The optimal structure of knowledge management efforts, however, is still very much unclear in our country. Some audience members worried about the creation of new bureaucracies led by chief knowledge officers and questioned whether they would be influential enough to meet today's business challenges.

Moreover, knowledge management often takes on many forms. It may come under the rubric of data warehousing, database marketing or even e-commerce.

As opposed to launching corporate-wide knowledge initiatives without clear business objectives at the outset, some speakers noted the importance of first identifying the business competencies that can be enhanced by knowledge management. Companies develop different knowledge strategies depending on whether they are competing on cost, product innovation or customer intimacy.

Another key challenge that knowledge strategists are attempting to address is the need for systems that integrate knowledge, technology and people. I note that we now have technologies at our fingertips, particularly the Web, that promise to enhance access to information. So companies must now incorporate people into their knowledge management efforts, enabling access, for instance, to experts. In this case firms should place more emphasis on capturing the knowledge and expertise of their employees, and then ensuring it is effectively shared.

Companies have a lot to gain by collaborating with their customers and encouraging them to subsidize "the innovation infrastructure". New modeling media and rapid prototyping can strengthen business relationships, improve decision-making processes and enhance the likelihood of successful outcome.

New computing capabilities can be expected to enhance prototyping and modeling processes - and thus, innovation- still further. As the marginal costs of computational modeling collapse, companies have an opportunity to engage in new experiments and create new prototypes that once would have been too costly and risky. They must manage a "portfolio of prototypes" and take advantage of the changing "risk profile" that the new abundance of computation cycles provides. "Once you have technologies and infrastructures that change your risk profile, you ask yourself a different set of questions and you begin to view your opportunities for creating value and managing risk differently.

There are many human-related difficulties associated with acquiring and managing customer knowledge. For one, customers are encountered by a wide range of organizational personnel. In addition, the motivation to capture, share and use this information is often weak. Some human knowledge is accurate, some is not. And it is difficult to summarize, synthesize, and update without being labor-intensive, but we must not forget that acquiring it requires a lot of direct interaction with customers.

"All of this stuff is pretty much useless unless IDRO instructs its companies to tailor products and services to customer needs, unless focus is done to profitable customers, unless a particular customer business problem is solved with the knowledge we have."

Finally, long-term customers are easier to serve and more profitable than new customers. How do you keep your best customers? Know who they are. Communicate with them. Use your best employees to service them. Build equity for them in the process. And do not stress price, stress the relationship. If you use customer knowledge properly, you will make customers happy and conscious of the benefits they are receiving. (Concluded)