Introduction: the problem

A curious, but perhaps not unusual problem besets knowledge management (KM). It is the problem of lack of agreement about what it is, and consequently about its effects, its successes, its failures, and its future. An illustration of the problem is provided by a recent exchange in the act-km list serve group begun when one of the participants contributed a lengthy post (Guerino, 2007), which reported on some recent interviews of 18 high-level executives of firms in the 15,000–75,000 employee range about their views on KM. According to his post, the executives expressed a number of 'common beliefs' and seemed to agree that the main goals of a general KM strategy are (Guerino, 2007):

1. Improve communications
2. Improve capture and persistence of content
3. Improve capture and persistence of work
4. Improve accessibility to data and information of all forms
5. Improve the capture, recognition, and exploitation of metrics and statistics
6. Improve automation

After stating these goals, he proceeded to outline the views of the 18 executives about 'KM strategies' in each of these areas, and even listed facilitating tools and technologies.

This post wasn't greeted enthusiastically by other KM practitioners on the list, but the problem of definition is illustrated in a particularly direct way by one reply (Bounds, 2007): Except for the Common Beliefs, that list of goals checks off pretty well against the issues I am looking into to design my organisation's 'Information' Management Framework. (And I have been explicitly told 'not' to worry about Knowledge Management at this point.)Really, all you've proved is that these high level execs only care about Information Management and not Knowledge Management.To borrow a quote from Robert Heinlein, calling a tail a leg does not make the name fit.

The problem of lack of agreement on what KM is, suggests four possibilities:

• people can be doing KM and calling it KM;
• people can be doing KM and calling it something else;
• people can be doing non-KM and calling it KM; or
• people can be doing non-KM and calling it non-KM.

These possibilities exist from whatever point of view KM is defined. The first and fourth represent no problem if one wants to evaluate KM, but clearly, without agreement on what KM is, the second and third introduce serious problems in any evaluation of KM's impact or effectiveness. And the more frequently these possibilities occur, the greater the error introduced into KM's track record, regardless of the truth of impact models developed to assess the impact of instances of the first possibility.

How frequently do the second and third possibilities occur? Clearly, the more there is disagreement about what KM is, the more second and third possibilities exist, and the more any track record evaluating KM, either formal or informal, will be distorted and misleading in telling us what percentage of KM efforts are successes.

Some objections

Some will say that the problem of agreement on a definition of KM is not the underlying problem in situations like the above, as lack of agreement only exists because there has been so little success in KM, and, as a result, no one use of the term, a use that might have been illustrated by its absent successes, has been able to dominate the field. Others will say that this is not a real problem because top executives don't care about disagreements among practitioners and academics about KM, or even about KM itself. They only care about results and specifically, whether an intervention, whatever it's called, solves an organizational problem. Who cares whether it's KM or information management or customer relationship management or quality management or any one of a number of other forms of management? While I think there's something to both of these objections; in the end, I think both are mistaken.

The objection attributing the failure to coalesce around a common understanding of KM to the failure to solve problems and be successful, neglects the possibility that even though the view that success would bring consensus may be correct, it may also be true that difficulty in making a record of success may be due, in part, to the failure to coalesce around a definition. Indeed, it is likely that consensus around a common view of KM, and success in its practice, may be related to one another and may co-evolve, rather than one being a precondition for the other.

The second objection that top executives don't care about names, but only about results, is typical of the short-sighted, 'realistic' nonsense that, these days, often passes for tough-minded thinking in certain quarters. The names we give the things that we do are important. Not in themselves, but because we use them to help us track the relationship of those things to other things with other names. If we use different names for the same thing, and also the same name for different things, then we can neither measure anything consistently, nor model impacts correctly, nor evaluate performance correctly, nor develop any cumulative knowledge at all, and there's nothing 'practical' or 'tough-minded' about that.

Many top executives in large organizations may well feel about KM, and also other forms of specialized management, exactly the way the second objection asserts. That is, they may not care what name people give to things, but only about the results they get by using those things. However, there are practitioners in every field, including management, who don't recognize the relationship between the results they want and what the people who work for them have to do in order to get them. So, their impatience with names and with conceptual discussions that try to image and specify important terms in KM, should not deter our efforts to arrive at good conceptual frameworks and agreement about them; especially since the very executives, who are skeptical of such activities, will not accept their skepticism as our excuse for not getting the results they seek.

