8 Ways to Measure Cloud ROI
You need more than capacity and utilization metrics to demonstrate cloud computing's ROI to the business. Consider these eight metrics to create a score card of your current and future business and IT needs relating to cloud computing.
Wed, May 26, 2010
CIO — An initiative from The Open Group has developed a set of key considerations for how to build and measure return on investment (ROI) for cloud computing initiatives from a business perspective. By examining the benefits cloud computing offers organizations and showing the potential return it can provide from the beginning, companies may find it easier to gain buy-in for cloud initiatives from the executive team, as well as the IT department.
Cloud computing has been described as a technological change brought about by the convergence of a number of new and existing technologies. The promise of cloud computing is identified primarily by the following key technical characteristics:
• The ability to create the illusion of infinite capacity performance is the same if scaled for one or one hundred or one thousand users with
consistent service-level characteristics.
• Abstraction of the infrastructure so applications are not locked into devices or locations.
• Pay-as-you-go usage of the IT service; you only pay for what you use, with no or minimal up-front investment costs. You typically just use the service through a connection and device.
• Service is on-demand and able to scale up and down with near instant availability. Typically, no forward planning forecast is required.
• Access to applications and information can be obtained from any access point.
But this is only half the story. These technical characteristics can also be found in many non-disruptive technology solutions. What sets the promise of cloud computing apart is that the rate of change, magnitude of cost reduction and specific technical performance impact that cloud computing can provide is not just incremental, but can give a five-to-ten times order of magnitude of improvement.
The Capacity-Utilization Curve
The famous graphic used by Amazon Web Services illustrating the capacity versus utilization curve has become an icon in cloud computing. The model illustrates the central idea around cloud-based services enabled through an on-demand business provisioning model to meet actual usage.
This model is important to businesses because one of the core precepts of cloud computing is to avoid the cost impact of over-provisioning and under-provisioning of computing resources. This benefit is in addition to the opportunity for cost, revenue, and margin advantages of business services enabled by rapid deployment of cloud services with low entry cost, as well as the potential to enter and exploit new markets.
We contend that years from now when cloud computing is seen in a historical context, the capacity versus utilization curve will be seen as an iconic model that had the same effect as previous well-known business models, such as Moore's Law, which, for example, has been seen as a major indicator of microprocessor speed in the computing industry and is now being applied to other industries, such as solar power, to define the rate of efficiency improvements.
8 Ways to Cloud Computing ROI
The problem with using the view of capacity and utilization alone is that it is a technology provider/seller viewpoint essentially based on key performance indicators (KPIs) rather than business benefit metrics. This model is primarily concerned with two specific measurements:
• IT capacity - measured by storage, CPU cycles, network bandwidth or workload memory capacity as indicators of
• IT utilization - measured by uptime availability and volume of usage as indicators of activity and usability.
But effective cost/performance ratios and levels of usage activity do not necessarily imply proportional business benefits. They are just indicators of business activity that are not in themselves more valuable than lower operating cost. What is needed instead is a set of business metrics that build on the cloud computing model.