Late last month, GTT closed the acquisition of One Source Networks, its seventh acquisition in the last six years. Over that time, the company has transformed itself into an enterprise-focused cloud networking and managed services provider backed by its own top-5 global internet backbone. With us today to tell us about the company’s plans for One Source Networks and future growth opportunitiesis GTT President and CEO Rick Calder.
TR: What was behind your decision to buy One Source Networks?
RC: We felt One Source Networks was very consistent with our growth strategy to expand cloud networking services to multinational clients. They fit right in that realm and had a broad scale of multinational enterprise clients. First, they helped our cloud networking service portfolio. We were not in the SIP trunking business, and that is one of the services our multinational clients have asked about. Second, one part of our strategy is to ensure that we provide secure network connectivity to any location in the world and with any application in the cloud. One Source Networks adds a couple of interesting Points of Presence (PoPs), one in Sydney, one in Bogotá, and well developed plans for another in São Paolo. We'll be expanding both of those existing PoPs as well to increase the scope of our core network footprint in Australia and Latin America. Finally, we felt the employees of One Source Networks lived our values. We'd looked at One Source Networks several times in the past and were very pleased to reach agreement.
TR: Other than SIP trunking, are there other cloud and managed services you’d like to add to your service portfolio?
RC: Both organically and through the past several acquisitions, we feel very good about the four pillars of our cloud networking portfolio. There's not a tremendous need to expand it. That is not to say that we wouldn't look for additions and enhancements, maybe in the security space or WAN acceleration. But we feel like we have a broad portfolio of options.
TR: GTT runs one of the top internet backbones as well, how does the transit business fit in your model?
RC: Clearly we run one of the top 5 AS's on the internet. As a function of that, we do have a broad portfolio of internet products. A big portion of our internet services are sold as dedicated internet access to client locations for high capacity enterprise connectivity. Another big part is Broadband IP, which is generally consolidated into big networks via the IPsec capability we sell on IP VPNs. So we have the capability to use our core backbone to provide different levels of IP connectivity to our clients. While we are a top 5 backbone by internet reach and routes, we are not by traffic. So we see a very attractive opportunity to continue to grow the scope and scale of our transit business with some of the largest service providers, legacy incumbent telcos, and content delivery companies around the globe.
TR: How much of your backbone traffic is enterprise versus wholesale? What is the pricing environment like?
RC: If you think about our mix, it is about 80% enterprise and government and 20% carrier, so we are first and foremost focused on the enterprise segment, but I would classify in that enterprise segment some of the largest content delivery companies, Google and Akamai for example. We see lots of interesting growth in that market. It is subject to continued price declines, but Moore's law is in effect and we see our costs come down faster than our price curve. That's why you see our margins continue to expand. As we scale, our cost economies are fantastic and we pass some of that on. In all of our market segments, we see the growth clearly outpacing client price-per-bit declines, and we don't see that trend stopping.
TR: Along with the backbone you also acquired the EtherCloud® business from Inteliquent. How has that fit into your model?
RC: Interestingly, while Inteliquent had owned the phrase EtherCloud, we borrowed that and put that as a label across all our wide area networking services. Prior to the acquisition, our depth and breadth in Ethernet transport services was actually much deeper than what they were doing. Their penetration in Ethernet transport/MPLS/VPLS was actually very small. We took the name and used it as the moniker across that whole part of our portfolio.
TR: How do you see enterprises using Ethernet access today?
RC: Depending on performance and cost criteria, clients use or select a different EtherCloud access option, the lowest end being secure broadband internet access roped in to an IP-VPN with secure connectivity through an IPSEC tunnel into a corporate LAN. Next would be MPLS and our ability to deliver that everywhere. The highest level would be dedicated Ethernet and VPLS technology. And we see them mixed and matched frequently, which is why we call the umbrella of services EtherCloud. But the bigger trend now is to put secure connectivity into cloud apps, being able to connect your users securely on your corporate EtherCloud WAN to IT apps that have been moved into cloud service providers. That's the biggest trend for us. We see that as a continuing evolution in wide area network for big internationals, and it's driving tremendous demand for bandwidth at the end user locations as applications migrate out of buildings into the cloud
TR: Where does software defined networking fit into your model?
RC: We clearly have the ability to very flexibly change existing bandwidth. It's usually limited by the tail, and so most of our larger clients will ask us to manage diverse access facilities into all of their larger locations with a managed router and frequently managed firewall as well. Given that there is a propensity for tails to fail or go down for periods of time, being able to manage multiple last mile facilities into a client premise is key for us -- being able to manage real time traffic flow across it, manage traffic flow across all their corporate locations and be able to provide flexible burst billing providers similar to what cloud service players do on the IT application side. We think we're building a cloud network and service that can parallel the as-a-service mechanism they are moving their services too. We think we are a better way to reach the cloud.
