A shift to the Cloud Computing model may seem an obvious decision for a sector as IT-intensive and data-centric as banking. But it's those very elements that generate the most significant adoption concerns for banks. 

As big spenders on enterprise technologies, banks have historically invested in developing capabilities on-premise; which means absolute control is a culture. And data is so sacrosanct that even within the enterprise it is buried under multiple layers of 'need-to' protocols.    

Data security, therefore, is paramount, which means that it is also an impediment to cloud adoption according to 70% of CIOs who were recently surveyed. The sector is subject to so many regulations that ensuring compliance within the as yet amorphous governance structures of the cloud is a huge challenge. Then there are also grey areas in ownership and accountability of data residing outside banks' systems, or when multiple transaction processors are involved.

CIOs, 79% of them according to the aforementioned survey, are also concerned about the implications of vendor lock-in and the situation is only bound to get stickier as banks turn to hybrid models and have to deal with multiple providers. Currently, providers' track records serve as credentials and the assurance of best effort represents commitment. But banks need to install more rigorous governance structures and watertight SLAs to demand higher accountability from them.

Data exchange is another aspect of the cloud model that warrants a bit more scrutiny than it generally gets. Beyond security, banks also need to be cognizant of the various cost scenarios arising from, for example, the accessing of data on mobile devices from outside home networks or the processing of data in overseas locations. Unless mutually productive contracts are struck with operators, the cost of data transfer could negate the savings of virtualization.

Banks also need to have a clear transition strategy for their move to the cloud, which has already mutated from the original 'Software as a Service' model into an 'Everything as a Service' model that includes infrastructure, platforms, business processes, networks etc. They need to be clear about the components that they want to migrate to the cloud. They can go modular – products, then lines of business, then universal banking operations, and so on. Or they could prioritize by tolerance to disruption, starting from least critical. Or go by customer segment, process or channel.      

The point is that they need to overcome their cloud concerns, not be stumped by them.