December 10, 2015

Finding the limits of open innovation

In preparing for #AOM2016 and another session on open innovation, one of the potential participants raised the issue of the limits of open innovation. I am certain sympathetic: since 2006, I have noticed “open innovation” used as though it were a sacred incantation or magic pixie dust that somehow could turn lead into gold or gruel into caviar. And as has long been said, a theory of everything is a theory of nothing.

It seems to me that idea of “limits” has at least three possible meanings.

1. When Does OI Work?

OI started as a normative managerial theory in a book from HBS Press. So one limit on OI is on when it works and when it doesn’t work.

I have heard Henry Chesbrough say for many years that we need to know more about failure of OI in practice, including at the 1st and 2nd annual World Open Innovation Conference. The oldest example (I know of) for this call is his 2012 interview with a Swedish website, where he said:
The question of unintended side effects for open innovation is one area where we need a lot more work since academics are still publishing open innovation success cases for the most part.  Companies are trumpeting their successes; consulting firms are packaging open innovation services for interested clients – none of these groups so far have had much to say about open innovation failures.
As Marcel Bogers and I surmised in citing this lack of open discussion of managerial failure:
Finally, little is known about the failures of open innovation. Chesbrough speculates that this is because companies and consultants are trumpeting their successes and hiding their failures, thus making it difficult for researchers and managers to learn from those failures (Pop, 2012).
2. What Is and Isn’t OI?

Not everything that is called OI is actually OI. This is a challenge I have faced on the program committee for both WOIC conferences and three special issues (RP, ICC, and now CMR) on OI.

Perhaps Chesbrough’s first OI definition was published in 2006:
Open Innovation is the use of purposive inflows and outflows of knowledge to accelerate internal innovation, and expand the markets for external use of innovation, respectively. Open Innovation is a paradigm that assumes that firms can and should use external ideas as well as internal ideas, and internal and external paths to market, as they look to advance their technology (Chesbrough 2006: 1)
The latest Chesbrough definition of OI is
open innovation is a distributed innovation process based on purposively managed knowledge flows across organizational boundaries, using pecuniary and non-pecuniary mechanisms in line with each organization’s business model. (Chesbrough & Bogers, 2014: 27).
It allows for non-firm organizations to utilize OI, although (unfortunately) it implies that the external partner must be an organization — ruling out out organizations sourcing innovations from individuals (e.g. crowdsourcing).

Still, from this, I would suggest at least two limits on the definition of open innovation.
  • Open. Knowledge flows across firm boundaries are needed for it to be “open”. By this measure internal contests may use OI (or crowdsourcing) principles, but do not involve firms being open (in the sense of having permeable firm boundaries).
  • Innovation. I sometimes see “open innovation” papers with no evidence of an actual innovation (at least as defined by Chris Freeman or the pioneers of innovation studies.) Getting a crowd to provide free user generated content on your website may be (or once have been) an innovative business model, but it’s not open innovation.
Conversely, just because something is not called open innovation doesn’t mean it isn’t: practices similar to open innovation has been around for decades (if not centuries). For example, much of the research on how firms use open source software in the period 2000-2010 is about open innovation, even though it isn’t called that. Similarly (as Frank Piller and I noted in our 2014 book chapter), some economic activity is both user innovation and open innovation, but neither is a proper subset of the other.

3. What are the Limits to OI’s Theoretical Predictions?

There is a minor debate whether OI is a theory, paradigm, framework, phenomenon, or something else. (A similar and much longer argument exists over the “resource-based view” vs. “resource-based theory”). That said, empirical (including qualitative) OI academic research over the past 10+ years has contributed to theory by making causal claims about what happens and what doesn’t.

Every theory in any field has boundary conditions — conditions when the theory applies and when it does not. These are the limits of the theoretical predictions that can be made by a theory (or paradigm etc.)

