Get email delivery of the Cadence blog featured here
At SEMICON West this year, I attended an interesting presentation by Steven Forrest of the University of Michigan on Moving Innovation from the Lab to the Marketplace. He said that one of the biggest problems is that universities all have offices of technology transfer who think that they are judged by how many royalties they bring in. But potential licensee companies hate that attitude and avoid engaging. He told his office of technology transfer that the only thing that he cared about was how much technology got transferred. The royalties would take care of themselves.
If you think that is bad, then it used to be worse. Prior to 1980, any technology developed under federal funding (such as the NSF or NIH) had to have its IP rights assigned to the federal government. But it was a one-way valve and there was no real attempt made to proliferate and commercialize any of it. So, in 1980, the Bayh-Dole Act was passed, which allowed universities (and anyone else receiving federal funding) to handle commercialization instead of the federal government. That act is where all those technology transfer offices came from.
On the way home recently, I listened to a recording of a meeting at Stanford at which Marc Andreessen talked. If you don't recognize his name you almost certainly use something that he pioneered. He was the main author of Mosaic, the first graphical browser. Once it was a runaway success, he and his colleagues were swamped with support issues and so they applied to the NSF to, basically, create a support team. The NSF turned him down since that is not what they do. Marc says he owes them a "huge debt of gratitude."
Instead of remaining at the University of Illinois, he moved to California and founded Netscape. He is now one of the figureheads at Andreessen Horowitz, a venture capital firm, often known by the geeky name a16z. (This is a computer science shorthand for long words like internationalization, which is abbreviated to i18n because there are 18 letters between the i and the n; apparently these are known as numeronyms.)
Marc pointed out that there are really two models of technology transfer from universities. The first is the traditional one that Steven Forrest described, where a licensing organization is set up and usually tries to maximize the revenue generated by the licenses. This seems to work reasonably well for certain types of IP such as a drug patent, but for most stuff to do with computers, it works really badly. The reality is that the patents don't contain the "secret sauce", which is actually generated by hard work over several years. In fact, Marc said that sometimes they are forced to license patents even though they are useless, and they regard it just as a cost of doing business, a cost of getting the company up and running.
The second model is the one practiced by Stanford and Berkeley. Give the technology away. The money comes back when companies are wildly successful and the founders give money back to the university that was a key component of their success. That is why the Stanford campus contains a Jim Clarke building, a Jerry Yang building, a Bill Gates building, and ones named after Hewlett and Packard. So instead of going for licensing fees, this is the wealth creation and philanthropy business model. It is not just money, either. Successful individuals also flow back to campus and bring up the next generation.
He was asked about success and failure factors. I have a saying I like to use, that I assume is not original, that "you can only transfer technology in people's heads." He agreed, although not in those words. It was absolutely critical that a lot of the students and/or professors who deeply understood the problem came across and worked full time for the company for an extended period, like five years. There is just no substitute for having the core team, especially the CEO, be people who really know the technology and are really full time. It is not necessary that the professor is full time provided the key grad students are, and the professor can sponsor, be on the board, or just return to academia and play no role.
In fact, the opposite is also true. Marc pointed out the dominant failure model. The professor starts the company, but only works part time. He raises funding from a second-tier VC, they hire a mediocre CEO. The company flies into the ground.
A final interesting point Marc made: You may well have read a famous essay by the Englishman CP Snow written in 1959. He was an odd mixture, both a chemist and a novelist. His essay is on the two cultures, which at the time were primarily physics on the one side and liberal arts on the other. He had a foot on both sides of the divide. The famous quote from it is:
A good many times I have been present at gatherings of people who, by the standards of the traditional culture, are thought highly educated and who have with considerable gusto been expressing their incredulity of scientists. Once or twice I have been provoked and have asked the company how many of them could describe the Second Law of Thermodynamics. The response was cold: it was also negative. Yet I was asking something which is the scientific equivalent of: Have you read a work of Shakespeare's?
Andreessen's interesting perspective is to read the essay and substitute "physics" with "computer science," and it reads very well over 50 years later.
You can listen to the podcast here.