Just listened to the podcast of the CoP Workshop at KM Australia 2007 run by Dave Snowden and Etienne Wenger
Here are some quick notes I took down.
“The lower the cost of codification, and the more I codify, the faster knowledge diffuses independently of the knowledge holder”
Abstraction is the formal/informal relationship you have with someone (language, education, morals, values, beliefs, experiences)
- high abstraction are those peers that you are in tune with, you think similar, same wavelength, the time you have to spend together spurs great stuff
Informal groups of eg 5 people you trust and gel with is a high abstraction community, and therefore a low cost of codification and high diffusion
- downside is this is only between a limited number of people
So we are limited to a narrow range (small amount of people) where we get efficient distribution of knowledge.
Formal group of eg 150 people (are low abstraction because of the degree of diffusion necessary, lots of money, still useful)
- “range of abstraction” is too high
- what happens is that the experts in these large groups don’t have time to answer all these questions (they don’t have time for codification at this level), they rather be in smaller same level abstraction groups where they can nuture ideas in good time with people in their trust circle of knowledge
- so others who are not as much experts in their field take on CoP more seriously as this can be their potential forte, so we get mediocrities running communities.
So you are better off with 5 CoPs on similar topics instead of 1, this way you have many groups with common levels of abstraction, you maximise perspective on dealing with problems, this is more naturally sustainable.
Plus people are more prone to share failures and learn from them with small groups (high abstraction), in a bigger group you are not as inclined to share certain information, and this is an opportunity loss.
IBM a while back (the power of self forming informal communities)
- 150,000 staff
- 50 formal communities
- once adhoc communities were easy enough to create, 64,000 informal rooms resulted
(important that documents are pointed to in the DMS and not housed in these rooms)
Formal communities become a force of conservatism when they live on too long (innovation comes not from experts but accidents and mavericks)
- if you set one up give it a limited time span
- if they want to stay together, they may become an informal group with high interdependency (this is where knowledge is generated, areas open to the outside universe)
Self organising communities
- top down (set boundaries)…otherwise dominant behaviours (manipulators) emerge
DO NOT DETERMINE
- outcome…naturally happens
Measure impact, not outcomes
- don’t want to say in advance what you want to achieve and set a measure
- complex system is based on uncertainty
- each time you change a target you change the system, so you are perpetually introducing new measures, creating new issues
Instead collect stories
- able to produce an object set of figures (context for what figures mean)
- then set target to change impact by changing ??? stories (don’t think I heard this right)
Community (cohesive common culture)
- more a ideation than a rule based culture, ie. grown up in it so you know how to behave, rather than explicit rules, common ways of looking at the world helps them to function, membership varies, you may not even know you are a member
- have more role based functions (role dependency expectations)
processes that ritualise the identity shift to perform a trained function
- more dynamix flexibility
- train people for specific functions
- create expectation or trust in function over people
After listening to this podcast, I clearly see the value of a tight informal network getting things done…sometimes an informal network is better staying that way, as it’s serving its purpose in the most optimal form, ie. a group of topic experts in a tight knit interdependency group of common abstraction.
It’s surprising to see a new report on formalising informal networks, when the informal nature is an act of getting away from formal procedures.
Noticing the value of informal networks as an alternative to hierarchy to get things done is great, and offering tools for that network to enhance their group may be beneficial.
But there is a difference between formalising existing groups and just giving them a few tools to help do their work easier.
Formalising groups is serious, you are messing with the structure and dynamics, it may never be the same…but I do agree that sometimes an informal group may welcome to grow and a formal approach may help it achieve it’s aims.
I just don’t think a clean sweep to formalise all informal groups is a good one, it must be done on a one by one basis, the best for the group must be considered, will formalising it weaken it, if so leave it; better that people are sharing in an informal way than not sharing at all.
If a group rather stay informal, and you really are set on formalising it, why not make a formal clone of their group and see if people take to it…
A few things I picked up from Dave Snowden’s podcast:
- usually it’s the other way around, a formal network may live on as an informal network
- some informal networks may be just that, due to trust reasons
- some informal networks may be just that, so they can talk about past failures
- some informal networks may be just that, because of the common level of abstraction
- some informal networks may be just that, because of the even level of interdependency
- some informal networks, once formalised are open to all eyes, resulting in less daring posts
- some informal networks, once formalised, are managed (to some extent) and the direction may change and content must lead to deliverables…the group is no longer what it used to be about, now it may have to perform for an outcome. Eventually the discussion is restricted to serve the end purpose, this restriction in content variance goes counter to the emergent nature of free-form discussion. Innovation is not going to come when your discussion is focused on an outcome, whereas a free-form environment goes with the flow, and the flow may give shape to things we wouldn’t of imagined as we have the space to let these ideas flourish.
The IBM example mentions 64,000 informal rooms, some might say, well aren’t these silos (firstly need to store documents in a DMS and point to them), isn’t a formal network better so more people can discuss the same topic. Well according to the podcast above, this high range of abstraction is not always the optimum scenario for people to flourish their ideas…information diffuses effectively when you have a group with a common level of abstraction, and there is high interdependency (compared to the guru being drained all day, and not learning much herself).
Hmmm, are the signs of the preference of informal networks to get things done and easier knowledge diffusion a sign of perfect timing for enterprise social networks. Here you would not have to visit any groups at all, instead you tap into your network and information comes to you.
Here are some comments I left on Jay Cross’s blog:
“I agree with Mike Gotta on some points, I think the idea of the report is the realisation that people connect in the enterprise, or are related other ways than org hierarchy…and they may do this effectively with informal networks, such as email groups.The report is wanting to capitalise on these networks, give them more exposure so others can be aware and have input…the more people connecting the more people know where to go to get the right information.
In some instances this may be handy as an email group may be too busy to amplify their network…but if some learning/km staff offer some tools, primers (guides), and training, they may be excited that it will be easier to network in their group.
But it’s only an offer…I agree with you, there is no need to formalise these groups if they don’t want to. Perhaps they may like networking tools as they are the right tools for the job, but this doesn’t mean in exchange they have to formalise (or open up) their group.
Like you say the moment this happens, it becomes a task group, especially having a performance leader…people once contributed for sharing interests and expertise, being formal may turn it into more work and deliverables…then people will slowly be less enthused as they have been hijacked to some extent.
The other issue is trust, whether a group or network takes on new tools to help connect better, they may not want to become formal as they only trust each other with the content that resides in the group.
It’s better leaving a group alone, than formalising…they may abandon the group, and then you are left with no-one sharing anything. It’s better that some share than none.
I got the trust point from Dave Snowden’s podcast:
Perfect example of what you posted about:
Management can kill a community of practice
Luis Suarez also picked this up a while back:
Instead of managing a CoP, he mentions leadership and sponsoring as much less intrusive”
[ADDED 29/11/07: Knowledge sharing in the new KM]