Sometimes these limits demark a frontier: I would argue that crossing firm (organization) boundaries is an essential prerequisite. At the same time, that frontier can change:
  • Chesbrough’s 2003 HBS book emphasized economic transactions between organizations, but West & Gallagher (2006) identified the importance of individuals as a source of innovation (and that they might have non-economic motives).
  • The HBS book assumed that OI only applied to profit-making firms, but Chesbrough and Di Minin (2014) have shown how the principles of OI have been applied in government and not-for-profit settings.
Finally, as in any theory, researchers seek to define the moderators of any causal effects. Yes, we believe the direct effects are true, but under what conditions? Tens of thousands of papers have been published by finding the mediators and moderators of well-accepted causal theories, and such papers have earned tenure for hundreds (if not thousands) of researchers.

November 26, 2015

WOIC: Final Thoughts

#WOIC2015 is now a memory as everyone has left Silicon Valley and back home. Now that I've had a chance to get home (and pack up and leave again), let me offer a few thoughts on the 2nd World Open Innovation Conference -- not as program chair, but as the owner of who blogs about open innovation related conferences.

Here are a few thoughts from Friday’s session.

I often find the final day of a conference a let down -- either because of the content or because I'm just plain worn out. Meanwhile, at good conferences, there's a twinge of disappointment at having to end what has been a great opportunity to hear and meet with like-minded scholars. On Friday, I definitely had more of the latter than the former, despite continuing to fight a sort throat that limited my ability to carry on a conversation.

Industry Sessions

The opening industry session was the best of all that I heard at WOIC. Yes, they focused on healthcare, which has been my main teaching and service focus since changing employers in 2011. It's also that (as my pharma friends told me for years) in the end saving lives is more important than (although not as profitable as) developing game software.

However, we were also fortunate to have a great panel of industry experts focusing on healthcare innovation — organized by Solomon Darwin. (I summarize the session on my BioBiz blog). While my industry focus has changed, there are two reasons I think I would have enjoyed it in my earlier life. First, Sanjita Reddy offered a very cogent explanation of why the superior efficacy and efficiency of her father's Madras-based hospital chain was about more than just lower labor costs. Second, chronic tech entrepreneur Pramod John provided a textbook explanation of how structural inefficiencies of pharma distribution are adding friction (i.e. cost) to the system and are ripe for disruption.

In the closing session, Henry Chesbrough led a panel discussion of three innovation managers: Lawrence Lee, Xerox Parc; Sammy Haroon, Baker Hughes; Amit Varma, HCL. He asked a series of questions of how they manage (and encourage) open innovation in their firms. Here are the discussion questions he asked:
  1. Innovation process. How does the transfer of ideas/projects work. Who is involved in the decision-making process
  2. Systems. What metrics do you use to measure the impact of (Open) innovation post-transfer? How do these align with overall company performance metrics?
  3. People & Culture. What are the incentives/motivations for employees to participate in open innovation? What are the cultural challenges during the transfer process and how do you try to overcome them?
Lee talked about Parc's new role as a stand-alone subsidiary of its (in)famous parent. Haroon described the $1 million internal innovation contest that was so successful, it ended up awarding four (instead of one) prizes. Varma described how HCL’s top innovators (in the US, India or elsewhere) win a Mercedes, and the cars in the parking lot serve as an ongoing reminder of their success.

Research Sessions

I was not presenting in the second day, but did attend three academic sessions, moderating one. (I actually was the one who assigned the moderators, and was quite pleased at both the fit and the value these experienced scholars provided).

The first session focused on open innovation in SMEs, a major theme in our 2014 book and the work of Vareska van de Vrande. The session was moderated by Sabine Brunswicker — co-author of the 2014 SME chapter — who emphasized the importance of value networks to the success of small and medium enterprises practicing open innovation. The first paper (presented by Chiara De Marco) looked at the strategies of eight Italian SMEs, while the third (presented by Katie Hyslop) examined those of three Austrian SMEs. In between, Freek Meulman examined something completely different: a new process by which an intermediary could identify potential SME suppliers (or collaborators) for inbound OI.
The second session was one that I moderated, on three papers that fit the network level of analysis for open innovation (as defined earlier by our 2006 and 2014 books). Using my chapter from the latter book, I summarized the prior research and encouraged scholars to study this area. (In retrospect, I should have also mentioned the 2006 chapter on local spillovers and open innovation, since two of the papers focused on local clusters).

