Report from KM Europe 2004

From all software vendors I met, two are worth mentioning for my company:

SAP: I spent some time with the SAP folks trying to figure out what they meant by entreprise collaboration and KM on SAP Netweaver. I saw a demo of SAP's own intranet using those collaborative features. I saw something that looked like a simplified version of Microsoft's Sharepoint with SAP look and feel and clean graphics. Nothing fancy, just a basic rugged collaboration tool completely integrated in the SAP environment. This should be great news for CIOs, and not so much for teams and communities, who are likely to be talked into using the company's official collaboration space instead of their own. It will be like being told to leave your cosy little table in the cafe down the street and rather meet your colleagues in the company cafeteria instead. I also found out that the SAP folks keep their e-mail and the Intranet totally separate, which came as a surprise to me.

Google (hey they have a blog) presented its new search appliance. It's basically a stand alone hardware + software system running the Google algorithms (100 different ones as we were told), which can index and crawl all the content of a company intranet having an http front-end (i.e. for example Notes databases can be crawled through the Domino server). Algorithms have been readjusted so as to reduce the importance of inbound links in the page ranking system. Security features are taken into account to make sure that the search result ony shows documents accessibe to the employee. Finally, the official voice of the company (i.e. those pages that the management wants to see appearing high in the ranking) are handled as "sponsored links" exactly the same way as advertised pages on the Google web search engine.

Keynote speakers were far more interesting.

Karl Wiig addressed the question of "people-focused KM". For Karl, the rule of the game today is no longer to learn but to innovate faster than competition. Thus, KM should essentially serve the purpose of continual Intellectual Capital (IC) development, and focus on "effective decision-making, problem solving and execution" (Mmmh...KM for Execution...my management will love this). Karl stressed that, in business settings, there is always a trade-off between what what is "best" and what is "easiest". Thus the natural tendency of some executives to view their organization as a newtonian machine and intervene in the organization as mechanics fixing an engine, without really understanding any of its underlying knowledge-related mechanisms. But by doing that, they are like a brainsurgeon that would have no interest in Neurology or Physiology, i.e. a menace. Karl stressed that "milking the organization", i.e. the productivity focus, is incompatible with the Intangible Asset Management mentality required to foster innovation...

Thus, quick and effective execution, which is of central concern to executives in many companies, is not about authority and chain of command. It's about making sure that every stakeholder has reached a personal understanding of the strategy and agrees to it. This means that unless the entire decision-making process is taken all the way down to the personal level, execution will not take place.

Nice quote to remember:

The purpose of information is description; the purpose of knowledge is action"

What I learned: KM people are often mislead by conceptions on how people make decisions. Only novices make rational decisions based on logic. Experts make decisions based on emotional feelings based on reference mental models that they have developed over time. So our KM practice should be primarily geared at building libraries of personal mental models. Hence the importance of simulation games and ontologies.

John Seely Brown, Chief Scientist at Xerox, probably gave the most inspiring keynote presentation, if we can evaluate by the number of disciples who gathered around him (including myself) after his speech

It was called -rather obscurely- "The Knowledge Management Dilemma: Do We Confuse Knowledge with Knowing?"

Now regardless of the title, and regardless of the fact that I downloaded this same presentation on paper last year, this was truly a high learning moment. Especially the gorilla experiment. At some point in time, we were asked to count the number of times players with white shirts passed a basketball in a video. Most of the subjects achieved a fairly accurate account of the passes (13 I think), but only a few saw something more important. A person in a black gorilla costume walks right into the center of the action, beats his chest and moves off. We were so engrossed in the task at hand that we couldn't see the gorilla. So the point is that our models and attention create blinders that limit what we see. In our daily lives, gorillas are moving through our field of vision and we fail to see them.

Here are some of the things I will remember:

• The cartesian concept "I think, therefore I am" separates the act of thinking and the act of doing, and induces an epistemology of possession, and knowledge becomes more important in mental models than knowing which cannot be separated from acting.(Personal comment: Hence the vertical thinking of many hierarchies regarding knowledge, which is equated to intelligence in the minds of many executives)


• A community of practice becomes necessary when capturing the context is essential to understand a piece of information. Otherwise a database is enough. In a CoP, every member learns how to read each other, and to be aware of what happens to each other. If an isolated individual cannot see the gorilla, chances are that someone in the community will and will raise a red flag: "Hey, guys, look at the gorilla!". And they will see it. So the first purpose of a CoP is to raise the level of awareness of its members about what is actually going on.


• In a thriving community of practice, the trust level between its members can be so high that communication becomes almost implicit. The sense of presence of the others is so high that high bandwidth modes of communication like face-to-face meetings or videoconferencing are dropped in favor of low-bandwidth modes like IM or e-mail. This illustrates that "Practice provides the rails on which knowledge travels" (John Seely Brown) and that "Trust is the bandwidth of communication" (Karl-Erik Sveiby)


• Knowledge is like a tree, with roots (tacit knowledge) and branches and leaves (explicit knowledge). Tacit knowledge is learning to be. Explicit knowledge is learning about. Transferring "best practices" is like transplanting the tree because you need to carry the tacit knowledge as well. So you disembed, carry away and reimbed. But oftentimes, the tree dies in the process, especially if it's a big one.


