Blog

Is there really such a thing as being cloud agnostic?

When companies moved away from huge monolithic systems to cloud apps, one of the key benefits was choice. But how much choice do they actually have?   And is there really such a thing as being cloud agnostic?

In my view, choice is driven by a number of factors such as price, resilience etc. The reality is most customers moving applications from data centres to cloud are already inhibited by previous choices, e.g. CRM or SFDC, in-house development vs as-a-service consumption.

The legacy decisions are the factors that really dictate future buying decisions, combined with the real cost and the predictability of costs when using cloud and either PaaS or SaaS offerings.

So, CIOs’ choices are unlimited – they are only fettered by the past which will dictate user expectations (e.g. I like SFDC because I have been using it for 5 years) and by costs. Cost of change will inhibit choices even more than user expectations – again it is the decisions of the past which will inhibit radical change rather than the available options.

CIOs don’t want to be locked in

CIOs of larger enterprises may feel like they want a choice of managed services to go along with the likes of Amazon EC2, so that when they want to switch choices they aren’t locked-in – but this can also prove costly. What is the right balance, and can you ever really be cloud agnostic?

My interpretation of “cloud agnostic” is that someone who is ‘cloud agnostic’ would be equally happy to use any cloud service or at least evaluate them all equally.

No organisation is agnostic as they will all come with preconceptions, legacy and requirements which will influence decision making and prevent them considering some cloud services. Only a start up with no bias other than the operating system of their laptop will by truly agnostic.

The reason is that putting mission critical applications into the cloud impacts so many parts of the organisation and infrastructure there is inevitably some in-built bias. Examples of how these forces will manifest themselves are;

  • EU and Data Protection rules will mean most UK organisations will not consider Cloud offerings which store data outside the EU
  • Organisations will have connectivity and access in place which will make access to some cloud services lower cost than others. For example, Interoute’s VDC offering will be a much harder sell to a new customer than to an existing consumer of Interoute connectivity and data centre services.
  • In talking to engineering colleagues looking to “spin up some compute to run a test environment” there is a bias toward AWS – simply because of previous use and therefore familiarity.