Standardizing Management of Knowledge

by John Berry

As the problem child of business intelligence, knowledge management (KM) has suffered an identity crisis for some time, which explains organizations' total lack of enthusiasm today for any technology or management strategy yoked to those two words. As I pointed out in a previous Executive Report [1], KM means almost anything, therefore it means almost nothing. Yet the cruel truth is that KM might just be the single most important management subject today given how the production, use (sharing), manipulation (creation, enhancement), and categorizing of information (knowledge) -- activities associated with KM -- are fundamental to the wealth creation of so many businesses. Rescue just might be at hand. KM is poised to extricate itself from the quicksand of its own contradictions, freed through a completely new way of looking at the subject that includes a concept historically alien to it: standards.

UNDERSTANDING KM

Andrew McAfee, a professor at Harvard Business School, has identified some patterns in the production, use, manipulation, and categorizing of information that may serve as the basis for a profound advancement in our understanding of how organizations might improve the productivity of their knowledge workers and accelerate innovation [2]. On a more practical level, this understanding would certainly improve the practice of knowledge management from what it is today.

This enhanced understanding is the result of observing how new technologies as well as new ways of using existing technologies improve knowledge processing. These usage innovations include corporate wikis and blogs, social bookmarking, and folksonomies, which I define later. Together, these Web-based technologies and the related approaches to using them potentially advance a new paradigm in understanding what KM can be -- a collaborative environment comprising technologies and management practices that will help knowledge workers get work done faster and more effectively than today. As Peter Drucker has argued, the productivity of knowledge workers is one of the greatest management challenges organizations face.

The components necessary to realize this new vision of knowledge management are at the core of our enhanced understanding. They could represent a set of management standards that would give some much-needed structure to the discipline of KM. Such standards, a mix of management practices and technologies, might actually draw practitioners' attention away from a focus on specific software packages called knowledge management software and toward a focus on how emerging technologies not tied to a specific vendor can be applied in the service of helping knowledge workers create, use, manipulate, and share information in their daily work. McAfee coined the acronym SLATES (search, links, authoring, tagging, extensions, signals) to describe the portfolio of technologies and behaviors necessary to deliver this vastly improved "knowledge worker" work environment [2].

UNDERSTANDING SLATES

The following section examines the technologies and behaviors inherent to the SLATES system.

• Search. As my previous report [1] made crystal clear, superior search functionality is central to effective KM and the effectiveness of the knowledge worker. Research about search functionality shows that people find hierarchies of structured information on a corporate intranet to be less helpful than a gateway into a robust search engine when they are trying to find information. This infers that organizations might do just as well to scrap the complicated navigation schemes of existing intranets and instead just put a Google search field on a Web page tied to the vast and chaotic store of information on company servers. Also, where the information is stored is less important than the ability of a search engine to access what a worker seeks.

• Links. Google's technology sits on the premise that hyperlinks between Web pages are an important characteristic for search engine indexing; pages and the links to other pages they contain reflect page or site popularity, and Google's search technology hones in on this. The practical implication for the enterprise is simply that organizations should take a grassroots approach to intranet construction; workers should be allowed to add links to intranet pages they believe are relevant and useful to the page in front of them. This not only improves search functionality but it also serves as a kind of knowledge sharing critical to effective knowledge management. The idea goes: "Hey, if the information on this page is useful, check out URL X and URL Y." Knowledge sharing's definition is broadened to include not just explicit information contributions but equally valuable suggestions to other knowledge sources the consumer might not know about. The next principle expands on this idea of a broadened definition.

• Authoring. In the historical parlance of knowledge management, authoring meant sharing a piece of information of some kind. This new construct shrewdly broadens the definition of authoring in the context of KM to include contributing information in the form of expertise, a comment, a fact, a link, or an edit; blogs are Web sites featuring running contributions from participants on a specific subject that can range from long-winded manifestos to just a link. Sharing does not explicitly mean providing big chunks of information per se anymore because of the emergence of wiki technology (the most obvious being Wikipedia). As a new kind of knowledge management platform, one of Wikipedia's core propositions is that anyone can contribute information or edit existing information. Correcting erroneous information can be just as important a knowledge contribution as providing the information in the first place. Broadening the definition of authoring in the context of sharing is good for KM; it helps us avoid thinking too narrowly about what sharing actually is [1].

