Sign up to our Newsletter
The cloud used to be viewed as a place of significant cost savings: rather than purchasing and maintaining dozens of server stacks, organizations could outsource this and purchase compute power on an as-needed basis. In the ensuing rush to cloud architecture, however, many companies simply lifted-and-shifted their old financial bad habits.
The sheer speed of cloud provisioning has moved engineers into the financial driving seat – and CFOs clinging on as passengers. As a result, cloud spend has accelerated to all-time highs. Financial DevOps (FinOps) aims to regain control by eradicating the heavy blockers between engineering and financial teams. To learn more about FinOps’ aims, see our in-depth guide.
While the practice has begun to build considerable momentum – business leaders spurred on by the multi-million potential savings – actually achieving the tight team collaboration and real-time visibility into cloud resources and spend is proving a considerable challenge. And finally, there’s the unyielding financial reality of FinOps projects: employee time is itself a cost. If teams are bogged down by endless FinOps meetings, your project’s net savings start to dwindle significantly. Knowing precisely what, when, and where savings can originate from demands laser-focused precision.
Thankfully, you don’t have to go it alone: this article will cover the key FinOps best practices that can actualize immediate change.
What are the FinOps principles?
FinOps implementation involves huge swathes of change across vast and disparate teams. To combat the analysis paralysis that plagues many early FinOps projects, a foundation of FinOps principles can illuminate the precise goals driving this change. for a deeper dive, check out our guide on how to enable key FinOps capabilities.
Teams Must Collaborate
Maturation is impossible without teamwork: nowhere is this more true than within FinOps. When engineers wind up new EC2 resources, their priority may be the long-term reliability of these instances. AWS is more than happy to supply this – for the higher fees of their On-Demand EC2 products. Cloud spend and engineering teams are already tightly linked – this interconnectivity now needs to be brought to the forefront of these teams.
In the same manner, driving costs down can only be achieved with collaboration. If a manager simply looks at their highest-cost instances and demands them to be cut, then an organization can expect an ensuing period of mayhem. Instead, successful FinOps combines the engineers’ day-to-day understanding of your cloud architecture with the ROI-focused philosophy of accountants.
Decisions are Driven by Measurable Value
Every decision needs to orbit around the cost data of your cloud – and have an on-the-ground understanding of what budget changes each decision will bring about.
To ensure that decisions are driven by measurable value, some organizations take an approach similar to a start-up. By re-contextualizing the cloud as a driver of innovation, rather than a constant background spend, it’s easier to conceptualize why unit economics are so important.
Unit economics is about breaking down revenues and costs into a single unit – this could be a transaction, a customer, or any other meaningful measure of business activity. Key metrics such as contribution margin, customer lifetime value (LTV), and customer acquisition cost (CAC) are key metrics in evaluating the profitability and efficiency of cloud spending. Every successful FinOps project requires enough visibility into the cloud to pull these real-time metrics on-demand.
Collective Ownership
Given that the provisioning of cloud resources now happens at the edge of your organization, a sense of ownership needs to follow suit. Keeping everyone on the same page needs to include more than once-quarterly, top-down meetings: your engineers need to fully embrace their own impact on the monthly invoice.
The team will work together to develop a unified vision and strategy for FinOps adoption. Effective communication is crucial – by consistently sharing updates on the objectives and advantages of FinOps, your organization will build trust and maintain engagement among its members. Establishing well-defined metrics and key performance indicators (KPIs) is essential for monitoring the advancement toward FinOps goals.
Moreover, leadership plays a vital role in nurturing a FinOps culture characterized by collective responsibility, accountability, and ongoing enhancement of cloud cost management. This culture acts as a driving force, guiding organizations towards seamless advancement in their FinOps journey. Thanks to this, as your organization becomes increasingly FinOps fluent, the decisions around cost-effective architecture will look increasingly decentralized.
Data Must Be as Real-Time as Possible
Cloud resources are intensely transient. This is why IT’s traditional capital-expenditure model breaks under any weight of cloud resources: while you used to be able to map out episodic increases in data center capacity, cloud resources can scale up and down in a matter of minutes.
In the cloud, it’s no longer possible to measure cost efficiency by comparing average cost against resource utilization levels. In fact, even measuring precisely what expense originated from where can be a challenge. Detailed cloud usage data spills across millions of lines of data – dividing costs between individual projects can be nearly impossible.
This is where two things need to be put into place: firstly, at the backend, your engineers need to be tagging as many resources as possible. Adding crucial info such as project, team, and person responsible can add considerable visibility right off the bat. Second to this is an automated tool that feeds this real-time cost data to each corresponding team.
Centralized FinOps Team
The cross-functional nature of FinOps is one area that many organizations get bogged down in. By consolidating a few experts into a central team, it’s possible to achieve vastly faster and superior cloud-cost decision-making. This then allows for the centralized FinOps team to track how its partner groups in accounting and development are getting on, achieving FinOps fairness while granting upstream decision makers rapid and actionable data.
