December 5, 2009

Open innovation and Silicon Valley

As California’s second most famous open innovation scholar — a distant second — I sometimes get emails from people asking about open innovation topics. This week a couple of visitors came from Japan to pick my brain, and we got talking about various open innovation topics, including markets for IP and innovation.

One question I’d heard before was “how is open innovation practiced in Silicon Valley?” It’s been my impression that the various Chesbrough books on open innovation had less of an impact on Silicon Valley than almost anywhere else. Yes, HP created an “Open Innovation Office,” but that’s just a new name (and broader responsibilities) for the existing university relations staff.

From my own consulting, research and speaking, I think open innovation has had a big impact in Europe, and in the Midwest. It’s also had an impact on South America (Brazil, Chile) and Asia, but it doesn’t seem to be as organized from an academic sense. I don’t see much of a groundswell here.

Why? One factor is that (in my decidedly unscientific anecdotal sampling) is that SV entrepreneurs don’t seem to read books. I don’t know if it’s because they are too busy or too arrogant, but business books seem to have less impact here than in say, Fortune 500 America, England or Japan.

But a more fundamental reason is that I think open innovation was practiced here before Chesbrough’s pathbreaking 2003 book was published. I’ve talked with Henry about it, and he sees my point — although he’s so polite that it was hard to tell if he was agreeing or being noncommittal.

After all, the first chapter of Chesbrough’s 2003 book is about Xerox’s Palo Alto Research Center. To put it bluntly, the copier executives from upstate NY didn’t know how to monetize all these great technologies, but eventually (as Chesbrough chronicles it) they use a combination of licensing and equity positions to capture value from the spin-off companies that are going to be created anyway.

It was an important and necessary adaptation by Xerox, since spin-off companies are a way of life here. The “silicon” in Silicon Valley came from semiconductor spinoffs (starting with Fairchild in 1957), and the idea of quitting to start a new company is well-entrenched. (A friend of mine, Martin Kenney, co-authored a 1990 book that argued that the spinoff-culture of Silicon Valley and Route 128 (still important then) were fragmenting American abilities to mass product new technologies; he’s since edited an important book praising the SV process: Understanding Silicon Valley.)

But more fundamentally, who were these companies that put the “silicon” in Silicon Valley? Merchant semiconductor manufacturers. These were companies that got all their revenues from component that went into the systems designed and built by other companies. And this was the norm in SV if not the US electronics industry.

Yes, IBM was vertically integrated (until the wrenching transformation chronicled by the 2003 Chesbrough book) — as was RCA before it committed seppaku — but such vertical integration is now long gone from US electronics companies. Meanwhile, NEC, Toshiba, Samsung, LG, Siemens, Ericsson, Nokia and other Asian (and European) companies remain among the most integrated of firms in the ICT sector.

Buying components is one of the three modes of acquiring external innovations, along with licensing knowledge/information/IP and custom solutions. Silicon Valley has thus been built around OI for 50 years, even if it wasn’t called that. Open innovation is relevant to SV, but you’d have a hard time selling it to local startups as a new practice.


Kai said...

I agree that the concept of "Open Innovation" is not new as Eric S. Raymond's 1997 essay will attest to.

I do believe however that Open Innovation has now become a core management discipline, and that is what's new. Business is now realising the benefits that the Open Source model has reaped for so long such as access to a wide assortment of expertise knowledge and the speed at which it can be transferred.

Although I haven't read any of Chesbrough's books, I am very familiar with OSS principles, so what I think is key to the success of Open Innovation is knowing how much to open, without of course, giving away too much. What OSS does very well, is that it delivers a base and foundation to build on, which in most cases is closed, meaning you can't change its fundamental principles, except to fix bugs. The focus is to build upon this framework (foundation).

Companies that embrace Open Innovation such as Apple and Google, do just that. They build a (proprietary) foundation and then open this up to the community for collaboration. What's great is that they avoid giving away the secret sauce that goes into the base, while reaping the benefits of all the different flavours that users add to the mix.

It's definitely a winning combination.

Joel West said...


This is a common misperception. Open source is not open innovation or even a precursor to it.

The free/open source software movements started about creating technology. Open innovation is about a business model to make money.

Sometimes the two overlap, sometimes they're completely different. Figure 5.1 in the 2006 open innovation book discusses all 4 possible cases (see Chapter 5 of the online manuscript. )

If you want to compare things to open innovation, you need to read some of the research that defines what it is, including at least one article or chapter by Chesbrough. See Chapter 1 of the 2006 book (also available online) or Chesbrough's 2003 articles in Harvard Business Review or Sloan Management Review.


Kai said...
This comment has been removed by a blog administrator.
KStark said...

There are a few things that are uniquely different about Silicon Valley. One is that job mobility is so common, that open innovation happens just by circling talent from one company to another (to paraphrase something a University Tech Transfer Exec once said "the most powerful and common form of Tech Transfer that Universities practice, is that of graduating students to be hired into companies").

Second, the semiconductor industry became increasingly flatter throughout the '90s because of foundries popping up everywhere (think the MOSIS process), and cheap availability of chip design software. Now, instead of investing $Billions$ to create a semiconductor fab, you could simply design your chip on your computer, and ship it off to a foundry, and for just a bit of coin, you could get back a handful of chips for testing. This caught on big for graduate programs, and extended to enable complete virtual device companies: design/fab/test/package could all be outsourced. Not exactly R&D outsourcing, not exactly open source, not exactly open innovation, but this helped create a sustainable and competitive semiconductor community where anyone with a niche idea/product could make it happen without heavy infrastructure investment.

Finally, there are still strong elements of myopic "not invented here" superiority within Silicon Valley. As a UC Berkeley Prof. friend told me "if SV companies have a question, they just send someone to Berkeley or Stanford to walk the halls and talk to students and faculty. They believe they can get any answer they need, but if they can't, then an answer doesn't exist (or it is not worth seeking out)." It will be interesting to see if the momentum that has carried Silicon Valley so far can continue, or if continued globalization of technology and talent will cause change.

-Kevin Stark, NineSigma, Inc.

Joel West said...


I would say only one of your examples is uniquely Silicon Valley — the extreme job hopping.

The two largest fabless semiconductor companies are Qualcomm and Broadcom, 300-400 miles away.

And NIH is a function of individual personality and corporate culture. There are certainly NIH companies located elsewhere, and I've seen SV companies very carefully studying competitors' offerings before starting their own efforts. So while NIH is a problem here, it’s not uniquely a SV problem.