Is Cloud really that special?

Earlier today I found myself commenting on yet another cloud strategy document from one of our clients. Over the last two years, I have seen countless documents about “cloud strategy”, “cloud implementation guidelines”, “cloud selection criteria”, “community cloud governance” and more.

Although most people would agree by now that – from an IT user perspective – cloud computing is nothing else than one out of several IT service acquisition and delivery methods, many still feel obliged to call out for a set of specific documents addressing cloud, often in isolation from anything else.

I have to say that this is particularly true in government, where there is an almost unbearable level of excitement about cloud computing. However, as it often happens with hugely hyped phenomena that get unduly levels of political attention, there seems to be a fair amount of wheel reinvention.

For instance, why should anybody need to build a business case for putting something into the cloud? I can understand the need for a business case for spending money on a new service or system or on a functional or non-functional upgrade made necessary by changes in requirements and context. So the business case should express the new functional requirements, what performance, security and compliance requirements need to be met and only then examine various alternatives to source the actual solution, each with its benefits, costs and risks. The choice would be determined by the best value for money: it may be cloud as much as it may be a completely insourced solution running on a dedicated infrastructure, or everything else in between.

Think about this. Are we being asked to developed an insourcing strategy? Or a COTS strategy? Nope. We should be thinking in terms of a sourcing strategy, and then take the best sourcing decision for each and every service or system we need to deploy, based on value for money considerations.

This reminds me of what was going on with open source a few years ago. People were asked to develop open source strategies, and in some cases forced to either go for open source or to prove why they were going for different approaches. Although open source (pretty much like cloud) has lots of advantages, they can only be realized if the right conditions apply. Forcing people to use either and ask them for justification in case they don't sounds to me like a reversal of the burden of proof: you are guilty (of not using open source or cloud) until you prove your innocence..

For the sake of IT users and vendors alike, it would be much better if cloud were given its place in the sourcing continuum. I was naively hoping this would happen already a year ago or so, but the cloud-rush seems to continue without a clear understanding that:

1. Not all IT requirements are best served by cloud;
2. Data sovereignty and other compliance requirements cannot be solved in a matter of months but will need policy-makers, users and vendors to work for years on these topics;
3. The huge financial benefits promised by cloud can only be realized if a sufficient scale is achieved, which often dwarfs the size of the community (and government) clouds some jurisdictions are thinking about;
4. The single fastest-growing market for cloud is Business Process as a Service, which is not even an IT department concern; and
5. The real transformative impact of cloud does not come from just sourcing traditional applications at an allegedly lower cost, but from developing applications that couldn't even be conceived without cloud  (however most of these will require public cloud scale).

How many more cloud strategies do we need before realising all this?