11:06 PM

Tech Watch: Silicon Valley's greatest success is failure...

....and copious amounts of failure.

What can be learned from failure in improving the process of commercializing innovation?

This was one of the questions posed to panel members at an SRI International event I was moderating. The event was organised by Alison Murdock, president of the IDB Network.

The question about the role of failure deserves more attention, because I'm convinced that it goes to the root of Silicon Valley's success. And it points to a fundamental quality that Silicon Valley possesses, and one that is rarely mentioned:

Silicon Valley is extremely good at failure.

In fact, it is extremely good at producing massive amounts of failure.

Curt Carlson, director of SRI, one of the largest research organisations in the US, pointed out that as few as one in 20 Silicon Valley startups succeed. And even then, they are unlikely to make it to their 13th birthday--the average lifespan of an S&P 500 corporation is 12.5 years.

With the pace of change accelerating in many industries and markets, Mr Carlson argues that we have to accelerate the process of successful innovation in order to keep up. But how?

SRI believes it has figured out part of the answer...

From years of studying successful innovation, SRI has developed a week long training course for its researchers and clients. And it seems to be working, judging by the increase in successful research projects coming out of SRI over the past two years. The course is so effective in raising productivity that Mr. Carlson has asked that every employee of SRI, not just researchers, take the course.

I wrote a column on this subject recently for the Financial Times.

The SRI training is no guarantee of success in commercializing innovation. But it will help to kill an unviable project.

Or as Dave Blakely from the research organization IDEO, puts it, "We have to get better at failing early and make our mistakes earlier."

Another question asked was if there was a way to document what goes wrong in early venture companies and create some sort of central database of what not to do.

Lee Burrows, vice president at VC firm VantagePoint Venture Partners said, "We perform our own post-mortems but it is impossible to collect all the information and document it, because people would not publicly reveal a lot of key mistakes that were made. Or name names."

Nand Mulchandani, CEO of Determina, said, "What does happen is that new ventures are formed that have teams of people that have worked together in the past. That way, you can try to aggregate the experience of failure in the hope that people have learned from past mistakes."

It reminded me of a recent conversation with Mike Sheridan, a partner at Mohr Davidow Ventures, "You get back into the trenches with the people that were in the foxhole with you the last time."

The panel couldn't provide easy answers to how best commercialize innovation. All of them said recruitment of good people is key. And finding people who can work as a team is essential. Easier said than done.

Other interesting points from the panel:

Norman Winarsky, vice president at SRI International.

--Make sure you are talking with potential customers as much as possible, as often as possible. There are too many examples of ventures emerging after 18 months and finding that their target market has moved on and has no use for their product.

--Don't invent a technology and then go looking for a market.

--We have seen a lot more success when researchers are involved with customers as early as possible in the development process.

David Blakey, IDEO.

--I've seen so many companies not bother using simple prototyping tools to create a mock-up of their product, or not use graphics tools to see what the user interface might be like. These simple procedures would have shown up problems early in the development process.

--You cannot innovate by sitting in a conference room. You have to get out of the room and onto the road.

--The most successful companies we've found have a strong, heartfelt connection with the customer.

Nand Mulchandani, CEO, Determina.

--It's all about team building. Cultural integration is a big part of my job, integrating new people into the team.

--When you start up a company, you round up the usual suspects, people you know because trust is very important. Then as you grow, you create extended networks of trusted people.

Laurie Yoler, chief development officer, Intellectual Ventures:

--I don't think it is possible to have a cookie-cutter approach to innovation. Although there is a cookie cutter approach in terms of getting funding, in terms of presenting that information.

--Researchers can have a difficult time transitioning into a startup venture because they were interested in solving a problem rather than being involved in marketing and business development.

Lee Burrows, VantagePoint Venture Partners:--You need people with different sets of skills at different points in the life of a venture. We've often seen that a startup CEO might not be the best person to lead the company after a couple of years. There has to be a willingness to step aside and learn, so that in a future venture, that person has those skills.

--Revenues are the only measure of success that we look at.

Interesting comments from conversations after the panel:

Clay Bullwinkel, of Bullwinkel Consulting--Poland is becoming an interesting place for outsourcing these days. Costs of outsourcing work to India are rising as wages are increasing, and there is a lot of staff turnover. Poland is being increasingly being recognized as a viable alternative.

Gregory Ruff, G.L. Ruff & Company:-Silicon Valley is too incestuous, we spend way too much time listening to ourselves and sucking up our own exhaust. There is a lot of innovation happening in other parts of the world.

Juan-Antonio Carballo, IBM Venture Group:

--We are keeping a very close eye on Silicon Valley because we need to know who to work with, who to partner with. We are in the solutions business, so we need to know what is out there that we can use.