Preserving and Using Institutional Memory Through Knowledge Management Practices (2007)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

102 The sources in this appendix are organized around the main knowledge management (KM) topics. For the most part, they concentrate on specific practices organizations can develop to manage institutional memory long-term. Sources were selected for their practical nature, their currency, and inclusion of case studies. The focus is not on theoretical works that justify or provide the underlying philosophical framework for institu- tional memory practices’ rather the focus is on a variety of con- temporary high-quality sources that, collectively, provide a window into a wide range of KM understandings and prac- tices. To some extent, sources were chosen because they are reasonably accessible to ordinary practitioners in state trans- portation agencies (STAs) regardless of the sophistication of the information services in their individual agency or avail- ability of the services of academic or research libraries or doc- ument repositories. A wide variety of types of sources were included. Among the selected choices listed here and else- where in this report are standards, guidance documents, a few scholarly works, business journals, outstanding websites, and selected writings from the trade and popular press. In short, a fairly large number of references have been provided that focus on the practical, and although no specific practices are recommended, it is hoped that these summaries and the many pragmatic ideas found in them will stimulate thinking in STAs. KNOWLEDGE AS AN ASSET In 1969, Peter F. Drucker in the book The Age of Disconti- nuity: Guidelines to Our Changing Society laid out the over- all management concept of the value of knowledge and knowledge workers. Thanks to him, the overall management concept—that of the value of knowledge and of knowledge workers—has been around for some time. Near the end of the book, after a complex, insightful discussion detailing what he saw as profound changes occurring in the nature of work and the global economy, he gave his view of the future: To make knowledge work productive will be the great manage- ment task of this century, just as to make manual work produc- tive was the great management task of the last century. The gap between knowledge work that is managed for productivity and knowledge work that is left unmanaged is probably a great deal wider than was the tremendous difference between manual work before and after the introduction of scientific management (1969, p. 290). This focus on the management of knowledge work was reinforced by Drucker himself in 1997, nearly 30 years later, when he wrote in the Harvard Business Review: Management will increasingly extend beyond business enter- prises, where it originated some 125 years ago as an attempt to organize the production of things. The most important area for developing new concepts, methods, and practices will be in the management of society’s knowledge resources. . . . Predictions? Not at all. Those are solely the reasonable implications of a future that has already happened (Drucker 1997). Ikujiro Nonaka and Hirotaka Takeuchi (1995) in The Knowledge-Creating Company: How Japanese Companies Create the Dynamics of Innovation, described myriad exam- ples from Japanese firms to support their view that “knowl- edge creation has been the most important source of their [Japanese firms] international competitiveness” (1995, p viii). They articulated the idea that KM is an organizational respon- sibility. Knowledge itself, for them, is an organizational asset and therefore to be managed not just by specific individual employees or selected managers, but also by the entire orga- nization at all levels. In the introduction to their book, they assert that . . . in the dominant Western philosophy, the individual is the principal agent who possesses and processes knowledge. In this study, however, we shall show that the individual interacts with the organization through knowledge. Knowledge creation takes place at three levels: the individual, the group, and the organiza- tional levels (1995, p. ix). They wrote within the context of Japanese manufacturing companies, but their book not only lays out a succinct ratio- nal for KM, but also details numerous specific practices and applications, especially using information technology (IT), but in a way that incorporates and enables better human inter- actions and content management. KNOWLEDGE MANAGEMENT AS A BUSINESS PROCESSES In Thinking for a Living (2005), Thomas H. Davenport, who began writing on the topic of KM at least as early as 1993, focuses not so much on knowledge as an asset, but on the knowledge worker and the knowledge work process. He warns not to simply impose KM on top of existing business processes. He cautions that few knowledge workers have the spare time to record lessons learned or to share their exper- tise with co-workers. Although he is, in his words, a “big supporter” of the idea of KM, he believes that knowledge behaviors must be “baked into” the job (2005, pp. 62–63). Davenport discusses knowledge workers, their antipathy to formalized processes, love of autonomy, and tendency to value their knowledge and not to share it easily. He describes interventions, measures, and experiments managers can use to make knowledge work more productive; knowledge work processes; organizational technology for workers; how to APPENDIX K Annotated Literature Survey on Knowledge Management Practices

