This article focuses on knowledge management in mid-size companies, generally considered to be enterprises with $50 million to $1 billion in revenues and overwhelmingly the largest business sector. We'll discuss the return on investment (ROI) of knowledge management; clear up the confusion between knowledge and its cousin, information; highlight an example of knowledge management at work; clarify the ideal components of successful knowledge management; and examine important considerations for deploying and automating a knowledge management system.

Knowledge Management ROI

Given the elusive nature of the knowledge management concept, it may not always be possible to attach conventional ROI metrics to the initiative. It may, however, be instructive consider a model developed by a content management consultant, Professor Ann Rockley, as an indicator of the savings that can result from establishing an organized content/knowledge base:

• Average cost per employee=$0.90–$1.90/minute
• Time spent searching for information=15–30% (or an hour or two per eight-hour workday)
• 400 employees at $0.90/min x 60 minutes (conservative estimate)=$21,600/day spent searching
• 240 working days per year =$5,184,000 lost annually
• Save 15 minutes per day per user (25%)=$1,296,000 recovered

And along with ROI, which wants specifics, there is value on investment (VOI), which considers the "value" derived from an initiative. The VOI factor considers various indefinables, the improvements that you perhaps cannot put a fence around: improved internal processes, user/customer satisfaction, better integration of resources and complementary activities, and so on –the myriad ways in which the overall operations might improve.

Confused About Knowledge vs. Information?

For a couple of decades or so, there has been an ongoing discussion as to whether knowledge can be precisely defined. The quest to isolate knowledge and to identify boundaries and applications for it seems to have evolved out of the fact that because technology consistently expands in power and application, it should facilitate a logical next step: data > information > knowledge, which would also bring into play the experience and intuition of the people who use the information.

This would seem to imply a separation between information and knowledge. At a practical level, they are often—at least in midsize companies—part of the same thing. As information processing matures, information technology executives find ways to automate internal processes through creation and application of processes that actually perform a certain level of "thinking," putting information in optimally usable form at the optimum place where it is needed. To that end, virtually anyone who holds a job in today's information-driven workplace is to one degree or another a "knowledge worker," whether a high-concept thinker/planner or a service or support person dealing with customers on the front lines.

Consider this useful entry in Wikipedia (itself an excellent example of shared knowledge): "Knowledge Management System" refers to a (generally IT based) system for managing knowledge in organizations, supporting creation, capture, storage and dissemination of information. It can comprise a part (neither necessary nor sufficient) of a Knowledge Management initiative. The idea of a KM system is to enable employees to have ready access to the organization's document-based facts, sources of information and solutions.

Objectives and benefits include, among others, these:

• Sharing of valuable organizational information
• Avoiding reinvention of the wheel, thereby reducing redundant work
• Reducing training time for new employees
• Retaining intellectual property if such knowledge can be codified

People working in the corporate trenches usually agree that debating the various terms achieves little and that terminology will always mean different things to different people. Such companies are finding enormous value in the establishment of a single, unified platform that can be readily exploited by the corporate knowledge workers, whatever their tasks.

Knowledge (or Information) at Work

Information accrues from many sources: internally through the company's productivity resources; externally through communications with vendors, suppliers, regulators, litigators, associations, and others; and via conventional (mail) and nonconventional (electronic) vehicles. Books, movies, music, and anything else that can be digitized today represents potential information that a company can use in whole or in part.

The commonality is that the information must be easily accessed—under authorization, of course—and that users be furnished with the training and skills to use it to support the corporate mission through their individual responsibilities. Normally, the corporate mission itself is well-defined, and the intellectual exercises necessary to exploit captured and stored information are clear-cut and often can be included in processes as "business rules."

In a professional call center that services clients of a government agency, the knowledge base takes the form of an extensive, browser-accessed library of documents in PDF format that are maintained by the agency. Agents in the call center refer to these documents when responding to calls, either by linking to the knowledge base directly through a Web browser or by typing in the Web address. As formatted documents, the information can also be distributed easily either electronically or as printed copies.

Both the agency and the call center maintain self-service portals that are accessible by several categories of clients or client-related individuals. The portals provide a spectrum of basic information, much of it available through Frequently Asked Questions, with the actual location of the portals transparent to the clients. Visitors to the call center's Web site can link directly to the portals, which also include interactive features that allow portal visitors to further define their requirements, including profile development and cost projections.

Elements of Knowledge Management in Midsize Enterprises

Whether you're a call center, where knowledge exploitation and delivery is the mission, or a mid-size firm, where knowledge occupies a position auxiliary to the delivery of a product or service, the basic mechanics are similar. The challenge is to harness processes to capture what you need and then to make use of it as needed, where needed.

