Open innovation sometimes refers to the strategy of bringing in concepts and inventions from the outside—or it can also mean making your company’s data, platform, or intellectual property accessible to others who’d like to build things with it.
But amidst all of the idea challenges and hackathons, the term “open innovation” has been getting overused and abused. (Not every meeting with a startup or visit to a university lab counts as open innovation.) And there’s a difference between the “outside-in” and “inside-out” approaches to open innovation.To clear things up, we consulted the guru of open innovation:
Henry Chesbrough, a professor at U.C. Berkeley’s Haas School of Business and faculty director of the Garwood Center for Corporate Innovation. In a recent conference call as part of our IL Live series, Scott Kirsner spoke with Chesbrough about why the “inside-out” approach is more challenging for most companies; how open innovation is being practiced at companies like Intel, Eli Lilly, Bayer, and United HealthCare; and some of the common barriers to making open innovation part of the way people work every day, rather than a one-off event. n Many large companies, Chesbrough argues, “have these fiefdoms or silos where one part of the company really doesn’t know what other parts of the company are doing. Open innovation really works best when you have open flows of information internally as well.”
Prior to his career in academia, Chesbrough spent over a decade working in the computer and hard disk drive industry, at companies including Quantum Corporation. The author of several books, including Open Innovation, Open Business Models, and Open Services Innovation, Chesbrough also received the Industrial Research Institute’s Medal of Achievement in August 2017.
You can read an edited transcript below, or listen to the complete audio of the call by clicking play below.
How I Define Open Innovation
I’ll tell you what I mean by it. You can think of it as the antithesis of a model of vertical integration or self-reliance, where you have to do everything yourself. Open innovation is about a distributed landscape where there’s a lot of useful information in lots of places.
Even the biggest and best companies would do well to make use of and incorporate external ideas and technologies in their own innovation, and allow things that they’re not using to go out for others to use in their innovation activity. There’s an outside in path of knowledge flows, and an inside out path of knowledge flows.
Outside-In Versus Inside-Out
Sure. It’s been my observation that companies have been quick to grasp the potential for bringing things from the outside into their own organization. That’s done to either reduce costs, or reduce time, or share risk in innovation activity. There are three dimensions of benefit — cost, time, and risk — to doing the outside in [path.]
There has been less uptake on the inside out. I think the reason is that doing inside out effectively means you’re placing a technology in a new business model. These are things that don’t fit with your current business model. We get into the whole domain of business model innovation, which typically touches many different parts of an organization to do it well, whereas outside in is something you can implement more or less inside your existing organizational structure. I think that’s the biggest difference for the uptake.
I have a case study of [an inside out example] with a medicine called Cubicin that was taken out of Eli Lilly and licensed to a specialty pharma company called Cubist. Subsequently, it became a blockbuster drug for anti-infectious diseases. Lilly has collected over $300 million in licensing and royalty income through the arrangement.
There are many other examples. One that I think is interesting and profound is Amazon’s Web Services, where they took their internal IT infrastructure and then created a business around opening it up for rent for others to use the same infrastructure to host their websites as well.
The reason I think that’s a very profound one is I think it shows anytime you have a high fixed-cost resource or asset where the utilization is not very high, there can be powerful economies by opening that up for others to use as well.
Open Innovation Gets More Digital
I think what’s driving best practices today in open innovation is that, in this domain where software is eating the world, and more and more things are moving to digital — big data, and blockchain, and machine learning, and all the rest — the best practices in open innovation are also moving that way.
One of the open innovation insights is all this data that you have to crunch to train your algorithms in this digital world — there’s nothing that says it has to be your data. You can be very successful crunching and analyzing other people’s data, too.
A company that does this well is part of United Healthcare. It’s a group called Optum, which is their internal IT organization who, like Amazon, has been unleashed to offer IT services to other hospital chains, other healthcare providers. Because they’re serving those IT needs as well, they aggregate all that data from all their clients, which is giving them more data to become more digital and more effective.
Barriers to the Spread of Open Innovation
There are many barriers to the adoption of open innovation and…also barriers to continuing and scaling it, even after you’ve adopted it internally. I’ve done two surveys now with the Fraunhofer Institute in Germany, where we surveyed large companies of $250 million [in revenue] a year, both in North America and in Europe.
In the two surveys, we find that roughly 80 percent of the companies we survey are practicing at least some parts of open innovation. Encouragingly, about 70 percent of those companies report that they’re doing it more, and management interest and support is increasing. The good news is, it doesn’t seem to be past its sell-by date yet.
Having said that, there are some things that really get in the way. One is the support organization, whether it’s legal, or brand, or procurement, often slow things down. You really do need to create new processes internally to use open innovation in collaborating with the external world.
A second thing that slows us down is a lot of companies, when you look inside how they’re organized, they have these fiefdoms or silos where one part of the company really doesn’t know what other parts of the company are doing. Open innovation really works best when you have open flows of information internally as well.
A lot of companies simply don’t do that. If I have knowledge that you don’t have, by hoarding that knowledge, I have more power. These are some of the cultural challenges that get in the way of using open innovation more effectively.
Making the Case for Open Innovation
When I’m talking to somebody at the top [about the rationale for open innovation], I usually don’t start with innovation. I usually start with growth and, in particular, organic growth. You can do acquisitions and you can buy growth, if you will, by acquiring somebody else who has executed and scaled up a new business model, a new set of technologies.
