VC legend Tom Perkins cofounded Silicon Valley's Kleiner Perkins Caufield & Byers. Long associated with Hewlett-Packard, he resigned from its board in 2006 in a dispute over the board spying on its own directors, sparking an SEC investigation and the exit of HP's chairman. His memoir Valley Boy published this month.

Tom Perkins

Everything I know about entrepreneurial management I learned from Dave Packard. Packard, a founder of Hewlett-Packard, was a friend, my mentor and the most important influence on my life. He and Bill Hewlett, also a friend, established Silicon Valley and the foundation for everything that transpired there. In my case, the most important thing was that they were entrepreneurs who ran HP as a venture capital enterprise. They did this by incubating new businesses within the company, the way venture capitalists have tried to do subsequently.

Bringing new technologies to market is about risk and timing. When an idea comes to you, it's always a mix of people, technology, marketing, even legal issues. You have to identify the most important risk element and structure the deal so your initial dollars get rid of it. If you do this properly, a modest amount of money will get rid of the maximum amount of risk. You can then pour the money into a situation with a high upside. However, you don't want to be so early in a technology that you are funding pure research. But you don't want to be too late. Timing is important.

I've made bets that didn't pay off. Many of them! KPCB gambled unsuccessfully on pen-based computing. We were too early. I hate to talk about our mistakes, but we've made them. You just have to back away and say there are other fish in the sea, so let's go catch them.

My experience with CIOs over the years is that they are essential. But my advice to them is to remember the old Russian proverb: "Better is the enemy of good." Huge sums can be spent on information technology. You don't want to end up being noncompetitive even though you're brilliantly informed about everything. You've got to have a cost consciousness as the CIO. Don't get swept away: There will always be more proposals to fund than you can do.

Sometimes you have to throw yourself on your sword. But the underlying question is this: Did I make a mistake when I resigned from HP's board of directors? Honestly, probably I did. I was emotionally involved with the company so this was an impassioned decision. Anytime you do something in anger you'd better reexamine it. Would I do it again? I probably would. I felt the board was making a profound mistake and that the nature of this investigation into media leaks from its own members was out of proportion. I thought the methods used were ethically questionable, if not illegal. If I had not resigned, maybe I could've changed things from within the company.But maybe not.