Maybe I’m not as open as I thought
16 December 2007 by David Bott
Earlier in the week I went to one of the always excellent NESTA Breakfasts to see Henry Chesbrough talk about Open Innovation and its effect on policy. Followers of this blog will know that I have talked about Open Innovation before and I continue to be fascinated about the use of models to explain organisational behaviour. Since I had actually eaten breakfast at the hotel, I was able to engage in discussion with Henry beforehand (everyone else had a mouthful of croissant) about how his “brand” has become more associated with open sourcing of technology for companies that have a profound understanding of their markets rather than those who have developed major “technology platforms” and require insight into new markets – sort of “Open-Out Innovation” rather than a more commonly cited “Open-In Innovation”. His book is a thorough piece of research for the best practice in modern innovation management and his early industrial experience informs his academic “discovery” and organisation of the work. However, I was unconvinced that what was essentially a best practice guide for modern research management had implications for policy – the book is weak in this area – and so was intrigued to see what spin he would put on his findings in this environment. I should also point out that there were precious few business people in the audience!
The story of the Xerox Palo Alto Research Centre and the number of amazingly good ideas it chose to reject for internal development that went on to shape our modern world is fairly well documented, so his assertion that false negatives are every bit as important as false positives in portfolio management went without challenge.
This being a workshop rather than a lecture, we had two other contributions. The first, from the Cambridge-MIT axis, managed to derive an argument that Open Innovation demanded major Government investment in prestigious universities with global connections was going well until its logic of self-promotion became apparent. The second started well, with an analysis of networks. From the one-to-many networks that follow Sarnoff’s law, through pair-wise connections that follow Metcalfe’s law to the many-to-many networks that follow Reed’s Law, the basic mathematics of networks are well documented. A Sarnoff network is essentially limited to transmission from a central point – good for coherence but bad for novelty. Our radio and television networks are good examples. A Metcalfe network is best exemplified by telephone and e-mail and works best when everyone is available to be connected to. A Reed network allows everyone to make Metcalfe connections but also allows groups of all sizes and structures and drives. The new social networks are good examples but I have met several “power networkers” in my career and thye use this approach. The speaker extolled the power of those who spanned more than one network, but my understanding on networks suggests that, in a Reed network, spanners are not needed!! It was also rather amusing that after showing plots of several “business based” Reed network, his example based on his own university was distinctly Sarnoff!!
I left with the distinct feeling that one of my heroes had been tarnished. Perhaps it was the attempt to extract implications for the unteachable from the inscrutable, but I felt that there had been little knowledge transfer but a reinforcement of established positions. I hope I was wrong.