Michael Feinstein joined the ranks of venture capitalists back in 1999, following stints in product marketing at remote access equipment makers Shiva and New Oak Communications, which were acquired by Intel and Nortel Networks, respectively. Feinstein, now a principal at Atlas Venture in Waltham, Mass., currently keeps his eyes open for the next big investment opportunity in the network industry while juggling his duties as board member for a handful of network processor and other product vendors. In this interview, he talks about about the VPN market's evolution, the current state of the network industry and more.
Q: Given that you were with New Oak right there at the start of the VPN market, what's your take on how VPNs have evolved since the mid-to-late '90s?VPNs have grown steadily, but not as fast as early projections indicated. However, the market is now significant. I am definitely happy to see carriers getting into the managed CPE business. This is what it will take for the market to scale and to reach small and medium business. The other development is that the network-based services [based on equipment from since-acquired start-ups such as Shasta and SpringTide] are not growing very quickly. They are more expensive to deploy, at least for the initial customers.
Q: What have turned out to be the biggest drivers behind VPN acceptance?The drivers have been 1) cost savings (vs. managing your own dial-up access), 2) the need to access the corporate network from users that have broadband (cable modem, DSL, broadband inside a hotel room, etc.), and 3) leveraging cheap public Internet dial-up infrastructure. VPNs have moved from something exotic and risky to mainstream.
Q: What about these new VPN alternatives based on existing Web clients?As far as SSL-type VPNs, this is an interesting idea, but a bit ahead of its time. Most enterprise applications don't have a Web interface, but when they do, SSL will be a way to ensure privacy. You still have to deal with strong authentication, authorization, and accounting. This will have to be built into the Web version of these enterprise apps. I imagine that this will happen, but it is still a few years off. When it does happen, I don't think a company based on SSL VPNs will have much of a business -- the apps will have the security built in.
Q: Given that you're invested heavily in companies whose products are designed for carriers' networks or designed for the products for the carriers' networks, I imagine you must watch the broadband deployment debate in the U.S. closely. Does the government need to step in, as some large industry players have proposed?I'm not one to design public policy, but I do know that I don't want the government to try to solve a business model problem. The market changes too fast for the government regulations to keep up. A model that might work -- except for special-interest pressures -- is one that is used in the auto industry for overall gas mileage. If the government wants to raise overall mileage, it tells the automakers that they need to manage their business to achieve some higher mileage across their fleet of cars. The automakers have to figure out what tactics are the most profitable to get there. Or, they try to lobby the government to change its mind. Maybe the government should mandate that the incumbent carriers must price access to their lines in a way that ensures that at least X percent are leased by competitive carriers.
Q: Is there any evidence of technologies coming along that would enable the cable companies to become bigger players in network services for enterprises?Look at Quantum Bridge [an Atlas investment]. Their biggest customers -- Comcast, Time Warner -- are cable companies that are branching out to serve business customers with unique offerings. It will take a while to become a significant share of the total business access market, but passive optical networking [Quantum's specialty] is the best place for the cable companies to start due to their architecture.
Q: Companies like Quantum Bridge and WaveSmith that are in your portfolio have been well received in the market, based on their technology. But given the bleak state of the carrier market, how are you trying to get your companies through these tough times?We are focused on getting the companies sized appropriate to today's market. We don't expect carriers to buy much of anything until at least early 2004. So, you have to wonder why a company needs to spend a lot on sales and marketing, or on accelerating feature development. It takes more finesse now to build a company in this space -- you have to time your investment to when the customers will start to buy. Each of the companies you mention have strong technology and should be survivors in their space. As a board member, my job is to work with the CEO to ensure that the company survives with a winning plan that can take hold when the market comes back a bit.
No. The carriers are focused heavily on squeezing costs out of their existing architectures, that means less capital and operating expense. They're focused on getting sounder balance sheets so they can make investments in the future. Although the wireless side is different, there is innovation going on there. Verizon recently rolled out its new data service, for example.
Q: It's often said that the processor market is one of the first places to look if you want to determine whether the health of the tech market is changing. Given that Atlas has invested in components companies such as Sandbridge and Telephotonics, what are you hearing from the processor industry?It was really dead for a while in terms of new customer projects. But now we're seeing some new development projects popping up, even at the bigger vendors. It's not going to go back to the way it was before, but most of the big networking vendors had shut down R&D on anything new in the last year to cut costs. I'm talking about the Nortels, Lucents, Ciscos and Junipers. The big vendors all had to do layoffs and they rationalized that by just dropping new product development. We're starting to see some signs that some of those companies are loosening the purse strings a bit and starting up some new projects. They really have to or they run the risk of their current technologies being behind in a couple of years. To be honest with you, as a venture capitalist with network equipment companies in my portfolio I don't mind so much since the bigger vendors will wind up having to buy somebody. But we have seen an interest in some of these component plays; there's been a steady stream of interest on the wireless side.
Q: Given that there were so many carrier suppliers started up in recent years, and that their potential customer base has dwindled through the bankruptcies of CLECs, etc., are you seeing any of these companies refocusing on enterprise networks?This is happening, with modest success. One of our companies, eDial, repackaged and repositioned itself to go after the enterprise voice conferencing market, rather than a carrier service. However, many of the hardware products have architectures that aren't well suited to enterprise cost points. In general, companies have to do some sort of restart to repackage themselves for enterprise customers. DWDM is definitely getting cheap enough that larger corporations are already deploying it. IBM is one of Nortel's largest resellers of DWDM equipment into the enterprise. Overall, there will be more investment -- either new companies or restarted older companies -- that will focus on the enterprise market. That has to be good for enterprise customers.
Q: I see you're on Vividon's board. I get the impression that streaming is one of those technologies that, while maybe not on the top of every enterprise's IT budget these days, is one that's paying off when people deploy it. What's your take on how streaming is being rolled out in enterprise networks?Streaming is definitely on the rise, but, like many things, has not grown at the rate people expected. Vividon has been selling to both enterprise and service provider customers. Where they have had success is in service providers doing large scale streaming (like their recent announcement at Radio Free Virgin) or in enterprises that need significant scale and quality of service (like the MIT Sloan School, which is streaming videos of class lectures to students).
Q: Where streaming is being deployed, who within these organizations is driving deployment?Streaming is being driven by several factors: 1) Content owners are moving to richer content (audio, video) to enhance their content experience and to continue to attract and retain viewers; 2) There are some emerging business models that show that rich content can actually generate an incremental revenue stream, such as with RealNetworks' Gold Pass and Major League Baseball's offerings; and 3) enterprises are leveraging their internal infrastructure to deliver content to their users on a worldwide basis for training, corporate communications, etc.
Q: Atlas has investments in a fair number of companies from outside the U.S. We don't tend to see a lot of new overseas companies hitting it big here in the network industry, aside from maybe in the security market, where companies with Israeli roots have done OK. What's behind your efforts to go overseas for investments?It is much more common for international companies to get acquired than to go public. So, they don't become as visible. But they can be very good investments. Atlas has always had an international strategy -- we started in Europe in 1980. We believe that there are different advances in different geographies and that we can leverage those differences to build better companies. For example, we have quite a few wireless investments in Europe, and relatively few in the U.S. Most of the U.S. wireless companies don't stand up against the international requirements, which are generally better understood by European companies. There are, of course, exceptions, but this is an example of one geography being pretty far ahead of another. During the optical boom, you could have said something similar about the U.S. being ahead of Europe. In general, we invest in companies that we think can be global players in their particular sector.