The Harvard Business School professor's most recent book, "The Innovator's Prescription" applies Christensen's widely discussed ideas to the health care industry where he projects significant changes in medical electronics and pharmaceuticals. But moving health care out of hospitals and clinics and into the home will require policy changes in insurance reimbursement, a fact that is increasingly taking him to Washington D.C.

Christensen spoke of his work in health care and his views on the current recession in an interview after a keynote at a supply chain summit sponsored by Microsoft.

His talk at the Microsoft event reflected his views that up-end conventional thinking about managing innovation in business. "It's the principals of good management that we teach at Harvard that cause innovation to be a crapshoot," he said, noting only 20 percent of new products by established companies and a similar fraction of startup companies ever find success.

Companies waste time segmenting markets by product and customer types, generating "one-size-fits-none products." Instead they should try to understand "the real job a customer hires a product to do," he said.

"That job is the fundamental unit of analysis, not the customer," he said. "The average customer doesn't exist."

Christensen also gave several examples of how companies fall into the trap of "going up the ladder to invest in high-profit margin products and wind up liquidating their business models."

PC makers such as Dell have fallen prey to that pratfall from Taiwan ODMs such as AsusTek. In the car industry, that's how Toyota overtook General Motors, he said, and the problem continues.

"Today the Koreans have taken the low-end of the car market, and now you have Cherry coming from China that no one has to worry about ever," he quipped.

He also reiterated his celebrated theme that companies get trapped into sustaining existing innovations and miss out on new ones. That's because existing products have their own set of resources, processes and financial models that people in the company get locked into defending.

To overcome that force, companies need to set up independent business units free to create their own processes and models, he said. IBM successfully did that when minicomputers threaten its mainframe business and again when PCs threatened its minis.

"Accountants want these new groups to share resources with the old ones because they have no way to quantify the value of focus," he said. "Creativity is rarely in short supply, it's the processes that force new ideas to conform to current practices," he added.