So far in this series, I have been focused on knowledge in the context of decisions, both for discovery and by implication purposes. I have always seen Knowledge Management as focused on decision support and innovation and nothing has emerged in the last two decades to change my mind. Nor for that matter my frustration with those who tend to focus on information management for its own sake linked to more abstract (and generally platitudinous) strategic goals. I’ll make a related point here; too many knowledge management programmes start with some abstract goal and then assemble the usual technological culprits to deliver them. Pretty soon they are bewailing the lack of participation and start talking about creating a knowledge-sharing culture. I’ve seen much-vaunted (and still celebrated) KM case studies where I know the team was subsequently closed down after an extended opening run. One of the main reasons is that they didn’t start bottom-up may make day-to-day reality easier for those making both operational and strategic decisions. There was much promise about what would happen when the system was in place and populated but you only get so long before reality catches up.
Now the original design of SenseMaker® came out of my experience in knowledge management but was mainly focused on risk assessment and horizon scanning, including weak signal detection and abductive insight. The fact that I had come at the narrative from the perspective of decision mapping, not communication was important here. A lucky accident of history in terms of utility and the more I discovered aspects of cognition such as conceptual blending the more I was grateful for said accident. It also allowed for the development of my three heuristics for managing a complex situation:
- Use finely-grained objects
- Distribute cognition
- Disintermediate decision-makers
Those rules, plus the conceptual blending aspects resulted in an approach to decision support in which the process of discovery also becomes the system itself. Something which also matches the need to create an eco-system which can change and adapt in use with only stable aspects of the system being put into the more traditional structures of KM practice. In Cynefin terms, only when the complex has been rendered complicated by socially evolved resilient practice over time. Something that happened a lot in the apprentice model by the way but more on that in the concluding post tomorrow.
So what do we do? Well, mapping the day-to-day narratives of decision-makers is the essence of this approach and it is what SenseMaker® was designed to do. In summary for those not familiar with the software, we can do the following:
- Allow web or app-based capture of day-to-day observations and narratives in the field under fire as well as in more reflective environments
- The capture of such observations can include self-recording and interviews within the same system. Capture can be as text, pictures, recordings or any combination thereof.
- Key to the approach is a high abstraction signifier set which introduces deliberate ambiguity into the recall mechanism. This is abductive in nature and matches the evolved recall function of the brain. This provides a qualitative framework that allows real-time use of the material. The signifiers lead to the observations and narrative which increases discovery. Yes, we can run traditional analysis on text. but we are not limited to it.
- We can also add in a traditional scale or multi-choice question but we recommend keeping it simple
- The resulting material can be recalled and used in real-time by the participants as needed to support decisions.
- We can also analyse the data, and suggest connections between silos without the need to share originating data which is available on a request basis (the key here is to recognise that permission to share content should be specific to request not general)
- Executive monitoring for weak signals, patterns of potential loss of knowledge etc. all come from the same capture system so the initial audit also becomes the first decision support system.
Now that is very brief, but there is more material on our site (I need to spend some time expanding that) or a lot on YouTube and I like the State of the Net ones. Day four of the Cynefin and Sense-making courses focus on this, so for KM people days three and four (which can be booked stand-alone) are especially relevant.
So given this what can we do? Well, we normally start these processes with some type of engagement or workshop, I reported on the Bradford one a few days ago which replicates an approach we have run in Belfast, Guildford, Cardiff, Bratislava, Bogota and many other locations! But examples of narrative capture relevant to this subject would include (but are not limited to) the following:
- New employees act as citizen journalists by being assigned groups of senior employees (and potentially retired employees) to interview about incidents in the past which they survived and learnt from, the advice they would give to themselves if they were starting again and so on. These are generally called non-hypothesis questions and are designed to elicit real experiences which are then self-interpreted by the originator into the signifier set.
- In an ethnographic variant of the above (which is more effective) the new member of staff spends a week or a month with different groups of older employees, gathering their own observations as well as those of the people they are observing. Ideally, they help with the work, carrying out menial tasks as needed to gain credibility and engagement of their subjects.
- As this is established existing workbooks and reporting can be replaced by this more fragmented approach. Real-time observations are more useful than reflective material after the event when you are under pressure to update systems (this does not apply to information needed in the complicated and obvious domains, by the way, I am focused on the complex). This means you reduce the reporting requirement rather than (as in most KM initiatives) increasing it.
- This type of system can be extended to those who have retired or left the company to encourage recall as needed with some reward mechanism based on the use of the material.
- Networks of such reporters can be activated to ask questions in real-time – check out my earlier posts on Crowdsensor and MassSense to get an idea of the capabilities here. I sometimes refer to this as building networks for ordinary purposes that you can activate for extraordinary needs. This can extend to external experts as well as staff.
- allow us to show patterns of meaning over these observations and also allow for rapid knowledge sharing over silos without the need to create a common taxonomy or depend on keyword searches or the like. Those have utility and we encourage it, but it is not enough for truly complex situations where you may not know what you need to know until something is presented to you that might have utility.
- Systems can be set up that request information discovered rather than automatically granting a right to the data. This is key in a lot of health, security and other areas.
- A developing use here is for micro-fault reports that build to a bigger picture, this allowing anticipatory audit or intervention. That is too big a subject to summarise here but it is of increasing importance
- Standard SenseMaker® offerings such as Culture Scan can run in parallel to identify which parts of the organisation are most susceptible to early-stage knowledge management initiatives. The ability to work in this way increases the early win successes that spread on the grapevine. Much more effective than the KM team lauding their interpretation of their perception of some use of their systems.
- Material can be incorporated into training programmes and linked to more formal documents – this is called narrative-enhanced doctrine or narrative-enhanced standard operating procedures. Multiple anecdotes (which can be vetted) linking into documents or discovered through some provide richer context and increase use.
So there is a lot we can do and I’ve only given a partial list above. Once these start they more or less immediately start to generate data that can have a practical use. With the ethnographic approach, we are drawing on the apprentice model of learning. By the way, I am making no apology here for selling SenseMaker®. I designed it based on my experience with KM over the years as well as other areas. I also made damn sure it can link symbiotically to content management systems and the like. I’ve shared enough methods and ideas over the years that other people have commercially exploited to be allowed this from time to time.
Tomorrow in the conclusions I want to come back to my original three types of knowledge and also give some advance notice of a standard offering we are creating here, with links to other software and service providers.