In the 1980s, as Henry Chesbrough studied companies as varied as computer disk manufacturer Quantum Corporation, Lucent, Cisco and Xerox Corporation and its Palo Alto Research Center , it dawned on him that the days of traditional firm-centric innovation were over. It was time for companies to look beyond their boundaries for new sources of innovation.
This approach led to ideas from a variety of sources, and proved to be more cost-effective compared to internal R&D. It also led to out-of-the-box thinking, more relevant products, and consequently became a source of competitive advantage. Chesbrough’s findings led to the 2003 book titled ‘Open Innovation: the New Imperative for Creating and Profiting from Technology‘ in which he coined the term Open Innovation. The book was a roaring success. It also helped that companies like P&G, Lucent and innovation intermediary InnoCentive were already practicing Open Innovation in some way or the other.
Adjunct Professor and Executive Director of the Garwood Center for Corporate Innovation at University of California, Berkeley’s Haas School of Business
Ever since, the Open Innovation movement has grown. For many companies like P&G and IBM, Open Innovation has become an integral part of their business model. Later Chesbrough advanced the idea of Open Innovation with two more books—’Open Business Models: How to Thrive in the New Innovation Landscape‘ and ‘Open Services Innovation: Rethinking your Business to Grow and Compete in a New Era‘ . While the first book was about how companies can navigate this new innovation landscape, the second one was on combining Open Innovation with a services approach.
In a freewheeling interview with CKGSB Knowledge’s Neelima Mahajan, Chesbrough, Adjunct Professor and Executive Director of the Garwood Center for Corporate Innovation at University of California, Berkeley’s Haas School of Business, elaborates on advances in his research.
Q. From the time when you first thought of ‘Open Innovation’ and published your first book on the theme, how has your work evolved? A. The original book was focused primarily on new models of research and development (R&D), and the shifting roles of internal central R&D labs, external relationships with entities like university research centers, and, generally, within the context of the current business models of the organizations. The second book that came out a few years later–‘Open Business Models’–was about innovating the business models that commercialize (the technology or the R&D). The book that came out last year–‘Open Services Innovation’– really looks not just at products and technologies, but also at the services side of the economy. One thing that I underestimated initially is just how hard it is to really adopt and practice Open Innovation. It isn’t simply changing the rules of your purchasing department, it isn’t even just changing the incentives for your R&D organization. It really affects many parts of your organization in a pretty systemic way. To my great delight, there are now job titles such as Director or Vice-President of Open Innovation in many companies. My most recent research is on innovation communities. So I am looking not just at the usual suspects in the value chain, or even the people who are making applications and services surrounding the value chain, but really at end-customers, end-users, and a more of a co-creation kind of a model. I think that there is one transition from products to services, the business models do change, the role of the customer does change and the innovation process has to change. Q. Can you elaborate some more on your work on innovation communities? A. I did a study on General Electric’s ecoimagination initiative, which was part of GE’s energy business. GE Energy’s traditional customers were power plant generation companies and utilities—big customers who bought megawatts of power from GE, turbines and things like that. GE Energy, a $40 billion part of the overall corporation, was noticing that there was a lot of venture investment going into the energy space, but not for the traditional, big power-generating stuff, but for things that were producing power in kilowatts rather than in megawatts. It was often sold off the grid to residents or commercial and industrial establishments, rather than.
They created a process and committed $100 million to put out a call for proposals for ideas that could be used to start up new ventures in these green and renewable spaces where they themselves were not well represented. So they saw a lot of activity, they knew what was going on, they knew they weren’t really a part of it, and their questions was: “How do we get into this and get engaged?” They did it in a more interesting and open way: instead of simply doing it themselves, they talked to four venture capital firms who put in another $100 million and they shared the evaluation and vetting process with them even though each of these five parties had their own independent investment authority. When they got all the submissions via a website they created for this, anybody could see the submissions. Over the course of the project, they had about 70,000 contributors—making comments, suggestions, submitting ideas etc.
