October 27, 2006
Authority, popularity, expert-ness - how to set your filter to get the right content?
Marry Hodder, CEO at Dabble, made a good point yesterday at the business blogging summit here in Seattle when she said that the Technorati rankings for authority are not really a measure of authority but a measure of popularity - and that measuring authority is not something that should be conceded to a web service but something that gets determined by the end user.
At one point or another everyone is struggling with how to filter content to find the relevant pieces of information. Some people believe that popularity may be the right filter. Others believe that you can only trust "real" journalists and should use that as your filter to get the right content. Some others still cling to the belief that only academic credentials, with publications and peer review, can result in the "experts" worth listening to.
The reality is that no one filter can work for all topics and at any point in time. If I am looking for customer feedback on a new product, I may not want the voice of the expert, and certainly have learned not to trust the voice of the "journalist" reviewer. If I am looking for information on cutting edge cancer treatments on the other hand, I may only trust academic types. And if I need information on a more popular topic, then popularity may well be the right filter - or ratio of blog posts to comments and trackbacks, or some other metric that determine how well a person is read, quoted, etc.
Now, another person may look at this in a whole different way. The bottom line is that the "right" filter is content-specific as well as reader-specific.
Making things even more complicated - the right filter is also time sensitive. If I have time to cull through a large amount of information then my filter may not be set as narrowly as when I am in a time-crunch.
So in a way, everybody has a personal profile that determines the right filters based on subject and moment in time. If somehow we could have web services that would match their results to my personal profile, then we would have a real cool solution.
It's really simple when you think about it...well...maybe not :)
January 24, 2006
Knowledge management - still an illusion
The Wall Street Journal had an article yesterday on the continued struggle that companies face with trying to pass knowledge from one worker to another (requires subscription).
The conundrum is that while people learn best from other people - a system that relies on employee word of mouth for knowledge transfer does not scale. In such a system, people only learn from those people they know, and in many cases there is just not enough opportunity for this type of knowledge transfer to occur (i.e., field repair workers, work at home folks, etc.).
Technology to help with knowledge management has had spotty track record so far - with the Wall Street journal citing an Bain study which ranked knowledge management tools near the bottom in effectiveness amongst 25 management tools.
Apparently some companies have had success with a brute force approach - by forcing people to document what they know in a central database. They claim a virtuous cycle with people contributing voluntarily once there is a critical mass of content in those knowledge repositories.
Perhaps the quote from Hadley Reynolds with the Delphi Group captures the real issue - "We don't necessarily understand enough yet about optimizing the conditions for knowledge work, even though we've been doing it for 25 years. Most organizations are still managing as if we were in the industrial era."
September 30, 2005
Is it time to revive Knowledge Management?
David Pollard over at How to Save the World has a good post on the old KM vs. the new KM. He summarizes the differences between the first wave and the second as:
"First-generation KM has vainly sought one-size fits-all integrated enterprise solutions, which are complicated to use and expensive to change, and which focus on content + collection; Second-generation KM must look instead to simple, lightweight, cheap, intuitive, stand-alone apps, which are easy to use, add or change, and which focus on context + connection. In the shift from first to second generation KM, the holy grail changes from cost savings to improvements in knowledge worker effectiveness."The article also contains a list of 23 human behaviors that impede the sharing of knowledge and collaboration, and how some recent organizational and technological changes do alleviate some of those impediments. The main message he conveys is:
"The challenges we face today in getting people to share what they know and to collaborate effectively are not caused or cured by technologies, they are cultural impediments. It's extremely difficult to change people's behaviours (they usually exist for a reason), so the solutions we find have to accommodate these behaviours, and these cultures, rather than trying to 'fix' them."While I buy most of what he's saying, I think that he is missing a few key points. There are two main reasons why KM has not worked in the past. The first one that in most organizations it was a top-down exercise with a disproportionate amount of "perceived" benefits for the organization vs. the individual (we will build a system to make sure that we capture all "your" knowledge if you walk out the door - or if we push you out the door). The second reason is that previous KM tools and processes (i.e., best practice teams, etc.) were never integrated with people's real work. That meant that KM became a "voluntary" extra-curricular activity - and guess what - most people don't do that.
For KM initiatives to work, they will have to be grassroots in nature (i.e., no taxonomy but folksonomy), and will indeed have to be based on lightweight tools (including Wiki's) that integrate with people's daily work. And most of the "perceived" benefits of the initiative have to be for the individual.