Aevia Inc. (www.aevia.net) expects potential customers to be so happy with its service they'll be singing its praises. The name of the company is a form of the word hallelujah created in the 15th century by Dominican monks. Aevia, which calls itself an "ASP-CLEC," this fall launched service in its hometown of San Francisco. The company expects to be overwhelmed by the demand for its 10mbps to 1gbps capacity and hosted services to individual business customers. The company has a proprietary technology called VIRTUALFIBER that will enable it to provision 100mbps to buildings over existing copper connections rather than doing a fiber buildout. Doug Picard, the company's president and CEO, recently spoke with xchange's executive editor Gail Lawyer about this technology and how it ties into the company's business plan.
X: What is Aevia's business plan?
Picard: We've targeted 35 metropolitan service areas in the U.S. Our goal is to lease dark fiber whenever possible, then build our metropolitan area rings using that existing fiber. We are applying for CLEC status in each state, so we have the ability to build our own fiber should it become necessary. But we're trying not to reinvent the wheel. We're avoiding building fiber whenever possible because of the cost. In metro areas, we've had vendors quote us as high as $650,000 per route mile to build fiber. The average runs about $250,000 a route mile, which is why there isn't a lot of fiber in the U.S. today.
X: From which companies are you purchasing dark fiber?
Picard: We haven't made those announcements yet. But we're working with practically everyone who is willing to lease dark fiber.
X: Metropolitan transport is a hot topic these days. How do you anticipate setting Aevia apart from other companies in this space?
Picard: There are some companies that started a couple of years ago and have tried to gain early market share. What they're finding is that fiber just goes by about 5 percent of the businesses, so they can't execute their plan in a timely manner because fiber doesn't follow Moore's Law. Fiber is a linear construction build. It's going to take five to 10 years to achieve 45 percent penetration in business areas to achieve direct fiber connections.
One of the biggest differentiators in our plan is a product that we call VIRTUALFIBER, where we can provision 100mbps to buildings in these metropolitan service areas today using a different methodology that is very low cost.
X: Is VIRTUALFIBER a flavor of DSL?
Picard: No. DSL is very slow. Even VDSL [virtual DSL] is very relatively slow compared to the types of connectivity we're talking about. We're talking about true broadband, 100mbps connectivity.
X: How does VIRTUALFIBER work?
Picard: We're using a proprietary method to deliver it over the existing ILEC copper infrastructure. It's technology that we have.
X: You refer to Aevia as an "ASP-CLEC." What does that mean?
Picard: We're breaking ground in a lot of areas now. One of the things we're doing to differentiate us is putting in services in our super PoP. Each of the metropolitan service area networks we'll build backhauls to a super PoP facility, where we're putting in our own and third-party ASP services in the voice, data and video categories.
Customers inevitably want the connectivity, but it's very difficult to get high-speed connectivity from anyone today. And they want services. Some of the hot ones are storage area networking and unified messaging. Down the road we're going to be delivering business cable TV. And we're close to signing an agreement with a portal company that will be able to develop a "My Aevia" type of portal.
X: What industry issues or customer problems is Aevia trying to remedy?
Picard: The way we've designed our network costs about two-thirds of the standard CLEC buildout for a metropolitan play. The reason we're able to achieve that is that we're taking a mix of expertise in our company. I come out of the enterprise networking space and am familiar with voice video and data. Our COO, Tom Glendenning, has 20 years in the CLEC and ILEC space. So we're taking these two sets of expertise, taking best practices from both and avoiding pitfalls from both to develop a next-generation best- of-breed network design.
We've put together a complete end- to-end Ethernet solution. We've built the entire network in the lab, with rolls of local loop, also using 23 GHz point-to-point radio and fiber. We've built out an entire DWDM metropolitan ring, a super PoP and end customer connections to lengths of local loop on giant rolls in the lab. And we've tested the thing end-to-end and it works. We're getting about 7 milliseconds of latency end-to-end, which is phenomenal for this size of network and the type of traffic we're putting over it.
X: What types of customers are you going after?
Picard: We're really looking at anyone who is paying $1,200 to $1,500 [per month]. We can give them the same speed for a much less expensive price. Once we bring our connection in, when the company wants to double the speed, we can provision that overnight, and the cost is about half.
X: Who are your prime competitors?
Picard: It's a question of execution. We've seen one of our competitors decide that the local loop was too hard, and they're trying to refocus their business model on the wholesale side of the game, carrying traffic between carrier hotels and central offices. Our competitors are most of the broadband companies that are trying to do data connectivity. The question is whether they can execute or not, whether or not they're really competition. Our sales team is not seeing them on the street. They don't really have a way to get to these customers, and it's going to take them a long time to get there.
X: The company will initially offer services in San Francisco?
Picard: Correct. We're constructing that right now. And we [expect to] start ramping up in New York and Los Angeles by the end of this year.
X: Do you already have customers in place?
Picard: They're signed. They understand what we're doing and want to get going. It's a very compelling offering for them. We're working on [lighting them up] right now.
X: How do you envision the broadband market evolving?
Picard: Nobody is going to be able to do anything until they get fiber in the ground. It's going to move slowly--a matter of sitting and waiting. Fiber takes time to get in the ground. It's very expensive to construct. We're wondering if our VIRTUALFIBER might not only allow us to grab a lot of market share really quickly and hold onto it through our services, but preempt some of these fiber installations because they won't be needed.
X: Is VIRTUALFIBER a technology you might license to others, or is it something that would be Aevia's competitive advantage?
Picard: Definitely a market advantage. It's highly unlikely we're going to share. I don't see the ILECs giving us any fiber. And the ILECs seem to be focused on the residential markets anyway. And the cable companies completely bypass the business markets. So the business markets get stuck with the standard TDM.
X: Is your function as an ASP another advantage?
Picard: Absolutely. That's what our first customer is doing.
The interesting thing about owning the last mile is that you have control over the services and bring bandwidth where it's needed and move services where they're needed. We can take a certain service the edge customer is looking for and move it closer to them, so the performance is excellent. If people are providing services to the world, we can move that closer to the peering or transit point. It's smart management of bandwidth utilization from end-to-end.
X: Have you had difficulties getting customers to understand the importance of the location of their hosted services?
Picard: People definitely get it. Right now folks go for collocation facilities, and that's probably the best way to get a controlled managed environment and push traffic to the edge. The problem is that it's sitting there in this cloud in the center, but the last mile loop to the customers is really low speed, faulty and slow to provision. So, you have this big core going on, all these services sitting on it and all these people sitting around it trying to get to it, but there's this gap called slow bandwidth and a long lead time to provision.
X: In addition to your less costly network design, how is Aevia able to make broadband access and services less expensive to its end users?
Picard: We're working on Internet time, and we're running a lean and mean crew. We have 12 employees now, adding 10 next month, and by the end of the year we should have 50. We're going to show that it doesn't take hundreds or thousands of people to build and run these networks. The way we put it together is extremely efficient, and we're able to operate this network with very low overhead with "five 9s" of reliability and superior throughput services. That means we can bring the customers better quality services at a better price.
