Duane Degler - Knowledge Maintenance Strategies

Knowledge Maintenance Strategies: Gaining User Involvement

by Duane Degler

What follows is a summary of ideas and approaches presented at Knowledge Management 2001, London, England, 3^rd April 2001.

Introduction

Over the recent years, Knowledge Management has been gaining significant momentum within organisations. Much of the emphasis, particularly on the part of the suppliers who are wrestling with the evolving technology, has been on the acquisition and storage of information, and how it can be retrieved more flexibly.

But think about this: if I see an article or item of information that I think is interesting, I am likely to copy it for other people. A physical article might trigger my making and circulating 10 copies, of which three get read, three get thrown out, and four get put in a pile “to look at later.” Valuable, yes. Insightful, yes. Available? No. Taking up storage space and making other things harder to find? Definitely!

E-mail is even worse—it is so easy to forward items to people in my address book, that hundreds of copies of the information can proliferate. I looked this morning—over 600 items in my inbox, and that’s with a recent cleaning. If I spend only 6 seconds evaluating each one to keep or delete, that is still one hour of my time that I am unlikely to spend cleaning my inbox. And it still is not effective shared information. If someone else wants the information that I hold there, I can search for it, but everyone’s time is affected. It is easier and faster to have a conversation and rely on my memory of the information, than to have both the information and the conversation—the combination of the two being a richer sharing of knowledge.

I have seen shared network drives filled with documents and presentations, many gigabytes worth, stretching back over many years. It is easier today to store and access that material. It needs to become easier to review or remove that material when it is no longer appropriate to the organisational mission.

How can organisations work to ensure that shared knowledge remains relevant over time? One key way is to involve the person using the information. In this paper I explore:

User-maintained and supported knowledge (maintaining knowledge in the foreground).
Stealth Knowledge Management (maintaining knowledge in the background).

What Knowledge Needs to be Maintained?

For the purposes of exploring the subject of maintenance, I will focus on explicit, formal knowledge that is able to be accessed by more than one person, and thus has value to an organisation generally rather than just a single individual. The maintenance of dialog and conversational threads, though very important as a part of knowledge sharing, is a more complex subject for another day.

Knowledge is often only apparent through the actions of individuals, and the outcomes they achieve—you demonstrate that you “know” something by the actions that you take. Shared knowledge is communicating something that helped one person achieve a quality outcome, so other people can achieve the same—or better—outcome.

Maintenance, in this context, is making sure that the information that is shared and used continues to encourage successful outcomes. To promote user involvement in maintenance, there has to be a shared belief in the value of the knowledge, and thus the value of keeping it relevant and accurate. If it isn’t seen to be valuable, people won’t waste their time on it.

Active User Involvement in Maintenance

Users should actively get involved in:

Specifying the initial approaches to Knowledge Management, to ensure that the approaches agreed meet the real needs of the working environment.
Ongoing feedback on the quality and relevance of information at the points where they come into contact with that information.

1. Information Usability

In the same way that software is tested for usability, so should the approach to handling information be tested. Comparing information usability with traditional usability testing for software applications and interfaces shows a number of similarities and differences:

Activity	Traditional software usability	Information usability
Paper prototypes	Using use cases and scenarios to ensure that the interaction/task flow in an application matches real-world needs, and that the application’s use is easily understood by people of varying abilities and experience levels.	Using use cases and scenarios to help the users articulate how well the information supports the tasks that align with it, also expanding on the understanding of real-world conditions such as information volatility, relevance to task, interruption, pressure to respond, that may affect ability to give feedback.
Heuristic evaluation	"Expert" evaluation of design approaches based on standard criteria, to identify where an application may not meet the needs or the users, or introduce error/risk.	Reviewing standards for information mapping/representation, mechanisms for soliciting maintenance feedback and testing the validity of how that feedback is responded to by the organisation and by content owners.
Usability tests	Formally testing the design of an interface or application using high-fidelity representations of what is actually intended to be built.	Formally testing formats, representations and the methods by which the user finds, uses and provides feedback on knowledge.
User acceptance	Simulating the actual environment of use and functional requirement to confirm finally that the finished developed product meets real-world requirements.	Less relevant and useful than some of the other items above because of the dynamic nature of content. It is still very important to engage the user in both migration of existing content, and “launch” of new facilities, to promote understanding and use.

Along with the above activities to support the design of the information and the knowledge-sharing applications, it is also important to assign responsibilities for information/knowledge ownership for both creation and maintenance—this may require very different skills and temperaments, and so may not be assigned to the same person.

Work to refine the quantity and quality of metadata captured. This is a critical task—if it isn’t done correctly, it will create a huge maintenance burden. The “information about information” has to support the maintenance task. For example, metadata should include anticipated life of an information element, review dates, the authors, owner, further subject experts, and any relevant date where the information was revised or feedback received. Metadata should also carry things that can hook it into applications that provide relevant information embedded in the screen.

2. Capturing Feedback

The simplest and most common way of asking the users to participate in the maintenance of knowledge is to put simple feedback mechanisms on the page where information is displayed. There are a number of ways this can appear—general guidelines include:

Be consistent when asking for feedback—don’t ask for feedback on only some pages, because users want to give feedback on things that are important to them at the time.
Keep it simple—ask for the least amount of information that gives meaning to the feedback, so as not to overburden users.
Provide for the unexpected—leave open space for users to type comments or changes.
Ask for structured input—check boxes, drop-downs and other constrained input makes sure you get feedback that you need, increased the consistency of feedback (allowing measurement), and is faster for the user to complete (so it feels less of an intrusion).
Keep the request for feedback out of the way of the information, so as not to diminish the usability of the information itself (for example, a sidebar next to a column of information will be visible but can be ignored when concentrating on the information).

