FinOps: The CFO’s magic wand to optimize Cloud spend

With cloud spending becoming a larger and more visible cost component in quarterly reports, the CFO is tasked with managing the cost of cloud utilization and ensuring that it is money wisely spent.

In fact, it has been determined that 31% organizations have an annual cloud spend topping $12 million, and 76% have an annual cloud spend above $1.2 million?*¹

Enter, FinOps, to help address the issue and enable spend optimization and Cloud spends, using necessary tools and operational practices.

What is FinOps?

FinOps stands for Financial Operations but is interchangeably used to imply Cloud Financial Operations, Cloud Financial Management, or Cloud Cost Optimizations. FinOps is described in the FinOps Framework by FinOps Foundation as

..the practice of bringing financial accountability to the variable spend model of cloud, enabling distributed teams to make business trade-offs between speed, cost, and quality.

While tools that gather, analyze, and advise on cloud consumption are an important part of FinOps, it’s really about fostering a culture and mindset where DevOps engineers and IT Operators collaborate and share greater financial responsibility. FinOps allows businesses to tag expenditures with the right department, follow them in real time, and manage expectations in the future. It’s all about assisting CFOs gain control over and optimize their company’s cloud spending.

A FinOps approach brings together cross-functional teams from IT operations, digital development, and finance to reduce the cost of designing and delivering cloud and digital services. For many companies, it’s an addition to their existing DevSecOps processes, with the finance dimension included, resulting in a DevSecFinOps methodology.

Benefits of implementing FinOps

• Optimised cloud spend: Ensure every penny is spent efficiently; wasted or unnecessary resources are optimised away.
• Improved cloud budgeting: Forecasting is simplified by understanding consumed cloud services grow by users and/or transactions. Budget can be set based on actual business targets.
• Empowered product teams: Providing teams the autonomy and responsibility to make their own financial decisions, while operating within established parameters.
• Reduce Carbon Emission: A “side-effect” of FinOps is that power consumption will go down as wasted or unnecessary compute and storage processing is removed, thus contributing to lower carbon emission.
• Improved Cost Control: By implementing guardrails, policies, and automation into self-service processes, an organisation can cost control the cloud consumption pro-actively.

FinOps Phases

The FinOps Foundation defines 3 phases of the FinOps: Inform, Optimize, and Operate.

Inform

Traditional Quality Assurance thinking has a lot in common with cloud cost optimization. The first step is to identify the source of the issue. When it comes to cloud cost management, it’s all about getting visibility into where the money is going. The organization uses common cloud best practices like account/subscription management and resource labelling in the “inform” phase to understand which departments/projects/services are consuming what resources at what expense.

The first step in increasing awareness is to make this information visible to the organization’s product teams. QA practices use a similar process to make performance metrics and functional acceptance for important services transparent to the development team. This allows them to be judged on their progress and be alerted if they are regressing from earlier releases. DevOps, on the other hand, is fully automated and incorporated into the release cycle. The goal is to make financial data as seamless as possible. Consider the following measure as an example. When compared to the previous release, Dollar$/Order has deteriorated by 30%. The cost of processing an order has increased, necessitating more investigation. An automatic defect ticket alerts the development team.

Optimize

It is possible to analyze and optimize cloud consumption and associated costs once they are visible. Typically, you begin with high-ticket items that stick out and they can be low-hanging fruit. Perhaps something has been misconfigured or misused.

In the second round, you can analyze IaaS and PaaS services for the chosen cloud provider using a variety of best-practice principles. Perhaps a disk is underutilized, with excessive IOPS for the job at hand. In this case, a more experienced FinOps SME for the chosen cloud provider can assist the product team with the investigation.

Unless you combine cost optimization with all the other quality parameters, the business service can destabilize. For example, selecting a single-region instance for a database service that lacks the disaster recovery functionality required for business continuity or switching to a spot-instance that slows down a batch-job. As a result, it is recommended that the cost optimization specialist collaborate with the product architect and QA lead to select services that meet all of the requirements. In a well-functioning FinOps model, all criteria are constantly balanced: functional, non-functional, security, and cost.

Operate

While the ‘inform’ and ‘optimize’ phases can be viewed as possible point-in-time actions, the ‘operate’ phase is the steady-state condition in which FinOps become part of day-to-day operations. Of course, the process iteratively loops back to continuously inform (and be informed) and apply cost-cutting optimizations. This is the most difficult aspect since it requires both a cultural shift and an operational model that supports it in order for it to stick in the organization and become repeatable.

