A paper on how cloud technologies increase business agility.


Click the image to download.

It’s a good paper for a developer to share with somebody from the business-side of an organisation to explain one aspect of the benefits of the cloud – agility.

There is another paper on the economics of the cloud here. What is emerging is that although many organisations initial motivation for moving technology to the cloud is economic, their continuing motivation to maintain and embrace the move is centred around business agility.


Here’s a short extract, but the paper itself is only 12 pages.

There are clear economic reasons why a business may want a pay-as-you-go and only-pay-for-what-you-use model for this type of processing. The cloud offers this. But how does the cloud affect agility in this example.

Let’s take the insurance company example. Imagine they use a large amount of computing resource every 50 days to perform these calculations. Somebody in the organization has already worked out they can’t afford to keep that amount of computing capacity idle for 49 days out of every 50. So they devise a usage schedule that sees the computing resources used by other departments in the intervening period. Sometimes this involves the heavy-lifting of manual reconfiguration and so on. No problem though, it’s all been compiled in to an operational procedures manual and reduced to a set of daily operations that all just happen like clockwork; executed by the IT department.

2 days after the risk calculations (which determines the pricing of policies sold in the future), let’s imagine a large natural disaster such as an earthquake or hurricane, means the insurance company needs to do another risk calculation. But according to the daily procedures manual, it’s not that department’s turn to use those compute resources for another 48 days. To try to change the schedule will be disruptive to the rest of the business. To try and re-write the daily procedures manual “on the fly” will be impossible.

This company now has to either disrupt the rest of the business with unpredictable consequences (remember we said “making predictions is difficult – especially about the future”), or simply wait their turn in the queue and let 48 days pass. It’s a dilemma that so many businesses are unwilling to get in to that they live with the uneconomic option of dedicating computing resources to departments that will sit idle, doing nothing for much of the time – this is an expensive way of buying agility. Or they forget agility and hope nothing ever occurs that will cause the procedures to be run out of their normal scheduled sequence.

When such a workload is moved to the cloud, not only is it more economical, it also permits greater agility. When the natural disaster occurs, it’s simply a matter of either bringing that period’s processing forward by 48 days and from that point on, operating on a new schedule. Or simply inserting a processing slot in to the unscheduled time period and then reverting to the normal schedule later. Because making the resources in the cloud available for such an unplanned eventuality is trivial, the company that adopts such a practice will be more agile than its competitors who still insist on doing things the traditional way. A huge calculation with new data such as I’ve mentioned could very probably be kicked off within 30 minutes or less of the disaster occurring.

Let’s continue the analogy of the natural disaster – as the story progresses, things may get better, or they may get worse. The organisation that is in a position to run such calculations let’s say every 24 hours, because the resources are instantly available to them can stay on top of developments. If things get worse, they can always have available, policy prices which reflect the true risk and still allow them to make a profit. If things get better, they can sell policies more cheaply and still maintain profit margins. When they sell policies more cheaply than their competitors they take market share – the fact they can do this in the full knowledge that the policies will hit the correct profitability numbers means the sales and marketing operations need not be hamstrung by delays. The last thing a large corporate customer wants to hear is “I’ll get back to you next week with a price”.

Let’s be clear here – the organisations involved in these kinds of workloads can of course buy and maintain a big enough datacentre to have the capacity they need to always be able to run the risk calculations. But they are using valuable capital to “buy agility” and this will negatively impact profitability. With cloud solutions, these organisations will “buy only the agility they need, when they need it”: such a prospect will positively impact profitability.

Planky – GBR-257