Everyone knows there is an ocean of infrastructure which will evaporate into the cloud over the next few years. The idea of a remote common infrastructure has been a long time coming. He quoted Larry Page and I quoted Scott McNeally on the subject.
However, at this early stage customers are struggling with the concept of a single, uniform platform. "We love your cloud" they say "but how about a little more security/support/resilience..." Requirements like these, and the inability of generic cloud solutions to meet them, are driving customers away from public cloud solutions towards private or hybrid clouds. Will the “one-size-fits-all-customers” approach that cloud providers would like to sell ever work?
I am certain common infrastructure can and will work. The so-called issues around reliability, security and legal compliance are entirely manageable. It can’t be long before custom-built infrastructure is only required for a handful of critical applications. However, there are reasons why it may take time to get there.
One barrier to achieving a common cloud infrastructure is that some of the biggest platforms have had a take-it-or-leave-it attitude to key customer concerns, for example refusing to share details of security measures or to give assurances over reliability with customers lest sharing creates vulnerabilities, or bad outages lead to liability. This might be a credible stance in the SME market but CIOs in larger organisations need more. CIOs have to know that their organisation is entrusting its precious IT and data to a trusted partner. Indeed, as well as the obvious commercial drive for CIOs to get this right or face extinction, their organisations will be looking over the CIO’s shoulder with a range of regulatory and commercial concerns. Think Sarbanes-Oxley, FSA, data protection, confidential information, leakage of customer details, trade secrets... How can any CIO be re-assured by an SLA on a web page which might change tomorrow?
This is creating anomalous situations where in fact many cloud platforms are secure and robust enough but corporates do not have the visibility to know that this is so and are presently reverting to internal virtualisation projects or hybrid clouds until their fears can be allayed. Vendors and customers must mature their engagement on cloud solutions for public cloud solutions to bloom.
So let's look forward to 2015 when all the issues around engagement have been solved?
In many ways, I do. Like others, I often liken the cloud to the electricity grid. Factories started with their own generators. But eventually they took most of their needs from the grid with only the mightiest plants having generators, and some residual complexity such as UPSs and back up generators remaining on site. Equally many foresee the vast majority of IT infrastructure in the future being run by utility providers with some specialist requirements kept on site.
However, IT infrastructure carries logical as well as engineering solutions and the former are often a matter of informed human choice. IT is harder to standardise and harder to optimise. Companies, I believe, will continue to have a significant proportion of customised infrastructure to meet particular requirements peculiar to them or their business.
Of course we can already see from sectors such as financial services that when common issues apply across the sector, it is possible to set up a platform with a common infrastructure for many sector players meeting the required standards. Why not a financial services cloud? After all, the sector did drive a lot of thinking behind XML, for example, and continues to innovate in cross-banking systems. Why not, in fact, EU data protection compliant platforms? Or government security cleared platforms? Or, or... Well, go choose!
So, while I believe common cloud infrastructure will work, I think we should be honest that there are in fact a number of clouds, with different vendors willing to build different resilience, security, compliance or scalability into their solutions depending on their market. The future may in fact be a number of clouds, fit for different markets, risks and applications; and a number of cloud solutions allowing robustness and flexibility to be built in. Within this supercloud, customers will have a choice of clouds and be able to find one or several clouds to fit their requirements.
For the outsourcing community, this should lead to an interesting shift in outlook. Might the day come when we move away from spending so much time being concerned with technology and instead focus on business outcomes, strategy and delivering value? After all, that’s what we all seek to do today - and common cloud infrastructures might just allow us more room to achieve it.
(This article was co-written with Huw Beverley-Smith, Senior Associate at Field Fisher Waterhouse specialising in outsourcing, IT procurement, intellectual property ownership and licensing, data protection, new media and e-commerce, and the commercial exploitation of personality rights.)