  Research also points out that the processes of contributing, fixing, adding, and so on that are inherent in Wikipedia actually lead to high-quality information, even though it would seem that too many contributions might contradict each other and lead to chaos. In fact, this kind of collaborative environment has shown that contributors are able to reach consensus and improve on each other's work without disagreement and acrimony. This is for two reasons. First, anyone can erase others' work. Therefore, the incentive for any single author to deface a contribution with derogatory comments disappears, because someone else (probably the target of the invective) will likely erase what the defamer wrote. Everyone seems to be on their best behavior, because anyone can edit or erase any other contribution. If contributors want their work to survive, they had better do it seriously, professionally, and, ideally, accurately.

  Second, senior members of a subject community keep contributors from flaming others [2]. A culture of polite respect amongst contributors rules -- at least so far.

• Tagging. Categorizing information is devilishly difficult because information is so malleable. For example, suppose you want to categorize a body of knowledge around human anatomy. There are many choices. Should information be categorized alphabetically? For example, brain, heart, spleen, and so on. Or maybe it should be categorized by anatomical system? Nervous, digestive, so forth. Then you face the task of subgrouping the organs in any system. What about diseases of these systems? Should they be subcategorized in the hierarchy under the organ associated with it? Should diseases have their own separate hierarchy linked in some way to the organ hierarchy?

  There is no right or wrong way to organize stores of knowledge. Yet hierarchies are rigid and not easily rearranged once a scheme is committed to. If users do not understand the logic of the hierarchy, they might get too frustrated to find anything. The new wisdom emerging out of this confining reality is that organizations should not try to arrange knowledge hierarchically. Instead, let the consumers of the information place a descriptive tag on it and then categorize around that tag. This methodology is called a folksonomy, because people do the work of organizing vast amounts of information. This approach is the opposite of a taxonomy in which an expert in a subject domain decides on an organization scheme that users will or will not like. Folksonomies reflect the ways in which people relate many pieces of information to each other, rather than having those relationships predetermined by an expert.

  Social bookmarking shares this kind of grassroots approach to information categorization. At the Web site del.icio.us., members can share bookmark preferences with others. The site's technology opts for a looser approach to information categorization than a top-down hierarchical scheme.

• Extensions. This refers not to file extensions but extensions by reasoning, such as the technology in Amazon.com's buying suggestion functionality: if you liked this, maybe you will like. McAfee refers to a technology called StumbleUpon that works in a similar way. Surfers "stumble" from one site to the next,while the technology builds understanding of personal information preferences. Over time, the StumbleUpon technology will suggest only those sites users are likely attracted to. In a work context, the sites a user prefers are most likely those that contain relevant subject matter to his or her job. This kind of functionality can speed information searches and retrieval by helping users avoid dead ends and irrelevant surfing.

• Signals. Who has the time to constantly check Web sites for added or updated content? E-mail alerts are one solution, but people are overwhelmed by this overused technology --information gets lost in a pile of unread e-mails. Real Simple Syndication (RSS) is an example of a technology that helps people keep up with new information in areas of interest. Effective knowledge management must include solutions to keep users apprised of new information as it is made available.

SLATES is an excellent management standard for approaching successful knowledge management. This new way of looking at KM is fundamentally driven by technology and will require a manager's full understanding of its functionality in deploying it internally if the management of knowledge is really to improve.

Technology and its capabilities might influence how KM is practiced. This is precisely the opposite of how other popular management methodologies and techniques work. The Balanced Scorecard, Total Quality Management, Six Sigma, and Economic Value Added are all widely used techniques that have nothing to do with technology. Technology aids their use, but it is not inherent in the functionality of any of them. Perhaps KM was never a true management discipline, and it took innovations in the use of Web-based technologies not only to unclothe this sheep in wolf's clothing, but to clarify what knowledge management -- if we can be so daring as to use that phrase anymore -- has the potential to deliver. Only time will tell if organizations can succeed in translating this new knowledge into meaningful productivity gains of its knowledge workers.

REFERENCES

1. Berry, John. "Knowledge Management Goes to Market." Cutter Consortium Business Intelligence Executive Report, Vol. 6, No. 2, February 2006.

2. McAfee, Andrew. "Enterprise 2.0: The Dawn of Emergent Collaboration." Sloan Management Review, Vol. 47, No. 3, 2006.

ABOUT THE AUTHOR

John Berry is a management consultant with extensive experience in helping organizations execute IT management strategies designed to deliver measurable value from IT and other investments. He is the inventor of a portfolio of strategic planning and value analysis methodologies designed for senior managers, IT organizations, and the vendors that sell to them. He is also the author of Tangible Strategies for Intangible Assets and Offshoring Opportunities: Strategies and Tactics for Global Competitiveness. Visit his Web site at www.nostrategies.com. He can be reached at vision@nostrategies.com or according2jb@earthlink.net.