Considering communities in open innovation

In July 2005, Hank Chesbrough, Wim Vanhaverbeke and I were wrapping up the manuscript for the Oxford book. We had two drafts of all the chapters from the authors, and were on track to finish the manuscript by August (which eventually proved to be October with hiccups merging all the references).

The problem was, we had no final chapter (Chapter 14). So in July we all met in Berkeley (followed by meetings the following month at Academy and again in Berkeley) to decide out how we wanted to conclude the book.

The original goal of the book was to provide an up-to-date and complete orientation to the academic study of open innovation. The three editors wanted to make the ideas and research approaches clear to those new to open innovation research — particularly doctoral students who were just beginning their careers as innovation scholars.

So for the concluding chapter, we tried to come up with a comprehensive framework for open innovation research by categorizing what had been done and what might be done. One categorization was by research design. I was particularly proud of the classification by level of analysis, as summarized by Table 14.1.

However, it turns out that table (and classification) had an important omission. Last spring, when I went to write my keynote for one of the EURAM tracks on open innovation, I realized that we’d left out communities.

There really was no excuse. I’d been researching open source since May 2000, and even had an open source chapter (Chapter 5) in the book. I’m guessing my problem was that my work (until my paper with Siobhán O’Mahony) was always about the firms, not the communities.

I recently got a chance to correct that omission. The best papers from one of the EURAM 2007 tracks are to be published this May in a special issue of Industry and Innovation (the CBS journal). Karim Lakhani and I were invited to write a concluding paper for the special issue based on our respective EURAM keynotes. In looking through the two PPT decks, the common theme that jumped out was that of community innovation — so that’s what we wrote about.

One goal of this year’s paper was to fill the gap of communities as a level of analysis in open innovation. Another was to make sure that those who study innovation communities will think about how it ties to open innovation. But mainly we wanted people to be more rigorous on how they defined and used the concept of innovation communities.

After Wim and the special issue editors gave us some feedback, we finished the paper a few weeks ago; it should be published later this year. We don’t claim it’s an AMR-style treatment, but hope it will stir thinking by innovation researchers on how to combine the ideas of communities and innovation — and to do so in an open innovation framework.

I’ve posted the PDF on my website and would welcome feedback about what we got wrong (or right).

EU does OI

The European Commission has launched its “Enterprise Europe Network” to help 23 million small/medium enterprises in the EU.

Parts of this seem to be standard EC industrial policy. But there is a wrinkle: recognizing the interdependence of these firms as part of the network of suppliers and complementors for Europe’s large multinationals. As PublicTechnology.net noted:

Finally, many multinational companies are opening up their innovation systems so that the 'Open Innovation' concept is now becoming a reality for business. Many multinationals have established structures to purchase or license innovative products with third parties. Brokering deals with these large companies requires skills, experience and a negotiating power that many SMEs do not possess, despite the potential of their innovations.

So this hits at the fundamental synergy between little companies and big companies, as with biotech startups and big pharma. The little companies (as in Teece 1986) lack the resources to commercialize their internally-generated innovations, while big firms need more and more innovations to grow and their corporate R&D labs are too big, too slow, or too inefficient to generate all that they need.

As a researcher sitting here in the US, a standard problem is that it’s hard to tell whether to take these government initiatives seriously. It reminds me a little of a paper I wrote back in 1994 (published in 1996): Japanese policymakers were reacting rapidly to US “National Information Infrastructure” policy, even though that policy was clearly going to have little impact on how US companies built the Internet.