Bill Dougherty, EVP Data Center Solutions, CBRE Group
Dallas, Singapore and Vantage: If I were in charge of my enterprise’s data centers, I would be paying attention to these three topics in 2017. While cloud adoption, hybrid IT, edge computing and IoT will dominate headlines, these three items are the barometers of industry trends and adoption that will impact your cost structure and relationships going forward.
There is universal agreement that companies, which provide the infrastructure platforms for cloud computing drove the majority of global data center absorption in 2016. Most industry experts predict that cloud infrastructure companies will continue to drive demand for data centers, and markets around the globe are quickly expanding to capture that demand. We now see new capital and business plans built around the idea of moving content a n d performance closer to the customer, referencing the internet of things (IoT) and hybrid IT as key demand drivers. The curiosity for this demand is in anticipation of future needs of the end-customer. Each cloud vendor and edge data center must invest the capital and deliver capacity in advance of the actual need. So how do we gain insight into those dynamics?
Dallas is my “canary in the coal mine”. Historically, Dallas has been one of the six national data center markets where Wall Street and our data center industry have concentrated their attention and investment. In 2016, Dallas trailed only Northern Virginia in the number of third party data centers delivered to the market (47 MW of supply) and the amount of data center space absorbed (38 MW of demand). However, while Northern Virginia and Chicago, the number three market, saw record absorption well above historical averages, Dallas saw a decline from the previous year. Another 50 MW of third party data center capacity should be delivered in Dallas in 2017.
Why concentrate on Dallas and not another market? Because Dallas is an established market with a wide array of operators, plentiful inventory, and is generally only average from a site selection standpoint. Dallas’ importance in the data center world cannot be attributed to just one or two factors. From a TCO (Total Costs of Ownership) standpoint, Texas is not the low cost leader.
Rental costs are certainly some of the lowest nationally, but the important data center incentives that drive site selection only benefit the largest single occupant facilities, and the utility cost structure can be beaten or replicated in many geographies. The companies and enterprises who lease these facilities in 2017 will provide a glimpse into the factors at work. Most data center transactions fall far below the threshold necessary for data center tax incentives. With low cost eliminated as a principal driver, we’ll be able to track if latency and performance are really ready to drive substantial demand. We’ll see if plentiful supply, ease of doing business and geography continue to be the major factors.
Like Dallas, the Singapore data center market will tell us much about what is happening on a global basis. Singapore has always been one of the safe choices when choosing a base of operations for multinational companies in Asia, along with Japan and Hong Kong. Pricing for all real estate, including data centers, is expensive because of finite resources, especially land. China’s increasing influence on Hong Kong coupled with infrastructure challenges in Japan has helped reinforce and elevate Singapore’s position. This isn’t a secret. Every cloud infrastructure provider has made huge bets on the marketplace, and the number of new operators who have entered the market and existing operators expanding is spectacular by any measure. If demand is really as deep as predicted, if China’s BAT companies’ (Baidu, Alibaba, and Tencent) platforms are really ready for global acceptance, and if adoption to cloud infrastructure is going to happen in Asia at the pace we have seen elsewhere, we are going to see its first evidence in Singapore. The plentiful supply of data center space and myriad operators means that established cloud infrastructure companies need only to turn to third party data centers when their customer demand is real and exceeds forecasts. Current pricing is showing dramatic weakness. We will soon find out if this is fueled by temporary oversupply and increased competition or fear that the demand for cloud services is more tempered than previously predicted.
Given the cost structure to develop data centers in Singapore, current pricing is not sustainable over the long term. Lastly, Vantage. Unless you are immersed in the data center industry, Vantage Data Centers may be an unknown entity. Vantage Data Centers is currently in the midst of being acquired by DigitalBridge for more than $1 billion. Why pay attention to this? Because it highlights a weakness within the data center site selection process. The operator you select today is likely to be owned by someone else in the future.
Most industry experts predict that cloud infrastructure companies will continue to drive demand for data centers, and markets around the globe are quickly expanding to capture that demand
Today the Vantage story has a happy ending. DigitalBridge appears to highly value the management and operational teams, and Vantage is a key component to their digital infrastructure master plan. However, the outcome could have been very different. While we use Vantage to highlight the issue, they are not unique. Recent examples include Ascent, C7, DataBank, Global Switch, Latisys, RagingWire, Sentinel, T5, and ViaWest. Public companies are not immune from divestitures and mergers, including CenturyLink, Tata, Telecity, Verizon, and rumored targets such as Interxion. Data center REITs (real estate investment trusts) are quick to point out that their structure fosters the stable, long-term ownership that customers covet. While more stable, there is certainly no guarantee. Both of the largest data center REITs have divested data centers or have announced intention to do so.
The reality is we cannot control who will own or operate our data center. There are no guarantees that staff will stay, that MOPs and SOPs will be followed and that required investments will be made to maintain resiliency. No amount of design excellence will make up for a neglected system. The only items that are guaranteed to survive a change of ownership are your colocation contract and its service level agreement. These contracts are highly negotiated when it comes to price, but are usually found lacking when it comes to guarantees around performance and remedies. Outage credits and termination rights create incentives but will not ensure uptime or performance. Next time you have a service order to add or renew services, take a look at your master service agreement, lease and service level agreement and ask yourself if it has the necessary protection if someone new owns and operates your data center.