103 develop individual workers’ networks and learning; how to set up the physical environment for best performance; and generally how to manage knowledge workers. de Holan, Phillips, and Lawrence (2004) describe how “failure to capture” is a form of organizational forgetting. To avoid loss, information must be captured from individuals and made institutional, a process that involves a range of activities to routinize, codify, and store knowledge. First is the process of making knowledge explicit, which they term “articulation.” Then it must be communicated, which the authors called knowledge institutionalization. They offered the following advice: • Avoid heroes because resident experts whose knowl- edge is not managed properly make the organization vulnerable and stifle institutional learning. • Structure the work to replicate knowledge among individuals. • Some individuals prefer not to share their knowledge— they prefer to try to remain indispensable. Management must correct that, perhaps by instituting a bonus system that rewards sharing. • Link the old to the new. Organizational knowledge is interconnected in complex ways; new knowledge has to fit in to the existing structure. The authors also discussed how to “forget”; learning is a double-edged sword because we can learn bad habits that are actually counterproductive. It is easy to learn, but hard to learn the right things. Make concerted efforts to break rou- tines and practices based on unquestioned assumptions. Restructure organizations to dismantle interconnections that make change difficult. The organization must have an adap- tive ability to recognize change and discontinuity and incor- porate it continuously. Gordon and Grant (2005) argued that the literature has insufficiently addressed the issue of power in KM. The prac- tice of KM can be enriched by research that recognizes how the struggle for power within an organization may influence the KM system and the KM system will, in turn, influence the power struggle. The authors performed very wide-ranging searches on KM in the ABI/Inform database and analyzed the resulting articles. Of the more than 4,000 articles found, only 138 contained the keyword “power.” Of those, most dis- cussed “knowledge as power,” and only four treated the rela- tionship between KM and power as problematic. With the knowledge-as-power approach, possession of knowledge implies possession of power. The authors contend that the power-as-strategy approach has been ignored by KM litera- ture and therefore does not reflect the potential problematic relationship between power and KM systems. They call for KM research that explores a more in-depth approach to power in organizations (pp. 1–9). Gupta and Govindarajan (2000) documented the expe- rience of Nucor Steel with KM. They wrote that although there is widespread awareness of the economic value of creat- ing and mobilizing intellectual capital, most companies do not realize the potential. Thinking of a sophisticated IT infrastruc- ture as the be all and end all of KM is mistaken. Effective KM depends on the social ecology of the organization as well. IT does play a central role because it is the only mechanism to connect effectively large numbers of geographically dispersed people; however, success relies on how people use the knowl- edge. Effective KM must create and acquire new knowledge and share and mobilize it throughout the organization. They discussed knowledge pathologies; for example, “knowledge as power,” which may result in a limited trans- fer of knowledge because the owner wishes to control it. They reported that Nucor’s success at knowledge creation sprang from superior human capital, high-powered incen- tives, and a high degree of empowerment. As KM business processes, Nucor: • Instituted continuous, on-the-job multifunctional training; • Acquired knowledge because every employee is driven to search for better ways to make steel and steel-related products in teams that included operational, engineer- ing, and management staff; • Retained knowledge by reducing the work week rather than the work force during difficult times; • Made performance data visible within the company; • Encouraged sharing best practices; • Implemented incentives that ensured that one individ- ual’s superior performance would have minimal impact on his or her bonus; • Exploited IT to develop rich transmission channels to transfer both codified and unstructured knowledge; and • Had a policy of keeping plant size at between 250 and 500 individuals to build social community and open communication. The authors also contended that investing in codifying and making tacit knowledge explicit can have high payoffs. All knowledge transfer occurs through a limited set of exchange mechanisms: exchange of documents, conversa- tion and coaching, and transfer of people and teams. The mechanisms must be tailored to the knowledge type being transferred (pp. 71–80). Hammer et al. (2004) argued that it is incorrect to focus on individual knowledge workers’ productivity. Rather, Hammer asserted in this three-part article that the goal is to get more out of the organization by improving the perfor- mance of the end-to-end business processes. The task is to eliminate non-value-adding work. It cannot be done by fiat and technology is not much help. Such work needs to be designed out of the process. Knowledge workers may be negative at the notion of process, seeing it as an intrusion on their creativity and individuality. However, that is a mis- understanding of process, which is not about the routinization

and bureaucratization of work but “about positioning all individual activities in the larger context in which they are performed” (p. 16). Leonard-Barton likened modern workers to cave dwellers. The cave walls are computers, but people still like story- telling and learning in communities. However, we do not have the time and technology cannot create shortcuts to the most valuable kinds of knowledge. She recommended that managers look for what she called “deep smarts” and pro- tect “shallow smarts”—their promising but inexperienced workers. People with deep smarts draw on a huge store of tacit knowledge. Computers help, but coaches (experts) are needed to help relative novices through guided experiences. They help their protégés build experiences through guided practice, observation, problem solving, and experimentation (pp. 16–17). Davenport, in the same article, described what we know about knowledge workers. Scientists and engineers need to work very near each other to be able to exchange ideas. Soft- ware engineers need process and practice combined. He also described what we do not know, but asserts that without too much additional effort, companies can resolve the issues. He also recommended experimental design as a tool to measure productivity, performance, or satisfaction of knowledge workers (pp. 17–18). Halladay and Burk (1998) presented a knowledge “prob- lem” for the reader to solve with FHWA. They reviewed the resources available including training, technical assistance, technical committees, and tapping into the academic com- munity. They went on to discuss what was in 1998 a rela- tively recent approach: KM. They described KM approaches, such as taking advantage of electronic communication and knowledge sharing repositories. The knowledge cycle