Centerfield

Why Big Research Labs Are Obsolete

This article was motivated by the appearance of a muddled and clueless New York Times article New York Times article about the loss of corporate research labs.

The real fact on the ground is that big corporate research labs are obsolete. That's because it's pretty rare for the company that finances the research to be able to take advantage of more than a handful of the new ideas. There are two reasons for that. Maybe the most important is that the management of companies big enough to have research labs is rarely agile enough to go spinning off new divisions actually ept at the new thing. That's why Xerox never made money on GUI tech, and why AT&T had the worst commercial UNIX.

The other reason, less well known, is that the researcher usually needs to be personally involved for commercialization to succeed; until very recently, lab hires were told they would stay in research.

Meanwhile, universities do as well or better at turning out supporting ideas and enlarging what high-tech companies can do in that kind of hit-or-miss way. And supported grad students are often persuaded by the support and the fact of related fields to at least take a look at joining them. But companies still have to overcome alot of internal hierarchicalness, NIH, and other problems to take best advantage of hires with new ideas. Mostly, companies get smart people with their ideas left on the floor.

As Ford's Model T, Intel, and a host of other companies have shown, innovators creating startups is the most efficient way to get ideas in customers' hands by far.