0 évaluation0% ont trouvé ce document utile (0 vote)
1K vues8 pages
Why have so few knowledge management (KM) systems met or exceeded expectations? Simply put, customers of KM systems are not getting what they want, need or were promised. This article examines the role of KM systems in behavioral change. It presents an analysis of the 'warehouse' model underlying most KM systems today, and contrasts that model with a more 'customer-focused' design that focuses on driving more productive behaviors.
Why have so few knowledge management (KM) systems met or exceeded expectations? Simply put, customers of KM systems are not getting what they want, need or were promised. This article examines the role of KM systems in behavioral change. It presents an analysis of the 'warehouse' model underlying most KM systems today, and contrasts that model with a more 'customer-focused' design that focuses on driving more productive behaviors.
Droits d'auteur :
Attribution Non-Commercial (BY-NC)
Formats disponibles
Téléchargez comme PDF, TXT ou lisez en ligne sur Scribd
Why have so few knowledge management (KM) systems met or exceeded expectations? Simply put, customers of KM systems are not getting what they want, need or were promised. This article examines the role of KM systems in behavioral change. It presents an analysis of the 'warehouse' model underlying most KM systems today, and contrasts that model with a more 'customer-focused' design that focuses on driving more productive behaviors.
Droits d'auteur :
Attribution Non-Commercial (BY-NC)
Formats disponibles
Téléchargez comme PDF, TXT ou lisez en ligne sur Scribd
Why have so few knowledge man- This article examines the role of KM WAREHOUSE MODELS OF KM agement (KM) systems met or systems in behavioral change. More Where have KM systems gone exceeded expectations? specifically, it presents an analysis wrong? of the underlying “warehouse Simply put, customers of KM sys- Most KM systems are well model” of most KM systems, con- tems are not getting what they designed, implemented, and trasts the warehouse model with want, need, or expect. In fact, the supported. However, most KM a customer-focused model of KM, idea of having a formal customer systems are also based on a deeply and presents guidelines for how for a KM system is pretty rare in flawed “warehouse” model of to make a KM system drive new KM/IT circles. Yet, as with all prod- knowledge management (see and more productive behaviors. ucts, services, and systems, effec- Figure 1). By reexamining this Companies using this behavioral tiveness at meeting expectations is underlying premise, we can directly approach have shown significant determined solely by the customer. and significantly enhance the over- financial results, including: What do customers expect from all effectiveness of KM. A $2 million per week per facil- KM? Most KM customers expect a ity savings for a semiconductor In a warehouse model of knowl- KM system to enable them to adopt manufacturing company edge management, there are a set more productive and efficient of knowledge inputs, a storage and A $2,000 per week per restau- behaviors, which in turn should transportation capability, and a set rant increase in sales at a fast improve an organization’s financial of knowledge outputs. Based on food company performance. In the customer’s this underlying model, the over- view, better knowledge drives bet- A 66% reduction in training whelming emphasis of knowledge ter behaviors, which drive better time in a federal agency management has been on the stor- results. It is an obvious cause-and- These results certainly exceed most age and transportation portion effect relationship. Management customer expectations for a KM (in the form of databases, portals, of the knowledge is only important system! and search engines), with relatively to the degree that the knowledge being managed contributes to a performance improvement. Data gathering Transport Reuse • Documents • Document • Periodic access Unfortunately, most KM systems appear search engines to documents focus more on the managing com- • Personnel • Personnel or people profiles are search engines ponent than on behavioral change. created • Report writers As a result, customers of KM sys- tems only occasionally change Where is the Where is behavior, and thus KM systems knowledge? the applied knowledge? only occasionally produce the expected results. Figure 1 — The warehouse model of knowledge management.