But what open innovation enables is, hopefully, sustainable organic growth. So yes, do M&A when it makes sense, but don’t make that the only tool in the arsenal. You also need to grow your skills in generating organic growth.
Open innovation is a great way to do that because it can save time, it can save money, you can share risks, and you get leverage on your internal R&D by collaborating with [partners on the] outside. That’s typically the starting point at the top of the organization.
In the middle of the organization, I tend to talk more about some of the viral resistance you’re going to encounter in a large company, and how you need to generate proof points to convince the organization that it can work here as well.
It’s fine to survey what other companies are doing and, indeed, that’s not a bad starting point, but you’re going to have to put it to work in your own organization. You better do that in a way that is contained, and not a huge investment of time and money at the outset.
Then, as you get some initial positive results, use those results as an early win to build momentum, spread the possibilities in your own organization, and only then start asking for additional resources and support.
Software Platforms for Open Innovation
I think the software tools do help to collect and provide information in one place about your innovation activities, both inside and outside the organization. I was being critical earlier in our conversation about people who hoard knowledge and keep it to themselves. By not sharing it with anybody else, there’s much less opportunity to leverage that knowledge.
These tools offer the promise of capturing and widely sharing what’s being done, and who’s doing it, and how might you connect to it. They can be, I think, a powerful tool, but I think these [software] companies will be the first to tell you that it’s just a tool.
You’ve got to put management processes around this to actually take the results of what’s in the tool, and get them into the marketplace, and generate a real business result, which means the front end of the innovation process needs to be connected to the back-end of the innovation process, where you can actually go to market through existing or new business units, and sell to real customers who pay real money.
‘What’s My Role? Who Am I’
Karim Lakhani [of Harvard Business School]…has a student of his who’s now working at NYU, named Hila Lifshitz-Assaf, who has does some wonderful work at NASA in how they’ve implemented some aspects of open innovation, particularly challenges.
One of the wonderful things in her work is that, when NASA put out these challenges and got some great results back from the external world, she could’ve just stopped there, and written up the paper, and got a nice publication talking about how powerful this is and how valuable it is to leverage outside knowledge.
But, because Harvard had a good research budget, and because she had some more time left in her doctorate, she kept going back to the Johnson Space Center in Houston for an additional year.
What she observed was that the impact inside Johnson Space Center on the scientists and engineers at NASA who were saying, “Gee, if all this stuff is coming in from outside, what’s my role? Who am I? I came to NASA to solve these difficult problems. It really is rocket science. That’s why I’m here. When we start taking all these ideas from outside the organization, what happens to me?”
I think that’s a very deep, human response that creates resistance to open innovation. I really credit Hira and Karim, who was advising her, for really digging into this… I think that’s something that’s going to happen in lots of organizations if you’re not managing it carefully.
Examples at Bayer and Eli Lilly
I’ve done some work with Bayer in Germany in the pharmaceutical industry. I think they’ve got a very well thought out, very well constructed open innovation program that has a number of moving parts in it.
One is they have grants for apps, where they actually will provide funding for people to look at their compounds and suggest new uses. They have another one where they’re looking for leads that has a similar kind of a crowd-sourced mechanism, but they still do a lot of internal R&D as well. They’re now increasingly sharing that with others.
Lilly does something I think the pharma industry would do well to copy, which is they offer to external entrepreneurs and research scientists a service where you can bring your compound to Lilly. They’ll run it through their internal assays and tests, and give you back the results at no charge.
This is really valuable information for you because these assays Lilly has built up over decades is really quite rare and valuable, but Lilly gets something out of this, too. They not only give you the results. Of course, they see the results as well, and they have something that you don’t. They have all the other results for compounds on those same assays.
They know where your assay falls in the distribution of results. For the ones that are really at the upper end of the total look really promising, they can then do further discussion with you about maybe collaborating or doing a joint development [arrangement]…It helps them scout and scan very effectively for new, promising compounds outside their own research labs.
Business Units and Budgeting
We need to be very thoughtful about linking the front end of the innovation process to the back end of the process. The back end — let’s assume that it is some kind of a stage-gate model that Robert Cooper and his colleagues have done a good job of articulating over the years. We need to link up the front end to that.
One thing I’ve seen in companies like Intel is, when they have their front end group come in with a promising new project, they’re looking for a business unit to take it and run with it. The business units are run on very tight budgets that are done annually, and there isn’t a lot of spare money lying around to pick up and run with something in the middle of the year.
Typically, you’d have to wait till the end of the year to make the budget request for the next fiscal year to have the resources to start to work with it. You can see right away there’s a bit of a mismatch. Not only a funding mismatch, but the people that have been working on the project typically get broken up and redeployed on new front end activities. So a lot of the knowledge that we actually assembled and put together gets diffused. It’s hard to put it all back together again when the funding does arrive for the next fiscal year.
What Intel did was they created a separate fund to provide funding during the current fiscal year for things that the businesses didn’t know about, and couldn’t specify at the beginning of the fiscal year.
These funds are there to tide over the project until the next budget justification. The second thing they do is they make at least one person on the front end team transfer over to the business unit for at least six months, to provide that continuity of knowledge… These are a couple of practices that help to connect the front-end to the back-end in a more practical way.