Spinout Doctors: How New Venture Partners Saved Freescale’s Magnetic Memory and Other Stranded Technologies
Magnetoresistive Random Access Memory, or MRAM, promises to change everything about how our computing devices work. It’s as fast as classical static RAM at the core of today’s microprocessors, but it doesn’t wear out, and it also holds data permanently, even when the power is off, like today’s flash memory. It could enable true “instant-on” information devices and speed up many types of computation.
Freescale Semiconductor, the former chipmaking division of Motorola, poured a lot of effort and money into making MRAM practical in the mid-2000s. But the technology almost wound up homeless, because Freescale is a microprocessor manufacturer, and wasn’t equipped to sell storage devices. The company had developed the technology without ever really intending to get into the fiercely competitive memory business.
Which is where New Venture Partners entered the picture. The San Mateo, CA-based venture firm purports to be one of the only venture firms in the world that specializes in finding promising but under-supported technologies inside corporate R&D labs and spinning them out as independent startups. Yesterday we published the first part of a conversation with David Tennenhouse, a former Amazon and Intel research executive who’s now a partner at the firm. In 2008, Tennenhouse and New Venture Partners helped to spin out the MRAM business from Freescale in the form of Everspin Technologies, which is now one of the world’s leading suppliers of MRAM for industrial, aerospace, and military applications. The details of the Everspin case, which illustrated many of the nuances and challenges of the spinout process, are among the highlights of Part 2 of our conversation, transcribed below.
During this part of our talk, Tennenhouse also described how the firm identifies teams and technologies that can be spun out profitably, what types of companies are likely to house potential spinoffs, why Google doesn’t do spinoffs but Microsoft does, and the roles that NVP’s partners must sometimes take on—including career counselor.
X: Can you walk me through a couple of your favorite examples of companies you’ve helped to spin out?
DT: One of my favorites is Everspin, where we spun out all the MRAM magnetic memory technology from Freescale. That’s an area where Motorola [which spun out Freescale in 2004] invested for many years and Freescale invested for many years. It’s the only group shipping real, working MRAM. It’s a great example, because one of the things you need is leadership in the team that is really headstrong and committed. Here you’re talking about a new semiconductor; a new material. They managed to get this all the way into customers’ hands. This was a hell of a dedicated team and they really wanted to foist this thing on the world.
Another critical ingredient is that the CTO at Freescale, Lisa Su, really wanted to see this thing happen also. But Freescale looked at the situation and said, ‘We don’t really want to be in the merchant memory business.’ If you could replace on-chip Flash memory with MRAM, that would be very interesting to Freescale, but first you would have to sell standalone memory parts, which improves your learning and gets everyone comfortable with you as a supplier, and they said they didn’t want to be in the memory business.
It truly was a strategic decision. If MRAM had already been a billion-dollar business, they might have made a different decision, but it was just getting started, and their sales force sold microprocessors, not memory. MRAM needed its own sales force. And there was a whole next-generation technology in the pipeline that needed to be pushed forward with more R&D and downstream resources. Lisa could see how this didn’t fit for Freescale currently but could be strategically important. A spinout was a way to get a lot more resources and attention focused on MRAM.
This deal came to us from the folks at Lux Capital. One of their folks had been a former Motorola research director, and knew the people. We did a lot of diligence and partnered up with them. And there were a couple of things we did. One was just financial engineering. If you look at the spinout, it’s a really complicated deal—it’s not just an IP deal. We bought the equipment from Freescale, but it still sits in their clean rooms. They perform some of the processing for us. There are Everspin people in a Freescale clean room that we rent back from them. Then there are products for which they are the customer. So I’d say having a track record and a playbook was important. Having negotiated 60 or more of these deals, our team knows that all sorts of barriers can get in the way. Most of us have worked in big companies in the past, and we remember what it’s like to be in a big company, so when a big company person comes and tells you about this problem, you don’t delegitimize it. It’s a real problem in the environment that person is working in, and you figure out how to work together to get it addressed.
The next big challenge was syndication. You’re talking about building not just a semiconductor company, but one that has equipment and a sales force. You want a lot of money on the table from day one. [NVP and its syndication partners Lux Capital, Sigma Partners, Draper Fisher Jurvetson, and Epic Ventures ultimately put a $20 million first tranche into Everspin, along with an unspecified second tranche.] We worked really hard to line up really strong investors. We and Lux were really proud of that, and it took a lot of time, and Freescale, to their credit, had the patience to wait until we did that. Then there was bringing in external board members to fill the gaps, recruiting a CEO, recruiting a sales and marketing director.
X: Do you have other examples that illustrate different models for spinning out stranded technologies?
DT: Sometimes there are opportunities to combine things from multiple places. We spun a piece out of Maxim Integrated Products [a maker of RFID technology]. We said, ‘These are good guys, we like the technology and we like the people, but to be honest it’s not a whole company.’ So we deliberately went out and found an existing company that was looking to raise funding, and we said, ‘Would you like to do this?’ So instead of spinning this out into an independent entity, we spun it out to an independent startup, and simultaneously invested in the startup, Intelleflex.
X: NVP is usually involved in these spinouts at the Series A stage, but do you make a point of staying around for the Series B round?
DT: Absolutely, we’re in it for the long haul. I give a lot of credit to the founding team here—they raised a big enough fund that we can stay in. We are, in many ways, like a big distributed incubator. The problem with incubators is that typically they can’t actually generate returns for their investors, because they get diluted out. So you really have to be in for the long term. Having said that, by the time companies get to Series B, they look more like classical ventures, and we play a role but we are one of a number of equal players. The spinout angle is no longer relevant and we’re all acting like typical VCs, as opposed to adding extra value.
