Interview: VC not giving up on carrier market

St. Paul Venture Capital General Partner Rod Randall built a resume within the network industry at product companies such as Ascend, which got snapped up by Lucent Technologies, and Teleos, which was bought out by Madge. With today's grim IPO and merger markets, who knows if his current investments -- carrier infrastructure players Avian, Kenetec, SnowShore and Teltier, will ever go public or be acquired. But Randall, who started his career at AT&T Bell Labs, isn't giving up on the carrier market these companies are trying to sell into. He told Network World News Editor Bob Brown last week that he still sees an opportunity for service providers -- especially wireless carriers -- to be successful by differentiating themselves through the offerings they deliver to corporate and residential customers.

Q: Given the bleak state of so many telecom carriers these days, do you have regrets about not focusing your investments more on enterprise networks?Ironically, we believe now is actually a good time to make investments into select spaces focused on service providers. The prices [to invest in these new companies] are down because the industry's kind of on its tail, so the opportunity there is to find those areas where the service providers are not trying to come up with whole new business models, but are trying to extend their business models for enterprise applications.

Q: Many of these new carrier equipment and software suppliers came out of the chute talking about all the advanced new services their technologies would enable carriers to offer. So, where are all these advanced services?A lot of the CLECs and DLECs failed their economic models with the installation of DSL. But there is still a lot of DSL being deployed, mostly by ILECs and much of that is from what, at least at one point, were new companies. The entire Lucent offering was something from its Ascend acquisition. To some extent, DSL was both new and not new at the same time. The first digital subscriber lines were ISDN lines. DSL just had a different twist with the always on ATM underlying layer and [the] focus [was] on Internet access. There was this twist of offering voice over DSL for relatively small businesses but that appealed to the CLECs who were very thinly capitalized and whose business plans could not be financed overall. So they either went bankrupt or morphed their business model. So now most of the DSL available is not focused on voice, but just high-speed Internet access for a broad array of customers.

Q: So, so much for the exciting new services? Is it true that carriers are focused more now on just better exploiting their current infrastructure?Nowadays the mantra is: How do you 'sweat' the assets of the network that is there. Instead of trying to radically increase the capacity of the network, they're asking: How can I radically increase the utilization of the network? One example of a category of new services will be at the intersection of where these broadband metro fiber networks have been installed and they touch large tall shiny buildings. Some of the more innovative services I've seen that may appeal to enterprises is this whole concept of getting an Ethernet interface from the server provider but over that interface getting a range of services, one of which is the equivalent of committed information rate. Let's say instead of paying $1,000 a month for 512K bit/sec access to the IP backbone, what if I could pay $1,500 a month as an enterprise and instead get a 2M bit/sec guaranteed pipe along with burstability up to 15M bit/sec or 20M bit/sec if I can share that with a bunch of other guys in the same building. The carrier can use a building aggregation switch that has the intelligence of allocating the committed chunks of bandwidth to the guys that buy it and allocating the remaining pool on an as-available and as-needed basis. It's 10 times the bandwidth at 20% to 50% more price and the operator gets to do it at a fraction of an increase to its existing infrastructure costs.

Q: Where are the wireless infrastructure vendors headed?What everyone is trying to do is figure out how to help decrease the rate of decay of the voice services provider's average revenue per user on a monthly basis. One strategy is to offer advanced data services as the networks gain more capabilities, as new phones and PDAs with improved capabilities come to market. Another strategy is to offer advanced voice services that cause people to simply use the voice network more or use some of these data services in a way that enhances the value of voice. One of our investments, Avian, aims to enhance these core wireless networks so carriers can offer specialized services. One example would be the ability to shape the data for an enterprise and offer very flexible billing models that allow an enterprise to pay for a user's data when they're getting corporate e-mail but allow the user to pay for his own ringer tone downloads. So you need to dig deep into the packet, it's not just a switching element. It's content sensitive and integrates deeply into the billing structure as well as the allocation of resources and quality of service.

Q: What's all this mean to corporate customers?If you look at how remote access happened, people were first dialing into their corporate networks over their own independent modem pools, then dialing into their companies' modem pools directly and then there became these virtual modem pools owned by the carriers. Well this is what's happening in wireless now. Increasingly, people have PDAs and wireless access from their PCs whether it's an 802.11 wireless access hot spot or [Global System for Mobile communications (GSM)], [General Packet Radio Service (GPRS] or 1xRTT [advanced Code Division Multiple Access], wireless access in the mobile network. Somebody's got to pay for the bits going through the wireless network that hit the corporate network to derive e-mail, and do SAP and other corporate applications. Historically, every individual user would have to pay his own bill. What new wireless infrastructure devices will do is create the equivalent of an 800 number for data in the corporate location. For the first time, it gives network executives some kind of control over wireless corporate usage, and this is just data.

Q: To hear it from carriers, they'll never be able to offer newfangled wireless services if they don't get more spectrum. What's your take on that?In the states, there has been no auction on the 3G spectrum and no designation by the government as to how it's going to handle 3G spectrum. So there is speculation that if it's overpriced it's going to ruin it for the service providers and either they won't buy it or if they do buy it, they'll bankrupt themselves. Right now the reality is that there does exist quite a bit of spectrum which can be the basis for quite a few services. All the service providers are now rolling out higher-speed data offerings whether it's 1xRTT on the CDMA networks which will go up to 144K bit/sec or GPRS for TDMA and GSM networks, which will go up to 64K bit/sec or arguably the high 40K bit/sec range. Then the 802.11 wireless LANs add a tremendous dimension of flexibility and many service providers are trying to figure out how to offer the individual users the best of both worlds. If they could have combo cards that offer 802.11 and licensed mobile interfaces, customers could get seamless roaming.

Q: Is St. Paul investing in 802.11 companies?All aspects of 802.11 are hot. Some people believe the plays for VCs have largely gone, except for maybe some advanced chips. But the reality is that some of the most interesting opportunities are in security [St. Paul has invested in a company called Bluesocket that addresses wireless LAN security]. The [Wired Equivalent Privacy] designed into 11b is such that it can be cracked relatively easily and that creates a huge security risk.

Q: Are we heading toward an overfunding situation in wireless similar to what we saw in optical networking?There's always that possibility of overfishing that pond. I think there are too many people trying to address [Gateway GPRS Service Node] issues. But we're at the beginning of the wireless evolution and there's still lots of expansion that needs to be done and problems that need to be solved.

Q: How many wireless devices do you own?I have two on me, one in my briefcase, two more sitting idle at home and one in my care. That brings you to some of the problems the industry still needs to solve.

Join the newsletter!


Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.

More about AT&TBell LabsLucentLucent TechnologiesSAP AustraliaSEC

Show Comments