is find/create, organize, share, and use/reuse, with the central theme being communication. Successful operation of KM requires balance between top management sponsorship and proactive participation of individuals. They detailed the role of “Knowledge Manager.” KM involves more than databases and networks. The key is to create a culture that is collabo- rative and open to innovation and knowledge sharing within and beyond FHWA. They described how communities of practice (COPs) can foster that culture (pp. 32–36). Timo Kucza (2001) did the KM community a major favor when he wrote his Knowledge Management Process Model. Kucza, of the Technical Research Centre of Finland, took the approach used in software process improvement projects. His goal was to create a KM process model that would allow a common understanding of KM and a possi- ble framework for analyzing KM (p. 13). Making liberal use of charts, graphs, tables, flowcharts, etc., he modeled the entire process in a highly structured manner familiar to computer scientists, software engineers, etc. It is a model of straight thinking. 104 Many authors recommend that an organization embark- ing on a KM program begin with a knowledge audit. A paper by Liebowitz et al. (2002) is a good resource for this process. It focuses on how an organization can determine what knowl- edge is needed, what is available and missing, who needs the knowledge, and how it will be applied. Although not quite a recipe book, it is a succinct, easily readable guide. The paper includes a useful case study of a knowledge audit in a behav- ior health care organization called ReVisions Behavioral Health Systems. It ends with KM recommendations for Re- Visions, based on the results of the audit (pp. 1–16). Scott Thurm (2006) wrote amusingly but seriously about the knowledge of a package-delivery courier. He asked the courier whether it was more efficient to start at the top of a very tall building and work down or begin at the bottom and work up. The courier had a definitive answer, based on experience—“it depends . . .” Thurm then wondered how long it would take the next courier, assuming the current one leaves, to figure out the best approach. In a more serious vein, Thurm wrote that businesses have dozens of solutions for managing knowledge, but most involve technology, and he argued that such a focus obscures the crucial human issues in learning and teaching. People need to have opportunities to trade the vital tricks of their trade. However, informal learning is limited in scope. Thurm referred to Dorothy Leonard-Barton’s recommendation to use a guided experience by which employees with extensive knowledge pass it on with the help of a coach. He cited Raytheon’s missile systems unit in Arizona as an example of this technique. Thurm discussed how companies struggle to create the critical mass in a database of knowledge such as might be found in repair databases for technicians. In some cases, companies “seed” these databases until finally critical mass is achieved where more people use the system, which leads more people to contribute. He reported that Xerox’s repair database now contains approximately 70,000 sugges- tions and saves the company millions of dollars per year. According to Thurm, executives surveyed by Bain & Co in 2005 increased their use of KM systems last year, in spite of their misgivings about their effectiveness. HUMAN RESOURCE AND ORGANIZATIONAL DEVELOPMENT IN KNOWLEDGE MANAGEMENT Dorothy Leonard-Barton (1995) articulated in Wellsprings of Knowledge how cultural climate affects an organization’s ability to take full advantage of knowledge as an asset. She moved us beyond the IT and document management para- digms common at the time to include the human resource dimension to knowledge-based organizations, and to clarify that, in her view, all managers and level of managers, as opposed to mostly IT managers, have responsibility for KM. Using scores of examples in the United States and interna- tional technology companies, she showed how a cultural cli- mate open to KM approaches can directly improve decision

105 making and, therefore, the chance for success. To fully express her vision of KM, Leonard-Barton, reached for a metaphor. A wellspring, the source of a stream, sustains life within and beyond the riverbanks or, by becoming damned up or polluted, denies its existence. The most useful wellsprings are constant, reliable, and their waters pure. As flows of water from such well- springs feed the biological systems around them, so in the same way, flows of appropriate knowledge into and within companies enable them to develop competitively advantageous capabilities. Managers at all levels of the organization are the keepers of the wellsprings of knowledge. To them falls the responsibility for selecting the correct knowledge sources, for understanding how knowledge is accessed and channeled, and for redirecting flows or fighting contamination (Leonard-Barton 1995, p. xiii). David Gilmour (2003) asserted that KM technology is not working. In his view, most organized corporate information sharing is based on publishing, which he saw as a failed par- adigm. Even the most organized effort collects only a fraction of what people know, plus the process is time-consuming and expensive, and retrospective, in that it captures what was use- ful in the past. He recommended a shift from the publishing model to a brokering model based on collaboration manage- ment. The challenge with the brokering model is to connect people who should be connected. Software can be used to sift through e-mail, network folders, and other sources to identify common information threads, which can then alert people about their shared interests. Parties can connect or confiden- tially decline. His experience showed that brokering works best when people can share information they want when they want, and the more privacy privileges are extended, the more people will choose to share (pp. 16–17). TRANS-DISCIPLINARY NATURE OF KNOWLEDGE MANAGEMENT According to Chang-Albitres and Krugler (2005), KM emerged in the mid-1970s, beginning with the implementa- tion of database management software, moving into data handling in the mid-1980s, and in the 1990s developing enterprise-wide database systems and document management systems. It emerged as a business process in the late 1990s. They list the disciplines having the most profound effect on the development of KM concepts