• On its assembly lines, Toyota does not "transfer best practices". It specifies all work at the lowest possible level and documents everything to the extreme. Every day, small experiences are made to improve the process, and the outcome is predicted in writing before the experiment takes place. Then the actual outcome is compared to predictions. When the two are consistent, it means that effective learning has taken place, and the improvement is accepted. The result is in JSB's own words a "ballet choreographed by a thousand learning cells". That's why if something goes wrong in the ballet, the assembly line line is frozen so that the learning context does not get lost.


• 95% of IT expenditure in companies supports business processes. Almost nothing goes into the social fabric. People acting as knowledge hubs, spanners or brokers in the company are by and large ignored. An advice for companies laying people off: Do not knock your knowledge hubs off!


• Changing mental models does not involve logic. All actions aiming at changing the way people see things must speak to the heart and guts. They must be about creating evocative objects like stories (storytelling). They must fuse memory and imagination. JSB talked about a movie shot at two untrained people trying to use the latest Xerox copier. They couldn't make it work and their conversation verbalized their inner thoughts as they were trying to figure it out. This movie was presented to Xerox executives. Initially they thought these two guys were slow witted, until they were told that they were two very respected professors. Only then were they able to accept the idea that the user interface of the copiers needed to be completely revisited.

Dan Holthouse and Richard Cross from Xerox gave the first keynote presentation I attended. It was called: "Knowledge at work: developing the ideal knowledge sharing environment". Like Karl Wiig, they started with the result of a study showing that the invisible capital of a company is explained #1 by its innovation capabilities and #2 by the talent of its people. The quality of management only comes in the third position. They reminded us things like:

1. 80% of what you learn in a company does not come from the intranet but from impromptu meetings around the coffee machine or in the cafeteria, or from casual office visits.


2. 2/3 of the America workforce are mobile workers, meaning that they spend at least one working day per week away from their office.


3. Place and working environments have a major importance in job satisfaction, and represent a relatively low share (5-8%) of the overall cost of doing work. And yet many companies engage into cost-cutting programs regarding real estate that drive morale down. (I agree from personal experience. It might be stupid, but it is depressing to be asked to move from a big office with a nice view to a smaller one with no light, because, no matter what the circumstances or the rationale are, you can't help feeling humiliated)

The core of the presentation was a space-centric view of working conditions as opposed to a simple workplace view. The idea is that employees perform best when their physical space (office) is adequate, but also when their information space (IT), organization space (collective layout?) and cognitive space (personal style) are adequate as well. Dan and Richard then went into some detailed examples of technologies and concepts for "great places to work".

What I learned:the importance of displays in future working
space: paper-like displays, large wall-mounted displays, curved displays...

What I question:Dan presented a Xerox survey showing the relative behavior of high-performing people vs low-performing people with respect to knowledge sharing habits. It appeared very clearly that high performers had a far more positive behavior regarding knowledge sharing activities than low performers. For example high performers value knowledge and people, whereas low performers value information and documents. High performers don't see the organization as an obstacle to knowledge sharing, whereas low performer complain about the working conditions preventing them from sharing their knowledge; etc.etc. Despite the fact that this is indeed an interesting survey which I could use for political purposes inside my company ;-), I tend to question the causality link. Is it really because they are high performers that the have a more positive attitude with respect to KM, or is it because they are trusted as high performers that they are given the "breathing space" needed to engage into knowledge sharing activities? In my company, managers who are considered as high potential are actually given the opportunity to network and learn from one another in a very casual mode within the corporate university. For those who do not belong to the club, restrictive policies are applied with respect to communication and travel, and even with respect to time management. So it's more like a chicken and egg (if not a Catch-22) situation.

The KnowledgeBoard's workshop on Personal KM was organized as an Open Space with Lilia (left)
and Ton as moderators. I left the workshop after 20 minutes and went to Karl Wiig's keynote presentation instead. I came back towards the end to hit the buffet and listen to the conclusions, but I was sort of hijacked by Ton who asked me to try and make sense of the post-its pasted on the wall. I had a hard time because I was totally out of context, and some of these post-its made absolutely no sense to me. So summarizing and rephrasing was almost impossible. What came out of it from my perspective though was a very fuzzy definition of what was actually meant by "personal knowledge management" asopposed to "corporate knowledge management", as well as the key concern of nearly all participants about how to manage cultural change in their respective organizations.

We had dinner on November 9. at an Australian restaurant downtown with a few cockroaches running here and there. I had a great conversation about EU politicy regarding KM with Peter Heisig, as well as a good chat with Andy Boyd (left, looking sideways enviously at Lilia's Touchscreen PC) about helping communities at Shell.

What I learned:

1. For helping communities, a few large communities is better than many small ones. Typically 10 communities of 1000 members each rather than the more traditional 100 CoPs of 100 members each. Andy explained that helping communities need to be large if you want questions to be answered in a matter of one or two days. Typically at Shell a 1000 member community generates 5 questions a day and 15-20 answers within 24 hours.


2. Face-to-face meetings are not necessary. It was deemed impossible to organize anyway. It appears that the Shell community is sufficiently bound together so that the trust level is high enough.

Later on, I learned during the conference that:

1. the Sitescape collaborative environment is used to support those communities.


2. a project is currently under way to develop ontologies as a collective mental model to pull all these Q/A together and make them reusable within a single knowledge base. Typically, it involves automatic clustering and highlighting of some of the meaningful words of questions and answers to facilitate navigation across similar Q & A of the knowledge base.

Personal comment: If this is the case, I wouldn't be surprised if some subgroups of the larger community became responsible for designing the sub-ontology within a certai domain of knowledge, thereby generating practice groups within a larger helping community.