In my last post, I opened a can of worms by taking the stand that cloud isn’t a panacea for companies’ infrastructure challenges. I’d like to expand on this a bit and discuss how we view the physical IT infrastructure world today.
Being in this business a lot longer than most (14 years), we at ServerCentral have seen this space evolve and mature from web hosting, dedicated servers and managed hosting to colocation, managed services, and now cloud computing (apps, platforms, and physical infrastructure). Like most new tech trends, with cloud we’ve seen an early adopter phase, initial mainstream acceptance, hype (where we are now), followed by rationalization (which we’re beginning to see now).
What we find when talking to customers – listening to both their needs and sometimes fears – is that most companies are not prepared to make the full leap to cloud. For many, it’s an application here and there, like CRM, VoIP, or Call Center that can comfortably be delivered via cloud. However, when it comes down to the “Crown Jewels” – their data, core apps, and customer facing systems – they are much more hesitant to throw everything onto a public cloud. Instead, they want a roadmap and migration plan, a level of engagement most providers aren’t interested in, or prepared to, deliver.
I talked about the concerns customers have regarding cloud performance and security in my last post. Another area of serious concern is control. We find our customer’s CIO’s and sysadmins aren’t quite ready to hand over the keys and make a long term commitment without a big dose of “show me.” This involves everything from financial due diligence to technical deep dives, proof of concepts, custom designs, and a popular evolutionary concept we’ve introduced called “cloud-enabled colocation.”
Infrastructure, to us, is not a product discussion. Infrastructure is more of a collaboration along a continuum which at one end can be as “simple” as pure space and power (colocation) and, at the opposite end, as complex as hybrid and public clouds. In between sit managed services (dedicated servers, firewalls, storage, and load balancers) and full-blown IaaS (custom designed, OPEX-based solutions designed for each customer’s totally private use). Every customer is different, and it’s unfairly presumptuous to lead with a product that serves the provider’s needs, not the customer’s. Yet, this is what most cloud providers do, because it’s all they have.
Cloud-enabled colocation provides customers a roadmap to where the market is headed. We start with good old-fashioned colocation and include both physical and contractual linkages into our custom infrastructure and cloud services. What this means to our customers is that as the cloud hype phase settles down, they no longer have to view cloud as a leap off a cliff. Cloud-enabled colocation customers can now ease into it step by step as their needs, applications, and trust in the cloud evolve.
Remember the initial rise of the “triple play” following the Telecom Deregulation Act of 1996, and the in-rush of the first ISPs in the early 1990s? I believe we’re seeing the same thing with cloud providers today: lots of startups, plenty of funding, and more supply than demand. Do you remember anyone who survived from those early days fully intact? It’s a short list. The survivors were the established providers, by and large, along with a small handful of new entrants. Most providers failed and became part of a littered landscape of cash burn, failed go-to-market plans, flopped IPOs, and unrealized potential. We’re already seeing the inevitable consolidation in cloud, with the biggest providers getting bigger, and established firms like ServerCentral taking a more steady approach.
If you search “cloud” today, you’ll still come up with hundreds of providers happy to sell you their hot cloud product. My advice is to take a step back, conduct some due diligence, and talk with companies who were here before the cloud hype started, and want to have an open and honest dialog about where you’re at and where you want to go—not a product discussion.