Thursday, June 25, 2009
Interview with Cliff Rees, XCast Labs
It seems that every week, a new, Voice-over-IP provider firm surfaces in the market, either providing PBX services, voicemail, or other similar services. Los Angeles-based XCast Labs (www.xcastlabs.com) announced earlier this month that it has raised a $2.7M round of funding from the Pasadena Angels and Frontera Capital for a voice-over-IP service, and we thought we'd talk with Cliff Rees, the firm's CEO, to understand how the firm is different from those many providers.
What is your VoIP service all about, and how is this different from what seems to be lots of Voice over IP service providers?
Cliff Rees: We're a software as a solutions company. We wrote our own code, which really distinguishes us. We wrote all of the core components of our technology, from the ground up--even our SIP stack, OIP application, session-border, softswitch, and media server--which runs on standards-based, off-the shelf servers. Running on a Dell pizza box, we can run 20 times as many calls as other software, giving us the lowest cost basis in the industry. Our major designer was the number two guy on the Lucent 5E development program, so we understand how to design scalability into telecom apps, and scale seamlessly, better than our competitors. Most of our competitors are Broadsoft resellers, or they might use some type of freeware like Asterisk. Asterisk is perfectly fine for a small business running an IP PBX, but if anyone thinks they can take Asterisk, plunk it in a garage, and become a VoIP reseller to many hundreds or thousands of clients, they're in for a rude awakening. Once you get to 300-500 customers, it blows up, because it was never intended to scale. Asterisk was not build with scalability in mind, and it was built by software engineers who don't know telecom.
One of the other things we do very well, and where we have engineering expertise, is we provide one-off, interactive IVR solutions for media companies in the Los Angeles Area. For example, we support promotions for Sony movies, a combination of database and telecom work for interactive TV shows like the Price is Right, Deal or No Deal, the Game Show Network, and I'm A Celebrity Get Me Out Of Here. Nobody in our competitive space can do all of those things, because they simply don't control their own destiny, because they don't own their own technology. They're led on whatever path Broadsoft, Acme Packet, or others want to do. Technology wise, we've got our own engineering expertise, and we've built our own video softphone with HD video--which we have in alpha--for HD videoconferencing on a standard CPU, as opposed to the extremely expensive, $20,000 systems that exist now in corporate conference rooms. Instead, we do this for about a $1000 for a site, in HD quality, supporting eight to 10 participants on a 48- or 52-inch screen. The technology is really what sets us apart from our competitors, who are mostly resellers of someone else's technology.
We understand your service also connects into Skype?
Cliff Rees: As far as I'm aware, we're unique in a VoIP service provider, as we've figured out how to do that on our own. We've figured out how to translate Skype calls into SIP, and vice versa. A Skype user can use us, not using SkypeOut, but as a free call. They can call us, we turn it into a IP call, and a 10 digit phone number, we then figure out where that user is on the planet, and deliver it over IP, making it a complete on-net call. The reverse also happens--if one of our users wants to call Skype users, they can do that purely on the net, without using the PSTN.
Can you talk about your team, and where it came from?
Cliff Rees: The core of our engineering team came together at Webley Systems, which developed the first commercial VoIP application in the last half of the 1990's. We created a find-me, follow-me application which rulled out on MCI to 60,000 seats. It was the first, commercial VoIP application, and was built to be scalable, because it was intentionally meant to be sold to the carrier marketplace. When the Dot Com bubble burst in 2000/2001, it became not as much fun to work as WebleySystems than in the past. Our CTO, Vladamir, the head of our engineering and core development team, went and formed his own garage R&D company. Starting truly from scratch, we built our own SIP stack, built much better than the other code out there in the market. We started with some fundamental assumptions -- it shouldn't demand a Sun Sparc CPU, it should be able to run on an off-the-shelf server, it needed to be scalable, and it should have no inherent gotchas that would limit it in any way to less than millions of users and tens of millions of calls per day. In 2004, I was introduced to Vlad through a mutual friend in the telecom business. I saw immediately that this was something interesting. I had been in telecom for 20 years, and I understand and know impressive work when I see it. I knew enough about buying and selling access, and teams, to be able to form a company around it, which is what we've done with XCast Labs. We've been able to triple our revenues over a year and a half, and are break-even profitable on a cash flow bases--no funny EBITDA numbers. Our future is looking very bright--in the last two months, we've signed up more customers than we ever had before.
How do you go to market--are you selling this directly to end users?
Cliff Rees: One of our major focuses is selling to cable MSOs or private cable operators--CLECs, RLECs, ILECs, and marketing companies, who already have an existing customer base and want to offer VoIP services of some sort, to residential single users or others--but don't have a couple of million dollars or engineering expertise. We will go to the cable MSO, and totally white label our technology so it looks and fells and smells to end users, customer service, technical support, and sales people, like their own brand. Avenue Broadband, one of our cable MSO customers, sells our service to existing end users of their cable TV and high speed Internet service. So, we see ourselves as a technology provider. We're much more interested in providing technology to people already with a customer base, than generating our own customer base. We're not actively going out, like Vonage, where they spend $300 per subscriber--that must take them a couple of centuries to amortize that investment in a subscriber. We have no interest in doing what I consider completely uneconomic.
Can you talk about how you connected with the Pasadena Angels and Frontera?
Cliff Rees: We are based in Los Angeles, although I work primarily out of my home office in Portland. Our headquarters is in Los Angeles, with VoIP, telecom, and network operations in Chicago--but Los Angeles is our home. Members of the Pasadena Angels and Frontera--you'r probably familiar with Eric Manlunas--were familiar with what was going on in the local area, so we made presentation to the angels back a number of months ago, and they were very supportive of what we do, both at the financial level, and with contacts and at the general business level. A couple of the Pasadena Angels and Eric Manlunas are also on our board. David Seimer of Seimer and Associates is not an angel, but a boutique venture capitalist and technology investor, and he and Eric Manlunas had shared previous investments. David got very interested in the uniqueness of our technology, because we can do things that no one else can do at this point. We've got patents on technology that helps save bandwidth utilized for VoIP calls and particularly for video. I personally am still the largest, single investors, because I believe in our technology, and put my money where my mouth is.
We mentioned Asterisk and open source earlier -- how does a team like yours compete against the many, many developers who contribute to an open source project?
Cliff Rees: Using open source code from something like Asterisk is perfectly reasonable for an individual, or a small or medium sized business, or maybe a really small service provider. But, it takes serious investment of time and energy to create and support something that can scale to hundred or millions of users. There is no incentive for someone in the Asterisk community to bother with taht. Why would they spend tens of thousands of man hours, when there is zero revenue possibility coming back to them? Open source communities can produce some very impressive software, but, to create truly commercial applications that suit the needs of businesses, you want to buy from people who are really committed and devoted to a project--and have a very strong, economic incentive to remain focused on that product and support it. We think Asterisk is an entirely different marketplace, and therefore not a competitive threat.