Why one VC firm puts its money into certain network and software companies

Azure Capital general partner high on VoIP, data center companies

Paul Weinstein, a general partner at Azure Capital Partners in San Francisco, says that one way the early-stage venture capital firm differs from others is that he and colleagues perform primary research on potential markets for investment along the lines of what they did during their days with financial giant Credit Suisse First Boston. Weinstein, who says his company has a little over $500 million under management and concentrates on communications and software from consumer applications to core infrastructure, serves on the boards of BroadLight, Fonality and World Wide Packets. He spoke with Executive News Editor Bob Brown about why Azure puts its money where it does.

What's your approach to investing in the VoIP market?

We look at the VoIP market being segmented into consumer, small business and large enterprise. When we did the analysis the thing that stood out about the enterprise market is that it was dominated by large players like Avaya, Cisco, Nortel and a few others. Large enterprises tend to make VoIP decisions in the context of an overall architecture decision related to their data products as well. When you're selling into that market it's helpful to have a legacy in your product set. It might not be the ideal place for entering with a venture-backed startup since the market is dominated by distribution and access to those large partners.

So we looked at the small business and consumer markets. On the consumer side, we saw it trending toward being a service. We looked at a lot of service and equipment providers and came away concerned that the business model didn't support enough margin for smaller players. We looked at [mobile virtual network operators (MVNO)] and other ways to get into that market.

On the small and medium business side, we've been in Sylantro [Systems] for a long time, which focuses on that market as well as on carriers that want to offer Centrex-like services to that market. We also looked at people who want premises-based solutions, which is where we came up with Fonality. It's an open-source-based IP telephony company capable of selling a PBX over the phone. Their price points are $5,000 to $7,000 for full PBX functionality.

That small business market is a multibillion-dollar one dominated by no one. There are lots of players in that SOHO market - Cisco, Nortel, Mitel, Avaya - but much more fragmented than the enterprise market. With the profile of companies we looked at we felt it would be easier for them to target and attack that market.

Where do you see the VoIP market headed?

The market is extremely hot and it is price-sensitive, which plays to our strength since we have a low-cost solution built around open source. Things are also changing in that, traditionally, interconnects or TDM VARs were the way you serviced that market, but you now have data VARs servicing that market. The other conclusion we came to was that there's this raging battle over whether you should be PSTN- or IP-focused, and what we found is that the hybrid model actually seems to dominating because there are certain applications where you need to have the quality of service around PSTN, for example. You might not want an all-IP solution because of the risk of downtime on the phone system. IP doesn't work all the time - one, because you have a dirty LAN connection, and two, because of problems in the WAN. One of the areas of investigation for Azure is technologies to clean up IP traffic in the WAN.

Are you talking about companies that do accelerators and that sort of thing?

Not so much acceleration but the guys who do packet shaping and management and have the ability to control your MOS (mean opinion score) that have an empirical impact on the quality of voice over the WAN. There are other guys who do diagnostics, which are important to get you set up to do IP telephony as well as to monitor the ongoing health of the network, which we think of as the insurance business. That's a pretty interesting business because there is a long recurring revenue component to it. Most people that get into IP telephony do so because there is some tremendous cost savings they think they'll achieve and we notice that they're willing to pay a slight premium.

So you're making investments here?

This space is still very much in its infancy and over time we'll be making investments here. One of the models that we think can work is if you try to build a platform company that can do multiple services across the same appliance. One thing VCs tend not to like is one-trick ponies. We love focus on solving a customer problem, but we need some vision to allow you to add services over time because that's the only way to protect your market and profitability. In Fonality's case they have services that are basic but can move to services like conferencing and IVR, then start to think about collaborative IP applications.

What about investing in service providers?

We haven’t done it, though we have looked at a bunch of hosted VoIP providers. Think of the SunRockets and Vonages and Covad look-alikes that are selling into enterprises. Under this business model customer acquisition costs are the biggest single determining factor in not investing. You can sort of see that in Vonage's numbers. They spent huge dollars for marketing and customer acquisitions. The good thing about service providers now though is the amount of capital required to build them is substantially less than it was five or 10 years ago. It's much more of a software-as-a-service mentality; there is a greater ability to offer bundled services. Though you only want to do this in environments where broadband has high penetration rates or there's the possibility for high penetration rates. What also happened in the early to mid-'90s is that most large service providers began re-architecting their networks for the consumer bundle as opposed to the business bundle, because a lot of the business services got commoditized. So if you're going to be a service provider you need to architect for the consumer.

Beyond VoIP, where are you looking?

We're looking at a noise suppression company. Originally they designed it around cellular opportunities but because of Skype you're seeing more of these Jabra headsets and we're taking a look at this market because the volumes are actually really interesting. The question is if you have cellular and VoIP connections then if you start to think about what's going on in the enterprise with fixed mobile convergence, people talk about Wi-Fi handsets transitioning between your PBX and cellular carrier seamlessly and remotely there's certainly going to be an aftermarket for peripherals.

What areas are business plans focusing on in communications?

IP TV. How get TV over some broadband pipe and all the associated stuff that goes with that: servers, middleware, set-top boxes. Our focus is around the home entertainment side of things, how do you shift video around the house.

VoIP has been around a long time now, but we still see a steady stream of packet telephony-related opportunities. The stuff that's getting more and more active is wireless access technologies, 802.11, some WiMAX, UWB is making a big comeback (there were two competing camps and now things are solidifying around one… a lot of the activity is around the silicon and radio technologies). Then there is the other kind of wireless, municipal and backhaul, both Wi-Fi and WiMAX.

It seems like we hear about another community setting up a Wi-Fi network every other day…

Yes, it's very hot. But there's a deployability question that still needs to be answered. One thing we've noticed with wireless and wireline is that these rollouts are nontrivial with respect to integration. The degree of success of a rollout will be a direct function of the partners in that ecosystem to roll it out. We've had experience with companies where they're trying to deploy citywide networks, and you cannot imagine the number of people who touch that from the public works department to the guy doing municipal bonds financing to your local telco and the guy who is producing a high school video to ship across that network. There are so many touch points that it just takes a long time. That's one area where things could slow a little bit. But it's a huge opportunity since there's another leg to the story I think around backhaul solutions that cellular carriers are going to want as they move to 3G and 4G applications. The market is dominated by incumbents but if you spend any time with the Nokias and Ericssons they know it's a huge pain point for their carrier customers. They spend an enormous amount of money just doing backhaul over T-1 lines.

Any other areas of interest?

The data center has kind of been re-emerging. From '01 to '04 it was very slow, with part of that being that capital just wasn't available. We're seeing technologies around blade server architectures, new security architectures, investigating how to speed up TCP/IP, edge routing. They're not huge markets on their own, but together if you look at how you re-invent the data center, there's a lot of stuff going on and it's all enterprise. Virtualization technology is creating a lot of submarkets; if you can virtualize machines for the OS, can you virtualize them for other things, too.

We also see optical strangely enough re-emerging. The companies that are established by and large are doing pretty well. Some are re-inventing themselves. The end drivers are that people are upgrading cellular networks, companies like Verizon are building out fiber-based networks, there's a reinvention on the DSL side. All of that stuff has optical backends. There's an ebb and flow to this, too. Service provider consolidation will have some impact on the negative side.

RELATED: See our recent write-up of the latest quarterly venture capital investment in network companies.

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.
Must read: 10 new UI features coming to Windows 10