At the end of the activity, they ended up with about 23 ventures getting funded. Obviously a lot more was submitted than got actually funded. They also had these 70,000 contributors–what could they do with that? In essence, it was a by-product of the initial intent in process, but the punch line in the story was that they actually decided to create a new title–Community Innovation Manager at GE to actually keep this community alive and engaged. The manager is doing things not simply to pull things from the community for GE, but to do things to get the community connected with each other, keep them active, allow them to help each other, or make introductions to others, particularly for the things that weren’t funded but had real promise. So this person’s job was to really take responsibility for making sure all got connected and hooked up, and to keep this as an active new resource for GE.
Q. When you talk about it, it make one wonder whether something like this within the GE system could eventually become as big or as important as InnoCentive (an Open Innovation intermediary spun out of pharmaceutical giant Eli Lilly)? Do you know what the internal thinking on this is within the GE? A. InnoCentive came out of Eli Lilly and the insight that led to them being spun out of the pharmaceutical company was that if you really want to engage great minds around the world in solving these difficult problems, you are going to get a better level of engagement and attract more people only if you open it up for others, even Eli Lilly competitors, to make use of the service. Once they realized that it was going to be much more powerful outside the company, they spun it out. So GE has a $40-billion energy business. They are not yet in the mode of saying, ‘You know, we need to be on top of addressing Siemens’ energy challenges as well and similarly engage’. So there is a difference there. But I think that GE also sees that they have, in their energy business, a deep insight into the stuff that is going to take these nascent ideas to the market. Companies like GE in the renewable and green energy space are really the path to market for many of these ideas. Since they have such a strong position there, it’s really in their long-term interest to have a very fertile growing set of ideas taking root upstream, because they will have a first look even though they cannot actually control what happens to them. Because they see them earlier than anybody else, they’ll have the ability to take action, and they see that as a real benefit. So they are looking at it in a different way than InnoCentive. InnoCentive is in essence an intermediary, it’s not trying to do anything downstream–that’s up to the clients. GE has some of the same search ideas that InnoCentive has, as you rightly note, but they are doing it not to intermediate, but to actively pull things into their own business. So there are different vantage points in this Open Innovation marketplace.
Q. This also brings up this really interesting point–for companies that have a long history, established processes and layers of management, is there a benefit in keeping the Open Innovation initiative at an arms’ length from the actual company? How has GE set this up within the organization? A. So, keep in mind, you’ve got four outside venture investors. One of the conditions that they established as a precondition to their participation is that they have the right to invest or not invest as they see fit in each and every opportunity. The second precondition was that GE could in no way limit the upside of the investment once made. For example, there’s no formal right of first refusal. GE doesn’t get to have the right to make an offer before anybody else can make an offer. And the reason you do that is if you are an investor in a company that’s got some promising technology and it matures to a point where it’s actually ready to go public or maybe be considered as an acquisition from a big company, you want as many active bidders as possible to get the price as high as possible so you get the biggest return. Anything that would give GE a preferential advantage in that stage will actually diminish the potential to get the best price and diminish your return.
GE agreed to consciously avoid anything that would do such a thing. They really didn’t have the ability to control downstream, as a precondition. They set up a process totally outside the internal corporate process, but run through their venture capital arm which does venture investing for GE businesses typically at the later stages, where it is much closer to the market and, therefore, easier to see how this is going to fit in various GE businesses. So they already had their venture capital organization, and what they are doing now is investing further upstream, further away from the market. It’s like planting a number of seeds, establishing a process that lets those seed grow in a very unimpeded way.
But they do make a number of connections. The GE internal research organization was part of the due diligence, reviewing these external technologies. So that was also like an infusion of a whole different set of perspectives to the research part of the organization who were traditionally serving very large companies. And so they knew those needs, insights and technologies very well, but now we see entirely different business models, entirely different customer sets, and entirely different technology bases. So it was really kind of a new education on the research side as well.
Q. Do you see this changing GE’s R&D machinery in the future? A. I had a chance to interview the head of the R&D organization for the project and I was very interested in how he would regard something like this. Even for a company of GE size, the $100 million committed to this initiative is a fairly big chunk of money for an R&D organization. To my surprise and pleasure, he was very enthusiastic about the program, he liked the diversity of ideas and getting a whole different set of inputs, and having his attention spread on a very different area than what he was focusing on usually. I did not talk to bench scientists in the R&D organization and maybe they would have a much more traditional view on this. I don’t know. But at the top of the R&D organization, they were quite positive about it.