A control process can be applied during the operate phase. It employs the same reasoning as that used to reduce travel expenses. As an employee arranging a work trip, you can consult your travel policy to determine the best route. You will need an exception or management clearance if you choose something that is against policy. The similar concept may be applied to cloud services, where some can be set to in-policy and others to out-of-policy, allowing you to default cloud usage to services that do not generate excessive expenses.

How to get started

Adopting a crawl, walk, and run approach where capabilities and improvements are continuously implemented is the way to start.

Rather than investing significant time on paper, developing the target operating model, bootstrap the team and get started, evolving the method-of-working along the way. Use the best practices established by the FinOps foundation and cloud providers to pick and choose what works best for your organization based on its size, environment, and objectives.

Here are three first actions to take.

Step 1 — Quick-wins: Identify low-hanging fruit

This is a tactical exercise designed to get quick wins as soon as possible. Assemble a SWAT squad of experts to examine the cloud provider’s invoice and look into any large-ticket items. Locate the owner of those items inside the organization and determine what changes can be made. It’s crucial to keep track of your savings. In many cases, these savings can be leveraged to persuade senior management to invest more and establish a strict FinOps practice. It becomes a very evident return on investment. The SWAT team approach is not a long-term, sustainable, or scalable solution. The longer-term implementation of FinOps is a shift-left approach supported by a central function of analysis tools and advisory services.

Step 2 — Select the analysis tools to create a single-source of truth

It’s critical that everyone in the company (finance, development, and operations) has access to relevant cost data and that it’s reported consistently. Visibility of data is another factor to consider. Exposing the total cloud consumption spend to the entire organization may not always be desirable. Teams can see spending data that is relevant to them.

A cloud cost management tool can simplify this task. Each cloud service provider has its unique set of tools, which typically is the starting point. Then it’s a matter of deciding whether to use additional third-party analytic tools. This blog article does not go further into the tooling selection. The cloud-native cost management tools provided by the hyperscaler vendors are getting better and better however a 3rd party tool should give an unbiased recommendation if that is a concern. There are multiple “Top 10 Cloud Cost Management Tools” out there but as always check who published the evaluation.

Cloud costs are not only about infrastructure and platform services used on hyperscalers. In fact, large cloud costs come from SaaS solutions and digital workplace services.

Step 3 — Implement a DevSec(Fin)Ops process

DevOps teams becoming self-sufficient in cloud cost management and optimization is the end goal: where development teams understand the financial consequences of their decisions, where the SRE/Ops team can act on real-time financial data and automate decisions, where QA regression testing can detect if a use-case consumes more cloud resources and credits, and where the finance team is involved in setting targets and objectives. This all points to a shift-left approach, wherein FinOps and security are integrated into the DevOps process, resulting in a DevSecFinOps process. This is the vision however the enablement and path to get can be long.

A good place to start is with a central team of expertise in the organization responsible for FinOps, which is usually housed in a cloud or DevOps Center of excellence. Even if you have central authority to empower, direct, guide, and advise, the goal is to gradually distribute responsibility to the teams, allowing them to become more self-sufficient. This can be accomplished through enablement, centrally supported tools and analytics, on-loan resources (FinOps specialists), and so on.

For an organization using a FinOps operating model, the FinOps Foundation defines common roles and responsibilities. Creating those roles and naming resources in the organization is an excellent place to start. The FinOps Foundation also offers certificates in FinOps, which can help practitioners who are both technically and financially motivated advance their careers.

The magic wand to cost optimization

The CFO is seeing the cost of cloud computing rise year after year and requires an explanation that at the very least demonstrates a knowledge and justification of that cost. A well-thought-out and implemented cloud spend optimization strategy based on the FinOps approach can help an organisation get the benefits of cloud computing at the most optimal cost. It’s a practice that needs to be integrated into the organisation, and it should start small, with central support from a competency team, and grow over time to become an integral part of the DevOps process.

If you’d like to talk more on this topic or would like to explore how we can help you with FinOps, please connect with me on LinkedIn.

About the author:

Mattias Persson
Chief Technology Officer, Nordic Cloud CoE