What to do about it

What we should not do is to force researchers and practitioners to agree on a definition of KM through premature efforts at standardization. While this might bring about the consensus we need in order to do evaluations of 'KM's' track record, any consensus forged in the political atmosphere of standards organizations may well be a consensus constructed around a compromise that has no conceptual unity, and which results in a version of 'KM' that is bound to fail.

Another thing we should not do is to evaluate KM projects without benefit of an effort at explicit definition or specification of what we mean by KM. It's surprising how frequently one sees this happening in practitioner circles, and how damaging it can be.

A recent good illustration of failing to make one's notion of KM explicit may be found in an HP presentation on 'Elements of KM' (Garfield, 2006). The terms 'KM' or 'knowledge' occur on the majority of pages in this presentation that outlines the people, process, and technology 'elements' of KM as the author and HP sees them. However, the presentation simply takes it for granted that the major 'KM' elements listed are KM elements.

And while, of course, HP certainly knows what the elements of its program are, and certainly has a perfect right to call it a 'KM' program, the decision to present things this way without explicitly defining or specifying KM and without explaining why their list of people, process, and technology elements, when taken together, constitutes a KM program rather than an information management program, or a collaboration management program, or a change management program in effect creates a denotative, rather than a conceptual definition of KM, and also creates the following problem for evaluating the program from a KM disciplinary perspective.

Let's assume that HP considers its program a success. Then the word goes forth that HP has had a substantial KM success. That's all well and good so far. If another organization learns about HPs program and decides to imitate it, the other organization knows what it's aiming for, what the elements of the program are, what it ought to expect as a result, and everything is fine. But what if the word goes forth from a number of companies that their 'KM' programs, also denotatively defined, were also successful, and the impression grows that KM is becoming successful in general, and what if, among these companies, there are many different denotative definitions and variants of KM being specified and practiced? Then, which is the 'KM' idea and associated 'KM' program that is being successful at the disciplinary level? Which is developing a substantial track record that deserves to be emulated? Which successful case is really instantiating 'KM' as a general class of activities and programs? Perhaps, none of them; and perhaps the impression of broadening success created by varying uses of the term 'KM' is a false and misleading one.

The situation would be even worse if the HP program is eventually considered a failure. Then the track record of 'KM' would record a failure, when it is, at best, debatable whether HP's program is a KM program at all. But picture a situation, where HP along with many other organizations present denotative definitions of KM to us in the form of lists of KM program elements, and a high percentage of these programs gets evaluated as failures. Then it could well be the case that the disciplinary track record of KM could read very poorly indeed, merely because the practitioners implementing KM programs are doing 'non-KM' and calling it 'KM' while implementing programs that fail. Although I don't have any data on this, I suspect that this, in fact, is one of the things that has happened to KM over the years. That is, people have called many things that have failed, 'KM,' even though there may have been very little 'KM' actually going on (Firestone, 2004, 2004a, 2004b; Firestone & McElroy, 2005).

What is essential from the standpoint of any researcher or practitioner trying to evaluate KM's track record is to frankly acknowledge that there is no disciplinary agreement on the nature of KM, and then to decide on their own explicit definition and/or specification of KM and apply it, both in doing KM and in evaluating KM's track record. If researchers and practitioners all were to do that, the discipline would gradually develop evaluations of different concepts of KM and differing track records of interventions arising out of each of the different concepts. Those who evaluate KM would become very self-conscious about distinguishing the variants of KM they were evaluating. At the end of the process, a consensus on leading definitions would self-organize around the concept associated with the greatest KM intervention success, provided there is one. And the current confusion about impact associated with the existence of many and differing concepts of KM will disappear.

The discussion so far indicates that we can do KM in two ways, either intentionally or unintentionally, without knowing that we're doing it. To do it intentionally, we can denotatively specify it in the way that HP and many others do, or we can develop a conceptual framework that defines and specifies KM, and apply that conceptual framework consistently to what we do. Once we've done that, we're in a much better position to do KM. Perhaps the KM that we do won't turn out to be the KM variant that the discipline eventually settles on. But at least it will be KM according to our best understanding of it, and it will be KM done in such a way that others can critically evaluate it, and coalesce around it, if it makes sense to them.

The remainder of this paper will do two things. First, I'll present my own conceptual viewpoint on the nature of KM. And second, I'll talk about two basic approaches to doing KM.

The nature of KM

Let's begin with routine decision-making. When we see a gap between the way the world is and the way we want it to be, we typically plan what we have to do to close the gap between the two. Once we develop a plan we then act. After acting, we monitor the results of our actions. Finally, we evaluate the results we've monitored, and then if we haven't reached our goal, and sometimes even if we have, we begin the cycle of decision-making again with a new round of planning. Let's call this pattern illustrated in Figure 1, the decision execution cycle (DEC) (Firestone, 2000).

