While officially Verizon remains quiet about the alleged auction for its massive data center portfolio, the report that it is looking to offload some $2.5 billion worth of data centers isn't far-fetched.
Other telecoms too have realized they aren’t prepared to spend as much as they learned was necessary to grow a data center business and stay competitive. This is generally considered a good time to sell, and at least some of the data centers in Verizon’s portfolio are highly valuable from a strategic point of view. There are plenty of companies that could benefit from taking them over, given that the price is right.
The Crown Jewel
The most valuable part of the portfolio consists of the Terremark data centers Verizon took over when it acquired the successful service provider in 2011 for $1.4 billion, and the big prize within that fleet, the crown jewel, is the Nap of the Americas in Miami. One of the world’s most important carrier hotels, it is the primary network interconnection hub between the US and Latin America.
“That’s the prize,” Kelly Morgan, research director at 451 Research who tracks the colocation market, said about the building. “A lot of people would want that.” Who are those people? “Any big interconnection provider should be at least interested in that asset,” she said.
The most obvious contender in that category is Equinix, the world’s largest colocation and interconnection company. Creating interconnection ecosystems within its data centers has always been at the core of its strategy, and controlling an interconnection asset as important as the carrier hotel in Miami would make a good fit.
Read more: What’s in Those Globes Atop the Miami NAP?
Another possible contender is Digital Realty Trust, which after a decade of focusing squarely on the wholesale data center business has recently changed gears and dove head-first into the retail colocation and interconnection market, acquiring last year Equinix’s major US rival Telx for $1.9 billion. Ownership of the Nap would help Digital make the case to the market that it can become a dominant force in interconnection.
Other obvious choices would be CoreSite, a long-time player in colocation and interconnection, as well as CyrusOne, the Texas-based data center provider that’s been emphasizing its interconnection play over the last several years.
Spokespeople from Equinix, Digital Realty, CoreSite, and CyrusOne declined to comment for this story.
The Other Prize: Culpeper
While Nap of the Americas is the prize in Verizon’s Terremark data center bundle, it’s not the only prize. Another important asset is the former Terremark campus in Culpeper, Virginia, called Nap of the Capital Region. The four-building campus is about 60 miles away from the big Northern Virginia data center cluster, but it has attracted a solid list of enterprise clients any provider would be happy to get its hands on.
Read more: Inside Terremark's Culpeper Data Fortress
Interestingly, one of the things that attracted Verizon to Terremark was the expectation that it would bring more government clients, and the Culpeper campus, being close to Washington, DC, was where those clients would presumably put their servers. In fact, Verizon was leasing a lot of space in the Culpeper campus to serve its government customers before the acquisition. “Verizon bought them to get a lot of federal business,” Morgan said. The campus did just OK with the feds – not as great as expected – but it did get a lot of traction with private enterprise customers, she said.
The Bad with the Good?
Another industry insider, who spoke on condition of anonymity, agreed that Terremark was an asset that would be of interest to a lot of companies. It’s too early in the process to try to predict what will happen, however, since it’s unclear whether Verizon wants to sell those key Terremark buildings at all. It’s also unclear whether it wants to sell its entire data center portfolio in one go or offload it piece by piece.
In all, Terremark had 13 data centers when Verizon bought it, which also included sites in Dallas, Silicon Valley, Sao Paulo, and Amsterdam – all major markets where providers generally don’t have trouble selling capacity. But there’s also the legacy Verizon data center portfolio, and if the telco will insist on selling the whole package, things will get complicated.
It’s unclear how many there are, but the company said at the time of the acquisition that it had “more than 220 data centers across 23 countries.” Today, it lists more than 40 data center locations around the world but doesn’t provide much detail about the amount of facilities or their size. If the reported value of the portfolio ($2.5 billion) is close to the truth, those older non-Terremark sites can’t be big given their large amount and value of the Terremark portfolio.
We don’t know what Verizon’s actual plans are – the company isn’t sharing those publicly – but if it offloads just the Terremark crown jewels on their own, selling the rest of the portfolio will be difficult. If it does insist on selling all of them at once, whether it can find a buyer that will want all the smaller Verizon sites in addition to the Terremark assets will be a big question, Morgan said.
Telcos and Colo – a Difficult Marriage
What is clear is that data center services turned out to not work as well as many of the telcos that expanded into the market thought they would. Both telecoms and colocation are extremely capital-intensive, and it’s possible that telcos that are now looking for alternatives to owning their data center portfolios – CenturyLink, Verizon, and also reportedly AT&T – had underestimated how much they would have to invest to grow their colocation businesses.
The expectation was that data center services would create additional revenue for network services and increase customer “stickiness,” Morgan said. They did to a certain extent, but possibly not to the degree that they expected.
Telecoms is also a very different business model from colocation, hosting, or cloud. “Hosting business and cloud business is dynamic, you need a lot of very qualified staff,” she said. “It’s a hard thing to sell, especially for telco sales people. It’s a very different sales process.”
As some telcos are either rethinking their strategy as it relates to data centers or have gotten out of the data center business (Windstream sold its data centers to TierPoint last year), others are forging full-speed ahead.
Examples of the latter are Japan’s NTT Communications, which has been buying data center companies in the US, Europe, and Asia, and Canada’s Shaw Communications, which bought the US data center provider ViaWest in 2014. The two telcos don’t appear to be reluctant to invest in expanding their new subsidiaries, and it will be interesting to see how these investments play out over the next several years.
It’s too early to pronounce that the telco sector’s data center ambitions have also been “f!@#ed by the cloud,” as Bloomberg’s Ashlee Vance put it when describing the strife of IBM, HP, Dell, EMC, and Cisco. One step in the right direction would be to let the data center companies they acquired do what they do best without dragging them into the quagmire of telecom bureaucracy, Morgan said. Another would be to continue investing in their growth, since scale is important for success in the business. And yes, the Cloud. Making sure customers can connect to as many cloud providers as they want from your data centers is crucial.