X: If you have to spend more hands-on time with your companies than typical venture partners, I’m not quite clear on how the economics of the firm work. How do you have time to do enough deals to put all the money to work?
DT: You’re correct that we are more labor-intensive. But I would say that we have more dollars per partner under management than would be typical. Without going into details about how our fund operates in terms of fees and things like that, I would say that our LPs appreciate that we are providing extra value from these deals having a head start. And the reality is that we are each willing to take a little bit less. We are going to divide up the fees and the carry into smaller bites. I don’t think anybody has ever really asked it this way, but we really believe in what we are doing, and we want to see great returns for the fund, and if we each get a little bit less, that’s okay. We’d rather bring more resources to the table and divvy up our share of the pie a little finer.
X: How do you find out about deals? Is it usually a researcher coming to you and asking for help extricating a technology from their parent company? Or is it the company coming to you? Or are you going out and finding these stranded technologies?
DT: It’s all of the above. Sometimes the company will bring it to us through their executive channels or research leadership. Often the entrepreneur will come to us and say ‘Can you help me out?’ In a lot of cases, it may be entrepreneurs that we are maintaining a relationship with over time. Or it might be somebody who just heard about us—maybe through reading your article in Xconomy. The other case I’m finding, increasingly, is VCs in the valley. Everybody in the valley has a neighbor who is a VC, and the neighbors talk to these partners and say, ‘You should create a spinout from X or Y.’ But they don’t really do that. So we get a lot of referrals, and if we do go spin it out, they may be very willing to co-invest with us.
X: I wonder if your job of finding spinouts is made a little easier by the heightened awareness around Silicon Valley of the examples where big companies let promising technologies slip away because they didn’t see the possibilities. I guess Xerox PARC and the personal computer is the classic example.
DT: I’d push back a little on the Xerox thing. I think they get a bum rap. They absolutely should have owned laser printing—sometime let’s have a separate conversation about that. But PARC has delivered a lot of technology to Xerox. People don’t talk about all the value PARC has created for the company; they just look at the one that got away.
But my sense is that there are a lot of jewels still buried in these companies. I also think that they believe that it’s just too hard [to commercialize disruptive ideas], that you are not really going to get it out to market, and even if you do the team won’t act like a startup; there’s just a litany of reasons not to do it. There is also a belief, and there is some truth to it, that a lot of these things will just leak out anyway–that people will just leave and start something new. That’s easier in software. But in technology, I think most of these things just get killed off—they go to the nonexistent “shelf” and then get reinvented many years later. Unfortunately, there isn’t the will to do anything about this in the mainstream venture community.
X: Do you come across companies that style themselves as so innovative that, basically, it’s impossible to get things spun out from them, even if that would be the sane thing to do?
DT: Absolutely, but that’s probably just a stage in a company’s development. When you are in rapid growth mode, you are trying to do everything. Google is in that space. I would not expect to see a lot of spinouts from Google at this point. You’re seeing a lot of people leave Google and trying to start things, but the view at Google is that if it’s a really good thing to do, we will do it ourselves and make it fit. I think there will be a swing of the pendulum, where they will decide their interests are best served by focusing on a smaller number of things.
Microsoft is interesting in this respect. They do have an IP ventures group that looks at spinning things out, but not with the original founding team. They’ll try to identify something they can spin out, and then recruit people to learn the technology, so they can retain the original team. That is not as attractive a model to us. We are not 100 percent opposed to it, but typically a startup has to reinvent and change directions a few times, and you really want to make sure that you know the technology.
X: You’ve probably seen cases where entrepreneurs inside big organizations had great ideas, and they got “shelved” rather than spun out. What happens to those people then?
DT: It depends on the orientation of the people. Some are hugely loyal to their companies. They work on a project, they love it, they want to foist it on the world, but as long as they feel the project got a fair hearing in the company, they are willing to accept the “no” and go on to do something great for the company. I saw that at Intel. The key part was to make sure they got a fair hearing. Another group of folks may have a different disposition, or their allegiance may be more to the technology, and when you try to take those people away from their dream, you run a high risk of breaking them. They may stay in the company, but with an attitude that may be bad for the company and for the people adjacent to them. Many executives tend to say, “I want to keep my people,” but ironically when they cancel a project they just put them into the redeployment pool and let them fend for themselves. So it’s very weird. As a research manager, you want to keep your people, but the right thing is probably to stop and say, “Where are these people going to do best?”
X: It sounds like there’s an element of career counseling to what you do.
DT: I wouldn’t say collectively that we necessarily do that, but individually I see that as part of what I’m doing. I’ve had circumstances where I’ve been talking to a corporate partner, even about something that we’re not going to spin off, and I’ll say, “We don’t think we can really do a spinoff, but on the other hand you have a problem—you have great technologists who care passionately about this, and is there a good answer here to get a good result for that organization and those people?” Then too, I’ve seen cases where somebody has a project canceled, got their hearing, went on, and did another project that was better and more closely related to the company’s agenda and then benefited from the learnings from the first project.
X: Are there any other venture firms that do what New Venture Partners does?
DT: There was one player in the past, Blueprint Ventures. They failed to raise another fund. To be honest, I was a little disappointed, in the sense that it was nice to have another player in the space. But it’s also nice to be the only one.