This sentiment may seem blasphemous to some of you, and others may be asking, “Why is Crosby biting the hand that feeds him?” To these admonitions, my response is that it isn’t, and I’m not. However, and let’s be honest with ourselves, in this age of rapid technological advancement didn’t we all just expect a little bit more? After all, the original iPod was marketed as the “iPod Classic” a little over a year after its original release, and we’re already on the fifth version of the iPhone,
As an industry, we have certainly gathered many of the trappings that one would expect to find in a business sector that has arrived. We have more companies in the market than ever before, are being attacked by special interest groups, and have been the subject of a scathing exposé in the New York Times (to put a human face on it, we could be Lindsay Lohan—and you just can’t buy publicity like that).
In retrospect, what have we really accomplished as an industry in the last five years? Let me give an example:
Have we defined what “modular” is yet? Hardly, if anything we’ve made the waters even murkier. At the present time you can have a modular offering by
• Building out a facility incrementally
• Packing equipment into a shipping crate
• Pre-building a structure at a remote location and shipping it to a site that requires a building to put it in
• Using the term frequently on your website or in your press releases
I’m not seeing a lot of uniformity here. None of these solutions reflect a customer-centric approach to the market (unless you think a company in Kansas City really wants to fly to Virginia to visit their data center or you think offerings that include an asterisk stating that “building sold separately” are just dandy).
Now we have seen strides in data center design. Our ability to build facilities with more square footage than the average shopping mall now characterizes our business. I’m all for economies of scale, but isn’t the real challenge to build them smaller at the same price points so the customer doesn’t have to take the monorail to get to his data center? If the current trend keeps up, customers are going to have to start paying freight charges to move equipment from the loading dock to their space within the building.
In terms of the applications that run inside our facilities, aren’t we really putting new names on old ideas? Sure virtualization is cool and necessary, but if Helen’s was the face that launched a thousand ships than virtualization is the technology that’s spawned a million articles. I mean when an airline overbooks seats on a flight, they call it overbooked—not “virtualized storage.” Really guys, is there an aspect of the cloud that hasn’t been written about in tones so euphoric that if the next article I read says it will cure male pattern baldness I won’t be surprised? I’m glad that we changed SaaS, ASP, and the other myriad acronyms to a real word like Cloud but that’s marketing, not innovation. And Big Data and HaDoop—nice but aren’t they really just data mining on steroids? Now don’t get me wrong, each of these is great for the data center business, but aren’t we all ready for something that actually is new and different?
Even as a collective industry it seems like we should be further along than we are. For example, aren’t we just dancing to someone else’s tune rather than leading in terms of energy use? Sure data centers require a lot of power, but that doesn’t mean that we should do more harm to the environment in our efforts to protect it. Does anyone really believe that denuding an entire forest to put up enough mirrors to let Saturn see its own reflection or establishing a wind turbine population big enough to initiate a chainsaw festival on migratory birds are the best solutions to the problem? Capitulation isn’t innovation. Let’s just tell the truth—we need more power plants to be built—and be done with it. I mean even the Sierra Club uses Facebook for crying out loud.
Seeing as the past five years have been great for business, but a little lacking on the innovation side, what should we be looking for in the next five years? Innovation no—Reuse yes. All the great ideas have been taken. As an industry, we should copy them. First, contrary to the current trend, data centers will get smaller. The need for more computing capacity to support the growing needs of both businesses and service providers will continue to drive the need for affordable data center space. This proliferation of demand and better usability through latency reduction will also necessitate that data centers will be required in areas outside of the major markets. Customers will seize on this need to insist that their data center space be closer to their base of operations, be it their headquarters or major satellite locations, or even “hub and spoke” configurations. These will be permanent facilities staffed either by customer or provider personnel.
The movement toward these “shrink-wrapped” data centers will force providers to become more efficient in their design and construction efforts. Rather than relying on the size of a facility, and the corresponding equipment requirements, to drive down costs, the new key to data center design will be the “right sizing” of materials and the standardization of procedures to lower costs through consistency and repetition. Due to the entrenchment of their existing business models, some of today’s providers will not be able to make this transition. Although they seemingly will be positioned to capitalize on this movement to customer-centric “standalone” facilities, current providers of pre-fabricated data centers will also be forced to adapt as customers will require that these new sites also be hardened for permanent usage. This will require pre-fabricated providers to develop cost-effective methods of housing their products in environmental element resistant shells at the customer’s location.
As an industry, data center providers will need to work together to effectively lobby for what is needed to continue to grow (e.g., new power plants) and to ward off unnecessary and burdensome government oversight. This requirement will be driven both by the inefficiency and immaturity of today’s sources of alternative energy and the digital economy’s unceasing need for additional data center capacity. The economics of the data center business will lead today’s providers to realize that the clear limits of efficiency measures and alternative methods for delivering power will not support the degree of growth required to meet customer demand or the profitable operation of new data center facilities. If only due to this economic necessity, we will see one or more industry consortiums develop in the next five years to effectively lobby to address these industry needs at the federal, state, and local levels.
While this five-year period has seen dramatic growth in the data center industry, it has been largely characterized by a “me too” mentality. As the industry continues to mature over the next half decade we can expect to see the natural evolutions that have historically defined all business sectors. Customer and industry demands will require today’s providers to adapt economically first, and technologically second. The next five years will see a much different data center industry than we’ve seen to date. History will be the key to success going forward. To put it more succinctly when I write this piece five years from now I expect that Karr’s quote will still apply, but not in the same way.