Tuesday, 1 September 2015

Why less is better in KM

Last week I had a conversation with a knowledge manager, who explained to me the incentive system they use. My heart sank.

In this particular organisation, people are incentivised to share knowledge. For each article they publish they are awarded "points". If they accumulate enough points, then they can trade them in for a reward.

"Is there any quality control on the publishing" I asked? No, was the answer.
"Do you give any points for re-using knowledge?". "No"
The result was an ever-expanding supply of poor quality material, with no demand.

This reminded me of another client, which required each person to publish ten items every year.  In an organisation of 100,000 people, that was a million items per year, with no quality control, no filtering, no re-use, no guidance on what to publish; no strategy other than to flood the knowledge base with documents.

Both of these organisations were falling into a common dual-aspect trap;

  • incentivising Push (supply of knowledge) and paying no attention to Pull (demand for knowledge), and
  • incentivising publishing quantity at the expense of quality.

More knowledge is not necessarily better. 

There are many cases where too much knowledge, especially where this knowledge is poor quality, is counter-productive. 

  • A situation where a user is already bombarded by a "Knowledge firehose" will not be made better by turning up the flow
  • As I found in another client recently, a lessons data base which is overly full of poor quality lessons is a frustrating experience for the user. "We cant find what we need, and when we do, its not helpful" was common feedback, and users seldom visited the database a second time. 
  • Another client, again rewarding publishing, but with no structure and guidance, was proud to see wikis popping up all over the place, not seeming to see the problem in having 5 competing wikis on the subject of Knowledge Management, for example. 

Davenport and Prusak, in their seminal "Working Knowledge" point out that over-volumed knowledge bases do not work. 

"Volume may be the friend of data management" they wrote - "but it is the enemy of knowledge management"
Too much knowledge requires the user - and each user in turn - to  filter the volume of material to find the relevant and timely knowledge, and as the volume increases, the task gets harder and harder, and people just give up. It's easier to reinvent the wheel, than to find the correct wheel in a junkyard.

Go for "less and better"

The lessons from these observations are relatively simple, though perhaps counter-intuitive in the early stages of KM.

  • Incentivise quality (usefulness and utility) of published knowledge, not quantity
  • Introduce a synthesis step, so new knowledge does not accumulate like a snowdrift, but is used to refine what is already known
Aim for less and better. 

No comments:

Blog Archive