Affordable tools are available to the mid-size company to do these things, whether they are called knowledge management or not. Central to any successful implementation is the establishment of a centralized repository, actual or virtual, for the relevant information. The company also needs a systematic program for producing and capturing information for the repository; storing and retrieving it; sharing, maintaining, and delivering it; and making it available in appropriate form to support corporate decisions.

Production

Information is always produced somewhere. Internally, it can be generated as paper or electronic documents, including checks, by the company's productivity software and then stored in an electronic archiving system—now increasingly referred to as an "electronic content management system." Externally, it can come from a host of sources in a variety of forms and format and then similarly captured and stored in the content management system and recoverable in its original form.

Capture and Formatting

The fundamental requirement is that the material be digitized and thus storable in a recoverable data format. Historically, there have been two methods for channeling documents into a content management system: raw data could be printed and then scanned into the archive system or the data could be exported as a PDF file. Both methods require manual indexing of key data fields. Now, all this can be accomplished with a software command. And it has become possible recently to use productivity software to capture documents automatically as they are produced, dramatically reducing the amount of manual activity required for the electronic storage process.

Subject matter experts can also capture digital content as it is produced by using traditional authoring tools like Microsoft Office. Already-digitized material, such as CDs, DVDs, emails, etc. can be stored easily by simply assigning them storage locations and attaching metadata to facilitate search. Paper documents, whatever their source, normally are scanned into storage using optical character recognition (OCR), image character recognition (ICR), and forms processing devices, with search parameters assigned as part of the process.

Information Storage and Access

Managed files can be checked out and versioned while retaining the benefit of being indexed and managed throughout the file's lifecycle. Files flagged for full-text indexing can be retrieved by searching the content of the files through an intuitive full-text search interface. Alternatively, files can be accessed by searching the metadata fields or "indices" in a traditional query by example search metaphor. Metadata essentially is data about data—in this context, data about a particular file and data that can be cataloged and searched for use and/or management.

Metadata structure is almost infinitely variable, totally flexible in its descriptions, and highly intelligent in that it can be structured hierarchically as meta-metadata. This is usually unnecessary in midsize enterprises, where the emphasis normally is placed on such things as information subject, type, and file elements (e.g., customer, documents, transaction records, etc.). Moreover, assigning metadata to information elements also enables searches beyond the specific item sought, since it can also bring up items related to it. For example, a legal search on a specific contract might be able to recover related correspondence. The way information is stored, indexed, and accessed represents the injection of knowledge management into what historically has been a static archiving process.

Assembling Knowledge Through Collaboration and Versioning

Few if any corporate activities exist in a vacuum, and centralized information storage can be very effective in capturing the kind of intellectual capital we regard as knowledge. Cases in point might include the collaborative development of marketing plans or engineering specifications. Such plans often build iteration by iteration, with involved individuals contributing their knowledge asynchronously, level by level, until a final version is arrived at. Intelligent versioning captures the input of each participant, storing each subsequent version independently under its own identifier, building on but not altering the original. This is "knowledge capture" both in the practical and the academic sense, and it brings with it the opportunity for almost limitless extension.

Process Management

The knowledge quotient of information within a centralized corporate repository rises significantly when various of its related functions can be automated, yielding a synergy unavailable to each document in isolation. Among many others, processes such as document submission, review, approval, routing, and task notification can all be organized and automated. Examples include these:

• Coordinating document-driven processes with stage-specific schedules and automated notifications that pace, prompt, and apprise designated participants throughout the process
• Routing documents to particular knowledge workers or project teams across multiple offices, departments, or divisions for review and revision
• Consulting graphical representations for current status and stage-by-stage timelines for participant activity within processes.

Process management is essentially a matter of automating the knowledge about how a process is supposed to work and making it work in a hands-off mode.

Decision Support

Applications can be employed that exploit information available within the company in order to provide choices between alternatives based on of the values assigned to those alternatives. In the book Decision Support Systems and Intelligent Systems, Ephraim Turban and his colleagues describe such systems as "interactive, flexible, and adaptable computer-based information system, especially developed for supporting the solution of a non-structured management problem for improved decision making." It utilizes data, provides an easy-to-use interface, and allows for the decision-maker's own insights.

Distribution

Knowledge about what to do with files—where they go within the repository and where and how to send them to external partners—can be programmed into a knowledge-oriented information system, again eliminating unnecessary handling while enhancing the overall "brains" of the corporation.

Retention Management

The end stage of any piece of corporate information/intelligence is its deletion, which frees up the resources needed to store and manage it as well as the personnel requirements associated with it. For example, achieving regulatory compliance relies on having an effective retention schedule for organizational files in addition to an effective and flexible security paradigm. Intelligent document management suggests that these files have a life cycle and should be held as long as they are useful and purged when their useful life is over. The retention process should be customized with archive and retention schedules assigned to files and folders to facilitate compliance with industry-specific standards and governmental regulations like Sarbanes-Oxley and HIPAA as well as company-specific policies for document types.