Get The Cutter Edge free: www.cutter.comwww.cutter.com/consortium/ Vol. 17, No. 12 11
little attention paid to either the to that document [5]. So why will need to find the right answers inputs or the outputs. As a result, bother with the search engine? to these questions or run the risk of great KM warehouses have been Even if the content is good, the undermining user confidence in developed. customer of a KM system has a the quality of the content. hard time finding it. Unfortunately, they suffer from Knowledge warehouses also have many of the same problems as Ask people who have used a KM problems with outputs. Most items older, industrial warehouses. For system about their search experi- retrieved through a knowledge example, both types of warehouses ences. They almost always have a management system have minimal have difficulty with quality assur- story about entering a common context associated with them. ance of the inputs. In most KM sys- word or phrase into a KM search “When should a document or tems, there is little or no significant engine and being “rewarded” with piece of data be used?” and “How quality review of the content posted thousands or even tens of thou- should it be used?” are questions to a database or added to a com- sands of search results. From a that the user must answer for her- munity of practice (CoP) bulletin technical perspective, finding so self with, at best, minimal support board. Thus, the customer has no much related information shows from the KM system. This means way of knowing if the content is the “power” of the KM system. the customer has to figure out actually correct or useful. From the user’s perspective, how- what the retrieved content actually ever, the overwhelming volume of means. For example, a professional Similarly, both types of warehouses the information renders that infor- services firm had a huge store of have problems finding the desired mation almost useless. sample project proposals and item. This is in part because both reports. Unfortunately, they were types of warehouses try to keep all The volume problem is made virtually useless because whenever products or knowledge for any pos- worse by difficulties with content the staff tried to use one of them, sible need, which produces a huge, that becomes obsolete. Naïve users they had to make significant modifi- essentially unmanageable volume will naturally assume that a docu- cations to it so that it would apply of products or knowledge. In indus- ment in the KM system is correct, to their particular situation. The trial warehouses, this problem even if in reality the content is out content lacked sufficient context was solved with “pick systems” of date. After all, how are they to to make it useful. that direct a worker to the exact know the difference? In industrial location of an item. In knowledge warehouses, many of the stored In summary, the very model under- warehouses, the equivalent of a products are no longer in use, but lying most KM systems is funda- pick system is the search engine. they still take up valuable shelf mentally flawed and is therefore However, in both cases the sheer space and capital. In this situation, unlikely to meet customer expecta- volume of items to be searched most companies have sales or tions, let alone exceed them. hinders the searching process. marketing promotions intended to close out items and get rid of In a lecture at KMWorld 2001, CUSTOMER-FOCUSED KM excess warehouse stock. In knowl- Dave Snowden, a distinguished edge warehouses, however, obso- Fortunately, many of the problems researcher with IBM’s Knowledge lete material just clutters searches with traditional warehouses were Management Institute, stated and is rarely removed. How much eventually resolved with the advent that a user of a typical KM search of the content of a typical KM sys- of “just-in-time” manufacturing. By engine can only find the correct tem is obsolete? How often is it applying similar just-in-time princi- document if he already knows purged? What criteria are used to ples to KM, it too can become sig- which document he is looking for purge the content? Organizations nificantly more productive. well enough to narrow the search
How can the just-in-time concept Once known, the knowledge is be applied to KM? To answer that stored and transported to the user Some knowledge is far more question, let’s first look at how man- in a form that is consistent with important to an organization ufacturing environments apply it. substantial behavior change. Just-in-time manufacturing begins than other knowledge. As with the application of just-in- with a thorough analysis of a spe- time to the traditional warehouse, cific station on the production line. customer-focused knowledge All of the work processes and flows management significantly reduces knowledge that most requires are identified for optimal perfor- the volume of content required and management: mance. This analysis includes spec- substantially increases its utility. ification of exactly which part is Identify a group of key execu- The KM system is no longer trying needed from the warehouse, when tives and managers across var- to provide everything to everybody. ious functions. it is needed, in what form it Instead, it is focused on the most is needed, and how it is delivered. Ask them, “Which of the frequently utilized and/or critical The warehouse is then modified to processes of the organization knowledge. By delivering focused, meet those specific requirements as a whole are most critical to trustworthy knowledge in a timely by emphasizing efficient delivery of your success?” They will typi- and efficient way, the KM system the more frequently utilized and/or cally mention only two to three optimizes behavioral change. core processes, even if the critical items. As a result, the num- ber of items kept in the warehouse executives are in different Defining the Customer’s functions. If the cumulative list drops significantly because the Knowledge Requirements has more than two or three warehouse isn’t expected to cover Is all knowledge of equal impor- processes, ask them to priori- all contingencies. Instead, the ware- tance to an organization? The tize the list until you have two house’s operations can be stream- assumptions behind typical KM or three that everyone agrees lined and focused on meeting the systems seem to say that the are the most important. key manufacturing priorities. answer is “yes.” The real answer is Suppose there were a new, that some knowledge is far more A similar model can be applied improved method available for important to an organization than to KM using the customer as each identified process. Ask other knowledge. In order to have these executives and man- the equivalent of the manufac- customer-focused KM, the organi- agers whether it would it be turing workstation. This customer- zation must identify which knowl- important enough for them, focused KM (see Figure 2) begins edge is most important to the under these circumstances, with a complete understanding of system’s customers. to allocate two hours of every- the customer’s function, including one’s time to learn and apply how he uses knowledge, when he The following process has been the new knowledge and uses it, and what form optimizes very successful in identifying the improve the organization’s behavioral change. Once this is known, the leadership and KM teams, working together, can iden- Reuse Harvesting Transport tify and qualify the best sources • Determine what • Only gather • Only transport of the knowledge. The required people really need what is needed content when it • Determine when • Only gather it is needed knowledge is then gathered from and how they need it in the form it • Minimize central these highly qualified sources in a • Provide local control is needed control • Provide local control way that promotes realism and credibility, thereby creating cus- tomer confidence in the content. Figure 2 — Customer-focused knowledge management.