The first paper (presented by Fathiro Putra) examined the tightly localized clustering dynamics of a specific industrial park, High Tech Campus Eindhoven. Instead of a cluster of firms 50 or 100 kilometers across, the authors examined the development of a local cluster of several square kilometers. As we showed in our 2006 chapter, local spillovers and other agglomeration effects are an important element of open innovation.

The second paper, by António Santos, examined the efforts by the Portuguese government to establish 16 industry-specific clusters. He came up with an interesting classification to contrast these clusters on two dimensions. One was the degree to which open innovation was considered important by the firms, and the second was the degree to which the cluster positively impacted (open) innovation practices.

Andy Zynga
The final paper, presented by Andy Zynga, sought to explain the timing of when firms build capabilities that aid their open innovation efforts. It used the responses of more than 700 firms to an online survey dubbed the “NineSigma Open Innovation Diagnostic Tool.”  (In retrospect, while the data was gathered by an intermediary, the actual study had little to do with how intermediaries manage a network of collaborators).

In the final session, John Ettlie moderated three papers on industry-specific studies. Jason Li-Ying explained licensing strategies of Chinese firms. This wwas followed by a study of nanotechnology collaboration presented by Arman Sadreddin and T.J. Hannigan discussing the hybrid joint venture of GM, Daimler and BMW. In this regard, this (and some other sessions) reminded me of trips to the Industry Studies Association, where the focus is on research informed by a deep understanding of industry dynamics.

Future of WOIC

Although there were some overlaps with our 2014 conference, in many ways this was a very different conference. There were less European academics, particularly senior ones -- who may have come last year to visit the Napa wineries. a the same time, our convenient Silicon Valley location brought a lot more industry participants, changing the character of the discussion in a way that was noticed by both the academic and industry attendees.

We are waiting to see the surveys before making plans for the 3rd World Open Innovation Conference. We expect it will be in December 2016, but have not picked a location. Ideally, we would have a plan that attracts both strong industry pariticpation and also the depth of academics we had last year.

November 24, 2015

Learning User Innovation by teaching it

They say you never learn something until you teach it. That is one reason why we academics love to teach in our own area of research (whether it’s our previous research, our in progress research, or merely in the field where we are research active). This month I spend part of my innovation strategy class at UCI on user innovation.

Thoughts in 2014

With more than 15 published papers (articles or chapters) about open innovation, I’m guessing many know me mostly for my OI work. (Clearly the majority of my cites have OI in the title). Of these articles,  two articles in 2009 and 2012 compare open and user innovation, the latter with Marcel Bogers (who did his diss at EPFL on UI).

My most recent effort comparing OI and UI was last year’s book chapter in New Frontiers in Open Innovation. I wrote it with Frank Piller, one of Europe’s best known open innovation scholars.

One goal of the chapter (entitled “Firms, Users, and Innovation: An Interactive Model of Coupled Open Innovation”) was to contrast open and user innovation. Our thoughts were summarized in Table 1 (before we went on to look at coupled OI).

Open Innovation
User Innovation
Core references
§  Chesbrough (2003, 2006)
§  Von Hippel (1988, 2005)
Focal actor of study
§  Firm (R&D Lab)
§  Individual user
Key principles
§  Knowledge is widely dispersed beyond any one firm
§  Innovations must be aligned to a firm’s business model
§  Firms should embrace both internal and external alternatives
§  Users have unique “sticky” information
§  When enabled, they will solve their own needs
§  Many will freely reveal to others
Focal object of transfer
§  Technological knowledge in form of IP or technologies
§  Information about needs and ideas how to transfer need into solution
Typical institutional arrangement for knowledge transfer
§  Research contracts
§  In- and out-licensing; IP transfer agreements
§  Tournament-based crowdsourcing for technical solutions
§  Lead user method
§  User communities
Representative IP practices
§  Patents
§  Licensing contracts
§  Free revealing
§  Open source or creative commons licenses
Governance of innovation process
§  Private model
§  Collective or private-collective model
 Motivations of actors to engage in distributed innovation
§  Monetary incentives
§  Innovation is seen as a “money market”
§  Incentives of self-use
§  Social incentives
§  Innovation is seen as a “social market”
Key managerial decision
§  Building absorptive capacity
§  Defining and defending IP
§  Internal organization for OI
§  Defining metrics for OI
§  Identifying lead users
§  Establishing bridging strategies to lead user innovation
§  Defining fair regimes of coordination
§  Opening- up IP
Other streams of related research
§  R&D networks / strategic alliances
§  University-firm research contracts
§  Absorptive capacity theory
§  “Voice of the customer” methods of market research in innovation
§  Participatory design
§  Social production
Table 1: Contrasting open and user innovation