Deploying and Implementing Knowledge Management

The type of information that comprises a knowledge base and the way it is distributed or presented should take into consideration a number of factors:

• Is the knowledge initiative an enterprise-wide effort or is it directed toward a departmental, mission-specific goal?
• What is the purpose of the initiative?
• Who needs the information?
• Who are the immediate users of the information?
• How will the information be created, captured, organized, stored, and distributed?

The type of information in a knowledge base can be relatively static, updated routinely, or highly current, with documents flowing automatically to the knowledge base whether obtained from external sources and scanned into the repository or generated on in-house productivity solutions, reporting systems, or specialized programs.

Readily apparent candidates for a knowledge management initiative include customer service operations and technical support. Both can be established in a hierarchical structure driven by levels of complexity. It is almost never a bad idea to start small and build out, as opposed to starting on a grand plan and having to cope with serious unforeseen obstacles along the way. In short, first prove it and then build on it.

Customer service is a good place to start, with possibly an even tighter focus on a specific function—for example, accounts payable. One of the most time-consuming and wasteful tasks within the accounting department is fielding queries from vendors on the status of particular invoices. With a centralized information/content repository installed, the payment software can direct a copy of a payment to the repository simultaneously with its issuance, along with supporting documentation. The payables clerk fielding the call can pull up the information with a keystroke or two, satisfy the query, and be back to his/her main task within a couple of minutes.

A similar case in point can be applied to technical support, perhaps an even more resource-intensive activity. By assigning personnel on a hierarchical scale of knowledge, the most complex issues can be technologically triaged, with the easy ones answered by lower-level techs who can pass the problems up the ladder of knowledge as necessary.

Automating Knowledge-Based Processes

Whether categorized as knowledge systems or not, significant advances have already been made using Internet sites and portals to screen out the easy stuff. Perhaps the most common and most fundamental of these efforts is the "frequently asked questions" approach, as mentioned in the example above. Through FAQs and their answers, many of the most common queries can be answered online through a simple knowledge base. Directions to more advanced resources, such as portals, can be posted for situations that are not dealt with in the FAQs.

Web portals provide easy-to-implement solutions for automating the delivery of information. For example, it is not difficult to set up a vendor self-service portal within the corporate Web site and then provide vendors with authorized access to their own information. Most questions relating to outstanding invoice status can be resolved with a visit to the portal, and for those that cannot, the vendor still has access to accounts payable staff.

Similarly, technical support personnel often find themselves resolving the same questions day after day, to the point that many of the answers are almost automatic. The solution: a customer portal where the common questions and the alternative solutions are posted. The lists of questions and answers can be assembled easily simply by tracking problem calls and posting them with possible solutions. Again, technical support personnel are available if they're needed.

The same philosophy can be applied in many areas through a company. For example, a university in California successfully automated much of its purchasing activity using the combination of a B2B EDI/XML solution and a knowledge bank consisting of catalogs from its major vendors. The purchasing process involves simply identifying the respective products and initiating a purchase through the appropriate electronic data interchange mechanism (e.g., X.12 EDI, XML, flat file). The entire process can be automated, all the way to the issuance of payment against an electronically filed invoice.

The point is that, while knowledge management is not solely a technological exercise, it relies heavily on technology for its execution. Corporate imagination is the other ingredient, and that involves identifying and employing the skills and experience of personnel involved with the many functions within the enterprise.

In activities perhaps less detail-based than the examples cited—for example, market planning, proposal development, or design engineering—similar methodologies can apply. The difference is in the content. Where the knowledge involved in responding to a vendor query or customer problem may be fact-based, these essentially collaborative processes may be more concept-based, with concepts of multiple contributors sharing and comparing their knowledge as the development process proceeds.

The Successful KM Initiative

Knowledge management initiatives in midsize enterprises must be centered on a business need, with identifiable objectives and foreseeable results, whether calculated in cost/revenue advantages or in judgments on potential process and related benefits that will accrue. Obviously, the expertise needed for the initiative relates directly to the tasks and functions involved. In mission-specific ventures, for example, that expertise might be highly localized, while in an enterprise-wide effort, a knowledge committee or task force would be appropriate.

In the end, the vital elements are planning, technology, and a commitment to the quality and maintenance of information. Powerful, affordable technology is available to midsize firms, and with the generous application of corporate wisdom and strong procedures for information generation and flow, knowledge management can be a productive, understandable, and down-to-earth reality.