Organizing for Knowledge Management

Adapted from "Organizing the Imagineers Corps", September 23, 1977

In the June Futurist, you called for ideas concerning how the "Imagineers Corps" should be organized. In passing, you described the concept as "the establishment of a special group to brainstorm problems."

At the time I proposed an organizational structure that embodies the basis principles required for what we now call knowledge management.  As I review that article and the developments in knowledge management, it is quite apparent that, aside from some dated terminology, the basic structure is immediately relevant to today's concerns.   Then I proposed the how.  Today I can add some of the why to it.

In preparing my proposal, I have integrated some ideas on creativity from the following references: Conceptual Blockbusting by James L. Adams, Synectics by William J. J. Gordon, Epistemics by J. Samuel Bois, and Joy by William C. Schutz. My effort is to synthesize the salient features in these works relating to a structure of a process resulting in creative output and to model this process in my proposed organizational structure and processes of "The Imagineers Corps." The psychological states are allegorized into the model processes. While the initial formation of the organization "will probably be" without benefit of Computers, the organizational structure is "biased" to facilitate "mechanization" of the communication, record, and display aspects of functioning as an organization.

At the time, personal computers had been available as build your own kits for a mere two years, and fledgling companies such as North Star, Vector Graphics, PolyMorphic Systems, Alpha Micro, and IMSIA were just getting started.  The kits hit the market in late 1975. The IBM Personal Computer would not make its debug for another four years. However, I anticipated the use of computers in my description:

An effort is made to integrate the interaction of human and computer in as complementary and effective a mode as possible. The 'bias' toward mechanization can be explicitly stated. I envision a stage in the development of the organization when the major functions are accomplished by individuals who sit at a consol with a VIP and interactively control data structures in the Processing System on a real time basis. In the beginning however, an organizational structure can be realized which can function without the Interactive Computer System, but will be compatible with its implementation.

In getting to the why to go along with the how, some background is appropriate.

I differentiate knowledge from information.  Knowledge is the capacity of an entity to act on information. Knowledge Engineering is the process of extracting knowledge from a person and embodying it in a system - such as an expert system program or an organizational structure. Entities that embody knowledge include persons, organizations, processes, and systems. Any such entity exists in a context or environment and has purposes and goals.  Knowledge management is the explicit organization and control of knowledge within an organization.

The function of knowledge management within organizations is to enhance the competitiveness of the organization within its environment by better enabling the organization to respond to opportunities and challenges.  These principles apply to organizations as wholes as well as to functional components within organizations.   Information, and acting on it, have always been part of the process of organizations carrying out their activities.  The difference that knowledge management makes is bringing the process under conscious and explicit control, thus enabling the application of modern principles of organization and efficiency.  It gives the company an increased awareness of how it uses information and knowledge to perform its chosen functions. 

An awareness of the theories of knowledge is not new.  Epistemology has been the subject of inquiry by philosophers for literally millennia.  Unfortunately, little or none of that informed perspective has been heretofore deemed relevant to business.  For the most part, it still is not so deemed. That is about to change.

With the advent of the computer age and the studies in artificial intelligence, explicit control of principles of organizing information and the knowledge to use that information is increasingly important to the effective functioning and survivability of the company.  (One philosopher noted that the study of artificial intelligence is just "philosophy with the tools".) It is the knowledge embodied in an organization and under its control - its intellectual capital - that enables the company to accomplish its chosen task, whether it is the manufacture of a product or the delivery of a service.  The ability to manage that knowledge effectively is critical to the long term viability of the company.

As more and more explicit and implicit knowledge engineering is carried out, information and the knowledge to use it is becoming more readily available.  The more often precise and detailed process information is embodied in information systems, the easier it is for employees to use that information to effect the organizations goals.  If the information about a process is sufficiently detailed, a person can perform the steps specified with little or no understanding of the knowledge involved.  In such cases, the knowledge is embodied more in the process structure than in the person.  Minimal experience is required on the part of the employee in such circumstances, because the knowledge has been effectively engineered into the information system processes.