as organizational science and human resource management, computer science and man- agement information systems, management science, psychol- ogy, and sociology (pp. 3–4). Note that the report also contains KM website and software reviews. A.B. Cleveland (1999), somewhat earlier, broadened our understanding of KM as being more than an IT issue. He acknowledged that we live in an information economy and that companies are investing increasingly greater levels of resources to maximize the benefits of KM. He cited a broad array of software companies and consulting firms offering technology-based products and services under the KM list- ings. Although these information technologies contribute sig- nificantly to the management of corporate knowledge, Cleve- land argued that KM is fundamentally not an IT issue and operating as if it were totally misses the point. He explained that IT projects often fail to meet expectations because the connection between the IT project and the expected business outcomes is not clearly articulated. The process is much more complex than the delivery of an IT system. IT systems are enablers. No IT can succeed without addressing the orga- nizational change issues. He recommended hiring good peo- ple, creating an environment where intuition, exploration, guesswork, and creativity can be exercised, and connecting people with one another, with customers, partners, and the world (p. 28). A. Cohen (2006) highlighted the role of the librarian in organizing research collections, without which the organiza- tion cannot be successful. Optimization of content influ- ences organizational growth. Organizations must integrate the library into its KM strategy, because in physical or digi- tal form it is the organization’s main research repository. Tra- ditionally, library collections were designed to support knowl- edge sharing. The librarian’s core competency is really at the heart of the matter—information can only flow freely when it is controlled and organized. Librarians manage the content for COPs to maximize COP effectiveness. Digital libraries created by COPs need maintenance and vetting. Physical libraries are spaces for meeting meet face-to-face and are a digital hub. El-Diraby and Kashif (2005) discussed the “semantic web” as an extension of the current web in which informa- tion is accessed based on meaning, not hypertext. Ontology is one of the main components of the semantic web. Ontol- ogy is a mechanism to categorize and classify domain knowl- edge items or information into interrelated concepts. The architecture they detailed classifies highway concepts into processes, projects, products, actors, resources, and techni- cal topics. The architecture was built as an extension to the e-COGNOS ontology and is the first ontology aimed at cov- ering the whole highway construction domain. It was devel- oped to allow for future expansion and customization of terms and relationships. This paper is valuable for illuminat- ing the need for categorizing and classification—typical domains of the librarian and the engineer, both of whom have much to learn from each other in this area of KM! Kenan Jarboe (2001) stressed the importance of a suitable IT infrastructure as a production as well as a consumption tool. He discussed the importance of “local knowledge,” sometimes known as “indigenous knowledge” (p. 7), and pointed out that this tacit knowledge is usually transmitted orally, experiential rather than theoretical, learned through repetition, and constantly changes. It has the advantage of speed. However, it is also generally location-specific, which limits its value. Capturing it is an important economic activ- ity. One way to do that is by geographically clustering people

(p. 8). He described the World Bank’s indigenous knowledge program by means of an Internet-based database. Jarboe con- tended that the social network, which included ad hoc inter- actions, was responsible for Silicon Valley’s success early on, as compared with Boston’s Route 128 geographical area, which did not have the same social network (p. 9). According to Jarboe, KM is often described using two tracks: the “IT Track” and the “People Track.” Although the IT Track is relatively new and evolving rapidly, the People Track is much older, more fragmented, and slower to incor- porate into the KM management process. He discussed four broad objectives in a KM project: • Creation of knowledge repositories, • Improved knowledge access, • Enhanced knowledge environment, and • Management of knowledge as an asset (p. 11). Success, Jarboe maintained, is a combination of IT tools and organizational techniques. IT tools include data captur- ing, document scanning and retrieval programs, expert sys- tems, data mining tools, and collaborative work software. However, Jarboe warned that IT tools should not take prece- dence over activities, including as face-to-face approaches, conferences, workshops, meetings, reports, or papers. He welcomed “e-government” initiatives, but also cautioned against simply imposing best practices from one location to another. Organizational changes may be needed, as con- trasted with IT approaches (p. 20). F. Mihai and J. Robertson (2003) described how Main Roads Western Australia replaced an old information system with a new Integrated Road Information System (IRIS). Prob- ably the most interesting and useful aspect of this paper is their description of specific KM principles and the practices related to those principles that were actually applied in the develop- ment of IRIS. The practices ranged from the development of a software solution to managing a large amount of documenta- tion, handling a diversity of users, and overcoming the human reluctance to change old work patterns. They clearly showed the multidisciplinary skills required for this project. Wallentine et al. (2000), Department of Computing and Information Sciences at Kansas State University, presented recommendations for technologies with the potential for the development and management of Kansas Department of Transportation’s (DOT) information systems. As one might expect, the recommendations fall in the area of the IT aspect of KM, including recommendations for object models, languages, and tools; network technologies; Visual Basic scripts for workflow processing in a network-based environ- ment; and a data warehouse structure to support research. Because this report is approximately six years old as of this