Cloud has become the ideal operations model for modern organizations. However, migrating processes and applications to the cloud entails many moving parts, departments, and resources. It is now imperative for enterprises that rely on the cloud to optimize their consumption of cloud resources and services. But that, in itself, is a challenge. This challenge has ushered in a new discipline: FinOps or financial operations.
FinOps is a term born from the union of two core disciplines – Finance and DevOps, or technology teams. It is a collaborative organizational approach to managing and optimizing cloud spending decisions across an organization.
According to The FinOps Foundation:
“FinOps is an evolving cloud financial management discipline and cultural practice that enables organizations to get maximum business value by helping engineering, finance, technology and business teams to collaborate on data-driven spending decisions.”
In practice, a FinOps approach will lead an organization to collaborate across disciplines to determine standardized best practices regarding the optimal utilization of cloud resources, the optimal development of application architecture for the best technological and financial outcomes, and how best to determine budgets and negotiate preferred pricing and optimal cloud spend.
In decades past, organizational teams too often worked in silos, with business leaders or finance leaders focused on bottom line costs and top line revenue in parallel to—but without integration with—the ongoing operational management decisions made by technology teams or the development decisions made by product teams. This lack of integrated decision making leads to spiraling or unpredictable business costs, or waste in human or technological resource utilization.
At its core, FinOps is an organizational practice to align disparate teams to common goals, a…
“…way for teams to manage their cloud costs, where everyone takes ownership of their cloud usage supported by a central best-practices group. Cross-functional teams in Engineering, Finance, Product, etc. work together to enable faster product delivery, while at the same time gaining more financial control and predictability.”
By gathering key cross-functional teams together into a central best-practices group, the FinOps practice enables critical cross-department collaboration. Bringing separate but equally essential organizational disciplines together in a centralized best-practices working group fundamentally shifts the way an organization plans and executes its product development cycle and its ongoing operating model.
By working collaboratively towards common goals, with each team taking ownership of its own cloud usage and providing timely reports, successful FinOps-enabled organizations are more nimble in product delivery, more effective in budgeting and planning in the deploying their human resources, and more efficient in predicting and managing cloud spend.
Gartner projects that cloud spending will grow to almost $600B by 2023. Yet while almost two-thirds of respondents expected to realize cost savings by moving to the cloud, only 46% were achieving that outcome. (Gartner’s Public Cloud Initiatives Study 2020)
Over one-third of organizations say they have exceeded their cloud spend by as much as 40%. Organizations migrate to the cloud with the promise of lower operational expenses, enhanced efficiencies, and more substantial output.
Unfortunately, while the move from on-premises to the public cloud has unleashed innovation and agility, it has also resulted in unpredictable bills and runaway costs.
Almost every factor that makes the cloud attractive for application development can also drive cost overruns. When rolled into high-level themes, such factors include:
IT teams struggle to optimize cloud performance and monitor costs when they don’t have visibility into their cloud infrastructure and processes. Containers further aggravate the cost visibility problem because their multi-tenant, ephemeral nature is a black box for cost apportioning.
Pay-as-you-go cloud computing is attractive to developers because it offers on-demand fast, and easy resource provisioning. Ordering new and extra instances is a simple process, accomplished in a few clicks. This enables applications to handle traffic uptick and provide users superior agility when launching new IT projects and services. But, unfortunately, this seamless on-demand ordering to meet pre-set resource allocation levels can lead to unused instances when demand is not at peak.
When instances stop functioning and become impaired, usually, these are tagged as unhealthy and are automatically deleted and replaced with new ones. The problem is that unhealthy instances still receive traffic and requests until the system’s load balancer recognizes its unhealthy state. If these instances are deleted without analyzing the traffic they received during its unhealthy state, then crucial, high-value data may have gone to waste.
Rightsizing ensures that cloud computing instances (containers, VMs, or bare metal) have sufficient compute resources, such as CPU, storage, RAM, and network. Reserving resources is crucial to achieving optimal performance in all instances.
