in information management without dropping a maturity level, even in the case of large external or internal changes.'

See the problem? This time instead of measuring knowledge management we were being assessed against information management!

Now jump forward to June 2008 and have a look at the Suresh and Mahesh model . Their maturity levels describe the position of knowledge management within an organisation, beginning with no knowledge management and finishing with a stable implementation. At least they use the words "knowledge management", but just what they are measuring is another thing all together.

And now look at APQC's knowledge management maturity model. They say their maturity model:

"provides a roadmap for moving from immature, inconsistent knowledge management activities to mature, disciplined approaches aligned to strategic business imperatives. The KMMM is integrated with APQC's Stages of Implementation™ so that implementation at each stage provides a foundation of success and a launching pad to the next stage."

Again they use the words "knowledge management", but it's light on the how. And again I just don't know what they are actually measuring - mainly processes it seems.

Until we have an agreed definition of just what "knowledge management" is, these generic frameworks provide at best some guides to how you might measure your initiative. They certainly don't provide the means for external bench-marking, and arguably as a knowledge management initiative matures and evolves they don't provide internal benchmarks - at least that's my TARDIS experience. Call me a cynic but I don't think knowledge management as a discipline needs a maturity model just yet. Definitions and what we are measuring matter, and none of these tools solve that impasse. Use them with extreme caution.

Regards, Graham