Q. Do you see other companies as well adopting similar models, maybe with slight differences? A. I would not describe this as a widespread trend yet. I think for most companies, they are still working with entities like InnoCentive, tapping into those external networks, on an ad hoc basis for individual problems they would like to see addressed. Most companies are not actively trying to foster their own innovation communities. But I did come across a company in Florida called Ocean Optics that does this.
Q. InnoCentive came with a bang about 10 years ago. Yet we still don’t see too many innovation intermediaries like them. A. I teach part time in Barcelona at a school called ESADE. In his dissertation, one of the Ph.D. students at ESADE found 48 organizations that are doing things like InnoCentive. Another one that is often mentioned is NineSigma. Yet2.com has a slightly different model, but at least at this level of discussion it kind of does the same things, but there are more than 40 other companies doing some variants of this. So, it remains I would say, a cottage industry. There’s been a lot of entry, a lot of experimentation.
Q. One big impediment to Open Innovation in a lot of organizations is really the issue of intellectual property (IP). How does it work out in GE’s case? A. It’s a very central issue in Open innovation. One area where it showed up in the GE case was when people were sending their ideas—who owns those ideas? They are on a public domain website that GE maintains, so does GE now own these ideas? GE decided, against the initial advice of their attorneys, that the ideas belong to those who submit them. Once the ventures are funded, those would be the property of the individual companies again, not GE. GE might be an investor in the company, so indirectly they would have some ownership, but because they are not the majority owner, they would still be in minority position. Then there is the issue of contamination. When you have your internal research staff, looking at these external ideas, how do you deal with the possibility that people think, ‘Gee, you’re gonna be using my ideas, even if you don’t give me money’.
Q. Or even take a part of it. If you see something every day, you might internalize it and unconsciously use it… A. So they had to work hard on that. They had some protocols they went through to make sure that the people who were doing the reviewing were trained on where the red lines were. A small number of people from the research organization was doing it and they were trained ahead of time on where they had to be careful and I guess, ultimately they took whatever risk was still remaining. These are again very far upstream and away from GE’s traditional strengths and they also may have judged that the likelihood that this was going to drive stuff within GE’s own product roadmap was sufficiently smaller than the risk they were taking. But there’s also a concern in any of these activities, where you inadvertently contaminate your own downstream development activities. Q. What are some of the other impediments to adoption? A. I think a big one is when you use an intermediary like InnoCentive or some scouting process and you identify something very interesting that you bring in. The technology or idea that’s brought in does not have an organic internal champion who is going to do the hard work to drive it through the rest of the process and into the market. At this point, it’s still just an idea or early technology, it’s not fully formed, it’s not commercially ready, it’s not sold into the market the next day. Months or years of additional work are needed to get it through. The ideas that come out through the internal R&D process have to go through a number of hurdles and screening and in the process, they do attract internal champions to carry them forward. These external ones often lack those champions. That is an organizational issue.
Q. How have some companies like Procter & Gamble (P&G) overcome this? P&G’s Open Innovation program, Connect+Develop, is huge after all. A. They deal with it by starting with a meeting with the businesses and getting a sort of “shopping list” of the things that are priorities for them, that are holding back the technology part of the business from delivering as much value as quickly as P&G would like. So that becomes the “shopping list” of things to look for.
Q. Also, what do you do with the ideas that are not used? One of the things you talk about is ‘inside-out’ and ‘outside-in’ innovation. How can that come into play here? A. A lot of people think of Open Innovation only as bringing external ideas in. That’s a really important part of the model, but it’s only one half. The other half is indeed the ‘inside-out’ and is the basis for the answer to your question, what do you do with the stuff you don’t use? There’s stuff you come up with during your internal R&D process, what do you do with that? In GE’s ecoimagination case, the stuff that came in from the outside and is not going to get used internally; how do you allow that to progress?