However, inaccurate sizing of instances will lead to two main consequences:
Autoscaling allows organizations to horizontally scale cloud services, such as instances, virtual machines, pods, and services depending on their traffic requirements, resource consumption forecasts, and more. However, ungoverned autoscaling based on non-optimal application configurations can lead to overprovisioning, further driving cloud spending.
The rise of cloud computing has disrupted the way modern businesses operate. Cloud computing and its sister technologies are imperative for today’s enterprises to grow, thrive, and succeed. Failing to adopt a cloud-first approach can leave an organization at a competitive disadvantage.
However, embracing the cloud also means a new emphasis on cloud cost management.
Because procuring services and resources in the cloud without control and oversight is possible, many organizations experience holes in financial accountability due to the loss of the benefits of centralized procurement.
Complex pricing by cloud service providers adds to these issues, leading to rapid financial waste due to overprovisioning and underprovisioning of cloud services, and resources, and redundant cloud computing usage.
FinOps helps enterprises achieve that balance. Moreover, as cloud computing grows and the adoption of hybrid and multi-cloud models intensifies, the role of FinOps in modern organizations will become more pronounced and crucial.
FinOps aims to help organizations enjoy the advantages of the cloud—agility, scalability, on-demand resource consumption, and decentralized control—while pushing productivity and innovation without unnecessary costs.
In an ideal enterprise, FinOps is not a cost center but rather a revenue center. This occurs when FinOps teams can accurately forecast the ROI of a given project. When FinOps can determine the difference between good and bad cloud spend, and the enterprise can calculate the financial benefits of a project relative to its cost, FinOps can pay for itself and then some.
It can be challenging to start a FinOps strategy from the ground up. FinOps is still an emerging methodology, and every strategy is iterative. There is no universal template for success.
Planning a FinOps strategy entails a significant investment of time, effort, and collaboration across different departments to create a transparent accountability process designed to maximize every dollar spent on the cloud. And every strategy begins with the team. In this section, we’ll examine what makes for a successful FinOps Team, explore the FinOps Framework and the FinOps Lifecycle.
FinOps is a team sport, where collaboration across the enterprise toward the enterprise’s goals—with a balance of empathy for one another’s goals—is essential to success.
While there are no fixed requirements for FinOps team members or its composition, TechTarget.com identifies the five major stakeholders, also referred to as personas, that typically collaborate in a FinOps environment:
The FinOps Foundation has laid out the principles of the FinOps framework through its members’ combined expertise and experience.
As the FinOps model is constantly changing, so are its principles. Initially announced in 2019, these principles have slightly changed over time as new technologies and trends emerged, like hybrid and multiple clouds.
Team Collaboration. Every department needs to communicate, and collaborate, and operate using a centralized set of best practices to achieve efficiency and continuous collaboration.
Accountability. Users should own their cloud usage and manage their consumption against their budget. Visibility into cloud spending at all levels is imperative—monitor team-level goals to empower accountability.
One FinOps Team. There should only be one FinOps team to govern and manage transactions, discounts, and other dealings with Cloud Providers.
Accessible and Timely Reports. Immediate access to feedback encourages efficient behavior, while visibility helps determine if resources are under or over-provisioned.
Business Value-Driven Decisions. All cloud-related decisions should be strategic and give more weight to achieving business goals. To achieve this, FinOps teams should perform trending and variance analysis to identify causes of increased costs. Internal team benchmarking is necessary to discover and implement best practices. Industry peer-level benchmarking helps gauge performance against the competition.
Use the Variable Cost Model of The Cloud. Rightsizing instances and services allow for more appropriate resourcing levels, similar to how pricing comparisons between services and resource types empower better decision making. Optimizing application spend is also critical. Resource waste can be significant in just one application across one cluster; this waste can expand to hundreds of thousands of dollars or more across a large data center with many applications running across hundreds of clusters.
Source: The FinOps Principles
There are three phases to the FinOps lifecycle—inform, optimize, and operate.
Inform. The first stage of the FinOps lifecycle is about empowerment through gathering and analyzing all the critical information to drive strategic and correct business decision-making. Next, the Inform phase ensures the enterprise has the accurate information and insights to fuel its allocation, benchmarking, budgeting, and forecasting.