writing, it is not that useful from a technical standpoint in the current web technology environment. However, it is highly useful for nontechnologists trying to get an understanding of 106 the nature of the work the IT team must accomplish to support KM processes. CHARACTERISTICS OF SUCCESSFUL KNOWLEDGE MANAGEMENT PROGRAMS Cross and Baird (2000) reported on their project on organiza- tional memory. They examined learning from 22 projects in professional services, financial services, and manufacturing organizations. These case studies were mostly from Fortune 250 organizations. They learned that organizations remember lessons from the past in various ways: • In the minds of employees and the relationships employ- ees tap into, • In repositories such as computer databases and file cabinets, and • As embedded in work processes or product and service offerings that have evolved over time. They contended that managers can improve performance by deliberately developing organizational memory and using the growing stores to guide organizational activities and deci- sion making (p. 70). The first step for managers is to deter- mine which experiences are worthy of learning from and then determine ways to maximize their inherent learning potential. Finally, the knowledge must be embedded into the organization. They discussed examples of how the U.S. Army and com- panies such as Chrysler, Ford, and Analog Devices go from individual experience to organizational knowledge. One exam- ple is the use of after-action reviews; they detailed how they are executed. The authors discussed how personal relation- ships can be turned into organizational know-how. They described the process by which British Petroleum uses peer reviews to tap into the knowledge of group members. They described how COPs are used by the exploration division of Shell Oil (pp. 69–72). Cross and Baird also asserted that organizations that suc- cessfully leverage IT to support organizational memory must: • Have the technology, policies, and procedures to ensure that reusable materials are screened by panels of experts and are entered rapidly into distributed databases so others can benefit; • Seek to leverage the knowledge contained in databases using technology that enables dialogue; • Provide structured learning processes so newly acquired knowledge can be integrated into daily activities; and • Embed the knowledge constantly into databases, work processes, support systems, products, and services. They used examples from a variety of consulting firms and also described how the U.S. Army uses an information exchange to promote collaboration and learning. At Chrysler,

107 product development teams capture lessons learned and trans- late them into suggested modifications to the product devel- opment process. Xerox created a distributed database to cap- ture knowledge about how to fix and service copiers, with rewards and potential for assignment to the coveted “Tiger Team.” Organizations must do more than just accrue and store knowledge (pp. 72–77). Davenport and Glaser (2002) found that the key to KM suc- cess is to implant specialized knowledge into the jobs of highly skilled workers, and the best way to do that is to embed it into the technology that knowledge workers need to do their job. That way KM is no longer a separate activity requiring addi- tional time and motivation. They compare this embedding process with just-in-time delivery systems that revolutionized inventory management. They admit that embedding knowl- edge into everyday work processes is time-consuming and expensive. They described a case study of a physicians’ order entry system at Partners HealthCare in Boston. Partners’ approach built on other key work processes as well, including an on-line referral and medical records system. All these sys- tems draw on a single database of clinical information and use a common logic engine. The process is difficult because infor- mation in the database cannot be untested or obsolete. There- fore, these solutions should be used only for the most critical processes. To justify such a system, a measure-oriented culture must be present. IT professionals must know the business as well as technology. The authors briefly describe similar proj- ects at Hewlett–Packard, Dell, Xerox, and GM’s Vehicle Engineering Centers (pp. 107–111). Davenport et al. (1998) studied 31 KM projects in 24 com- panies. In these case studies they identified eight factors that characterize a successful project. To summarize (p. 1): • The project involves money saved or earned. • The project uses a broad infrastructure of both technol- ogy and organization. • The project has a balanced structure that while flexible and evolutionary, makes knowledge easy to access. • People are positive about creating, using, and sharing knowledge. • The purpose of the project is clear in language common to the company’s culture. • The project contains motivators. • The project may use multiple channels to transfer knowl- edge, such as the Internet, communication systems, and face-to-face communication. • Senior managers support the project. The authors identified four broad types of objectives for these KM projects: • Create knowledge repositories • Improve knowledge access • Enhance knowledge environment • Manage knowledge as an asset (pp. 1–2). Many projects had multiple characteristics. Success indi- cators included: • Growth in the project resources • Growth in volume of content • Likelihood project would survive without support of par- ticular individuals • Some evidence of financial gain. Projects were more likely to succeed when they used both technology and organization. Some important factors for this dual approach were: • Pervasive desktop and communications technology; • Establishing a set of roles and organizational groups whose members have the skills to serve as resources for individual projects; • Finding the right balance for knowledge structure; • The repository has to have structure, including categories and key terms; • Identifying someone to control decisions about knowl- edge structure; and • Allowing for continual evolution of knowledge structure. Some other important management aspects were: • Establishing a “knowledge friendly” culture; • Avoiding “hero” mentalities or the strong need to be seen as very creative; • Using language common to the company culture; • In some cases excluding the term “data” to ensure that raw, undistilled