In teaching earlier this month

Insight #1: Three Types of User Innovators

My class used the Bogers & West (2012) paper to discuss how open innovation compares to vertical integration and user innovation. It’s a short conceptual article that’s relatively light on jargon; the article also provides a bridge between a brief OI lit review and the Lettl et al (2006) article on medical device UI by doctors.

One idea is that there are two types of user innovators. The article emphasizes UI is by individuals but “A limited amount of research considers innovations by user firms.” The Piller & West (2014) chapter (which they did not see) also positions OI as being about firms and UI being about individuals.

In discussing UI, I realized that this concept in the earlier article was inaccurate. In the end, firms don’t create inventions; people do. The IP world recognizes that patents and other creative works are created by individuals — not firms — even it ends up being assigned to their employers. (This is not to ignore that the work to bring it to market is done by a firm — only to focus on where the discovery or idea comes from originally.)

So really, research on user innovation has considered (at least) three different types of user innovators:
  1. Individuals acting on their own behalf (citizens, consumers, etc.). This is the focus of most of the UI work of von Hippel and others of the past 20 years.
  2. Employees working on their own behalf. I would expect this would cover most research on user innovators who become user entrepreneurs in the course of their job, such as that the doctors in Lett et al (2006). (Since firms control the IP of their employees, I imagine this will mainly be in a university or other academic setting).
  3. Employees who are innovating on their employer’s behalf. The exemplar here are the IT system administrators who decided to patch the NCSA HTTPd server, creating the open source Apache server (Behlendorf, 2006); since the sysadmins worked for companies that operated web servers, this was not only within the scope of their work but — as it turned out — crucial to the success of their web hosting efforts in the late 1990s. (From a quick read, I think this is probably the main story of on Hippel’s original 1988 work on engineers innovating test instruments.)
I don’t recall ever seeing this latter distinction in the UI literature. However, some of the early open source software work (including the work of Dahlander, O’Mahony and others) distinguished between IT employees who contribute in their spare time (or in unmonitored slack) and those that are working on the company dime.

Some additional work needs to be done to distinguish these cases. For example, I recall seeing a conference paper (can’t find the cite) about how nurses can be user innovators. So if (as I recall) they are doing local sub optimizations to make their daily routine more efficient (or less tedious), is this to benefit the firm? Or is this for their own benefit? (Are there cases that combine both?)

Insight #2: Commercializing OI vs UI

My students asked if (or what) the difference was between the innovations that come from OI vs. UI. We talked about firm economic utility vs. individual personal utility (or intrinsic reward), but that seemed incomplete.

Instead, I thought about the fraction of user innovations (vs. OI innovations) that will be diffused to others.

My thinking was the transaction costs of bringing a (vertically integrated) producer innovation to market will mean that only the most valuable innovations will come to market.  Similar, for open innovation (external innovations), firms will generally not contract to acquire an innovation unless it’s valuable enough to be worth the transaction costs.

Conversely, the relatively low transaction costs of many forms of user innovation (such as those envisioned in von Hippel 2005) would mean that more (if not most) of these innovations will be disseminated. So trivial innovations that occur inside a firm will be the ones that never see the light of day, while all but the simplest user innovations have the opportunity for diffusion.

At first glance, this seems related to the Baldwin & von Hippel (2011) model of user vs. producer innovation. But their variables are on the costs of the innovation, whereas I’m suggesting that if we assume a random (exogenous) variation in the value of an innovation, some will make it to market and others won’t. I guess theirs emphasizes rationale discounting of the future cost of innovation, whereas mine assumes more of a serendipitous conception that gets pursued based on its value (rather than its cost).