In the preceding, some terms are not explicitly defined - 'understanding' and 'experience'.  Understanding, we can tentatively explain as an ability of a person or entity to articulate what is being done and provide some explanation as to why it needs to be done that way. Experience, we can tentatively explain as a measure of how much a person has performed a function (successfully).  One does not necessarily presuppose the other.

Key to the understanding of the necessary structures is the definition of knowledge as the capacity to act (effectively) on information.

Acting on information has its own presuppositions.

Incoming information.

1. Incoming information must be detected - received - sensed, etc.
2. And it must be differentiated among
   • relevant
   • irrelevant
3. Relevant information must be processed in the knowledge base.
4. Actions required in response to the information must be identified.
   In the context of a knowledge management branch in an organization, the identified actions must be delivered to the persons and/or branches within the organization that has the capacity to act, and it must be delivered in a timely manner.
5. Appropriate actions must be taken.

Let us restrict the current discussion to the context of a knowledge management branch or division within an organization which has the goal of collecting, organizing, updating, and making available information about and knowledge of processes used by the company to perform its services. 

Such a group would need to have procedures and requirements promulgated within the company for collecting information about relevant processes.  These would serve as the group's senses for collecting information. For the moment, we'll defer discussion of the form and content of such reports until after we have talked about the organization of the collected body of information. 

The group itself would need to process these incoming reports by relating them to the existing information store.  We are presuming that there is already a store of procedures in use by employees while carrying out the company's mission.  Incoming reports must, among other requirements, identify the process that was used as well as include information to enable judging how effective the use of that process was.  In order for this judgment to be relevant, the challenge or opportunity must also be identified in sufficient detail to permit comparing it to other challenges and opportunities previously responded to.  Any assessment needs to take into consideration the degree of similarity or dissimilarity between the current opportunity or challenge and the ones related to the procedure used, as well as the degree of similarity or dissimilarity between the actual procedure as implemented on this occasion and the procedure associated with the previous response process.  This requirement dictates a minimal structure to the stored knowledge repository.

At a minimum, the knowledge base must include a pairing between an opportunity and a response.  Such a paring must also be associated with an evaluation of the resulting performance in relation to the company's chosen product or service.

Opportunity - Process - Result (OPR).

1. What was the client's need we responded to?
2. How did we solve that need?
3. What did we gain as a result?

Each of these needs to be "broken down" into small description units, and, as much as possible, correlated with each other.  Such correlations require a compatible format for comparison.  As a consequence of this requirement, there needs to be some standardized formats established for documenting the opportunities, the process for handling them, and the results achieved.  However, in order for the organization to remain viable, there also needs to be a process for the standardized formats to evolve to meet changing situations and needs.

For most organizations with sufficient critical mass to actively pursue a program of knowledge management, there will have already been established numerous processes describing the development of opportunities, the processes for responding to them, as well as criteria for evaluating the effectiveness of the process. In such cases, it will be necessary for explicit knowledge management to build upon these pre-existing structures. The "business of conducting business" is already the general province of managing the company and its operations. The "Knowledge Management" branch or group  is not allowed to take these functions over, consequently the niche for explicit knowledge management is as a support function that delivers processes and procedures and provides training to facilitate the transfer of knowledge from person to system to person. The knowledge that is targeted for explicit knowledge management is the technical information and know-how associated with the product or service delivered by the company.  In reality, all aspects of the operation of any organization or group could be codified and subject to explicit knowledge management, but for the reasons stated above, that is unlikely to happen.  Knowledge management will refer primarily to the support function of managing a knowledge data base of information used for the delivery of products and services in the company.  The effectiveness of a company manager in his or her conduct of day to day operations will normally not be subject to explicit knowledge management, whereas the use by that same company manager of a company process for delivering a service or product, and its subsequent resulting profit, will be.

In the light of the foregoing, the knowledge management support function needs to be organized to perform the following functions.