information did not get into the repository; • Framing the project in business terms understandable to employees; • Instituting incentives and motivational aids that were not trivial and were long-term and tied into the overall evaluation and compensation system; and • Senior managers had to send the message that the KM and organizational learning are critical to organization, pro- vide funding and other resources, and had to clarify what types of knowledge were most important (pp. 44–54). Michael Zack (2003) found that the knowledge-based organization pays attention to the application of existing knowledge and the creation of new knowledge. He used two companies as case studies. Holcim Limited, a cement, aggre- gates, and gravel company headquartered in Zurich, Switzer- land, is highly decentralized, but it reorganized its company to develop, identify, transfer, and apply strategic knowledge among all its entities worldwide. Although its main product has remained unchanged for approximately 100 years, it is clearly operating as a knowledge-based organization. In contrast, Zack documented an unnamed company as hav- ing not leveraged the information and expertise residing within the company as a whole. Although its product, economic

forecasts, is a “knowledge” product, it did not have the pro- cesses in place to make it a knowledge-based organization. Zack maintained that the knowledge-based organization is a collection of people and supporting resources that create and apply knowledge by means of continued interaction. Such an organization seeks knowledge where it exists. Organizations that are truly knowledge-based stop worrying about “who works for whom and focus instead on who needs to work with whom” (p. 69). Knowledge communities at Holcim tran- scended formal boundaries. The company also made invest- ments to learn from customers. In contrast, the unnamed company did a good job of extracting information from out- side the organization; however, it could not surmount the boundaries raised among the many mini-companies it had created (pp. 68–69). Zack described a case study of Polaroid. Although it had a strong culture of sharing knowledge, that knowledge was focused entirely on analog products. The KM activities were not aligned with the company’s strategy, and eventually the company declared bankruptcy. Organizations have to close the knowledge gaps, both externally and internally. Holcim engaged the entire organization in determining more effi- cient, sustainable, and environmentally friendly processes; the unnamed company never made the link between knowledge and strategy (p. 69). The knowledge-based organization holds a knowledge-oriented image of itself. It takes knowledge into account in every aspect of operations, including how it is organized, what it does, where it locates, whom it hires, how it relates to customers, and its image. Zack also documented in the same reference a brief case study of Buckman Labs as a company with a knowledge perspective. Zack compiled key actions managers need to take to turn their organizations into knowledge-based organizations (pp. 69–71): • Define the organization’s mission and purpose in terms of knowledge. • Define the organization’s industry and position within it in terms of knowledge. • Formulate strategy with knowledge in mind. • Implement knowledge-management processes and struc- tures that directly support the company’s strategic knowl- edge requirements. • Transform the company into a strategic learning organization. • Segment the company’s customers and markets not only on the basis of products and services but also according to how much can be learned from them. • Treat the cost of learning as an investment, not an expense. • Rethink the business model. • Take human resource management seriously. • Reinforce the organization’s mission through coordi- nated internal and external communication. 108 SPECIFIC KNOWLEDGE MANAGEMENT PRACTICES Bob Boiko (2002) focused on content management. In his view, content management is important because it makes e-business real and workable, it is an antidote to today’s infor- mation frenzy, and it can create and manage pieces of infor- mation and tag them with all the information one needs to fig- ure out what they are worth (p. xxxiv). Part philosophy, part handbook, he answers for us: • What is content? • What is content management? • How does one actually create a content management system? • What is the metadata framework (logical design) that is needed to tie all the content together? • How does one actually build a content management system? This excellent source balances the emphasis of many authors on the human or organizational aspects of KM. In this glimpse into his thinking, he described his own organization, the Institute for Advanced Metatation, as a system of institutes; schemas; taxonomies; and methods to acquire, remember, and deliver facts and stories on demand. Job titles include hierar- chists, indicists, and associationists (pp. 922–923). Boiko’s book is useful for practical day-to-day applications as well as for stimulating the imagination. M. Burk (1999) explained the concept of KM to the trans- portation community, using the experience at FHWA. He argued that the transportation community has always put great value on sharing of knowledge, including an emphasis on the continuous gathering and sharing of information through informal, person-to-person contacts, conferences, and the like. KM, he asserted, offers the opportunity to improve the effec- tiveness and efficiency of this sharing process. He described the “knowledge cycle”: find/create, orga- nize, share, use/reuse. He described the tasks of knowledge managers: • Indexing and cataloguing new information as it comes in. • Serving as information brokers by assisting people to obtain the information they need, and • Advocating for knowledge-sharing practices. Burk wrote that much knowledge can be made available through the Internet; however, KM is more than databases and networks. He described COPs, which are professional net- works that identify issues, share approaches, and make results available to others. The “output” of the COP—the research papers, technical briefings, product evaluations, identification of experts, good practices, and so forth—is made available by means of the knowledge repository. An important task of the COP is to identify knowledge gaps and create new knowledge as needed to fill it.

