you are here: articles
Notes from KM Asia 2003First published November 11, 2003
The KM Asia 2003 conference was held in Singapore on 4-6 November. Thanks to the Ark Group, the conference organizers, I got a press pass to attend the conference. Here are my notes on the conference.
The conference hall was smaller and there were fewer vendors as compared to last year. Even the usual big banners and free magazines were sadly missing. I guess this was to be expected, given the economic battering the Asia-Pacific region has gone through lately. When asked the on the reasons for holding the conference, the business development manager of the Ark Group said, "We thought it necessary to keep the conversations going." And in that respect the conference was a success.
A large percentage of the audience, it seemed, had some KM project in place, and were attending the conference either to seek legitimacy for what they were doing, or to learn about what they ought to be doing. This was evident in the more-than-usual number questions asked at the end of some sessions. Those sessions that did not evoke such reactions were the ones where the audience was paying deference only to the perceived "expert-like hairstyle."
A couple of things were quite apparent at the conference:
- More speakers used stories in their presentations, which is always a good thing
- The audience asked more questions on KM measures
- More speakers talked about embedding KM in business processes (from where the measures can be ascertained)
- And many speakers used spidergrams to visualize KM measures
Here are some notes from key presentations.
Tom Davenport gave his presentation via a video link from Boston. His presentation was on embedding KM into business processes. The classic example, which he quotes often, is the Partners HealthCare case. The case goes something like this:
- Studies in the early 90s found high levels of medical error
- The errors were related to information overload. Tons of new information is added to the biomedical literature each year (60,000 articles are added each year)
- Instead of asking docs to keep abreast of constantly changing information, all the information is kept in a database which is linked to the process docs use to prescribe patient medication. So if the docs prescribe a medication, the system looks it up and compares it with the patient history and either confirms the prescription or recommends alternate ones when it finds discrepancies.
- The errors have reduced by almost 55% because the system now keeps track of all the changes and recommends them to the docs at the right time and under the right contexts
Tom's HBR article "Just–in–Time Delivery Comes to KM" describes this case at length.
Tom also commented on the importance of structure when content proliferates. His advise: understand content architecture, taxonomies, thesauri, meta-knowledge, and multimedia. Here Tom seems to promote the marriage of KM and Information Architecture.
The other illustrious Tom, Thomas Stewart, was present in the flesh and presented using only transparencies. His presentation -- Intellectual Capital and the Real 'New Economy' -- was based largely on his book, The Wealth of Knowledge.
Here are some key points from his presentation:
- Ideas are Capital. Everything Else is Just Money. (From Deutsche Bank Advertisement, Wall Street Journal, April 2001.) At the bottom of it all, its all about making money with knowledge
- We have to be focused on taking knowledge to the market and selling it over and over again (Never Sell Anything Once -- Art Buchwald)
- Knowledge is as fundamental and versatile as electricity; it can do almost anything you want it to do
There are 3 knowledge-strategy families:
- Instilled Knowledge (smart products)
- Distilled Knowledge (knowledge as a product)
- Black Box Strategies (knowledge services)
His intellectual capital model was as follows:
- Tangible Assets
- Intangible Assets
- Human Capital (e.g. a restaurant famous for its chef; if chef leaves, restaurant loses value)
- Structural Capital (e.g. McDonalds and the way they've tailored their fast-food business using both technology and know-how)
- Customer Capital (e.g. bar like Cheers, where everybody knows your name. The value here is in the customer interactions)
HSBC's Steve Ellis made a fantastic presentation. Unfortunately, I missed that presentation, but Patrick Lambe of Straits Knowledge (also check out his Green Chameleon website) agreed to share his thoughts on the presentation.
Who really tells the truth in knowledge management these days? Steve Ellis of HSBC took the prize of the whole event for gut wrenching honesty. He started by pointing out that his slides held no HSBC logo because he was presenting his personal views and not the official ones of HSBC. The reason for that cautionary clause became immediately transparent as Steve launched into a striking account of how his KM initiative at HSBC had not exactly failed miserably, but had stumbled badly. Battered, but still passionate about KM, Steve positively incarnated the ethic of knowledge sharing and learning from failure, in the most powerful KM presentation at a conference I've seen.
- Don't try to sell KM, sell the benefits
- The culture has to be ready for KM, so find a bit of it that is ready
- Winning hearts and minds is easy – getting action is tougher
- Nobody wants to hear about the long term – move fast and get quick wins
Steve also presented the most quotable version of the undercurrent of techno-scepticism present throughout the conference: "If another technology vendor tells me they can solve all my KM problems with their solution, I'll punch them."
Thanks for the write up Patrick.
The other presentations I found useful were related to taxonomies. Siemens presented their taxonomy for managing 50 million pages, and the National Library Board Singapore presented theirs for managing 100 libraries.
The taxonomies in both cases were quite simple and basic, which is true for taxonomies in general, but as both presentations rightly pointed out, it is the taxonomy building process that is the most difficult and complex.
Some guidelines were presented:
- Plan for the usage model (browse or search + browse)
- Develop Controlled Vocabulary
- Build Thesaurus
- Select Taxonomy
- Create/Edit/Update Metadata
I guess this was the IA connection that Tom Davenport was talking about.
For more information, check out Boxes and Arrow's in-depth treatment of Controlled Vocabs.
On the whole, KM Asia did manage to keep the conversations going. I'm pretty sure that the audience got a few actionable points to work on. The reason I'm so sure is because the "core group" of companies attended this conference, unlike the powerless types that attend e-learning conferences (at least here in Asia). These core group decision makers will take the lessons learnt from this conference and put them into practice. And the discipline as a whole will benefit from it.