An intake process for collecting detailed information on the processes identified as relevant.

An analysis process for analyzing and organizing the collected information into the knowledge base.

A delivery process for providing the results of the analysis to the users.

The analysis process must inform the intake process so as to facilitate the rapid assimilation of information into the knowledge base.

It 1977 I described such an organization 

 The organization I envision should be 'driven' by some basic requirements.

1. A Formative medium or media to abstract concerns and formulate 'problems'.
2. A Processing medium or media to relate the problems among themselves and to record the action relating to problems.
3. A Coordinating medium or media to synthesize 'all' knowledge relating to problems.
4. A Resolving medium or media to brainstorm resolutions and formulate action-solution proposals to problems.

The purpose of the organization was to "imagine" new solutions to problems in the context of the World Future Society. 

As you can see from the scope of my driving requirements, I am proposing an organizational structure which will do much more than provide a forum for idea users to review. The function of this organization will be to develop problems, develop and integrate 'all' knowledge, coordinate problems and knowledge, and to develop resolutions for the problems developed. As the organization grows and matures, its sophistication and complexity will permit dealing with the problems or metaproblems classed 'the unexpected'. For a start however, a simple system is best.

In the limited context today - a company's branch or group providing explicit knowledge management services - "'all' knowledge" would refer to information and knowledge restricted to the processes and services the company delivers.  Any information or knowledge resource relating to the appropriate OPR's must be incorporated into the "knowledge base".  As part of the incorporation process, two aspects must be taken into consideration.

In order to arrive at any evaluations of "goodness-of-fit", there must necessarily be a means to compare OPR's among themselves.  In theory, we would look for repetitions of "the same opportunity" with different processes used, and then use the results to judge the effectiveness of the processes used to service the opportunity.  In practice, opportunities are nearly never "the same". The general description of an opportunity must be sufficiently detailed (broken down into factors) to allow comparing different opportunities in a measurable fashion.  There needs to be a metric or measurement function that can be applied to two descriptions of opportunities and produce a measure of similarity.  In a like manner, the similarity in the processes applied to respond to the opportunity must also be measurable.  Also the result or benefits resulting must be capable of being compared.

As the context of life in general and business in particular is constantly changing (as was pointed out by Heraclitus of Ephesus - fifth century B.C.), provisions will need to be made for the changing nature of the opportunities responded to, the processes used to respond to them, and the results achieved. In this regard, the knowledge management branch or group will need to be continually evaluating the OPR descriptions that make up the data base to determine the effectiveness of the information that is collected to describe them.  It will be a responsibility of the knowledge management branch or group to periodically revise the "reporting requirements" for documenting the OPR's.  Standard report formats will need to be developed, promulgated, reviewed, and enforced. Consider the organization I proposed in 1977:

The organization  focuses around four branches, one for each of the driving requirements, two models, and several records.

The four branches are:

1. A Formative branch
2. A Processing branch
3. A Coordinating branch
4. A Resolving branch

Each of the branches will be responsible for certain functions and processes. To begin with, each portion of my description presumes the rest.

The Formative Branch shall be responsible for identification of concerns and asserted problems, soliciting concerns and problems, and to develop "problems as understood". this entails the preparation of a problem formulation which meets the Problem Formulation Requirement Specification. The initial Problem Formulation Requirement Specification requires that each Problem Formulation consist of a series of problem statements. Each statement shell be referred to as a "level of precision" statement or "level" for short. Each successive level statement is to be a succinct abstraction of the previous level statement. The "highest" level statement shall be from one to six words and will serve as the "Name" of the problem. The Problem Formulation shall consist of at least three, but not more than ten levels of precision statements. Problems formulated will be passed to the Processing Branch via an "Awaiting Processing" queue.

That was in 1977.  In 2001, some of this is subsumed under the company's mission and their business targets.   Essentially, the product or service that the company delivers is fairly well defined.  These represent the kinds of solutions (the resolving branch)  that have been delivered in the past.  "Problem statements" become the Requests for Proposals (RFP) that the company receives as well as the sales force's descriptions of potential sales opportunities.