Get The Cutter Edge free: www.cutter.com Vol. 17, No. 12 13
performance in that process. the entire spectrum of possible professionals uncomfortable. But If they say “yes,” continue to knowledge, thereby contradicting there is a true 80-20 rule here. By the next step. If they say “no,” the idea of focus. In reality, the concentrating on the knowledge there is no point in going fur- knowledge required to perform that really determines long-term ther. If the executives and core functions is always very lim- organizational success, you ensure management team don’t think ited, focused, and tightly organized. that the core functionality is supe- using the knowledge is impor- In order to perform a core function rior, which actually reduces the tant enough to spend two well, one only has to master a few occurrence of low-probability hours learning it, then it isn’t important enough to manage! underlying principles. The knowl- events. By doing the core processes edge in the system needs to include correctly, many of the exceptions If they pass the “two-hour only these core principles and spe- that disrupt a system are pre- test,” drill down by next asking cific supporting content that is empted. Therefore, the KM system them, “Within this key process, directly relevant to the principles. doesn’t need to be designed to what are the two or three most handle them. critical subprocesses?” For example, when a manufactur- Repeat the two-hour test with ing company’s personnel focused Furthermore, the top performers the other key processes and on how to fix a particular type of (see below) tend to include low- drill down until they say an machine (a goal that failed the two- probability exception handling identified process or sub- hour test), they loaded more than as part of their expert knowledge. process isn’t worth two hours 100 reports the machine produced They do this by specifying an of time. into their database. However, when exception-handling process (e.g., they reconceptualized the process “If you have this situation, call into “how to optimize throughput Bob…”), rather than trying to pro- In order to perform a core in a section of the manufacturing vide all of the information required line,” they realized that only three for every situation. At the manufac- function well, one only has of the reports were relevant to their turing company discussed above, to master a few underlying real goal: improved efficiency. They the top performers put it this way: principles. went so far as to explicitly exclude “If you can’t improve the efficiency the 97 other reports from the sys- of the manufacturing line using the tem so they wouldn’t clutter the three reports, you have a serious You have now identified the man- most important content. Processes problem and need to contact your agement areas where knowledge that are too small tend to generate technical specialist immediately.” management is most critical. The tremendous amounts of extraneous No effort is made to include the idea is to rigorously prioritize the content because the organization specific knowledge in the KM sys- knowledge to be managed accord- loses sight of what really matters. tem because it is so rarely used and ing to the customer’s priorities and Conversely, by becoming extremely so specialized. Instead, a process is not try to manage all available clear about what really matters, the defined for handling the exception. knowledge. This prioritization is the organization reduces the knowl- Thus, the KM system only has to foundation for ensuring both that edge requirement. cover the 20% of the content that knowledge focuses on only core makes up 80% of the performance Of course, this means that a functions and that the knowledge of a function plus exception- tremendous amount of what KM will be used. handling processes for anything professionals typically think of as else. The KM system does not have Management processes that pass an organization’s knowledge is not to include all the knowledge for the two-hour test may seem so going to be included in the KM sys- every situation, which can’t be large that they will encompass tem — a fact that makes many KM done anyway.