Baldwin, Carliss, and Eric von Hippel. "Modeling a paradigm shift: From producer innovation to user and open collaborative innovation." Organization Science 22, 6 (2011): 1399-1417.

Behlendorf, Brian. "Open source as a business strategy." In Chris DiBona, Sam Ockman, and Mark Stone (Editors), Open Sources: Voices of the Open Source Revolution, Sebastapol; O’Reilly and Associates, 1999, pp. 149-170.

Bogers, Marcel, and Joel West. "Managing distributed innovation: Strategic utilization of open and user innovation." Creativity and Innovation Management 21, 1 (2012): 61-75.

Lettl, Christopher, Cornelius Herstatt, and Hans Georg Gemuenden. "Users' contributions to radical innovation: evidence from four cases in the field of medical equipment technology." R&D Management 36, 3 (2006): 251-272.

Piller, Frank and Joel West, “Firms, Users, and Innovation: An Interactive Model of Coupled Open Innovation,” in Henry Chesbrough, Wim Vanhaverbeke and Joel West, eds., New Frontiers in Open Innovation, Oxford: Oxford, 2014, pp 29-49.

Shah, Sonali K., and Mary Tripsas. "The accidental entrepreneur: The emergent and collective process of user entrepreneurship." Strategic Entrepreneurship Journal 1, 1‐2 (2007): 123-140.

Von Hippel, Eric. The Sources of Innovation. New York: Oxford:, 1988.

Von Hippel, Eric A. Democratizing Innovation. Cambridge, Mass.: MIT Press, 2005

November 20, 2015

WOIC 2015, Day 1

Today we kicked off our first full day of #WOIC2015 here in Santa Clara. There were some good paper sessions and great interactions with the industry participants.

From the stats recited by Henry Chesbrough, the audience was unusually balanced compared to 2014. With 163 attendees, we had 69 industry, 67 academics and 27 students; 84 North America, 60 Europe and 19 RoW. Chesbrough’s opening talk was followed by industry speakers (on the Internet of Things) from GE, IBM and Deloitte.

Modified from last year is the use of the corporate problem challenge: we have six successive challenges, which appear to be attracting a reasonable audience (the academic sessions appear to be attracting 2-4 industry people). I was sorry not to be able to attend the first problem session, which talked about how Pfizer is managing its use innovation intermediaries.

Dries Faems
Instead, I heard an excellent presentation by Dries Faems on how one firm negotiates the agreement that leads to an open innovation collaboration: they conclude that any such deal must have executive, business and technical consensus before signing or it won't succeed. This was one of the top-reviewed papers, and even so was surprisingly good in both its theoretical framing and the novelty of its insights.

In the second session, I moderated three papers on crowdsourcing. One was on internal crowdsourcing, one was on crowdfunding, and one of defining different forms of crowds. In the following session, I presented my own paper (moderated by Faems) which looked at pharma, open source biology and open source software.

Interestingly, two of the strongest papers were co-authored by senior academics with company executives who provided access to their firms.  The Faems paper was co-authored with Willem Posthouwer, the open innovation manager of FrieslandCampina (a Netherlands-based global dairy firm). In my crowdsourcing session, the best discussion was for a presentation by Ann Majchrzak on internal crowdsourcing (with very busy operational employees), but all the questions went to  co-author William Bonfield, who was Chief Medical Officer at OptumHealth.

In my own session, we had a paper by Christoph Hienerth (of WHU) and Monika Lessl (VP at Bayer AG). However, it was presented by David Tamoschus, both a WHU Ph.D. student and a Bayer manager. The paper won for David the award for the top doctoral student paper. (The best “emerging scholar” paper came through the collaboration of Francesca Di Pietro with two senior academics: Andrea Principe and Majchrzak).
David Tamoschus receives award for best doctoral paper.
The day’s program ended the poster session, which borrowed Kevin Crowston’s (2011) innovation having the authors controlling drink tickets for the attendees. As with 2014, the presenters and attendees both found it a worthwhile endeavor.

Ana Paula Barbosa

Gabriele Santoro