Chuck Goolsbee wrote this article entitled ‘Don’t buy cloud computing hype – business model will evaporate‘. Rather doom and gloom but catchy. I guess that is half the art to pay the bills. Unfortunately I think the hype of the title should be the thing evaporating.
The article examines the exercise of “What would I put in the cloud(as IT dude) if I had the opportunity to use it?” So Chuck walks through a couple applications he’d like to push(payment card/ecommerce apps) out there and realizes that security and audit requirements will prevent any meaningful app from being deployed. Case closed, clouds are hype. But what Chuck doesn’t spend time on is innovation of auditing, processes, and all the parts in between when an overwhelming value proposition is materialized.
I can’t help but refer to one of my favorite books, The Innovator’s Dilemma by Clayton Christensen. The book points out that,
“Disruptive products are simpler and cheaper; they generally promise lower margins, not greater profits. Second, disruptive technologies typically are first commercialized in emerging or insignificant markets”
The intrinsic value of an elastic compute platform is too early in its development to throw the whole concept out the window. The cloud is of huge value for several (non) ‘insignificant’ markets(ie. google, facebook, adobe). But the cloud for the enterprise? I’d say thats a market that is in search of insignificance. The use cases of apps to be deployed in the cloud for the enterprise market appears small.
But, value drives change, including those stodgy auditor processes and ITIL bigots, both influencers of the enterprise market. Today, most cloud use cases for the enterprise seem to fall into scientific domains(drug computations, weather models, energy models) or SaaS applications. In the scientific realm, most of these applications require bursts of compute power for relatively short periods of time. This is not the market that will be responsible for the broader adoption of cloud services.
Emerging technologies take time for customers to understand and adopt. In software, the applications are always the last to emerge once the HW & software tools have been absorbed. When the cloud is viewed as a new “software tool”, the innovative customers will define the use cases either through trial and error or real planned innovation. We know enterprises move slowly and in times of financial uncertainty, this only adds to the uncertainty of “when will the enterprise use the cloud?”. However, there are always first movers, and if you listen to what Tim O’Reilly is saying about cloud vendors selling the service as a commodity, trying it out might not be too painful. The first movers will find the edge conditions, the ‘insignificant use cases’ that become real competitive or cost avoidance value and mainstream adoption begins.
Thus, if a compute cloud reaches a degree of value where it becomes a standard through industry examples or pure cost reduction value, how could a security process derail this inevitability? Or even better, why wouldn’t a new feature enable this value from being realized? Its way too early to throw the babe out. There is also the concept of hybrid clouds where users can switch or be split(Data tier/application tier) between an on or off premise facility. Options are wide and varied.
My interest in cloud computing is directly related to this question, the use case for enterprises. I’m less interested to know how well Google can scale up and down or Amazon can provision Xen hypervisors for developers or Facebook can support photo sharing. I’m more interested in how companies like FedEx, ETrade, and Walgreens will leverage the cloud. The cloud-as-a-feature model might have some legs, which is what I’m currently researching.
So Chuck stirred the pot with a provacative title, here here! Unfortunately, whats in the pot kind of smells.