Determining and Gathering words, real stories contain all Specific Knowledge of the information needed to Getting people to quickly Once the organization has identi- completely and effectively per- absorb knowledge and imme- fied its most important manage- form the function, including diately change their behav- ment processes, it can begin to the handling of occasional and low-probability events. iors is extremely difficult. gather the specific knowledge required to perform the function. Using digital coaching technology (A detailed description of the (DCT) [1], a single KM professional knowledge-gathering process immediately change their behav- can gather this type of knowledge can be found in “Harvesting the iors is extremely difficult. We are from just six to eight top performers Expert’s Secret Sauce” [2].) In brief, bombarded every day by such a in as little as three days, even for use the following process: variety of stimuli that people have extraordinarily complex processes. become experts at resisting new Ask the management team to For example, it took only three days knowledge. identify the six to eight people to gather the key expert knowledge who are most respected for on how to design a microprocessor. How can KM overcome this barrier their expertise in performing It is a very efficient process. and bring about the expected the top process. Notice we behavior changes? To achieve this, said “respected.” Respect Knowledge Storage and Transport the KM system must contain the means that everyone accepts Of course, before a customer can key process knowledge identified these experts’ guidance on use knowledge to change behav- by the organization in a form that how to perform the function iors, she must be able to find and can be readily applied to new situa- without second-guessing. If access it. This is the primary focus tions. The two things to remember they say, “Do it this way,” you of current KM systems. But as we here are just the key knowledge believe them! have said, these systems often and immediate application. Use “naïve new person ques- deliver poor search results because Experience has shown that when tions” to prompt the experts of the huge volume of information the knowledge is recognized as to tell what we call the “real they contain. The customer- coming from a respected source stories” (as opposed to “offi- focused approach to KM reduces and it can easily be applied to cial” stories) of their function. the search requirement because new situations, behavioral change Official stories are the ones knowledge is stored for only a lim- is possible. you find in procedure manuals ited number of processes and the and training binders. Real sto- There are at least three key ele- ries always contain the tacit content is well defined by the experts. It is rare for even a very ments required for converting knowledge most critical to large organization to have more knowledge into behavior change success. They include the top than 10-12 key management (see Figure 3): performers’ mental models or vision of the function, their processes. Consequently the data- 1. Clear expectations. The exec- organizing principles with bases are much smaller and the utives must set a clear expec- detailed supporting informa- searching capability can be much tation that performance of the tion, specific roles and respon- less sophisticated. identified function is important sibilities, milestone and enough that all personnel per- detailed schedules, compre- Knowledge Drives Immediate forming the function will take hensive risk management Behavioral Change at least two hours to learn the plans, and identification and Unlike storage and search capa- knowledge and develop a plan use of supporting resources bilities, which are fairly simple to for actively applying it to their such as training programs and implement, getting people to personal situation (thus the process manuals. In other quickly absorb knowledge and importance of selecting a
Get The Cutter Edge free: www.cutter.com Vol. 17, No. 12 15
they visibly changed, many cus- Clear tomers have responded with some expectations variation of: “This is great informa- tion. It will really help me be suc- cessful.” A district manager of a large insurance company observed that with his agents’ use of their Instant Behavior Instant new KM system, “You can actually credibility change application see how people have organized their behavior around the expert content. It is very apparent.” In this case, knowledge has driven behav- Basis of ior, and it is behavior that is based planning on the most consistently successful people in the organization.
Sustaining Use of Knowledge
Figure 3 — Converting knowledge into behavior change. While these immediate impacts are a tremendous step forward in process that the executives comes from the realism of the think is critical for the organi- experts’ “secret sauce,” but creating value from KM, they are zation’s success). also from its comprehensive- not sufficient to produce a long- ness. The mental models, term benefit from the knowledge. 2. Instant credibility. The con- Several other processes are organizing principles, and risk tent must have “instant credi- management elements are required to ensure that the initial bility,” such that the customer particularly powerful for creat- use of the knowledge for behav- can recognize the realism ing instant credibility. Together, ioral change continues until it is immediately. This comes from these components lead to fully internalized. (These monitor- the linguistic properties of intense use of the content and ing and learning functions are the real stories and the list its integration into a behavior described in detail in “The of respected personnel who pattern in about two hours. created them. Studies have Performance Improvement shown that when presented (These processes are described in Multiplier” [4].) with new information, the detail in “8 Minutes to Performance recipient resists it reflexively In brief, these key processes are Improvement” [3]). for about 90 seconds. In our as follows: experience, instant credibility When all three elements are pres- The immediate supervisor overcomes the 90-second ent, it is possible to see the impact must regularly and systemati- skepticism barrier, leading to of knowledge on customer behav- cally monitor use of the knowl- intense engagement in four to ior. At the four- to eight-minute edge while the function is eight minutes. mark mentioned above, there is being performed, and she 3. Instant application. The a pronounced physical change in must do it in a way that does content must have “instant the customers. They lean forward not produce alienation. (In our application,” such that the toward the system, squint their method, she would ask four customer can immediately eyes into a tighter focus, breathe simple questions that interact see how this applies to his more rapidly, and change the lan- with a set of natural self- situation and how use of the guage they use with anyone sharing monitoring processes used by knowledge will make him the people she supervises.) the experience. When asked why more successful. This too
The organization’s executives must also monitor use of the Successful Key knowledge within their respec- key initiatives Initiati implementations tive functions. If they don’t monitor the new process, no Setting Executive one else will, and the knowl- expectations monitors for planned edge will rapidly fall into dis- implementation excellence use. Again, if the process isn’t important enough to monitor, why bother doing it? The organization must fre- quently refresh the expert Expert Field content, adding new learnings “secret sauce” monitors from new real stories and rig- for implementation implementation orously purging obsolete mate- rial. Everyone must believe the Guided content is absolutely the best coaching creates plan available up to that minute.