Take Advantage of Cloud Cost Variability
Finally, once the structures are in place to measure your FinOps progress and let your teams act on real-time cloud data, the process of cutting costs can begin. While cloud variability was once the bane of cost forecasting attempts, FinOps seizes every chance to cut back cloud cost. Whether it’s through your current cloud provider’s reserved capacity – or by switching up the architecture of your services – each project is prioritized via the potential cost savings and your new cross-team capabilities.
The 6 principles of FinOps can offer incredible insight into what fully-matured FinOps programs look like. Getting there, on the other hand, is a constant effort. This process can be considerably expedited by sticking to the following best practices.
8 FinOps Best Practices
The best practices driving successful FinOps projects are numerous: to help break them down into practical bite-sized pieces, consider the three stages of the FinOps lifecycle:
- Stage 1 | Inform: FinOps requires a baseline of cloud understanding. This is why the inform phase is first: it builds and reinforces your team’s visibility and experience in allocating cloud costs. The inform stage also focuses on developing a detailed map of your cloud resources, budgets, and corresponding business value.
- Stage 2 | Optimize: The process continues with the optimization phase: having uncovered every node of your cloudscape, it’s time to start actualizing cost savings. This can look like purchasing savings plans or turning off unused dev resources – knowing what to prioritize and what to cut builds off the visibility from stage 1.
- Stage 3 | Operate: The final phase of FInOps incorporates tooling to ensure continued savings, focusing on seamless improvements and qualifying goals.
Best Practices to Inform
The inform phase is critical: the key goal of this stage is to gain accurate visibility into your true cloud usage and expenses. This allows your FinOps teams to build a picture of each project’s ROI. From a cultural perspective, this first phase allows you to not just pull a high-performance team together – but grant them the ability to collaborate on your cloud’s financial benchmarks.
#1. Adopt a Single Source of Cost Truth
With cost data being so important to FinOps’ success, it’s critical to nail its implementation into your organization. Unfortunately, the sheer scale and variability of cloud products has led to a lack of standardization between cost data – even major platforms don’t support billing standardization. This risks your first step toward FinOps being riddled with avoidable stumbling blocks.
A single, universal cost data tool is ideal for this analysis, as it provides a foundation of universal data for every team member. While it’s technically possible to rely on a cloud provider’s native tooling, multi-cloud cost tools have evolved far beyond AWS’ own dashboard. For multi-cloud setups, it becomes even more vital to establish a tool that can correlate costs across every instance and project.
This approach is so pivotal thanks to the fact that all teams need access to the same underlying cost data. Its consistency eliminates any discrepancies that could lead to misinterpretation, while enabling more accurate budgeting, forecasting, and cost allocation.
With a unified view of cloud costs, decision-makers can actually make informed choices about resource allocation, scaling, and optimization. It allows for identifying underutilized resources, making decisions on reserved instances, or considering spot instances to reduce costs without compromising performance.
Furthermore, with everyone on the same page, it can further foster a culture of cost accountability. It encourages teams to start understanding their wider impact on their organization’s overall budget, promoting a culture of cost-effectiveness. In this way, a single tool lays the groundwork for a more mature and optimized cloud strategy, while also scaling with the organization as its cloud usage grows and becomes more complex.
#2. Make the Most of Tagging
Connecting your cloud resources with their invoice counterparts is one of the biggest hurdles to FinOps’ success. Thankfully, almost every cloud provider has a built-in solution to this: tags. From the cloud provider’s POV, tags are strings of characters that are essentially devoid of meaning. For your organization, however, they’re a boon of granular insight. Your tagging strategy needs to be built around your own organization and workflows, but some critical best practices apply across the board.
Firstly – audit what’s already there. Defining how mature your tagging currently is can help prevent time wastage, as well as give you a better idea of what’s actually being tagged.
From there, define your tagging project’s key goals – is it to define what business unit is responsible for each cost? To identify unused resources? Or to find out how much a product you’re responsible for costs? By validating your tagging approach through the lens of several different FinOps stakeholders, it becomes possible to build multi-faceted capability into the very core of your cloud visibility.
When the project’s pedal hits the metal, keep engineers supported via easy-to-understand guidelines. Start with up to five clear areas – such as business unit, product, and owner. Overwhelming engineers with a dozen-long list can upset the balance before the project has even started seeing a return. At the same time, development policies can enforce that new resources are being wound up with this extra info.
#3. Ensure Relevant Cost Visibility
To truly make a move toward collective responsibility, a single source of cost truth is only the beginning: to really support the informing phase of FinOps, every relevant team needs to be supported with data that reflects their own contributions. Customizable and dynamic dashboards play a key role within this, as engineers are empowered to hone in on the cost data relevant to them. This way, divvying up your larger cloud bill among accounts and teams, providers, features, teams, and customers is key to understanding the monthly cloud bill.
