Wednesday, April 19, 2006

Knowledge management problems in a large organization.

I've been thinking about knowledge management lately. I thought I share some of the ideas I have.

The intended use is in the organization I work for. It is a large government ministry. The personnel consists mostly of academics, together with support staff functions, like finance, personnel and IT. We mostly make policies and laws, but don't execute them. (Others have to work with the mess we create) So all in all it is not too different from what you would see in the private sector in think tanks, consultancies etc.

Work revolves around the processing of information. There are developments in government, politics, society, courts, etc. to which we react. Reactions consist mostly out of rules, regulations, information campaigns, subsidies, taxes. The actual execution of those is mostly the work of other organizations. So my work consists of projects (that can run indefinitely) and going concern issues, which can linger dormant for a while, until somebody in parliament sees an item on the news. Information in, information out.

The problem
Big problem in an organization like ours is getting access and distributing information that exists in one part to other parts. It is amazing how hard it is to first of all find out who is responsible for certain topics. Of course we have lists, who does what, but finding them is not always easy. Some departments publish them on the corporate intranet and with the new Google Search Appliance, we can actually find them. But finding out what their projects entail, the intended output, project plans, drafts, introductions into the subject, is quite simply impossible.

E-mail is the tool of choice
Work in projects is done through mail. Recently for a project I made summaries of the opinions of various stakeholders. All the project members mailed the summaries to each other, but how other people should get access to it? I don't know. Could it be of interest to others? Probably. Same goes for a list of all the research conducted financed by the ministry etc. All this information is somewhere, but getting it depends on finding the right colleague at the right moment. The hundreds of professionals create so much information on so many topics and so often they explain the same thing, multiple times, it is almost amazing we get any work done.

One would think that if e-mail is the tool of choice, it would be easy to find that one e-mail that contains the required information. Of course not! First of all we use Outlook. Microsoft should be beaten over the head for not properly letting people search quickly in all their mailboxes. The IT-department makes things worse by requiring people to have a maximum mailbox size of 150 megabytes. Amnesia by regulation! Desktop search programs (Lookout and Google) are only used by a very limited pilot group to whom I belong fortunately. It has been a blessing, it has made searching 3 years of mail for e-mail relevant to a court case a breeze. Instead of performing a regular lobotomy, it lets the memory grow manageable.

Why is e-mail such a success? Isaac Garcia gave the following reasons and I agree with him:
  • Email is Easy To Understand
  • Email is Universal
  • Email is Accessible from Anywhere
  • Email can be Personalized
  • Email is Manageable/Configurable
  • Email is Searchable
  • Email is In Your Face
  • Email Just Works
The Solution
So now I'm looking for the solution. The solution would ideally be the following. It would have to be a tool that facilitates collaboration and as a result of that collaboration yield information that is available and relevant to other people. So a collaboration tool that has a byproduct knowledge management.

Document Management Systems?
The solution is not Document Management systems. I've been trying to use a Document Management System now for some months. It has been pushed from the top down into the organization. It is supposed to replace the group directories on a central disk. It is wrong in so many ways, both the product and the implementation, just read and cry:
  • It's not easy
  • Everything requires multiple steps.
  • The dictated structure doesn't follow users expectations
  • Only the archivists are allowed to change the tree. Every change is certainly one e-mail. Sometimes more.
  • It supports the archivists, not the policy advisors.
  • Just dropping stuff in it doesn't work. Every item you put into it requires a description, a location and a name. Average time certainly a minute. Try archiving just the important e-mails of a day (that adds up to at least 30 minutes a day).
  • Structure comes through a tree hierarchy, not a weblike structure.
  • The servers underperformed, breaking search and more
  • The way it works is different for each application of the MSoffice suite.
  • There are no direct bonuses in storing information in this system.
  • It doesn't encourage people to store everything in it. It only encourages to store official documents, that need an official number.
  • Collaboration is quicker done through sending an e-mail and later editing manually than through using the system.
Basically it is only useful to archive important documents There were three good design decisions:
  • Everything is open to all.
  • There are multiple ways of working with the software: Outlook, stand alone and through "Save as.."-dialogues in Office applications
  • All incoming physical mail is scanned into the system.
Intranets are for staff, not for line.
The intranet is stale and not up to date. This is a common complaint for many corporate intranets. The main reason the professionals don't update the intranet are:
  • It is extra work. The information is not the result of me doing my work, it is the result of me doing extra work.
  • It is something you do for the rest of the organization; not for your project, not for your direct colleagues, but for the 'others'.
  • There is no return for doing it. It is an investment where all the returns seem to fall somewhere else and you hardly if ever get something back for it.
  • The medium is the message, so you try to tune your message for the medium. It's not a simple cut and paste operation.
  • It is made hard, because you have to jump through many hoops to get the right to edit pages and the structure you can edit is rigid. (example: a diary (blog) I wrote during one week I wrote in Word. I sent it to a secretary, she published it.)
  • The person that is supposed to update the page is dead, gone, overworked, ill, retired, uninterested or all of the above.
  • It is not an intranet, it is a corporate almanac. So all the troubles of publishing a paper almanac apply.
So all in all the professionals whose work the intranet should support are not using it to support their work, nor to support the work of others. The only ones that do make use of the intranet are the staff departments. They publish information so that the professionals don't bother them for every small thing like forms, personnel information etc.

Online collaboration tools
I have very great misgivings against these tools, though it has been years since I have actually used them. They are all very task oriented and generally very hierarchical. Furthermore they often exhibit very cumbersome and clumsy interfaces. Worse still, they generally shield all information in the workspace from the rest of the organization, requiring extra work to publish the information.

My current hope is in Wiki's like Socialtext and Jotspot. I have been fiddling with free trial accounts now for these and I hope to post some first impressions later on. But here are my reasons to think of these wiki's:
  • They center work on a topic around a group of webpages
  • They are easy to use. Socialtext is just a double click on a page
  • They open up information to the entire organization through simple searches
  • Information entered into them for the benefit of the project group is immediately also of benefit to others. So when doing my job, I unintended also help others
  • They enable sending e-mail to and from pages, enabling e-mail repositories and lists of useful links on the relevant page.
  • By sending an e-mail to the relevant project page, you add both metadata to the page and to the e-mail.
  • They are free form, but can be structured
  • If one co-worker doesn't update his page, because of time constraints or just being dead, others can.
  • They can be about such highly critical information as: Best restaurants in Berlin, travel suggestions to Kiev, the latest law and its implications, biographies of important people, a list of insultants, the next project meeting or the office Christmas party, without requiring a central command and control structure.
  • They don't assume where knowledge is in the organization.
Later on I will post some impressions of Wiki like software for these purposes. Current ideas are Socialtext: cool and easy, but expensive. Jotspot: cool, loads of features, cheap. Confluence: interesting, bit techy. Plain standard Wiki: Too primitive, too much like an intranet that anybody can edit.