Figure 1.

Routine decision-making (the decision execution cycle) (adapted from Firestone & McElroy, 2007).

Full figure and legend (34K)

DECs, of course, produce decisions and decisions, actions. And actions – activities – are the stuff that social processes, social networks, and (complex adaptive) social systems are made of. These are all built up from activities that are inter-related by their objectives, goals, effects, and the values that are associated with them.

Now let's distinguish three categories of business processes: operational business processes, knowledge processes, and KM processes. Operational processes are those that are comprised of routine DECs. Examples are sales, marketing, logistics, accounting, etc. They use knowledge but, apart from knowledge about specific events and conditions, do not produce or integrate it.

Knowledge processes are also composed of DECs. But these DECs are primarily motivated by the need to guard against and solve problems that arise in operational business processes; and while there are some routine DECs in knowledge processing, there are also creative DECs in which new ideas are created. There are three knowledge processes: problem seeking, recognition, and formulation, the process that transitions processing from operational processing to knowledge processing and produces the problems that drive other knowledge processes; knowledge production, the process an agent executes that produces new general knowledge; and knowledge integration, the process that presents this new knowledge to storage containers and agents comprising the system.

Knowledge production is a process made up of four sub-processes:

• information acquisition,
• individual and group learning,
• knowledge claim formulation, and
• knowledge claim evaluation.

Knowledge integration is made up of four more sub-processes, all of which may use interpersonal, electronic, or both types of methods in execution:

• knowledge and information broadcasting ,
• searching/retrieving,
• knowledge sharing (peer-to-peer presentation of previously produced knowledge), and
• teaching (hierarchical presentation of previously produced knowledge).

Knowledge processes, of course, produce outcomes. Chief among these is knowledge, which I've defined and specified at length elsewhere (Firestone & McElroy, 2003, chapter 1; Firestone, 2006). The various outcomes of knowledge processes may be viewed as part of an abstraction called the distributed organizational knowledge base (DOKB) (Firestone, 2000). The DOKB has electronic storage components. But it is more than that, because it contains all of the outcomes of knowledge processing in electronic and non-electronic media. And since it includes beliefs and belief predispositions, and memories as well, it also includes all of the mental knowledge in the organization, as well as the changed synaptic structures that result from organizational learning processes.

Keeping the above notions in mind, here is how things work in organizations. Routine DECs and operational business processes are performed by agents who use previous knowledge in the DOKB: synaptic knowledge, mental knowledge, and knowledge in organizational repositories to make decisions. Sometimes the DOKB and an agent's perceived situation doesn't provide the answers it needs, and the agent recognizes that, and goes further to formulate the problem that has arisen, consciously or in words. The problem is an epistemic gap between what an agent knows and what it needs to know to participate successfully in the operational business process (Firestone & McElroy, 2003, chapter 2). Such a problem initiates a new knowledge production process. Once the problem is perceived, there is a need to formulate tentative solutions. These can come from new individual and group learning addressing the problem, or they can come from external sources through information acquisition, or they can come from entirely creative knowledge claim formulation, or, of course, they can come from all three.

Where the tentative solutions (Popper, 1972) come from, and in what sequence, is of no importance to the self-organizing knowledge-processing pattern of knowledge production. The only important thing about sequence here is that knowledge is not produced until the tentative solutions, the previously formulated knowledge claims, have been tested and evaluated in the knowledge claim evaluation sub-process (Firestone, 2004c). And that sub-process, knowledge claim evaluation (KCE), is the way in which agents select among tentative solutions, competitive alternatives, by comparing them against each other in the context of perspectives, criteria, or newly created ideas for selecting among them to arrive at the solution to the problem motivating knowledge production (Firestone & McElroy, 2003, chapter 5).

KCE is at the very center of knowledge processing and knowledge management. Think about it. Without KCE, what is the difference between information and knowledge? How do we know that we are integrating (broadcasting, searching/retrieving, sharing, or teaching) knowledge rather than just information? And finally, how do we know that we are doing knowledge management and not just information management? (Firestone & McElroy, 2003, chapter 3)

Once knowledge and other tested and evaluated information is produced by KCE, the process of knowledge integration of the solution begins. There is no particular sequence to the integration of sub-processes listed earlier. One or all of them may be used to present what has been produced to the enterprise's agents, or to store what has been produced in the various repositories in the enterprise.

