Anyway, I have really enjoyed the role because it has allowed me to become reengaged in that part of my brain that is reserved for logical and critical thinking in areas not just reserved for eggheaded technology stuff. Specifically, I am now helping a myriad of tech-oriented companies figure out what their role is in this whacky, brave new world in IT we call, "Cloud Computing".
As you know from my previous posts as well as tech press coverage that puts Presidential election coverage to shame, there is a shift occurring out there where many businesses want to stop owning their own IT infrastructure and throw the mess over the fence to a company that will take it all on for them. Think about it - why have the hassle of maintaining servers, storage, networking, software licensing, and people to maintain an in house e-mail system if you could just toss it all Google or Microsoft's way for $5.00 per mailbox per month? It's a pretty compelling proposition; especially if you are lucky enough to offload other in house applications in a manageable, secure way...
For many technology-oriented companies I work with, there is a natural allure to want to take on the Cloud SP role. From big systems integrators, to consulting firms, to vendors who make technologies that work in a specific part of a Cloud infrastructure - there is a greedy, over zealous desire among many of them to build out datacenters, offer up access portals, and let the end users flock in to roost. I was recently in a meeting where a program manager declared with some kind of faux vision, "We have smart tech people and make a lot of our own pieces to build a cloud. How hard could building one out be?" Notice they failed to add, "...and run it successfully," to the mix of what they said.
Running a Cloud SP business, IMHO, is the LAST and most difficult thing ANYBODY would want to do in today's IT world, because it involves more things than just marrying a bunch of smart people with a bunch of smart technologies. Operating a service provider is a lot like operating a utility company - you exist to oversubscribe your infrastructure with the hopes your customer base doesn't all show up at once to use everything; all while scraping in profitability over a longer period of time that most companies aren't patient enough to stomach. Excess capacity on the datacenter floor is lost money, and your usage reporting needs to be tight enough to allow billing for services in dimensions that address resource consumption in a more granular fashion.
Did you notice my description didn't use the phrases, "Smart Tech People" or "Really Awesome Technology"? I sure did, since I wrote it and all :-).
Seriously - operating a Cloud Service Provider involves creating and adopting a infrastructure as well as a series of workflows and processes that most organizations aren't equipped to handle or even comprehend...no matter HOW good their people or technologies are. This is not being rude; this is just reality.
Remember my last post about how Apple creating their own mapping service wasn't the smartest idea? We're pretty much in the same arena with this SP topic...
Don't believe me? Take a trip with me in the Wayback Machine to the mid 1990s, when Congress passed the Freedom of Telecommunications Act of 1996. To summarize - the act was a long overdue bill overhaul that lifted restrictions on allowing more companies whose names didn't contain letters like "AT&T" and "MCI" to get into the telecommunications business. And, because of it, companies known as CLECS - Competitive Local Exchange Carriers cropped up with the dream of making big big bucks by displacing the fat cat providers like Verizon and others.
The challenges to make a CLEC successful were thorny - they were still at the mercy of Verizon for the "last mile", ended up collocating their equipment in existing carriers' network centers because of the exorbitant costs to build out their own facility, and faced trying to provide a level of service and pricing that organizations were used to getting through the big providers. As you may have guessed, as the dust settled, being in the CLEC business ended up not being a land of rainbows and puppy dogs. As a matter of fact, HUNDREDS upon HUNDREDS of CLECs that were started in the gold rush days of 1996 have been reduced to about 10.
See the parallels here? Starting a Cloud Service Provider is as big of a gamble as a CLEC, and the chances of success are 1000:1 at best - mainly because the dust has already settled over a group of companies that have proven to show they know what they're doing over a much longer timeframe. Amazon AWS, Google, Rackspace, Terremark - these are juggernaut organizations that were designed with SERVICE DELIVERY in mind, and have the infrastructure to back it all. Granted, there are other smaller companies that have some good momentum in keeping pace with the mainstream providers, but they more than likely service a specific niche, and will be acquisition targets once they get big enough.
To those tech companies who want to gamble on being Cloud SPs...respect the history of those who marched before you in the good ol' days before you dive into the pool. Otherwise, the Deja Vu all over again may kill you...