A Closed-Loop Behavioral Figure 4 — A behavioral KM system.
KM System We now have a fully closed-loop, causes knowledge to drive behav- A women’s apparel chain self-regenerating, behavioral KM ior change on a mass scale, increased customer conver- system (Figure 4). We begin with enabling organizations to substan- sions 6% in six weeks. the executive team defining the tially improve performance even management processes of greatest Converting your warehouse KM sys- if the organization is large and geo- tem to a customer-focused one is importance, progress to gathering graphically dispersed. the expert knowledge needed to easy. Just think from the customer’s most effectively perform those perspective. functions, guide personnel into EXCEEDING EXPECTATIONS immediate then sustained appli- The value of any KM system is the REFERENCES cation of this knowledge, monitor impact it has on productivity and 1. Seidman, W.H. “Humanistic its use, and learn from the entire the profitability of the organization. Knowledge Technology.” Cutter IT process. Executives expect KM systems Journal, Vol. 15, No. 3, March 2002, This may sound as though it is quite to clearly generate gains in these pp. 25-31. hard to do, but it really isn’t. Using areas. Customer-focused KM sys- tems consistently improve organi- 2. Seidman, W.H., and M. McCauley. DCT, an organization can go from zational performance. For example: “Harvesting the Experts’ initial definition to complete inter- ‘Secret Sauce’ and Closing the nalization of any key management A restaurant chain was able to Performance Gap.” Performance process in as little as six weeks. For achieve 100% utilization of a Improvement, Vol. 42, No. 1, example, a fast food company new KM system, which led to January 2003, pp. 32-37. introduced three new knowledge- a 10% increase in sales. based initiatives into 1,400 restau- A federal agency reduced 3. Seidman, W.H., and M. McCauley. rants in just over six months, with training time for new per- “8 Minutes to Performance almost 100% applying the knowl- sonnel from three years to Improvement.” Performance edge more systematically and six months. Improvement, Vol. 42, No. 6, July completely than expected. DCT 2003, pp. 23-28.
Get The Cutter Edge free: www.cutter.com Vol. 17, No. 12 17
4. Seidman, W.H., and M. McCauley. Over the past 20 years, Michael McCauley “The Performance Improvement has worked with some of the industry’s best and brightest to significantly Multiplier.” Performance enhance the way organizations plan Improvement, Vol. 42, No. 9, and manage their product development October 2003, pp. 33-37. efforts. He has helped improve the development processes of numerous 5. Snowden, Dave. “Innovation, KM, leading-edge companies, including Hewlett-Packard, Dell Computer, Novell, and Corporate Success.” Keynote Motorola, Chevron, Alpha Therapeutic, address to KMWorld 2001, Santa DuPont, Pacific Gas & Electric, Ameriserv, Clara, CA, December 2001. Core Media, LSI Logic, Xerox, and Microsoft. William Seidman is a recognized thought Mr. McCauley’s background as an engi- leader and expert on management deci- neer and management consultant, along sion making, intellectual capital man- with his experience in a wide range agement, and executive leadership. In of organizations and industries, has particular, Dr. Seidman is renowned for resulted in a unique ability to translate understanding the processes required to complex best practices into logical harvest and use expert knowledge. His sequences of “guided” steps that can be doctoral work at Stanford resulted in the systematically implemented in software. development of ground-breaking tech- He also has a demonstrated ability to niques for analyzing management deci- lead high-performance product develop- sion making that became the genesis of ment organizations. While vice president the Cerebyte Infinos System. Dr. Seidman of product development at Integrated is currently leading Cerebyte into a domi- Product Systems, a Silicon Valley start-up, nant position in the executive leadership he led the development of the company’s software market. His particular interest flagship products. Currently he heads is the use of technology to ensure the product development for Cerebyte, Inc. implementation of an organization’s key strategic initiatives. Mr. McCauley holds an undergraduate engineering degree from Cal Poly–San Dr. Seidman has more than 20 years’ Luis Obispo and an MBA from Golden experience as a manager of profit and State University. He also maintains a loss centers in high-technology compa- Project Management Professional certifi- nies, including Hewlett-Packard, Silicon cation from the Project Management Graphics, Mentor Graphics, and Institute. Integrated Project Systems, a Silicon Valley consulting company. He is an expe- Mr. McCauley can be reached at rienced consultant to and manager of Michael.McCauley@Cerebyte.com. fast-growth, high-technology environ- ments and a frequent speaker and author on knowledge management topics.