A potential sale - challenge or opportunity - will naturally go through a process of expanding its description from the most abstract to the more detailed as it is researched and/or negotiated, while an RFP is already presented in a detailed format.  Resolutions will also need similar expansion.  At Keane, the resolution phase expansion has already been highly developed in the form of the Frameworks for software development.  The "solution space" for IT development and related services have been well developed at several levels of detail.  And since the frameworks are identified in the process of responding to RFP's and/or negotiated "deals", the ability to compare the delivered service is already highly comparable.

 Caution - Construction Zone 

The Processing Branch shall be responsible for relating problems among themselves and developing and revising a model "Problem Space". Each problem is a point in the problem space. The structure of the Problem Space is determined by the relations among problems. The Processing Branch, in conjunction with the Coordinating Branch, shall be responsible for revision and maintenance of the Problem Formulation Requirements Specification. For each problem, a problem space point record will be created by the processing branch and amended as applicable by all branches. The initial problem space point record shall provide for the name of the problem; the names and relationships of all directly related problems, and the date and identity of the individual making the relation; the names, a qualifier of "all" related knowledge and the identity and date of the individual making the correlation; the urgency of need for a solution and the date and identity of the person making the determination; the degree of resolution and identity of the person making the determination and the Queue Status.

The initial Problem Space Model shall be a discrete point space with the initial relation of "subordinate" (or superordinate). The general structure of the problem space is hierarchically ordered with relative "levels".

The Processing Branch shall be responsible for revising the Problem Space Point Record as necessitated by revisions to the Problem Space Model. Problems processed will be passed to the Coordinating Branch via an "Awaiting Coordinating" queue.

The Coordinating Branch shall be responsible to coordinate "all knowledge" to each problem. This involves selecting relevant knowledge and "naming" the facet of knowledge as well as identifying the relevancy qualifier, and appending these remarks to the Problem Space Point Record. Additionally, the Coordinating Branch is responsible to develop and maintain a Knowledge Space Model. The Epistemological considerations of the model will influence the Problem Formulation Requirements Specification. The initial Knowledge Space Model is a hierarchically ordered set with "levels" and an initial relation of "is a specialization of". Problems coordinated will be passed to the Resolving Branch via an "Awaiting Resolution" queue.

The Resolving Branch is responsible for brainstorming resolution solution-action for problems. In the process, requests for related problems and their point records may be requested of the Processing Branch. Requests for additional knowledge may be requested of the Coordinating Branch. In these cases, the respective Queue Status will be called "Related Problems" or "Related Knowledge Queue". The Resolving Branch will be responsible for developing a problem solution action (Resolution) Formulation and appending this Resolution Formulation to the Problem Formulation. Additionally, a "degree of resolution" estimate shall be appended to the Problem Point Record. This problem resolution, if deemed appropriate, will be returned to the Formative Branch as an "answered" or partial answer. If the answer is adequate, the Problem Space Point record will be referred to the Processing Branch for de-integration and to the Coordinating branch as a possible knowledge input. If the answer is deemed not adequate, the problem will be reformulated. Problems with resolutions awaiting Formative Branch action will be in the "Solutions" Queue.

In summary, the "Imagineers Corps" organization consists of four interrelated branches which maintains two "models" or images of the environment external to the organization. One model of the way "problems" are related and one model of the way knowledge is related. Problems are "formulated" consistent with the requirements of the structure of knowledge and the structure of problems, i.e. within the epistemological bias of the organization. All these facets are allowed to evolve in a natural way to provide for dynamic growth and increased sophistication and maturity of the organization. By having the various records on line and allowing them to be called up and viewed on a VIP screen, the entire process can be mechanized.

A significant function of the Formative Branch is to seek out "problems" for formulation. The Resolving Branch could be extended to include WFS members at large, e.g. via the medium of a special studies division of the WFS.

For further information, feel free to contact me.