Embracing the cloud can have a significant impact on your business’ bottom line; however, it’s important to appreciate that the models used to define Total Cost of Ownership for public cloud environments differ in many respects from traditional data centres.

An analysis of the economics of the cloud often begins by focusing on the differences between traditional on-premises data centre capital expenses and cloud-based operational expenses, where resource costs often vary as customer or processing demand fluctuates.

While this is a good starting point, there is a lot more to cloud economics than just the obvious dollar-figure comparisons of CapEx versus OpEx. The secret to truly understanding cloud economics lies in the aspects that often can’t be directly traced to a line item on a data centre invoice.

So how do cloud economics work in your favour? First, they take a holistic that extends beyond bare-metal comparisons and raw compute. They factor in demand fluctuation, realworld resource utilisation and managed services opportunities, helping your business determine the most suitable cloud migration approach.

Evaluating the full benefits of cloud economics extends to quantifying aspects such as staff productivity, considering that the elasticity of the cloud – combined with resources on demand capabilities – can result in thousands or hundreds of thousands of dollars of gained opportunity cost.

On top of this is the uplift in operational resilience that the cloud delivers compared to traditional data centre deployments. Options to enhance redundancy and increase availability come without impacting operational cost – for example, executing an application on lightweight, right-sized infrastructure while balancing the load between two or more nodes. This approach delivers immediate benefits by improving operational resilience, reducing unplanned maintenance and ensuring SLAs (service level agreements) are met.

Improved staff productivity and increased operational resilience create business agility. As such, cloud economics must also consider the overall business benefits derived from shorter time to market opportunities – often associated with adopting a DevOps and agile culture within the organisation.

When weighing up cloud economics, it is important to appreciate the difference between perceived costs and actual costs. Common pitfalls when comparing traditional data centre operations and cloud operations fall under two categories: scenarios that overestimate the expense of cloud operations and those that underestimate.

Overestimations can include lift-and-shift infrastructure sizing, as traditional data centre infrastructure normally only operates at around 30 percent utilisation, primarily due to purchase considerations for future or perceived demand. Meanwhile, cloud infrastructure allows a very close match to current demand and can be easily scaled.

The facilities cost of a data centre, such as the physical space for the infrastructure and power consumption, are rarely taken into account when performing a like-for-like cost comparison with the cloud. Maintaining bare metal infrastructure also requires dedicated and specialist resources, either as permanent or contract staff to deploy and operate on-premises infrastructure. On top of this are aspects such as legal costs, contract management, vendor management and incident support personnel, all of which play a role in the total cost of ownership (TCO) of a traditional data centre.

Aspirations of being cloud agnostic can also result in overestimations, as businesses aim to retain the ability to deploy complete solutions in any public cloud, due to fear of vendor lock-in. This approach limits the cloud services that can be exploited for next generation architecture, which can in turn reduce the cloud benefits to almost zero. In some use cases, it also increases the design, implementation and maintenance of such solutions exponentially.

A more feasible approach to a cloud agnostic model is to choose the best services from each cloud provider and then perform the necessary integrations to support the workload – in essence defining a multi-cloud ecosystem.

When it comes to underestimating the expenses of the cloud, some businesses overlook the need to adhere to a number of key architectural principles in order to take full advantage of cloud-based computing models.

Failure to properly estimate a possible re-architecture as part of a target state solution can result in an underestimation of the TCO. In contrast, where commercial off-the-shelf (COTS) applications are utilised, they may require re-platforming.

Operational costs can also be underestimated, as cloud providers mostly work on a Pay-AsYou-Go (PAYG) model for every service or application component. Where engineering teams are exposed to this PAYG model without appropriate boundaries or monitoring, nonproduction environment costs may quickly exceed those of production environments – resulting in the dreaded OpEx bill shock that many CIOs fear.

The appropriate controls, automated processes and monitoring to manage these costs are not always factored in. Nor are the costs of reskilling or augmenting an existing staff talent pool.

During a migration project, cloud economics guide the process of evaluating the current IT infrastructure cost, establishing the future cloud costs and drawing conclusions in support of a cloud migration. There are typically three phases of migration analysis, each producing a cost comparison or business case to support decisionmaking and migration planning:

  • Phase 1 consists of high-level infrastructure estimates, with a ‘do something’ and ‘do nothing’ approach over a five-year cash flow model, with consideration for growth. Comparisons are performed using virtual infrastructure reports, such as VMware or application catalogues, to produce a highlevel TCO model.
  • Phase 2 is data driven and often requires the deployment of agents within the on-premises data centre to collect vital infrastructure capacity and utilisation metrics. High-level architecture and target states are defined, while application clustering and suitability for cloud deployment are investigated. Cloud alternatives are evaluated, and infrastructure is right-sized, plus backup and disaster recovery scenarios are evaluated and costed to expand on the holistic view and prepare a detailed TCO model.
  • Phase 3 is a detailed business case that includes a high-level migration plan. It’s best if this business case includes application migration assessments, application grouping and dependency mapping. This phase establishes a migration schedule, including a transition architecture, and details the labour costs to support the migration project.

Cloud economics should not be treated as a one-off exercise in order to define a great business case and migrate applications to the cloud. New technologies and cloud services are launched at an astonishing rate, while customer engagement may fluctuate and new cost optimisation opportunities may be had through simple re-platforming exercises.

Cloud providers offer a variety of tooling to assist customers in identifying cost optimisation opportunities, with third-party vendor tooling and cloud partners offering expert advice to assist businesses in making the most of cloud economics.