Knowledge Management in a (Coco)Nutshell

August 31, 2002

What is Knowledge Management?

My definition is that Knowledge Management is “a way to find relevant data about something”. Note that’s not the relevant data, just relevant data and that’s not the thing it is just something.

One of the central issues for implementing KM is an understanding that it is intended to be qualitative not quantitative. This is a tool for finding documents you didn’t even know existed. Unfortunately it turns out to be just as good at finding specific documents and as a result gets perceived as something it is not, and is often evaluated on the purely qualitative term of how many of the knowledge workers ‘own’ documents are returned. Add to this the fact that every KM system has at least three types of user with different expectations of what it will supply them and it can be hard to get consensus on what needs to be done, in what order and to what degree.

One key feature is the browser-based nature of nearly all KM systems. Although a KM system can be written as a desktop application, like Word or Excel, its only benefit over a browser based system is that it feels slicker than a Web application which refreshes all the time. Even this benefit can be reduced if the Web developer invests time and care, and a few cunning tricks, in their development. On the other hand a Web application is available from any browser, anywhere. There is no need for rollout to desktop PCs, instead all the Information Technology department needs to do is e-mail a link to the address the system is on. Similarly, upgrades are achieved by just changing the address of the new system to the current address. This provides for a system that is available anytime, anywhere and can be easily changed and upgraded.

Who Uses KM?

Fee Earners

Fee earners are the real users of the system. Faced with the need for information, especially late at night, the weekend or when away from the office, they need a simple system they can type a couple of words into, and receive a set of suitable documents that contain what they need to know. They may also need to use it for research when finding out about a new area or case – here they need to be able to isolate the areas they are interested in quickly and simply and then manipulate the results that come back, often in their hundreds, so they can get a good overview before drilling down to the ones they want. One use for these systems would be in discovery, when presented with a couple of hundred thousand documents as part of a case, they could be loaded into the KM system and this could be used to analyse and interrogate the data. In this case, features like automatic summarisation and classification come into their own.

Knowledge Workers

Knowledge workers use the system in a radically different way from fee earners. They are concerned with populating it, checking it and retrieving specific items from it. It must be simple to add value, initially to controlled, internal data sources like document repositories and to control what is available. In the right hands this can be a great tool for encouraging knowledge sharing. Some firms use the number of documents accepted and approved for use in the KM system as part of their employees’ annual review procedure or make awards for top contributors. Ultimately the real benefit for these people is that the Fee Earners will get the information they need from the system rather than contacting them directly, thus reducing the demands on their time and allowing them to handle a greater quantity of knowledge. How knowledge workers add classifications and grade or approve knowledge is one of the major distinguishing factors between KM systems even more important than the software used.

Information Technology Workers

The Information Technology department may see KM as the next step of their Document Management strategy. After all, their current system gives them full-text searching and comments. It’s just a case of adding another interface and a couple more data sources, isn’t it? In practice, while it will cope with structured, internal knowledge, that is all it can cope with. IT is just as concerned with documents as law practice, but in the former the manuals, notes and procedures on the products are used alongside their producer’s Web sites. An IT department can benefit greatly from a KM system. We use one in house with all the logged issues, with our software, its manuals, training guides, product definitions, and those of its third parties available through it, and being programmers we have delighted in using our own product as a tool to help us improve it. However, most of the time IT is concerned with the nuts and bolts of running the system: the production of usage statistics and reports, adding new data sources, backup and disaster recovery, and delivery of new versions to the users.

What is “Knowledge”?

I am frequently surprised when I receive a firm’s conformation of requirements. So far every one has included different sources and ways of using them. It is gratifying that nearly all of them can be included in a well-designed KM system.


In an attempt to simplify the number of sources I divided them into simple categories. Whilst a data source can be categorised in many ways, the classifications I find useful are:

  • metadata or content based
  • controlled or uncontrolled
  • internal or external

The basic component in a KM system I refer to as a document. This may not actually be a physical or electronic document but is a set of data that can be considered as a document-like thing. Each thing returned may have one or two types of data, content and metadata.

Here are some common data sources and how they fit this document paradigm:

  • Electronic Documents

    • Always have content (ie the text)
    • Always have metadata (eg title, author, last updated, etc); is either held as document properties within the document or comes from a document management system
    • Are normally controlled because they are in a document management system
    • Are rarely uncontrolled because most electronic documents conform to some specification
    • Normally internal if they reside within the firm’s document management system
    • Sometimes external such as documents held on public Web sites

  • E-mail

    • Always have content (ie the body of the email)
    • Always have metadata (eg sender, recipient/s, date sent, importance, etc)
    • Are always controlled because they reside in an e-mail system
    • Are always internal because no one would ever let anyone else see the ramblings, jokes and banter that goes on in such documents

  • Database Records

    • Sometimes have content in the form of free form notes columns
    • Always have metadata in the form of all the other columns
    • Always controlled because database management systems require all data to conform to a schema
    • Often internal, as the database resides within the firm
    • Sometimes external, such as public subscription databases

  • Web Sites

    • Always have content in the form of the text on the Web pages
    • Often have metadata in the form of defined ‘tags’ which may or may not be visible such as the