A recent series of cloud briefings sought to address the technology, legal and change management aspects of cloud. They prompted some thought-provoking discussions around the longevity and value of cloud computing.
With so much noise around 'cloud' and so many opinions, it can be difficult to get consensus in the true benefits of cloud. There are many cloud-sceptics, quite understandably from the overload of cloud hype out there.
During these briefings, we wanted to offer some insights into how to migrate IT services to a cloud-based model and some of the pitfalls to avoid — and to field questions on the challenges cloud might be creating.
Topics of discussion included:
- Is innovation being held back by a traditional mindset towards technology? – How can IT leaders create the right internal culture to encourage innovation? – What are the interactions between law and technology?
Attendees spanned most verticals in the corporate and government sectors around the country, and one of the most thought-provoking questions was whether cloud is a passing fad — with the inference that it was — and therefore is it worth adopting?
Mainframes and outsourcing
I spent many years in the outsourcing industry, which in many ways is the pre-cursor to cloud computing. As we know, outsourcing took off when people started slicing up mainframe capacity to effectively rent out space and time from this technology resource to multiple customers and to recoup the considerable investment made.
In the early days, mainframes were common in most industries and represented a significant investment, with far greater capacity than most organisations would consume.
This worked well for the outsourcer, through the ability to aggregate customers on a single mainframe platform. It was ideal for the customer who managed to transform their business faster and easier through access to better technology. Sound familiar?
The first threat to the outsourcing model was the rise of the midrange computer, with one application per machine, and then the x86 stack.
This eliminated the significant upfront investment in mainframe computing and allowed organisations to scale computing to their requirements but it was very inefficient, and many midrange computers were never run at more than 20 per cent of capacity.
Server sprawl and a significant rise in IT administration overhead was the result. As outsourcing grew it encompassed these platforms and required considerable upfront investment in consolidating, rationalising and standardising the infrastructure to obtain the expense reductions.
This investment was amortised over the life of the contract to minimise the upfront cost to the customer. Outsourcers typically signed ten year contracts, taking on the technology and investment risk.
This model worked well for the early years of a contract, but as time passed and the customers needs changed they requested contract changes, but had forgotten that the upfront investment was still being amortised. This resulted in considerable conflict between the customer and the outsourcer.
This changed with virtualisation of the x86 stack and the ability to slice the compute resource by time and space allocation. People realised they had computing requirements for 20 per cent instead of 100 per cent of the time and started to look at more efficient compute delivery.
It's here the death knell for outsourcing sounded and the birth of cloud computing came about. Just as the economics of mainframes sustained a profitable industry and made sense for customers for many decades, cloud computing has the same characteristics.
These are fundamental economic drivers that provide economic value for the supplier and customer. The economics are fuelled by organisations in vertical industry sectors needing different computing processing at different times of the day, month, season and year.
For example, spikes in December typically see the highest demand from the banks and retailers, but very low requirements from manufacturers.
This leads us to the "pay-as-you-go" and "pay-for-what-you-use" cloud computing models today that provide the best economies of scale. For now, at least.
Time, space and economics
So back to the question: "Is cloud a passing fad"? Well, I found this question particularly valuable because it challenged the room to look at the macro picture, instead of the day-to-day technical demand of getting up and running on cloud.
Mainframe, outsourcing, ASP, on demand, cloud — all of these are simply iterations of ways to enable your business using the latest technology and investment models.
Cloud, mobility, BYOD, collaboration plus the hot new technology yet to be invented — all of these will be superseded by something, just like mainframes and outsourcing have been.
The point is to take a macro view on embracing the dual operational and cost benefits while they're available and be ready to journey into the future when it's time. Because that time will surely come given that change is a constant.
