As the practice gains traction across organizations of all shapes and sizes, what kind of challenges do FinOps practitioners face and how are they growing the discipline?
At the end of 2020, the FinOps Foundation began sending a detailed survey to FinOps practitioners to share their learnings and collected hundreds of responses.
The takeaways were clear: that cloud financial management (i.e., FinOps) has become a mainstream practice in large organizations across all sizes of cloud spend.
Over time, we will look to slice and pivot the data by various dimensions like clouds used, size of spend, persona surveyed, and more.
This survey report is meant to be a living snapshot of the industry. We will continue to collect additional responses from our membership and update this report as new insights come in.
$30.9 billion of spend
This report sheds light on the following areas:
We also collected hundreds of open ended answers on ways to get engineers to take action and the most important KPIs used for FinOps. We’ll be sharing more details on those throughout 2021.
Now it’s time to dig into the findings.
Throughout the report, we’ll be referring to our respondents as how they rated their maturity level: Crawlers (beginner), Walkers (intermediate), and Runners (experts). In many cases, Runners had very different targets, goals and challenges. Look for examples in the report where we break out different data points based on maturity level.
Our findings indicate a trend that larger companies drive the adoption of FinOps with nearly half of our respondents representing companies with 10,000 or more employees, and 70% being over 1,000 employees. Why so many big companies tend to be heavier adopters of FinOps? Due to the complexity of their cloud environments, reporting requirements and the sheer number of disparate teams requiring collaboration than found at smaller companies.
Practitioner experience ranged from 0 to 8 years:
They see FinOps/Cloud Financial Management as…
When asked if FinOps is part of their career path, practitioners responded:
Runners tended to have 3 years experience with FinOps being their primary job 41% of the time, while walkers had 1 year experience and only 19% of them had FinOps as the primary focus of their job.
Long-time practitioners understand that the more complex the company and cloud usage—e.g., number of teams, workloads and clouds—the greater the FinOps need. Data showed a large portion of practitioners feeling the need for FinOps as spend reaches $1 to $10 million annually and up.
While a few organizations have found their stride, a majority are starting out or beginning to establish FinOps practices.
Respondents on average have about four people on their team dedicated to FinOps and see they expect this number to double! Our analysis also shows teams growing 47% from last year and expecting 75% growth next year.
In addition to cloud and platforms investing in cloud financial management, we’ve also seen major service providers (e.g., HCL, Tata, Accenture, Deloitte, KPMG, Capgemini, Wipro, etc) building FinOps advisory services and many have begun certifying their teams as FinOps Certified Practitioners.
So, what does a FinOps team look like? Before we look at where in the organization they report, let’s start with where FinOps practitioners are around the world.
Our data also shows that enterprises have different names for the practice: 37% call their practice “FinOps”, 25% “Cloud Cost Management”, 16% “Cloud Financial Management” followed by a long tail of other similar names like Cloud Economics, Cost Optimizaton, and our favorite write-in response: “It doesn’t have a name yet.”
FinOps practitioners come from many different roles and backgrounds. Our respondents illustrate how much of a variety a given team could have.
Most indicated that they report to a technical executive. The top answers were CTO (31.38%) & CIO (26.82%), followed by CFO, (12.41%), with a mix of ‘other’ responses including COO and Procurement. About 32.75% are part of dedicated FinOps teams, and 43.56% report that their FinOps team is a part of a greater Cloud Center of Excellence .
One known FinOps challenge is to not only start the practice up, but to encourage and incentivize cloud users (like devs and engineers) to participate in cloud cost management. Here are themes of strategies and tactics from our respondents. We’ll be sharing hundreds more of these in the coming year.
There’s no surprises here that a cross-functional team, like FinOps, will have a variety of key performance indicators (KPIs) to track successful outcomes. Between tracking utilization metrics, specific services, and usage of savings instruments (like AWS Reserved Instances), here is a sample of the hundreds of answers shared.
