Saving Stranded Technologies: Talking with Spinout Expert David Tennenhouse at New Venture Partners
The “two guys in a garage” story of how technology companies get started is a powerful one in Silicon Valley. And it does happen: there really was a garage at Hewlett-Packard in the 1930s, Apple in the 1970s, and Google in the 1990s. The Mountain View, CA-based Y Combinator startup incubator, where most companies consist of two or three co-founders, is essentially one big garage. But startups can also come from other places—including large technology companies, where there’s often a surfeit of creative researchers and product developers, but a shortage of vision among company leadership about how to bring their promising ideas to market. (Or, to put it more charitably, a disconnect between the researchers’ ideas and the companies’ existing markets.)
That’s where New Venture Partners comes in. At this one-of-a-kind firm, which has offices in San Mateo, CA, Murray Hill, NJ, Ipswich, England, and Bussum, The Netherlands, the specialty of the house is finding promising technologies that aren’t going anywhere inside their native corporate R&D labs and spinning them out into successful companies. With more than $700 million under management, including a $300 million fund that closed in 2006, NVP has unearthed buried jewels at more than 50 technology, cleantech, and healthcare organizations, including big names like Boeing, British Telecom, Intel, Lucent, and Philips. Many of the spinouts it has helped to birth have gone on to lucrative exits, including Flarion Technologies (acquired by Qualcomm), Celiant (acquired by Andrew Corporation), SyChip (acquired by Murata), and Vallent (acquired by IBM).
NVP is a spinout itself. It was originally formed in 1997 as the New Ventures Group at Lucent Technologies, the telecom giant that had inherited most of Bell Labs from AT&T. Many technology corporations were experimenting at the time with corporate venture groups as a way to manage disruptive innovation—in fact, Xconomy CEO and editor-in-chief Bob Buderi wrote a feature article about the New Ventures Group and its implications for Lucent and the wider industry for the November 2000 issue of Technology Review. But the group’s tenure inside Lucent turned out to be short-lived. In 2001, with the telecom industry flailing and Lucent in financial disarray, a “special situations” fund called Coller Capital bought most of Lucent’s equity stake in the group’s portfolio, and spun out the core team as New Venture Partners.
NVP went on to raise its own funds and to absorb the corporate venture groups at British Telecom in 2003 and Philips in 2004. In 2007, the firm reeled in one of its biggest catches: David Tennenhouse. A computer scientist who trained at the University of Toronto and the University of Cambridge, Tennenhouse has a ridiculously broad resume that spans the worlds of academia, defense, corporate research, and e-commerce. Most recently, he’s served as chief scientist and director of the Information Technology office at the Defense Advanced Research Projects Agency; vice president and director of research at Intel Corporation; and vice president of platform strategy at Amazon and CEO of its A9.com research subsidiary.
I got to know Tennenhouse in the early 2000s when he was at Intel. When I returned to the Bay Area this summer to open Xconomy San Francisco and learned that he was at NVP, I contacted him right away to arrange an interview. I wanted to know how the spinout business is going, and how NVP’s challenges differ from those encountered by the average venture capital firm.
Our conversation was in-depth and fascinating, so I’ve broken it up below into two parts. In Part 1 today, Tennenhouse talks about the firm’s basic mission, why large corporations so often fail to find a commercial home for their own promising technologies, and how NVP operates. In Part 2, coming later this week, Tennenhouse talks about specific spinout examples that illustrate his firm’s approach.
Xconomy: What is the mission of New Venture Partners?
David Tennenhouse: The mission is pretty straightforward. To me, it is a bit of a personal mission. It’s identifying “stranded” research and trying to spin that out into stand-alone ventures that can create value for the company that sponsored it, for the employees, and more importantly, at some level, for society as a whole. People always talk about putting research “on the shelf.” There is no shelf. Teams dissipate. Some number of years later, another team comes along and reinvents it.
There are a lot of different reasons why research gets stranded. Applied research is not curiosity-driven, so we usually have a goal of figuring out how to build something or provide a service. Often, you figure out how to do it, but not how to do it affordably. The research still has value, and the people will go on and do great things and benefit from the insights. But only a fraction of those projects transition into the parent company. Basically, the stuff was invented in the wrong place. We are trying to create more of a marketplace for information to flow out of the organization that invented it, but chooses not to invest in it, and move it into a startup where the venture community can invest in it.
X: Why do you think big companies so often fail to commercialize promising R&D projects in-house?
DT: Some fraction of things are technically successful, but you can’t find a business case. Other things are technically successful and do have a good business case, but still don’t transfer. Why not? I can give you a list of 10 different reasons, but most of them can be converted into “Not big enough soon enough” or “Won’t move the needle.” There is an energy barrier that might be overcome if it’s apparent that something is going to be really big really soon. But you have vice presidents running $10 billion businesses, and if they can’t see a way to make a project into a $1 billion business in four to five years, it’s not going to be delivered on their watch. In addition, for a lot of folks at the senior VP level, their expertise is actually in scaling. It’s a much better investment of their time and energy to figure out how to grow their existing business. So they look at these things that are only going to be $100 million businesses in four to five years, and they don’t want to get distracted.