Make sure that the mechanisms for capturing feedback at the point of use do not conflict with the task or the environment. For example, I had a conversation with one help desk application developer whose product included a “rating” section on each page of information. As much as I applaud the recognition that this feedback could help the quality of the knowledge base, the idea ran aground on two points:

The users—help desk staff—are focused on talking to the customer and resolving complex problems, not on providing feedback by submitting a form. The staff who are willing to submit that form may not be representative of the whole user base.
The task of customer problem-solving may involve many different pieces of information, and the insight or resolution may not relate directly to a single item of information on the screen. So rating the usefulness of one item of information independent of the process and the synthesis of knowledge may produce a biased and inaccurate picture.

The answer may lie in a more holistic approach:

Closer integration with application usage tracking, where you use the application to help you interpret the value of use.
Ongoing analysis, interviews, questionnaires, etc to get feedback from the individuals during slow periods—this also encourages thoughtful reflection on the value of certain types of knowledge.
Using subject matter experts to clarify the feedback, and follow up if necessary.

Decide how much feedback you need. If you don’t intend to use it, don’t capture it! Users who are motivated to provide information are easily de-motivated if they feel they are being ignored or the contribution is not valued. The best ways to make sure people feel their contribution is valued are:

Use the information you receive, and make sure that the improvements are visible, or signposted.
Communicate what you have learned, and where you have learned it.

Incentivizing people to involve themselves in “knowledge quality” is a good thing. Incentives are those things that motivate people to become involved and commit their attention to something. This doesn’t necessarily mean bottles of champagne or big bonuses—research has shown that this is not always as effective as people think. The biggest incentive for involvement in knowledge maintenance is often the improvement in the available, useful knowledge itself.

Stealth Knowledge Management – Learning from Use

A useful concept I and some colleagues began developing a few years ago is “Stealth Knowledge Management.” When computers are used to carry out job tasks and make decisions, there is a growing record of the way decisions are made – embodied in the way the applications are used. The rules of decision-making are not artificially defined in advance, but based on actual performance—whether that performance is carrying out a task or seeking knowledge from a knowledge base. The rules are as dynamic as the experienced people who perform the tasks day-in and day-out. The capturing of this experience (“knowledge from actions”) can be valuable to less experienced people performing the same tasks, and can also help content owners understand the value of the information in their care.

The key factors of Stealth Knowledge Management are:

The data is a representation of the business knowledge of experienced users.
The knowledge is dynamic: it changes as the business context and application use change.
Decision-making patterns define rules that can be modified or weighted based on the qualitative outcomes of the tasks which generated the patterns.
The knowledge can be made available immediately when a decision is to be made, presented as recommendations for action.

For example, for one client the challenge was to receive requests for information from the public, and route these requests to the most suitable departments for providing the reply. The knowledge of how to interpret a request and identify the department(s) most likely to hold relevant information was rooted in the experience of one person who had been doing it for over ten years. When that person was away, routing virtually stopped. This is a classic problem in knowledge management, that could not be solved by informing or training a second person—the job circumstances were too dynamic. However, by tracking the nature of the decisions made when categorizing and routing, a dynamic picture of those decisions could be created. The rules could change as the types of information being requested by the public changed. The routing challenge and part of the solution are illustrated below.

Beginning to map the actual flow of requests over time.

Tracking the dynamic nature of requests, which affects the rules to be applied.

Application dialog designed to help users make decisions in their work.

"Tuning" helps users manage presentation of knowledge.

Using stealth knowledge management, the hidden business rules (tacit knowledge) become overt (explicit knowledge), while not becoming so rigid that they lose their meaning. The user remains in control of the use of knowledge, and becomes an active contributor without being distracted from the task or having to make an extra effort to participate.

Stealth knowledge management contributes to an ecological model for maintenance, where information is self-sustaining and begins to have a life of its own. The model is appearing primarily within communities of practice and online group interaction, where there are no formal rules for knowledge management, but the group’s contribution to the knowledge base and use of the information identifies the value—and again, this is done dynamically. This also allows the group or the information to die naturally when it no longer has value.

Conclusions

As mentioned above, active maintenance that involves users in the process requires the combination of:

Background techniques for dynamically identifying where information has value.
Foreground techniques for promoting direct user involvement and responsibility.

Without both of those elements, the burden for maintenance is probably going to fall back to a select few, whose job will be much harder. When users become actively involved, you begin to learn the real value that knowledge has to the people who use it.

Copyright, Reuse and Citation

The content of this article may be referenced with the appropriate citation information included (see below). I would appreciate your notifying me if you intend to use these concepts or images, as I am curious to know where they prove valuable. The entirety of the article must not be reproduced without written communication with Duane Degler at IPGems.

To cite the material, please include the following information. I recommend the format:

Degler, Duane (2001). Knowledge Maintenance Strategies: Gaining User Involvement. Knowledge Management 2001, London, England. Online: www.ipgems.com/present/knowmaint.htm.

Return to the top of the page