FinOps continues to expand and grow at the rate it does due to the variety of challenges that the practice can tackle to help bring order to the chaos that can be cloud cost management. We asked our respondents to rank these common challenges and here’s how they responded:
Getting engineers to take action is the clear winner here— it’s a challenge all FinOps teams seem to face. Other big challenges include allocating shared and unallocated costs, forecasting variable spend, aligning tech and finance teams, reporting on container costs, and reducing waste.
Beyond sharing a common challenge of getting engineers to take action and dealing with shared costs, respondent’s top challenges differ depending on FinOps maturity level.
Our research shows that challenges faced by those spending $5 million per year are surprisingly similar to those spending $500 million per year. This supports the notion that developing and standardized best practices earlier on build habits to help teams navigate cloud cost management as spending increases.
Across all respondents, “Getting engineers to take action on cost optimization” is noted as the top challenge. Take a look at ways FinOps teams encourage engineers and developers to practice better cloud cost management.
Getting engineers to take action on cost opportunities is the hardest challenge because they are not accustomed considering spend and have other competing priorities.
We’re proud to see that most of our respondents find the FinOps Foundation to be a great source of information (78.33%). Practitioners also catch up on FinOps best practices on Last Week in AWS, AWS’s blog , Google Cloud’s blog, and from other sources as well.
However, self-learning off the internet isn’t enough. Practitioners also let us know which kinds of FinOps certification programs they want to see to better inform and educate their teams. The immediate three types of roles to train are primary FinOps roles, engineers, and general finance roles.
So you have your FinOps team, and you understand the challenges at hand-- what kind of platforms and tooling do you equip yourself with to run the practice?
We asked respondents to let us know what’s in their FinOps tool kit:
Of those tools in their kit, respondents let us know which ones were their top, go-to solution to manage cloud costs.
AWS Cost Explorer, Cloudability (Apptio), CloudHealth (VMWare), Azure Cost Management, GCP Cost Tools, and Cloudcheckr were the top choices.
46% use cloud native tooling as their primary technology, 43% use a 3rd party platform, and 11% use home grown tools or spreadsheets.
Primary tools by FinOps maturity level
Even with a variety of tools available, many FinOps practitioners still rely on tried-and-true data collection, collation, and analysis via spreadsheet. Almost all practitioners from all demographics use a combination of tooling, while still relying on spreadsheets for some tasks-- with forecasting being the biggest Excel use.
Our respondents report a broad range of volumes of virtual server (EC2, Compute, etc.) that they run. When asked to estimate how many they would run in the next year, they reported increasing that number, with nearly 1 in 10 respondents thinking they’ll need to operate 25000+ servers!
It’s no surprise that AWS, Azure, and GCP top our list of cloud service providers used by FinOps practitioners. We asked to list a rough breakdown of all the clouds they used. Newer FinOps practitioners report using mainly AWS, Azure, and on-premise solutions (53%). Our advanced respondents indicate much more AWS and GCP usage with a drastic reduction in on-premise use (27%).
Now that we know their tooling choices, how do our respondents make the most of savings instruments (like AWS Reserved Instances and Savings Plans) versus on-demand pricing or Spot?
Keep in mind totals can be greater than 100% as ‘spot’ usage is not typically reservable under and RI or Savings Plan.
No surprises here: the more advanced the practice the more likely to target higher commitment rates to clouds. In this question we asked what experts thought they should be targeting in terms of commitment programs, on-demand and spot. Note: the totals don’t add up to 100% as these are median targets proposed.
The details get interesting if you split it by cloud maturity: All in Public Cloud companies said 80% (have fully migrated and may still have some static workloads), Born in public cloud said 70% (built to be more cloud native so inherently more elastic with no more migrations coming), Hybrid but Public Cloud First said 60-70% (depending on whether they were in a datacenter or private cloud), where as primarily on-prem said 20-50% (depending on whether they were in a datacenter or private cloud).