Those agents receiving knowledge or information don't receive it passively. For them, it represents an input that may create a knowledge gap and initiate a new round of knowledge production at the level of the agent receiving it. Integration of the knowledge therefore, doesn't signal its acceptance. It only signals that the instance of knowledge processing initiated by the first problem is over and that new problems have been initiated for some by the solution. While for others the knowledge integrated is knowledge to be used: either to continue with executing the business process that initiated the problem, or at a later time, when the situation calls for it.

Either way, the original problem that motivated knowledge processing is gone. It was born in the operational business process, solved in the knowledge production process, and its solution was spread throughout the organization during knowledge integration, and in this way, it ceased to be a problem – that is, it died. This pattern is a life cycle, a birth-and-death cycle for problems arising from business processes.

The life cycle gives rise to knowledge, synaptic, mental, and cultural (linguistic), and so I call it the knowledge life cycle (KLC) (McElroy, 1999, 2003; Firestone, 2000, 2003; Firestone & McElroy, 2003, 2003a, 2003b). Every organization produces its knowledge through the myriad KLCs that respond to its problems: KLCs at the organizational level, and KLCs at every level of social interaction and individual functioning in the organization. It is through these KLCs that knowledge is produced, and the organization acquires the solutions it needs to adapt to its environment.

Organizations differ in the profile of their KLCs. They acquire information in different ways. They formulate solutions in different ways. They integrate them in different ways. And, above all, they evaluate tentative solutions in different ways. Organizations also differ in the patterning of their knowledge outcomes. They have different procedures for doing things, different software capabilities, different sales forecasting models, and different performance monitoring schemes.

Knowledge Management is the set of activities and/or processes that seeks to change the organization's present pattern of knowledge processing to enhance both it and its knowledge outcomes. This implies that KM doesn't directly manage knowledge outcomes, but only impacts processes, which in turn impact outcomes. For example, if one changes the rules affecting knowledge production, the quality of knowledge claims may improve, or if a KM intervention supplies a new search technology based on semantic analysis of knowledge bases, then that may result in improvement in the quality of models. There are at least 10 types of knowledge management activities, which, however, need not be listed here (see Firestone, 2000; Firestone & McElroy, 2003; Firestone & McElroy, 2005 for a listing and discussion). The relationships among operational business processing, knowledge processing, and knowledge management are summarized in the three-tier model in Figure 2.

Figure 2.

The three–tier model (adapted From Firestone & McElroy, 2007).

Full figure and legend (116K)

Doing KM

According to the conceptual framework just presented, KM refers to activities aimed at enhancing knowledge processing. These activities are interventions designed to affect how knowledge processing is done. Interventions may be classified by the knowledge sub-process and the type of organizational agent they are targeted at. Tables 1 and 2 illustrate such a classification. At the simplest level, doing KM is a matter of carrying out activities targeted at, at least one knowledge sub-process, and at least one type of organizational agent.

Table 1 - Classification of knowledge production targets of KM.

Full table

Table 2 - Classification of knowledge integration targets of KM.

Full table

There are two basic approaches that may encompass all KM interventions. First, there are interventions introducing strategies, policies, programs, techniques, and tools that enhance knowledge processing by attempting to enhance the background conditions affecting both operational and knowledge processing DECs in such a way as to enable more effective knowledge processing by people. And second, there are interventions introducing strategies, policies, programs, techniques, and tools that enhance knowledge processing by interrupting the DECs of people. Let's cover the second approach first.

The DEC interruption approach

Keeping the DEC (Figure 1) in mind, this cycle can be interrupted by providing feedback from an enterprise knowledge base questioning contemplated decisions, after they are made, but before they are implemented, that, based on past results, appear to be in error, and offer 'correct' alternatives to the decision maker from the knowledge base. Such an interruption can make a decision maker (a) think twice about their own decision, but stick with it, (b) change their decision to correspond with the feedback, or (c) change their decision after solving a new problem they formulated in response to the feedback.

The last alternative implies creative learning and implementing a KLC (which is the same thing); which brings us to the important point that feedback introduced after a decision is made, but before it is implemented, is likely to increase the amount of problem seeking, recognition, and formulation, because it forces the decision maker to consider whether her decision is based on an error and may be associated with a knowledge gap. In addition, since new problems initiate KLCs, such feedback also increases the creative learning (knowledge claim formulation and evaluation) that occurs in the business decision area where we introduce the feedback.