TR: Do you have plans to use NFV-based routing to augment your strategy in the future?
RC: For those that want to move their router functionality into the cloud, sure. However, we have not found it to be a major movement yet. We have found that having premise based routers continues to be the norm. But as one of the larger cloud networking service providers in the world, integrating those capabilities into the cloud similar to what we do with network-based firewalls, intrusion detection and prevention -- absolutely we can embed router services in the cloud as well.
TR: How long do you expect the integration process to take?
RC: We have a very proven formula. An acquisition must be strategic, and it must be integrated within 1-2 quarters. Sometimes they move beyond 2 quarters due to existing term arrangements, but most of the major integration is done within the first two quarters. We've demonstrated that capability for past integrations. And we see the same thing here with One Source Networks.
TR: You have deliberately chosen a low capex, wavelength-based business model rather than seek to directly own the underlying infrastructure. What’s the thinking behind that?
RC: Our business strategy is to provide cloud networking services to multinationals everywhere in the world. As a function of that, our core backbone is on most of the major fiber paths in the world and is in most of the carrier neutral data centers. Because we want to be on all of them, and we need access to last mile facilities to all major carriers around the globe, currently 2000, owning any particular fiber route or building is non-strategic to us. Being agnostic as to the last mile is an advantage to us, not a disadvantage. I'm not saying those are bad businesses, it's just not our business. We happen to be one of the major buyers of lit capacity and colocation space. The intelligence we own is the switching, routing, and security infrastructure to make sure we can control client traffic from any location to any location in the cloud. That happens to be less capital intensive than a traditional player who has a regional fiber network or data center.
If you are a CLEC trying to serve buildings where you want to serve SMEs and you don't own access facilities into those, competing against the incumbent ILEC or cable company becomes very difficult. But we're not a CLEC. We serve a completely different segment where our clients have locations everywhere in the world. We find our unlevered free cash flow tends to be one of the highest in the industry. We like our strategy, and we don't see deviating from it.
TR: With the One Source Networks deal, you will have achieved the targets you publicly announced two years ago. Do you have a new target in mind?
RC: We announced our first growth objective in January 2011 of $200M in revenue and $30M in EBITDA. 11 quarters later, we reached $30M in EBITDA and I think $186M in revenue. Nonetheless, we set a new financial objective in November 2013 to grow to $400M in revenue and $100M in EBITDA. We will achieve that with the closing of the One Source Networks deal. We are committed to growing very aggressively; we believe the growth prospects for GTT are better than they have ever been. That’s why we established our next set of financial objectives to reach $1 billion in revenue and $250 million in Adjusted EBITDA within the next five years.
TR: Is there an organic growth path open to you from here?
RC: Actually, we think over the past 2-3 years we've demonstrated it. Between the acquisition of Tinet and the acquisition of UNSi, we achieved about a 9% organic growth rate. Over the past couple of quarters as we integrated a couple of businesses that weren't growing (UNSi, Megapath’s managed services business), we achieved 5-6% organic growth. We expect in the second half of the year to be back on an 8-10% growth pace and moving forward. We will continue to complement that with selected, strategic acquisitions as well.
TR: What types of M&A would interest you from here?
RC: I would say that if I think about the parts of our strategy - expanding our cloud networking portfolio, expanding reach across the world, and focusing on multinationals, it's probably the third that is of most interest to us now. One of the more difficult things is to crack a new account, and we have found tremendous cross-sell/up-sell opportunities for accounts we add via acquisition. Given the depth, breadth, and reach of our portfolio, what we're looking for are those that add great international accounts. That’s not to say we wouldn't add a little bit of reach to the network, or a little bit of cloud networking focused in the managed service realm. That said, we don't have to do anything of course.
TR: How do you view asset prices in today’s M&A market?
RC: The market is always competitive, and we are never the sole bidder. We are less focused on pre-synergy price, but rather on a post-synergy price after two quarters. What we have been able to do is maintain a disciplined approach to yield post-synergy EBITDA multiples in the 5-6x range or better, and every deal so far has been 5x or better within two quarters. Because One Source Networks is a high growth business, we said that the post-synergy would be 7x or better, but of course we'll be shooting for 'or better'. We will find out as we project our earnings over the next couple of quarters and see how that integration proceeds.
TR: Thank you for talking with Telecom Ramblings!
RC: Thank you for talking to GTT!Ethernet · Industry Spotlight · Internet Backbones · Managed Services