Road to the cloud: FinOps as the key to efficiency and cost savings for companies
The path to the cloud promises efficiency gains, shorter development times and more security. In practice, however, many companies are faced with increasing costs. Against this backdrop, FinOps provides processes and tools for identifying, monitoring, controlling and optimizing cloud-related costs and expenses – and works in two ways: In the short term, it enables rapid cost reductions by eliminating “cloud waste” or long-running commitments. In the long term, it provides a comprehensive framework for building effective cloud financial management.
Companies moving their business to the cloud are facing numerous challenges, from selecting the right cloud supplier to the appropriate migration strategy to sizing their cloud capacities. To speed up migration, many companies still rely on a lift and shift approach, migrating existing applications to the cloud without refactoring them to optimize them for the cloud platform. The result: skyrocketing cloud costs. According to recent studies, 45 percent of respondents are spending more on cloud than originally planned.[1]
Decentralization of cost responsibility
One reason for the cost pressure: lack of transparency and cost control in the changed IT landscape: Whereas IT budgets tended to be static and plannable in the on-premises world, pricing structures in the cloud are dynamic and volatile in the cloud because of usage-based billing models. This is clearly illustrated by the transition from monolithic architectures to microservices: Where companies used to buy software monoliths and operate them in their own data center with maximum control over all run and change costs, today they find microservice-based architectures consisting of many individual components and operated decentrally in the cloud – whether by a team abroad or by an external service provider integrated via an SLA. This increased level of complexity requires solid governance and a decentralized control model.
“The more fragmented the IT architecture and the greater the reliance on decentralized structures and the autonomy of the product or application owner, the greater the risk that cloud costs will explode if appropriate governance has not been established.”
Dr. Manuel Audi
Changed mechanisms due to usage-based pricing models
In contrast to purchased infrastructure, cloud costs need to be considered as recurring service costs that are incurred on a usage basis. In this context, long-term mechanisms of the cloud service providers offer the possibility of purchasing resources at a lower price if you know exactly which specific resources are required in the long term and in a stable manner. The downside to this: Such short-term and granular decentralized control of costs at the product or project level cannot be mapped in classic centralized cost controlling models.
"The more individual pieces are in motion, the less a central team is able to keep track of everything."
Christian Malke
FinOps as a framework for holistic cloud financial management
This is exactly where FinOps comes in: It provides a comprehensive framework for efficient cost and resource management in the cloud and describes the governance required for this with corresponding roles, processes, and tools. In terms of the DevOps approach, this has a strong organizational transformation aspect in addition to pure cost management. In this respect, the FinOps culture refers to set of values, principles and practices that are intended to permeate the entire organization and help achieve maximum business value from cloud investments.
Essential to the FinOps approach is the decentralized nature of the governance model, which places responsibility for costs where they actually accrue – at the service or application owner level. Strong governance is intended to enable them to manage costs on their own responsibility, depending on the business value of their application. At the same time, FinOps provide a range of tools and measures to achieve transparency of ongoing costs, identify requirements and to eliminate potential cost drivers and inefficiencies, so-called "cloud waste".
Quick wins for recognizable cost optimization in the cloud
Key drivers of cost increases in the cloud are cloud services that are paid for but not used productively. Especially for companies that started into the cloud with a lift and shift approach and have not yet implemented FinOps practices, there are significant potential savings to be had by reducing cloud waste and optimizing plans. The following five fast-acting immediate actions offer the greatest leverage for cost savings in the cloud:
- Rightsizing
The average data center utilization is 8-15%. Transferring this excess capacity to the cloud results in enormous savings potential by redimensioning the cloud resources provided. Regular analysis of cloud usage and performance metrics can identify instances that are underutilized or have excess capacity. Reducing these instances to an appropriate size can result in immediate cost savings.
- Reserved instances (RIs) & saving plans:
Under reserved instances or analogous constructs, a customer agrees to rent a certain amount of cloud resources (e.g., virtual machines or databases) for a contractually defined period of time, regardless of actual usage. In return, the cloud service provider offers a substantial discount on regular on-demand pricing, which can result in significant cost savings. In addition, depending on the cloud service provider, saving plans can also be used, which reduce the basic price for on-demand use. A prerequisite for both models is the identification of stable workloads that are expected to run consistently over the relevant time period.
- Automated start/stop:
Implementing automated shutdown of non-productive resources, such as development and test environments, enables cloud resources to be sized to meet demand. By setting up policies that start instances only when they are needed and automatically shut them down when they are not, costs for unused resources can be avoided.
- Optimized data transfer:
Cloud suppliers often charge for data transfer between different regions and services. These data transfer costs can be minimized by sending compressed data and by designing the architectures so that the data is transferred within the same region or virtual private cloud whenever possible. In addition, content delivery networks (CDNs) can be used to reduce the data transfer cost of delivering content to end users.
- Using spot instances and preemtible VMs:
Spot instances (AWS) or preemtible VMs (GCP) offer cost benefits for non-critical, fault-tolerant workloads. These instances are significantly less expensive than on-demand or reserved instances, but can be terminated with short notice. They are especially suitable for tasks that can tolerate interruptions because of appropriate architecture.
“Where no cost optimization measures have been implemented before, up to 30 percent of cloud costs can often be eliminated in a short time with little effort.”
Christian Malke
Realize quick wins now and reduce cloud costs
Have you moved existing applications to the cloud and are now looking for ways to quickly and effectively reduce operating costs? Get in touch: By eliminating cloud waste and other quick wins, we ensure noticeable reductions in costs in a matter of weeks.
Have you already implemented initial cost optimizations and would like to establish sustainable cost transparency and control in your organization? msg advisors can help you establish comprehensive cloud financial management – from developing a robust governance structure to implementing suitable tools for cost and usage monitoring to establishing and empowering the necessary roles in the company.