Now add to this pattern by requiring that if the eventual decision using the result of creative learning conflicts with the feedback from the knowledge base provided earlier, the decision maker must record the criticisms that led her to either stick with her first decision, or develop a third alternative, against the recommendation recorded in the earlier feedback. Recording these criticisms in the knowledge base creates a growing historical record of the new cultural knowledge that is produced through problem solving.

In this pattern, feedback to decision makers comes from a knowledge base in the organization. But, where does the content come from in the first place? Who is the decision maker in the envisioned pattern actually interacting with? Whoever determines what goes into the knowledge base. But, what sort of content representing 'the organization's best knowledge,' created by what sort of human structures, would be automatically fed back in case it conflicted with the initial decision? The sort of knowledge claims that have survived evaluation in the past and have performed well. Human beings must decide on what the surviving organizational knowledge claims in a particular decision area are, and I think this suggests that you need committees, or perhaps communities, of domain experts in the various decision areas continuously evaluating developing knowledge, and then feeding the record of what they produce into the organizational knowledge base.

What I've been gradually developing here, is a pattern in which individual decision makers are interacting with organizational knowledge (however, measured, determined, or created), and in the process are, if you allow the further enhancement of tracking and recording their evaluations of that organizational knowledge, both receiving criticism from it, and also contributing to its critical evaluation over time, by recording their reasons for decisions that conflict with that knowledge. So, there are two levels of social process in this pattern: the level of individual problem solving and decision models and the level of organizational problem solving and decision models. The organization surfaces problems in individual knowledge underlying individual decisions, while individuals surface problems in organizational knowledge. Surfacing at both levels occurs in response to contradictions with their knowledge presented to individuals and to organizational communities or committees, as the case may be.

This pattern requires IT application support, since one can't very well provide feedback between a decision and its implementation without either mandating an onerous human review, or providing an IT application that will present the required feedback of organizational knowledge to an individual decision maker. Also, one can't very well track and record individual decision maker's critiques of organizational knowledge without IT applications support. The pattern is both 'top-down' and 'bottom-up.' It involves both episodic and distributed creative learning at the individual level, and assuming that many individuals are contributing feedback to the organizational level, pretty much continuous learning and critical review at the organizational level, at least in those areas where we implement the pattern. As the pattern also potentially draws everyone into a distributed creative learning process, I've called it the 'Open Enterprise Pattern' or the OEP (Firestone & McElroy, 2005) (see Figure 3).

Figure 3.

The open enterprise pattern (adapted from Firestone & McElroy, 2007).

Full figure and legend (161K)

It is an important characteristic of the OEP that it is comparatively easy to measure its success. If the pattern is applied to frequently occurring decisions of a specific type, then one can easily track the pattern of results before an OEP intervention compared to the pattern afterwards. If there is a material change, positive or negative, in the outcomes of the decisions being influenced by the OEP, it is hard to attribute the impact to anything else but the direct intervention in the DEC by the system.

What kind of IT system do we need to support (a) feedback to individuals, (b) creative learning by the individual, (c) claims and meta-claims presented to the organization by the individual, (d) creative learning by communities or committees responding to problems surfaced by individuals, and (e) changes in the results of decisions being influenced by the OEP?

First, a minimal system to support the OEP would require a database application to store the knowledge and meta-claims produced by individuals, committees, and communities, and also the record of individual decisions and outcomes both before and after the intervention establishing the OEP. Second, we need a web application for the minimal system to: access the database, enter decisions, meta-claims, and also organizational knowledge developed by committees or communities, and alert individuals to contradictions with their decisions present in organizational knowledge. The feedback applications will record and track decision models and meta-claims at the individual level. At the committee level that can be done by using applications that track and record content in 'collaboration spaces,' including evaluations by the committee of the track record of knowledge claims, decisions, and outcomes.

The long-term effect of attempting to implement the OEP systematically across many domain areas will reduce the risk of error in decision models and increase the quality of knowledge in key decision areas. But also, many more problems will be discovered in routine knowledge, much more creative learning will occur and also much more sustainable innovation will result. The character of an organization applying the OEP will also change somewhat. There will be more and more 'openness,' after the OEP is broadly implemented, and the likelihood is that there will be a general enhancing of distributed problem solving, creative learning, and the quality of decision models in the enterprise.