109 He directed the reader to the FHWA website on rumble strips as an example. Burk asserted that an organization does not need a wholesale transformation of its culture to imple- ment KM. Some individuals are proprietary about knowl- edge; some managers fear loss of control; some staff mem- bers resent, at least initially, the perceived extra effort. Burk recommended examining rewards and recognition programs. He reiterates that implementation of a KM program is not a one-time project, but an evolving process. Burk (2003) described COPs as networks of people that identify issues, share approaches, and make results available to others. COPs can exist solely within an organizational unit or can cross divisional and geographical boundaries. They can even span several different companies or organizations. How- ever, they all have a core group of participants who provide intellectual and social leadership. Burk described how COPs differ from work teams: teams are formed by management and report to a boss, have a defined membership, and specific deliverables. COPs can be voluntary, usually have longer life spans than teams, have no specific deliverables imposed, and are largely responsi- ble to themselves. Burk reported that FHWA is facilitating COPs. He main- tained that their value will increase as a relatively large per- centage of FHWA’s technical and operational staff near retirement. He warned against • Too much official scrutiny, • Unsuitable IT systems, • Unsupportive reward structures, • Lack of legitimacy, and • Inadequate budgets. He discussed an example where a group of quality coordi- nators had been meeting formally a few times a year to discuss best practices and share information. However, they realized they needed to meet more frequently; therefore, they started a virtual COP. A COP does need maintenance, and it may be dif- ficult to integrate new people. Burk pointed out that technical support is needed including: • Web-enabled software for online discussion, • Document sharing and storage, • Community member information, • Group e-mail lists, • E-mail notification of new information, and • Online meeting spaces. He asserted that even with a positive corporate culture and good technology in place, it still really is the COP itself that determines its own success. Bush and Tiwana (2005) discussed collaborative knowl- edge networks, which they described as peer-to-peer digital networks connecting individuals with relevant experience to their peers who need it. Individual users frequently abandon such networks; to be sustaining, knowledge networks must be “sticky.” They define stickiness as “the users desire to continue using a knowledge network system” (p. 68). Networks must enhance the interplay of people, relationships, and systems with careful attention paid to intersections of IT and human psychology. They traced three key drivers of stickiness: • Individual relationship capital, • Individual user reputation, and • Personalization. Additionally, individuals with a long history of using the network tend to continue using it; however, the intensity increases only marginally over time. They described peer-to- peer networks using the “café” metaphor, as opposed to repos- itories, for which they used the “library” metaphor. There is a need for both, but their research focused on the former. Increasing stickiness requires that the network be designed so that users can perceive the costs of not using the system (pp. 68–71). John Carroll (2004) wrote that attention or, as he put it, “heedfulness” is the first ingredient in making a KM system capable. He asserted that: • Reporting systems are an institutionalization of heed- fulness. • Storytelling is an important way to capture knowledge. • Knowledge-developing COPs with specific boundary- spanning or bridging practices are useful for capturing knowledge. • Incident investigations and root cause analyses not only include techniques for looking beyond the immediate, but also for conversations with shared purposes. • Mixing people of differing occupational and educational backgrounds and cognitive styles can lead to informa- tional diversity. He discussed the stock versus the flow models of KM and emphasized that both models are needed (pp. 129–130). “Stock” refers to codified repositories of explicit infor- mation and know-how including: • Databases, • Procedural manuals, • Drawings, and • Planning documents. The stock model works especially well for routine opera- tions and problems. The key issue for the stock model is where the knowledge resides. However, most problems are local and contextual; there- fore, additional knowledge is needed and that’s where tacit

knowledge comes in, which he called the “flow” model. That model relies on such practices as: • Liaisons, • Job rotations, • Temporary exchanges, • Informal contacts, • Broad personal networks, and • Strong connections between individuals and groups. Coffey et al. (2004) described a case study of how concept maps were used in a knowledge-retention study at a nuclear power plant. The technique was used with a nuclear physicist over a 2.5 day site visit. A final wrap-up visit incorporated tan- gibles from the work into training materials, policies, and pro- cedures, etc. Concept mapping has proven to be an effective means of capturing an expert’s key concepts of a knowledge domain. The experts for the study were selected by plant man- agement based on a survey of employees who held critical positions and would shortly be eligible for retirement. The actual techniques used to implement the interview sessions are described in detail. A detailed knowledge model of critical knowledge and activity maps was created to capture the basic activities and relationships associated with a job. The author encourages the reader to consult this paper for its step-by-step guide to how to achieve concept mapping with a single expert. DiCesare and Demers (2000) reported on reengineering the bridge inspection process, eliminating all non-value added steps, and enabling the new process with a knowledge management-based inspection solution that integrates an enterprise document management system (EDMS) and an electronic forms system. They described the two distinct infor- mation worlds: structured and unstructured. The challenge