They did a few things. One, they created a ‘people’s choice’ award and they awarded five projects that they didn’t fund as investments. I think they were each given $100,000 to keep going. This is not equity and there’s no ownership attached. It’s simply about: ‘This is a great idea, but not something we want to fund. But we think it’s got real promise, so go do it’. Two, through the community innovation manager, they used their social networks—their social capital, rather than financial capital—to help some of these projects advance. They tried to make sure that within the community, people were connected to one another. So people who commented on ideas that didn’t get funded got introduced to the ideas that were going to continue. A lot of social capital was deployed in that process. More generally, in the ‘inside-out’, you know there’s also licensing, you can donate things to the Commons, you can put things into a university context, maybe give it some funding, so the project continues to evolve, not under your control, but in a university setting, and again you can follow with progress reports etc.
A lot of these things you can do to take things from the inside to the outside, in essence become options for the future. You have the opportunity, but not the requirement to do something. So, in many states of the world, the option is not worth anything and all it will cost you is whatever it took to write the option initially. But in other states of the world, that option may prove to be very valuable. So if water becomes very scarce and expensive, a waterless technology might become very valuable, and it wouldn’t in other states of the world. Different states of the world have different outcomes, and different options are going to be useful. A company like GE can and should have multiple options in play, so that whatever state of the world emerges, they’ve got some ways that advantage them, that they wouldn’t have had if they’d put those out there.
Q. I find that a lot of companies would prefer to waste an idea than let it go out. A. It’s true. I refer to this as potential ‘false negatives’. When you’re making an evaluation of a potential project, in my view, you’re usually evaluating that project based on its fit with your current business model. If it looks promising in your current business model and you’re likely to pursue it, then off you go, and then you find out whether the market likes it or doesn’t like it. But at other times, projects come through that get rejected: was it because the project was bad or was it because the project did not fit with your business model? And this is where my work on Xerox and the Palo Alto Research Center (PARC) drove this home for me. A lot of the PARC technologies that ended up going out of Xerox were really false negatives. They actually had very interesting technical merit, but they didn’t contribute to a copy or a printing company that was making its business on those kinds of products. When they were able to go outside (in many cases, they went out and failed, so this is not true for all of them), a number of them found a different business model. And actually in these cases, they adopted a more open business model that allowed them to find value that wasn’t obvious within Xerox’s evaluation, because of how Xerox was looking at the projects.
Q. But one can also argue that many years down the line, PARC is seen as a failure in many ways. A. Well, we have to figure out how we define success and failure. PARC did deliver some wonderful printing technology to Xerox that probably more than paid for the 30 years of investment that PARC made. So if it’s just a matter of pure return on investment, you can say PARC was a success. But in another sense, if you think about the value that was created externally, and how relatively little of that was captured by Xerox, then you must say, ‘Gee, that was at best a missed opportunity–couldn’t they have figured a way to get more from it?’ But you triggered me by saying a lot of companies would rather have a lot of ideas waste away than do something like this. And that mentality is very anti-innovative if you think about it. I think it is a test of a company’s innovation culture: how do you treat the projects that you don’t continue yourself? Do you allow them to go out? And do you have processes to learn from the experience once they go out? Or do you bury them and let them waste away?
Q. So in your opinion, PARC is a success? A. Well, PARC’s business model changed about a decade ago. Xerox actually made it a free-standing organization (still owned by Xerox) that had to sort of make its own way in the world, no longer subsidized by major corporate R&D from Xerox. So the idea was: ‘Can we create a business at PARC that can fund itself on the basis of corporate research?’ And they’re still around, so they’ve survived. At the high water mark they had 300 or so employees, and now they have somewhere between 225 and 250, so they’re smaller, though not radically smaller. They have partnerships now not just with Xerox, but with some other companies, and not only in the IT sector, but, interestingly, in the automotive sector and the life sciences sector. So they have had some diversification, but they certainly haven’t been a raging success. I think there are real challenges to being a free-standing corporate research organization and trying to build that as a business. The business models for that I think are very tough.
By contrast, if you look at Intellectual Ventures (a company that focuses on developing a large patent portfolio) that Nathan Myhrvold and his colleagues are running, that is also a free-standing organization. It also does some research, in a very different kind of a model, and they have a very intellectual property-driven business model, much more so than PARC does. I don’t have access to their internal financials, but I’m been told they’ve raised billions in capital and they have already collected hundreds of millions in licenses and royalty income. And that appears to be, on a pure financial basis, a bigger success than PARC.
