January 2005

Knowledge Management Doesn't Come In A Box

By Harold Hernandez

In day-to-day life, there are many examples of how components can be interconnected to create a desired solution. Although Apple Computer has now hidden the computer inside the monitor, there is still a host of external components that will likely come together before a basic desktop configuration is complete. Along the same lines, a home entertainment center that includes a TV, multimedia receiver, VCR, CD/DVD player and speakers is im-practical and generally not available as one all-inclusive unit because the cost would be relatively high and odds are that one or more of the component parts would not be exactly what you were looking for. The key is that with these consumer electronics components, interoperability is virtually guaranteed. This allows you to choose from a wide range of component parts and feel confident that everything will work well together.

With enterprise-class solutions, the challenge of interoperability is so great at almost every level, most organizations are willing to make substantial concessions in terms of requirements, business processes, ease of use and every other buying criterion in exchange for the assurance that all the component parts will work together. Organizations understand that CRM, ERP and other enterprise-class implementations will require compromise, and these companies are now reasonably prepared to accept that and move forward with those types of solutions. Unfortunately there is a knowledge gap when it comes to knowledge management processes and solutions, and that has resulted in a string of implementations that, at best, have failed to deliver on expectations.

The core challenges are that knowledge management is relatively new and remains a broadly defined concept. The variables that should be taken into account when considering a KM implementation require a level of understanding, vision and coordination that goes back to the early days of ERP implementations. Those projects consisted of tying together independent components that did not have the interoperability advantages of today’s consumer electronics, or the still-debated benefits of today’s all-in-one packages. Getting it right from a KM perspective remains a huge challenge. Even getting to the point where you can articulate exactly what KM opportunity you are tackling can be daunting. But even rocket science can be explained, so let’s take a look at a basic checklist of considerations that go into a KM decision.

At the highest level, you have organizational considerations and KM components to think about. On the organizational side, you may have to consider one or more of the following:

• Strategic plan. What are you doing, and for whom? Are you improving search functionality for a specific type of user or are you tying together all enterprise KM initiatives into a holistic vision?
• Executive commitment. Does the organization understand the risks, rewards, requirements and timeline and agree that KM must become part of the organizational fabric? Is there an executive sponsor driving this initiative?
• Stakeholder commitment. Are the users, contributors and support staff well informed and supportive of the respective requirements and proposed changes? Does anyone feel his or her job is threatened or that he or she will be burdened with additional undesired work as a result of your KM project?
• Type of knowledge required. Will your users be dealing with reasonably predictable and routine questions that have known answers, or are the users typically working in a more unique research mode? Are they engineers or end users?
• Knowledge delivery approaches. What are the intended delivery channels (bots, search, push, collaborative browsing, library, portal, machine-to-machine, etc.), and what is the process to properly format the knowledge for the channel and user?
• Required versus available enterprise resources. Identify your required resources early in the process. Compare required time, budget, subject matter experts, various IT staff, knowledge base engineers, etc. to the available resources, and identify your gaps. Please note that content management and authoring requirements are routinely underestimated, and this results in substantial content, data and/or knowledge-related challenges.
• Business process redesign plan. In the event that you are able to implement a KM solution using all of your current processes, you can skip this step. It is more likely that the full benefit of your proposed solution will only be realized through some changes to the existing processes. New processes should be defined to your best ability early in the design phase and revisited at regular intervals.
• Organizational change management plan. How do you implement your proposed KM vision while sustaining current requirements and operations?

For an understanding of the KM components needed, you’ll have to look at your organizational considerations relative to the following:

• Type of knowledge users you will be supporting: creative, routine questions, internal, external, etc.
• Data/content sources: single or multiple sources? Nature of content? Content submission rate? Batch, stream, pull or other content acquisition process?
• Content source formats: What formats will the process be required to accept and process?
• Text mining: Will the source content be manually processed? To what extent? Will text mining be leveraged in order to substantially automate the content management requirements?
• Knowledge/rules authoring: Can any content be directly published from the automated content management process? Who will the knowledge authors and reviewers be? What is the project authoring workload? Will those be full-time assignments or collateral duties? From where do these resources come?
• Knowledge utility/management: What is the content review process? Is it manual or automated? How will knowledge value be measured?
• Workflow management: Is workflow management a module of the proposed KM suite or will it be externally imposed? Who owns workflow design and oversight?
• Reporting: What tools will be used to report on KM processes and overall performance? What metrics will be applied? Who will require reports, which reports and at what frequency? What actions/reactions can be anticipated or required as a result of the report date?
• Metadata: Is there already a metadata schema in place? Can the proposed KM solution conform to the existing standard? Could or should anyone outside the scope of the current KM initiative leverage the proposed metadata?
• Application administration: What are the required resources for ongoing application administration? Will they reside within IT or operations?
• Hardware requirements/admin: Does additional hardware need to be purchased in support of your KM initiative? Who will own and maintain the hardware? Do the required skills already exist within the organization?
• Knowledge base development and maintenance: Is this required for your particular KM design? Will it be a one-time or ongoing need? Should the resources be internal or contract? What is the rate of change in the key variables within your target knowledge universe?
• Knowledge delivery channels: How will knowledge reach the intended users? Will bots drive the knowledge to the users? Will the knowledge be leveraged in a self-help environment, for live agent support, by machine-to-machine interaction or via some other configuration?

Now that you’ve answered all these questions, you are ready to go. Depending on your final objectives, you may be able to leverage one of the available KM packages that are out there. If you require some combination of packages or point solutions, rest assured that both the technology and skills are available to ensure a successful implementation from proof of concept to final delivery. Try to avoid shopping for a solution until your requirements are well defined, or you will likely end up stuffing your organization into a box.

Harold Hernandez is an independent consultant with 14 years’ experience in the design, implementation and optimization of customer contact operations. His technical and operational experience supports the delivery of cost-effective results across all areas of a modern call center. He can be contacted at [email protected].

[ Return To January 2005 Table Of Contents ]