Best Practices to Optimize
In the “Optimize” stage, organizations and teams, now empowered with clear visibility and understanding of their cloud spend, move to actively optimize their cloud footprint. This stage involves the utilization of various levers provided by cloud providers to maximize efficiency and cost-effectiveness.
Notably, this stage includes transitioning from the more expensive on-demand capacity to more cost-effective options, such as Reserved Instances or Committed Use Discounts. It also encompasses rightsizing and automating the elimination of wasteful resource usage. Whether on AWS or Azure, these can be actioned rapidly – giving your newly-formed FinOps collective a jumpstart.
#4. Rightsize Your Instances
It’s incredibly easy to purchase compute power. However, sometimes the CPU and memory usage you’re paying for just isn’t being used that much – rightsizing aims to fix that by bringing instance types and sizes in line with the actual requirements of a workload.
The measurement phase will have helped identify precisely what workloads are your biggest drivers of cost. This tool would ideally have shed further light on some key metrics such as memory, disk, and network usage across your volumes and instances. Furthermore, by gaining insight into corresponding performance and usage patterns, it’s possible to quickly and easily scale resource sizes to actual usage. This process does have some variability among your chosen cloud provider – if you’re primarily on AWS, make sure to read more about AWS FinOps Best Practices.
While AWS offers a tool to rightsize EC2 instances on their cost dashboard, other resource providers aren’t quite so forthcoming. Kubernetes has traditionally been difficult to rightsize, but an increasing number of tools are offering automated rightsizing even for deployed applications. Azure also offers a number of ways to rightsize and save – read more about Azure FinOps Best Practices here.
#5. Save Costs by Thinking Ahead
Taking a few relatively simple cost-saving measures can be pivotal for seeing quick returns on cloud cost. A key practice is simply winding down resources that are accessed on a set schedule: for applications that see usage mostly during the day, schedule resources to be dropped at night. This can be supported and automated with tagging support, wherein engineers can identify such instances at deployment. This targeted approach ensures that resources are only active when needed, preventing unnecessary expenditure and contributing to a more cost-effective management of cloud environments.
Similar to winding down resources when they’re not used – finding and discontinuing unused legacy resources can lead to substantial savings. Consider a storage bucket costing $100 monthly that remains untouched for four years – this equates to an annual expense of $1,200, accumulating nearly $5,000 over that period. Though this amount may seem minor in the context of a multi-million dollar enterprise, the cumulative cost of several such underutilized resources can represent a noteworthy financial burden.
#6. Leverage Reserved Instances
With fluctuating deployments sorted, any workloads that require a constant amount of compute can now be addressed. Utilizing Reserved Instances (RIs) involves committing to cloud services for a set duration, usually spanning years, to benefit from considerable discounts over on-demand rates. This approach is ideal for deployments expected to maintain stable performance demands over time, facilitating significant reductions in costs. Additionally, exploring savings plans and potential private pricing agreements, based on the specific services utilized, can further optimize cloud expenditure. These strategies allow for tailored cost management, ensuring that long-term, predictable workloads are run in the most financially efficient manner possible.
Best Practices to Operate
In the final phase of the FinOps lifecycle, FinOps teams must engage in continuous monitoring and assessment of cloud operations to ensure they align with the organization’s business objectives and established metrics. This involves closely observing how cloud utilization and performance meet the evolving needs of the business – and communicating these insights to stakeholders.
#7. Optimize
Optimization represents a critical opportunity for cost reduction at this stage. Many organizations initially migrate to the cloud through a “lift and shift” approach, but achieving deeper cost savings requires a more sophisticated strategy. Once a central governing body has been established, re-architecting applications often becomes a pivotal step. By redesigning applications to leverage native cloud services, organizations can unlock further savings.
#8. Automate
Remember how the cost of employee wages further plays into FinOps savings? FinOps maturity demands that automation go beyond simple, short-term reactivity. The goal is to proactively manage cloud resources through automated optimization. This proactive approach, supported by appropriate tooling, not only enhances efficiency but also ensures that cloud operations are continually aligned with business goals and financial strategies.
Jumpstart your FinOps Straight to the Edge
GlobalDots’ approach is meticulously designed to handle your cloudscape’s dynamic complexities. Decades of industry experience has built a framework of transparency and efficiency – so your project can hit the ground running no matter where your organization is in the FinOps lifecycle.
It’s not just our methodology that champions this: as industry leaders, our view into cutting-edge FinOps tooling helps mold the ever-expanding options into the unique FinOps suite required by your organization. This grants you the headspace to truly maximize the cultural and educational opportunities for your team. See how we can start actioning your cloud cost optimization today.