Allocating all costs in cloud is a perennial challenge. We asked our respondents how close they thought a mature FinOps Runners should operate in terms of 1) cost allocation and tagging and 2) budget forecasting accuracy.
Our respondents indicated that Crawlers should be able to allocate at least 80% of their spending, whereas a run stage around 90%.
In terms of forecast to actual accuracy, Crawlers said a 20% variance was ideal whereas Runners said 12%. Some individuals said as low as 2% variance was ideal. Keep in mind that these are aspirational, not actual...
We asked how accurate FinOps teams actual spend should be to estimates or forecast cloud spend. Run stage practitioners said to within 12%. In other words a $100k forecast would need to be less than $112k of actual spend to be considered accurate. Some said as low as 2%.
Nearly half of our respondents (and almost 70% of walk stagers) have little or no automation, and only 18% actually automate infrastructure changes (7% for crawlers, 29% for runners).
For those who do automate, they report sending recommendations to teams (31%) and tagging hygiene (29%), whereas automating rightsizing and savings plans/CUD/RI management both came in at 12%.
Implementing cloud cost optimization automation is no small feat. Once it gets started, FinOps practitioners tend to reap many benefits. What did our respondents report in as barriers to automation?
“Our team is unaware of automation options.”
“We don’t have the resources or time for automation.”
“Our team size holds us back.”
“Our vendor tool limits us here.”
“We don’t have the governance to support this.”
“Temporary workloads and constant size changing of EC2 and RDS instances.”
“Getting the basics right first by manually identifying gaps.”
“...with manual analysis we get to know more and stay on top of our infrastructure.”
“Rapidly changing technologies and uses that complicate automation.”
We received some interesting feedback on what tools or solutions are lacking in the industry (pay attention service providers and vendors!).
“Real-time cost analysis between the various cloud providers.”
“Foundational education for engineering teams.”
“There probably is something but we are young in the process so unknown at this time.”
“Visualization of architecture vs cost of those parts of the infrastructure.”
“Robust tag enforcement Easy ability to export cost and present it to teams.”
“...reporting total spend, cloud spend, service management spend and contract spend, the entire cloud spend picture.”
“Cloud Management tool are now adressing too many fields ( Compliance, etc.) but not only the cost part of it.”
“Container control, network / hardware integration at scale & speed of cloud.”
“Highlighting and fostering conversations around cloud cost management from a behavior change perspective.”
There’s a lot more to come from deeper analysis of this data. This first report has given us a deeper look at how FinOps teams practice adoption, how they build teams, what common challenges they face, and how they solve them. In the spirit of iterative improvement, here’s what we’re going to do with these insights.
To address these needs, the FinOps Foundation will create content and a standards roadmap to solve for each key challenge area this year.
Our analysis indicates that the top 3 needs to support our practitioners are:
All of this will be supported by monthly member discussions and deep dives into specific areas of the data and learnings from the report.
Beginning near the end of Q1, we’ll release iterations of the FinOps Framework to provide a roadmap for build a practice, KPI and benchmarking guidance, map FinOps Certified Platforms to Framework capabilities, and offer persona-based training and certification FinOps curriculums.
Together these items will support the FinOps for Enterprises certification to ensure an organization has built the right capabilities for effective cloud financial management. Enterprises will be able to self-assess or work with a FinOps Certified Service Provider to audit and remediate their practices.
The FinOps Foundation (a program of the Linux Foundation alongside Kubernetes, Cloud Native Computing Foundation, Hyperledger, and others) has a mission to advance the people who do cloud financial management through community, best practices, education and standards. It has over 3,000 practitioner members, representing more than 1,400 companies. Its standards are vendor & cloud neutral agreed by a Technical Advisory Council that includes Google, VMware, Apptio, Fidelity, Nationwide, and Atlassian.
We encourage the re-use of data, charts and text published in this report under the terms of this Creative Commons Attribution 4.0 International License.You are free to share and make commercial use of this work as long as you attribute the FinOps Foundation State of FinOps Report 2021 as stipulated in the terms of the license.