Q. But they are also fairly controversial. A. Oh yes, undoubtedly.
Q. What do you think about that? Are they really the patent troll they are accused of being? A. I have a chapter on Intellectual Ventures in my 2006 book. I spent a day there for researching the book and I observed one of their invention sessions all day. It was a high-energy physics discussion about an interesting problem: if we take Moore’s Law and project forward another decade in time (that would be five or six more iterations of Moore’s Law), what kind of things will be enabled by that kind of evolution and what could be done with that? So this is long-range stuff and they would argue that this is the kind of work that large corporations used to do and increasingly don’t. They think they’re actually filling a gap there, and that the things that they are discovering and inventing therefore are actually making a real contribution. So that’s one piece of what they are doing. They’re also actively buying up lots of other patents in the secondary market to build big portfolios that strengthen their bargaining position. They are consciously not going to practice these technologies. They are built to license. If you’re implicit in that licensing conversation, what happens if you don’t license? And although they claim that they don’t sue, they’ve done enough threatening of lawsuits. In a few cases, they’ve even sold some of their portfolios to companies like Acacia that are active patent trolls, so that they don’t do it directly, but they can benefit from it indirectly. I think one would say that they are very close to a ‘patent troll’, if not a patent troll. Unlike patent trolls though, they do have internal invention activities as well. Q. Talking of the Chinese context, can China adopt Open Innovation models, given the current IP environment? A. In the first book, I said that there were some enabling factors that really helped Open Innovation. One of them was the mobility of the workforce. The second was the availability of capital to start thing things up. And the third one was the willingness and ability of customers to switch from one supplier to another. These things are strongly present in China to the best of my knowledge. I hear from multinational companies (MNCs) that have built R&D laboratories in China, that the turnover rate in the labs is quite high. I hear of numbers like 20-25% of the workforce each year is turning over. Every 4-5 years it’s a different lab. The knowledge that these people earn in the lab goes with them. I’m not talking about reverse engineering and copying, I’m talking about the sort of diffusion that goes on in Silicon Valley is also going on in China, from the MNCs out. All that is very positive for Open Innovation. IP protection is not there yet and this complicates the ability to do truly innovative versus imitative things. I think this is holding back, to some extent, the performance of innovation in China.
I do see some companies like Huawei and Lenovo becoming much stronger in their R&D and technology in China. As these companies grow and strengthen, they are going to want protection themselves from the Chinese government. I am actually optimistic about IP protection in China, not because it will attract MNCs from the West, which it will, but because that’s going to make it work. Indigenous technology suppliers are going to demand it because they suffer as well from the rapid imitation and piracy. I think that’s what is going to drive it. Take Japan or Taiwan. In their history over the last 50 years, they went from imitation and piracy to an environment where IP protection is quite strong now. I think again it was the indigenous technology base and the development that drove that.
I did some work in the semiconductor industry in China, with funding from the Sloan Foundation. In that industry, we ended up finding two very different segments: former state-owned enterprises (SOEs) and start-ups. They really behave very differently. The former SOEs were selling almost exclusively in China. As for the start-ups, 80% or more of their revenues was from outside of China. The capital for the former SOEs was from national, provincial and city sources. The funding for the start-ups often came from outside of China. The leadership of the organizations in SOEs was very much from China. In the start-ups it was returning ‘sea turtles’–from the US, Singapore and Taiwan—so they were very international, and that’s why most of the sales were outside of China too. They were already part of that global network, versus that national one. So this means that in one industry you have two segments behaving very differently. The employment was concentrated in SOEs, but the growth was coming from the start-ups. The profit was much more coming from the start-ups than from SOEs. The rate of new technology introduction was much faster in the start-ups, than in the SOEs.
I can’t generalize that to all of China but my hunch is that that dynamic is present in many industries in China. One mistake China is making is that the SOEs are growing in strength and this is actually bad for innovation in China, just as it was bad in the US to have big oligopolies. Having these big economic concentrations discourages innovation. That is something that has to be corrected in China for the innovation potential to be realized. I don’t see signs in the near term that that’s happening.