A case exemplifying the DEC interruption approach and the OEP is the well-known Partners HealthCare case (Davenport & Glaser, 2002; Firestone & McElroy, 2005). At Partners, the KM intervention implemented a system that monitored the order entry decisions of physicians and delivered alerts when orders were at variance with those recorded in a knowledge base matching diagnoses to preferred treatments. The knowledge base was initially constituted by expert committees in each major treatment domain. The committees, in effect, were viewed as the original sources of organizational knowledge. The physicians were not required to accept the recommendations of the committees. But if they decided to order a prescription at variance to that recommended in the knowledge base, they were required to record their reasons for disagreement in the system. These reasons were then evaluated by the committees leading to completion of the OEP pattern of interaction between the individual and organizational levels of the Partners system.

The Partners Healthcare KM intervention was an enormously successful one combining social process and IT techniques, and illustrates the ease with which it was possible to arrive at outcome metrics in this type of approach. Three of the most important results were: (1) serious medical errors in order entry were reduced by 55%; (2) use of beneficial drugs increased sevenfold; and successful dosages increased 11-fold. A detailed discussion of results is in Davenport & Glaser (2002), and a detailed interpretation using the above framework and the OEP pattern is in Firestone & McElroy (2005).

The background conditions or ecological approach

This approach is the most popular one in KM. There are many variants of it. Some are IT tools-based. Some are social process-focused, but use IT tools to enable or implement social interventions. Still others employ varying mixes of social techniques and IT tools. What unites all of these variants is the idea that some aspect of knowledge processing, knowledge sharing, knowledge creation, problem solving can be influenced and enhanced by focusing on some aspects of individual ecology, so that when individuals need knowledge to make a decision or to solve a problem they can find or make the right knowledge on demand.

However, note that in this approach, the burden is on individuals to self-organize, to take action, and to use the transformed ecology to improve their knowledge processing. They may not do it. You might build it and they won't come, and then the KM intervention might fail. So, one thing that's indicated in this type of approach is the need for marketing. While the KM intervention is being implemented, a participatory culture must be set up for the implementation, and this culture must be maintained for the effort to be successful.

Another very important consideration in the ecological approach is its design. That is, the social/IT intervention combinations should be planned with reference to the knowledge processing targets the interventions are supposed to enhance. However, this is frequently not done, and interventions designed to introduce storytelling, or communities of practice, or anecdote circles, or after action reviews, or portals, or blogs, or wikis, or collaborative team spaces are often introduced using vague knowledge transfer/sharing or problem solving justifications. Rarely are such interventions modeled with reference to their anticipated effects on business processes, decisions, or Return-on-Investment (ROI), or evaluated relative to whether their post-intervention effects meet expectations. In part, this failure is due to the absence of good conceptual frameworks for KM and knowledge processing. To think about the impact of KM interventions on knowledge processing targets, knowledge outcomes, and downstream business processes and outcomes, the first thing we need is a clear conception of knowledge processing and of various targets. Without that, how can we begin to think about impact?

And that brings us to the question of measurement. To be successful, any variant of the ecological approach should incorporate a careful measurement structure for generating metrics, as we can't begin to measure impact without such metrics. However, many KM programs have been remiss in developing metrics or have developed metrics that are not very important to others who have later evaluated KM work based on the ecological approach.

A good example of an anti-pattern in this respect is provided by the World Bank Knowledge Sharing Program. From FY97–FY02, the bank ran a high-profile KM program characterized as a knowledge sharing program, spending $280 million in the process. The program encouraged development of some 125 thematic groups (communities of practice), 24 advisory services, widespread web site use, and training in and use of storytelling (Denning, 2001). In the field of KM itself, this program was widely viewed as a successful, even a flagship, model, and many other organizations emulated its emphasis on Community of Practice (CoP) and storytelling.

However, a World Bank Review (Gwin, 2003), while providing a perfunctory nod to the success of the program in fostering a new knowledge-sharing culture and a wide variety of new activities for aggregating and sharing knowledge, also concluded that 'the Bank's new activities have not been well-integrated into core lending and non-lending processes' (p. ix). And the report mentions management shortfalls as accounting for this state of affairs. Specifically, management did not define roles and responsibilities for making knowledge sharing a way of doing business; nor did it provide incentives for incorporating knowledge sharing into operational processes. Further, there was 'no systematic monitoring and evaluation of knowledge-sharing programs and activities' (p. ix). In other words, no structure of metrics was developed for the project, and no connection between the accomplishments of the program and the bank's operational activities and day-to-day business could be established.