Now, if you look at something that is at zero revenue today, but will be at $100 million in four to six years, that is a great venture investment. We in the venture community also want to know that it’s got the capacity to go from $100 million to $1 billion. But for us, that is a good time frame. So there is a window in time where having a project be in the venture community is the right place to be.
X: If so few of these R&D projects find homes within the sponsoring companies, it makes you wonder why big companies continue to do applied research.
DT: At Intel, I was always careful to differentiate between road map-driven research and non-road map. A lot of corporate R&D is feeding the road map, and there is no problem there, if it’s well managed and efficient. It’s directly feeding the road map, and in certain cases creating options along the road map. Then there is the work that’s trying to create whole new businesses, and that part is very difficult, because the reality is that there is almost nothing that goes from zero to a couple billion dollars overnight. Even Google didn’t do that. So you are not going to get over this “not big enough soon enough” phenomenon unless you have some very long-term guidance.
That guidance can come from a few sources. Founder-based companies are different. One of the things that attracted me to Amazon was that in many ways Jeff Bezos is still running the shop. If Jeff likes it and believes in it, that will sustain it for a very long time. Microsoft will stick with things like the Xbox and Windows NT to make them successful. Some of these companies can still do a good job because the founder is an icon and will stick to his guns and has a feel for what it will take to grow something from that first couple million to $100 million. But once you get past the founder-led companies, that’s a bit of an exception.
X: Are there certain companies, or types of companies, that you keep an eye on because they’re more likely to have these stranded R&D projects that you could spin out?
DT: I don’t think we’re targeted in that way. Certain companies find that they can turn the ship. At Intel we called that “strategic forcing”—you’ve got your research agenda, now can you come up with a strategy that will fit the company? Arguably that happened with digital health at Intel. But it’s pretty unpredictable where that strategic forcing is going to happen. So we really don’t choose companies that way.
What you want to do is keep in touch with the technology teams that are working on interesting products that might someday be a spinoff. Often, it’s mentoring the teams on how they can be successful within their company. My number one passion is about seeing great technologies see the light of day, so I’m quite happy to give people advice that helps them navigate their own companies. That’s a way, meanwhile, for us to stay informed. And if they are not able to make that internal transfer, it doesn’t end up looking like a Hail Mary pass. If we look like we were Plan B all along, that is my ideal situation.
X: The team here must be pretty different from your usual group of venture partners.
DT: I joined after the spinout [of New Venture Partners from Lucent]. The people here very much have a passion about this technology transfer thing. Andy Garman, one of the managing partners, was at PARC. [Garman also managed spinouts at Bankers Trust and Lucent's Bell Labs.] So he’s been butting his head against tech transfer for years. Steve Socolof was closely involved in the spinoff from Lucent and was in management consulting. Tom Uhlman, before Lucent, had been at HP, so he saw the problems there. Those are the three managing partners. So there is a deep commitment about this issue of spinouts and trying to help teams move these technologies downstream.
X: How does the pitch that NVP makes to potential investors differ from other venture firms?
DT: The limited partner base is not as atypical as you might think. But we are offering them a different value proposition. For one thing, it’s a differentiated deal flow that’s coming to us. The guys on Sand Hill Road are all seeing the same deals and vying amongst themselves, so if you are thinking about investing, should you do your n-plus-first investment on Sand Hill Road, or should you invest in NVP? The second part of the proposition is that these are technologies where somebody has invested a lot to give us a head start. This is university research, plus some corporate research on top usually, and it’s much more de-risked and ruggedized and all that.
I don’t kid myself that I’m smarter than the folks on Sand Hill Road, but we are looking to do deals that they would not normally do—and once we package these deals, they will often participate. We like to syndicate, but most other VCs don’t really like to lead the spinout.
X: How hands-on do you get with your portfolio companies?
DT: We are very hands-on. Part of the reason most VCs won’t do these deals is that you have to spend a lot of time looking at the deal, and then you have to negotiate the IP, and then you have to go find a CEO and sales and marketing people to go with the founding team. During that first six to nine months before you do the spinout, and for a year or two afterward, there is a lot more heavy lifting by the investors. Once you get to the Series B, it’s pretty much the same. In the classical VC world, partners are really good at looking at a huge deal flow, picking the ones they want, and moving on. We are able and willing to spend a lot more time. It also means we’re going to look at far fewer deals, but as I said, the stuff we’re looking at is a lot more mature. Because we’re getting highly selected deal flow, we can afford to put in a lot more time per deal.
X: What do you spend the most time on with your portfolio companies?
DT: Probably putting teams together, and thinking about what else they could do with the technology. In the classic VC world, what happens is that business plans get serially improved as entrepreneurs shop them around. The entrepreneur shows up, the VC tells them why they aren’t going to invest, the entrepreneur improves the plan, they go to the next VC, and the original VC doesn’t benefit. We stick with the deal and go the whole life cycle. The entrepreneurs have typically been thinking with blinders on, because their assumption has been that they are going to commercialize their technology within the context of their company, and that causes them to think in a very limited way about channels. We get to exercise a lot of creativity. The traditional VC has the same creative capability, they just tend not to use it. They tend to say yes or no.
Continue to Part 2: The spinout story at Everspin, and what happens when companies when companies don’t give their researcher-entrepreneurs’ best ideas a fair hearing.