Cloud Cost Optimisation: Best practices

Flexera’s 2022 State of the Cloud report finds that on average, organisations have spent 13% more on public cloud annually than their planned budget..

The spending on public cloud worldwide is poised to increase to 591.8 billion USD in 2023 from 490.3 billion USD in 2022 as per the latest forecast from Gartner. Public cloud adoption is the fulcrum around which the digital transformation initiatives of enterprises revolve. The agility, flexibility, ability to rapidly scale up, and scale down capacity, and the innovation potential public cloud offers have ensured that enterprises move more of their workloads onto the public cloud. On the flip side, this rapid adoption of the public cloud across the enterprise has led to an increase in cloud bills, often more than their on-premises IT spending. Flexera’s 2022 State of the Cloud report finds that on average, organisations have spent 13% more on public cloud annually than their planned budget. It is no wonder that “optimising cloud cost” and “better financial reporting on cloud costs” trends are among the top cloud initiatives of the CXOs today.

Factors contributing to the high cloud cost:

Software development teams tend to overprovision resources on the cloud and design cloud-native solutions without keeping cost in mind. Enterprises that have adopted multiple public clouds lack the tooling to have a consolidated view of the cloud cost, usage, and resource consumption across the enterprise. Many enterprises lack centralised cloud governance, resulting in different units within the same organisation spinning up multiple instances of cloud. They find the granular nature of cloud bills complex and are unable to adopt the right pricing model and thereby leverage the discounts provided by the cloud provider.

Optimising cloud cost:

Using a mix of tools and frameworks and having the right operational structure can help organisations in their journey toward optimising cloud costs.

Public cloud providers support a “well-architected framework” or “cloud-architecture framework”, a set of recommendations and best practices for designing, implementing, and operating a secure, fast-performing, resilient, and cost-efficient cloud solution. Native cost calculators and cost-advisory tools help in estimating the cost that will be incurred while running the solution. Leveraging these tools enables “right-sizing” and “right-pricing” of the cloud resources.

Embedding cloud-cost-conscious practices across the software development lifecycle of cloud-native solutions plays a big role in optimising the operational expense incurred in running these solutions. “serverless” components, a mechanism that allows software development teams to build and execute cloud-native services without the need to manage infrastructure, can be leveraged to ensure that cost is incurred only when the cloud components are being executed.

Managing cloud costs is a continuous process. Post-deployment of the solution on the public cloud, monitoring the usage of unused or idle services on the cloud and shutting them down can play a big role in reducing cloud bills.

Mature Cloud adopters not only monitor and forecast cloud costs but also leverage automation and data science models to optimise cloud costs. Real-time analytics and data science models help detect usage and cost anomalies and trigger remedial actions.

Implementing “tagging”, a process of adding metadata to identify a cloud resource can help in distributing the costs to individual departments within an enterprise as per their cloud consumption.

Tools that can provide a single pane of glass across multiple cloud provider resources enable centralised cloud governance.

FinOps:

As per FinOps Foundation (www.finops.org), the typical cloud optimisation challenges most organisations face are getting the cloud engineers to take action on identified anomalies, forecasting the accuracy of cloud costs, alignment of technology and finance teams, dealing with shared costs of the cloud services and resources, reducing waste, shutting down idle resources and managing the costs of running containers.

A centralised “FinOps” team, consisting of people with the right mix of cloud, IT costing and financial skills are increasingly seen as an answer to foster the “cost-conscious and responsible cloud-consumption culture” across the organisation and get the “cloud operating model” optimised and standardised across the enterprise. This team works as a bridge between the Finance, IT and Business teams to understand cloud consumption patterns, manage cloud budgets, track cloud costs incurred against the budgeted cost, identify opportunities for optimising cloud costs and alert errant teams within the organisation.

Public cloud providers support multiple pricing models to cater to different types of consumers and usage patterns. Their models include “pay-as-you-go”, savings and discounts based on “upfront commitment” and “pay-less-as you use more” models. The pricing can vary based on the hours the services are provisioned for, the size and type of resources, storage consumed, data-out transactions performed, the number of API requests as well as the need for security and privacy services like encryption. FinOps team plays a big role in continuously monitoring the cloud consumption patterns across the enterprise and identifying opportunities for making use of discounts and savings plans offered by cloud providers.

To summarise, FinOps teams can help enforce fiscal discipline, improvise financial reporting on cloud costs and centralise governance while enabling enterprises to innovate and transform by leveraging public cloud, while keeping the costs aligned within the budget.

Disclaimer: The views expressed in this article are those of the author and do not necessarily reflect the views of ET Edge Insights, its management, or its members

Scroll to Top