A second case illustrating a successful ecological approach was implemented at Halliburton, and provides a great contrast to the World Bank model. Halliburton's KM effort began in its energy services group (ESG), encompassing 35,000 employees in 100 countries (Ash, 2005, p. 24). Later on it was placed in a new business functions unit, which allows it to serve both ESG and Kellogg-Brown-Root (65,000 employees). ESG established a core team of five knowledge managers under the direction of Michael Behounek. The core team's role was to approve and monitor a much larger network of KM projects. By 2005, Halliburton had established 19 projects (p. 25). All were built around a CoP, supplemented with one or more tools, including portal, social network analysis, survey, collaboration, document, and content management tools. Training was also used to integrate people into communities, and newsletters and white papers were used to publicize activities. Finally, quarterly reports are used to periodically maintain the attention of management and to communicate ROI results attributable to the KM program.

Each of the 19 projects was viewed as a business investment from the beginning of its planning and design activities (p. 26). Executives wanting to establish a community needed a problem or class of problems they wanted to address related to operational business processes and outcomes. They worked with the KM core team to plan the intervention and to pre-specify measures of success/failure and ROI within a balanced scorecard framework. Multiple measures were used because of the difficulty of establishing cause and effect in a social process context. Impact was inferred when multiple measures pointed in the same direction.

Projects were never approved by the KM core team without a plan and a specific business case that specified the impact expected from the intervention and also the measures that would be used to evaluate whether the expected impact had occurred (Behounek, 2003). As the focus was on problems or classes of problems in business processes, the interventions were not simply about knowledge sharing, they also attempted to and did enhance problem solving in the specific domains of each intervention.

As each intervention had a fairly specific purpose and, from the beginning, was related to expected ROI, profit-and-loss responsibilities, and a balanced scorecard, each community project was able to justify either a part-time, or full-time knowledge broker(s) at the local project level to moderate and facilitate discussion threads, coordinate with the KM core team as well as other knowledge brokers, and to steward repositories (Behounek, 2003). Knowledge brokers (sometimes as many as five) were selected for their breadth of subject matter knowledge in their domain areas, but not for deep expertise.

Each community worked in the following way. Membership was open to anyone, unless, as was true in three cases, the community was closed due to Intellectual Property issue considerations. Unlike CoP programs in other organizations, however, community members couldn't self-organize the overall subject matter of the CoP, since that subject matter was already set by the business plan and the process for establishing the community. Within these constraints, users entered questions and issues of concern to them into the collaboration tool, and other users responded. Brokers added to the process by engaging subject matter experts, who both validated previously expressed ideas and also contributed new ideas. Brokers also tracked the use of ideas produced by the community and compared impact with prior expectations (Ash, 2005, p. 25). In a program such as Halliburton's, each community may develop knowledge that is important for another community. Halliburton's core team handles this possibility by networking the communities, including bringing knowledge brokers, champions, consultants, and the core team together three times per year.

Halliburton's KM/CoP-based program is a great success for the ecological approach to KM. It is clearly a KM program, since its focus is on enhancing both problem solving and knowledge sharing in the business process domains served by the communities (Behounek, 2003). Every one of the 19 programs approved by 2005 delivered positive ROI (Ash, 2005, pp. 25–26). Some outstanding results are: a 564% annual ROI figure; an estimated $20 million in benefits per year for Halliburton's 'Employee Central Portal'; and a $71.5 million cost savings in procurement and materials(P&M) during the first year of operation of the community P&M portal (partly attributable to the community) (p. 27).

Summary and conclusions

There is too little agreement on the nature of KM among researchers and practitioners. This is important at a disciplinary level when one is concerned with evaluating the overall success of KM because, from the point of view of any of the different definitions of KM, people are (1) doing KM and calling it something else (non-KM); and (2) doing non-KM and calling it KM. This makes it very difficult for them to accurately evaluate the level of success or failure of even their own version of KM, much less anyone else's, and it makes it likely that any overall evaluation of KM, from whatever source, is inaccurate because of failure to take account of the record in categories (1) and (2) just above.

We should not attempt to work through standard organizations or, in evaluation projects, to define KM denotatively, in order to handle this problem of definition. Instead, we should each offer a conceptual definition and framework of our own and consistently apply it in evaluation. By itself, this won't remove definitional problems, but it will allow clear evaluations of KM impact to be completed for each of the contending concepts of KM. Eventually KM will self-organize around the more successful concepts and the definitional problem will end.

To illustrate a conceptual definition and specification, I presented an abbreviated version of my own framework, and then used aspects of it to analyze two primary approaches to KM: the DEC Interruption Approach, and the Background Conditions, or Ecological Approach. I analyzed the DEC Interruption Approach by sketching out an ideal pattern called the OEP, and presented a successful example of it in the Partners Healthcare Case. I did not present an ideal pattern for the Ecological Approach, but contrasted two cases: the World Bank case, where a huge KM effort failed, in the end, to demonstrate a connection to core lending and non-lending bank business processes; and the Halliburton case, where a smaller, but still substantial effort, demonstrated success in all 19 of its interventions by emphasizing a business case philosophy, a close tie to business processes and decisions, a vigorous effort at measurement of impact, using multiple indicators and balanced scorecards, and ROI.