Optimize. Equipped with the correct information and fueled by highly actionable insights, organizations and teams proceed to optimize their cloud footprint. This is done by entering into better deals with cloud providers to rightsizing instances and automating the elimination of wasteful use of resources.
Operate. Enterprises implement a basic FinOps strategy at first and then constantly measure their performance and alignment based on speed, quality, and spending. As they go along, they build a culture of FinOps. This includes creating and maintaining a Cloud Cost Center of Excellence built around business, financial, and operational stakeholders who also describe and set the most viable and appropriate governance policies and models.
Today’s FinOps solutions enable enterprises to align the rigor of financial accounting with the variable cost model of the cloud. Although such tools are an excellent place for organizations struggling to get started on the ‘inform’ phase of the life cycle, they fall short in driving usage reduction because of engineering pushback for spending their precious time on cost optimization.
FinOps 2.0 is emerging as the next generation of FinOps to overcome these challenges. There is a need for remediation that can run autonomously and continuously to optimize data infrastructure. This capability is called automated remediation by Forrester or Augmented FinOps by Gartner.
FinOps 2.0 automates traditional DevOps concepts of agility, continuous integration and deployment, and end-user feedback to financial governance, budgeting, and cost optimization through the application of artificial intelligence (AI) and machine learning (ML) practices (Gartner Hype Cycle for Emerging Technologies 2022).
Key capabilities of FinOps 2.0 include:
Intelligent observability of cloud clusters resulting in tailored, intelligent recommendations to optimize cloud spend.
Autonomous, real-time remediation of cloud spend at both the infrastructure and the application level to ensure that business objectives can be achieved without budget overruns.
A shifting of an organization’s posture from reacting to monthly cloud bills to proactive determination of true spend and identification of areas of waste.
An ability to calculate the true price/performance ratio of a cloud project, which Gartner has called “the only metric that matters.”.
Automated, real time cost optimization is essential to supporting and improving your organization’s FinOps strategy. Your enterprise’s legacy monitoring tools can’t handle the constant tuning of cloud infrastructure, applications, and processes. To effectively manage your cloud application spend and avoid cloud bill shock, your organization requires more than just traditional cloud cost monitoring.
Kubernetes, in particular, delivers operations teams unparalleled control and automation. It aids in on-prem-to-cloud migration and is rapidly becoming essential for cloud performance. By 2025, more than 75% of global organizations will be running containerized applications in production, significantly increasing from fewer than 35% in 2019.
The combination of FinOps and Kubernetes offers enormous potential benefits for application development but also poses massive new challenges regarding cloud cost visibility, allocation, analysis, planning, and reporting.
Pepperdata provides solutions that help businesses regain control of runaway cloud costs. It is the first company to offer a continuous, autonomous, and real-time solution for optimizing Big Data workloads on-premises and in public clouds.
As enterprises move more workloads to containers on Kubernetes, they must have a way to automate away some of the tedium associated with ensuring those workloads run efficiently.
FinOps 1.0 tools provide a barrage of optimization suggestions, but these recommendations are often unrealized because engineering teams find them too burdensome to implement at scale.
With a FinOps 2.0 solution like Pepperdata’s Capacity Optimizer, such concerns are no longer an issue. Instead, the Pepperdata ML engine takes care of all the hard work—identifying opportunities for savings, and implementing those changes autonomously in real time across an organization’s entire Kubernetes infrastructure.
With an Augmented FinOps solution such as Pepperdata Capacity Optimizer, FinOps teams can facilitate their FinOps lifecycle—simply enable this autonomous solution at a cluster level and achieve immediate cost savings. Nothing could be simpler to cost-effectively fuel optimization efforts and achieve business goals.
A FinOps culture is now imperative as cloud computing grows. Discover how Pepperdata can support your FinOps strategy and help your organization maximize the cloud to its fullest potential without compromising control, speed, and innovation.
To learn more about how Pepperdata can help you achieve these results immediately, please visit pepperdata.com/kubernetes