is to envision a physical KM infrastructure that bridges these two worlds. They described how the EDMS, built on a relational database model with built-in workflow, enables team members to coordinate and control the document life-cycle process, from creating to archiving. Documents are stored as objects. Version control is assured. The resulting KM system required a culture change and shift in work methods to be successfully integrated. This KM implementation can be achieved by adopting the methodology of process reengineering, including a rethinking of jobs, orga- nization, structure, management systems, values, and beliefs. Maureen Hammer (2005) reported on the Virginia DOT Knowledge Management Office’s effort to establish COPs. By the end of June 2005 there were 10 such communities, with four more under development. Hammer provided myriad use- ful details about how specifically these COPs operate. No two communities developed similarly, but they all provided shar- ing of lessons learned. She noted that each COP has an exec- utive sponsor to ensure that participation is supported and that the Virginia DOT will get a return on its investment. One of the COPs was surveyed on what value they had gained from 110 their COP and from the support of the Knowledge Manage- ment Office. Values included: • Ongoing improved communication; • Improved processes and/or integration of people and ideas; • Lessons learned, best practices, and effective process model for use elsewhere in the Virginia DOT; • Effective meeting facilitation; • Effective communication support; • Collection of useful information; • Analysis of information; • Integration that respected differences among group members; • Neutral perspective; • Access to decision makers; and • Better understanding of how to increase collaboration. Madanmohan Rao (2004) wrote a wonderfully rich intro- duction to the book Knowledge Management Tools and Tech- niques: Practitioners and Experts Evaluate KM Solutions titled “Overview: the Social Life of KM Tools.” In it, he asserted that the concepts and practices of KM are approach- ing mainstream adoption, and that the focus must now be on practical applications of tools and technologies, which are essentially people-centric. IT appears to be a supportive enabler in his scheme, but is not in itself a tool or technique. Rao’s tool categories clearly show his understanding of the trans-disciplinary nature of KM and how KM as a business management process flows across all levels and work units of an organization. We encourage the reader to consult the wealth of useful ideas in this book, which contains 32 chapters writ- ten by numerous authors providing case studies of the use of KM tools within real organizations, with examples from business, non-profit, and government. William M. Snyder and Xavier de Souza Briggs (2003) provide an excellent overview of COPs. The authors doc- ument federal experience with COPs, explain how federal agencies can cultivate COPs, and describe how COPs are implemented. They highlight the FHWA Rumble Strips Ini- tiative in a case study and provide reflections on other case studies. This very useful guide is especially applicable to government agencies. Russell C. Walters and Lifeng Li (2003) described the development of an electronic reference library (ERL) that pro- vides a virtual library for construction design and management of state highway projects in Iowa. The ERL is a very large document which uses hyperlinks as navigation tool (p. 132). L. Yu (2005) reported in his synopsis of How to Support Knowledge Creation in New Product Development: An Inves- tigation of Knowledge Management Methods (2004) that researchers Anja Schulze and Martin Hoegl found that man- agers are familiar with a range of KM techniques and rate

111 some higher than others. The 10 highest ranked for sharing knowledge were: • Informal events, • Experience workshops where team members review completed projects, • Communities of practice, • Project briefings before beginning new projects, • Expert interviews, • Best practice cases, • Knowledge brokering by a third party connecting knowl- edge seekers to resources, • Reports documenting positive and negative experiences on projects, • Databases, and • Professional research services. According to Yu, the researchers assert that the effective- ness of any KM effort must balance how often users draw on the particular activity as measured against how much the method actually contributed to knowledge sharing or creating (p. 5). (Note: unfortunately the author was unable to locate the original working paper by Schulze and Hoegl, but decided to include the synopsis anyway.) Michael Zack (1999) asserted that explicit knowledge is becoming more important in organizations. He suggested that organizations face a fundamental challenge in determining which knowledge should be made explicit and which should remain tacit. Certain knowledge might not be articulated because to do so might be politically incorrect or not cultur- ally legitimate. If knowledge is made explicit, it can be rou- tinely integrated and applied. Some knowledge may be inherently inarticulable; a balance must be maintained. Man- agers should not accept tacitness blindly. Process knowledge is more valuable at the explicable level; when imagination and flexibility are important; perhaps the tacit form is better (pp. 46–47). MEASURING KNOWLEDGE MANAGEMENT EFFECTIVENESS J. Liebowitz (2005) asserted that part of the possible failure of KM initiatives, or management’s skepticism towards KM, is the result of the inability to develop metrics to measure KM success. Because KM deals with intangible assets, metrics may be more difficult to obtain than from other assets. How- ever, for-profit and not-for-profit organizations are developing such metrics. Liebowitz’s article contains a highly useful chart detailing KM performance measures developed by the U.S. Navy. He detailed key factors as to why organizations embark on KM initiatives under the general categories of adaptabil- ity and agility, creativity, institutional memory building, orga- nization internal effectiveness, and organizational external effectiveness. He concluded that fuzzy logic is a reasonable approach to explore for determining KM effectiveness mea- sures (pp. 36–39).