References

1. Ash J (2005) A sustained commitment to collaboration. Inside Knowledge 8(6), 24–28.
2. Behounek M (2003) Verbal Communication.
3. Bounds S (2007) Post on Achieving better KM without dedicated KM people, KM tools, or KM processes. 09/02/07, available at: http://actkm.org/mailman/private/actkm_actkm.org/2007-September/004472.html.
4. Davenport T and Glaser J (2002) Just-in-time delivery comes to knowledge management. Harvard Business Review 80(7), 107–111. | PubMed |
5. Denning S (2001) The Springboard. KMCI Press/Butterworth Heinemann, Boston, MA.
6. Firestone J (2000) Knowledge management: a framework for analysis and measurement. White Paper No. 17, Executive Information Systems, Inc., Wilmington, DE, 1 October 2000, available at: http://www.dkms.com/White_Papers.htm.
7. Firestone J (2003) Enterprise Information Portals and Knowledge Management. KMCI Press/Butterworth Heinemann, Burlington, MA.
8. Firestone J (2004) Has KM been done? Part 1, available at: http://radio.weblogs.com/0135950/2004/04/11.html#a6.
9. Firestone J (2004a) Has KM been done? Part 2, available at: http://radio.weblogs.com/0135950/2004/04/13.html#a8.
10. Firestone J (2004b) Has KM been done? Part 3, available at: http://radio.weblogs.com/0135950/2004/04/14.html.
11. Firestone J (2004c) The importance of knowledge claim evaluation. Available at: http://radio.weblogs.com/0135950/2004/05/09.html#a16.
12. Firestone J (2006) What knowledge is, an excerpt from Riskonomics: Reducing Risk By Killing Your Worst Ideas, Alexandria, VA, KMCI Online Press, available at: http://www.kmci.org/media/Whatknowledgeis%20(non-fiction%20version).pdf.
13. Firestone J and McElroy M (2003) Key Issues in the New Knowledge Management. KMCI Press/Butterworth Heinemann, Burlington, MA.
14. Firestone J and McElroy M (2003a) The new knowledge management. Knowledge Management Magazine 6(10), 12–16.
15. Firestone J and McElroy M (2003b) Excerpt #1 From The Open Enterprise: Building Business Architectures for Openness and Sustainable Innovation (with Mark W. McElroy) (Hartland Four Corners, VT, KMCI Online Press, 2003), available at: http://www.dkms/com, http://www.macroinnovation.com, and http://www.kmci.org.
16. Firestone J and McElroy M (2005) Doing knowledge management. The Learning Organization 12(2), 189–212. | Article |
17. Firestone J and McElroy M (2007) KMCI CKIM Certificate Workshop Notes. KMCI Publications, Alexandria, VA.
18. Garfield S (2006) Initiating & running a successful worldwide KM program: examples from HP. Available at: http://www.kmworld.com/kmw06/presentations/A102_Garfield.pps.
19. Guerino F (2007) Post on Achieving better KM without dedicated KM people, KM tools, or KM processes. 09/01/07, available at: http://actkm.org/mailman/private/actkm_actkm.org/2007-September/004456.html.
20. Gwin C (2003) Sharing Knowledge: Innovations and Remaining Challenges, An OED Evaluation. The World Bank, Washington, DC.
21. McElroy MW (1999) The second generation of KM. Knowledge Management 2(10), 86–88.
22. McElroy MW (2003) The New Knowledge Management: Complexity, Learning, and Sustainable Innovation. KMCI Press/Butterworth-Heinemann, Burlington, MA.
23. Popper K (1972) Objective Knowledge. Oxford University Press, London, England.

About the author

Joseph M. Firestone is author of Enterprise Information Portals and Knowledge Management, co-author of Key Issues in the New Knowledge Management (both KMCI Press/Butterworth-Heinemann, 2003), Excerpt # 1 from The Open Enterprise: (KMCI Online Press, 2003), more than 175 articles, papers, and reports, and is developer of the web sites, www.kmci.org, www.dkms.com, www.adaptivemetricscenter.com, and the KM Blog "All Life is Problem Solving" at: http://www.radio.weblogs.com/0135950.