One of David Snowden's principles is that "Knowledge can't be conscripted, it can only be volunteered". However waiting passively for voluntary contributions is the wrong way to populate a KM repository.
This is an outsome of the problem of the Unknown Knowns. Often people don't don't realise they have learned something, until they are asked about it, or have the chance to discuss it.
Sometimes you find organisations who have set up a system whereby people are required to identify lessons or new knowledge themselves, and then to add them into a knowledge repository. I am not a huge fan of volunteer systems like this; I don't even like them for collecting innovation suggestions. I think you capture only a small proportion of the lessons this way, because people are not aware that they have learned anything, and if they are aware, they often discount the learning as "not important". Also, self-written knowledge are often superficial, because there hasn't been the depth of dialogue and questioning to get to the lesson.
I am not arguing for forcing people to share knowledge, but I am suggesting that you don't wait for the volunteers to come to you. Instead you give people scheduled facilitated conversation-based opportunities where they can become aware of what they know, and which provide a safe and encouraging environment for them to volunteer the knowledge when asked.
There are two main approaches for doing this; reactive, and scheduled.
The reactive approach requires someone to identify particular successes and failures from which to learn. The failures can be obvious, such as HSE incidents or significant project overruns, and many companies have mandatory processes for reviewing these failures. But how do you spot the successes? Maybe you can use your company benchmark metrics, and pick the best performing units for review. Perhaps you could work with the knowledge from the manufacturing plant that never had an accident, as well as from the one with frequent accidents. Maybe you can look for the best sales team, and look to learn the secrets of their success. Or maybe you can do both successes and failures - I did a very interesting study not long ago for an organisation that measures staff engagement using the Gallup survey. We picked the ten top scoring sales teams, the ten bottom scoring teams and the ten teams which had shown the most improvement over the previous year, and interviewed the team leader and a team member of each one, to pick out the secrets of successful staff engagement.
Another organisation we have worked with uses global consultants and Technical Directors to identify opportunities for learning and knowledge transfer. They travel the world, reviewing activity at different centres, and will identify good practice which needs to be repeated, as well as opportunities to learn from mistakes.
An alternative approach, common within project-based organisations, is to schedule learning reviews and knowledge exchange within the activity framework. These could be
- After Action Reviews on a daily basis during high-intensity learning, or after each significant task
- Peer Assists early in each project stage, or during project set-up
- Retrospects (or some other form of Post Project review) at the end of each project stage, or at each project review gate
- A Knowledge Handover meeting at the end of a project, to discuss new knowledge with other projects
- A Technical Limit meeting during the detailed planning stage, to bring in knowledge from people with detailed experience
- A Retrospect (or some other form of review) at the end of a bid process, when the company knows if the bid has been successful or unsuccessful.
There are many advantages to the scheduled approach. Firstly, success and failure are components of every project, and if every project is reviewed, lessons may be identified which can avoid the big mistakes later on. Secondly, if lessons identification is scheduled, it becomes a clear expectation, and the company can monitor if the expectation is being met. This expectation is common in many organisations, thought the rigour with which the expectation is met seems to vary. Finally, by scheduling and facilitating the learning dialogue, you can uncover the knowledge that nobody knows they know, until they start to discuss it.
Don't rely on people volunteering their knowledge spontaneously - instead set up scheduled processes which provide a request and a context for volunteering.