Most large and midsize companies decided long ago that cloud infrastructure and applications would play an important role in how they work, innovate and support customers. Gartner has estimated the scope of this commitment: worldwide, end users will spend $332 billion on cloud software in 2021, up from $270 billion in 2020.
And yet, despite the magnitude of this investment, most companies struggle to keep track of how much they’re spending on cloud technology. Worse still, most don’t measure key profitability metrics like cloud cost by product/service, cost by customer, or the return on their cloud investment relative to legacy on-premise environments.
FinOps practitioners around the world are working to tackle these challenges and today we are seeing exciting progress in the evolution of FinOps practices. Any enterprise consuming cloud services at scale can benefit from these practices.
Scoping the Problem
Prior to 2019, FinOps practitioners found themselves trying to find solutions to cloud financial management challenges in isolation, most without the ability to benefit from standardized best practices. The FinOps Foundation was formed as a non-profit, standard-setting and knowledge-sharing organization to solve this problem by helping practitioners document and share best practices.
Today the organization is part of The Linux Foundation’s non-profit technology consortium and it has become the global standard-setting body for best-in-class FinOps practices and FinOps training. Google, VMWare and SADA are Premier members of the FinOps Foundation.
FinOps is shorthand for “Cloud Financial Operations” or the practice “Cloud Financial Management,” which ensures organizations get the most value out of every dollar they spend in the public cloud. According to research from the FinOps Foundation, only 15% of companies regard their FinOps program as being “mature.” So, while exciting progress is being made by enterprises across the globe to develop, adopt and share FinOps best practices, there is much work to be done.
The Evolution of FinOps
The speed of business and the accessibility of cloud resources make cost management a significant challenge for IT organizations, particularly during early migration phases as new cloud workloads are coming online. Many organizations starting off on their cloud journey don’t have a coherent plan for managing the spend that results.
At this stage they don’t have the people or processes in place to accurately charge cloud costs back to internal consumers, to consistently apply optimization best practices, or to measure the value of their cloud services based on gross margin of products or services. Organizations operating in this stage are “driving without headlights” when it comes to their cloud spend, and often optimization practices are applied sporadically and on a reactive or even a crisis basis when costs bloom unexpectedly.
Today the establishment of a FinOps team is the best means of addressing these challenges. These are normally cross-functional teams composed of fractional shares of individuals from accounting, finance and technical or engineering teams. The best of these organizations combine these “timeshare” team members with 1-2 additional individuals dedicated 100% to FinOps.
When properly trained and equipped, these teams can often prove highly effective at applying iterated “efficiency sweeps” that apply FinOps optimization frameworks to eliminate waste. At this stage they are often able to provide good reporting to both business managers and the technical teams consuming cloud resources. Finally, they are usually able to provide clear “charge back” of cloud costs for accounting purposes.
These are good outcomes, but where do such organizations need to go from there? We believe the biggest opportunities for improvement in today’s FinOps teams lie in (1) better management accounting for business decisions and (2) more accurate measurement of the value of cloud versus legacy on-premise environments.
Management Accounting for Business Decisions: Phase I
Being able to measure and charge back costs is good, but measures of cost in isolation are not as meaningful as unit costs – i.e. costs as a percentage of other metrics such as revenue.
Once cost tracking is under control, FinOps practitioners should team up with management to find out what unit costs would be most helpful to make good business decisions. Two prime examples would be cost by product or service for gross margin calculations and cost by customer as a part of a wider customer profitability measure.
The reporting required for such calculations is often more complex than it may first appear, particularly when cloud resources are shared by more than one product, service, customer, or all three. In these cases, close collaboration with technical teams will be required to appropriately trace unit costs.
The Economic Value (or ROI) of the Cloud
Analysis of cloud costs versus on-premise costs are most often conducted for business cases justifying the initial migration of workloads. These analyses most often focus almost entirely on a Total Cost of Ownership (TCO) comparison and there are rarely “look backs” to validate the assumptions of the TCO projections once workloads are migrated.
Most cloud migrations don’t happen all at once. Rather, the migrations occur in waves. Ideally, FinOps teams will not only conduct “look backs” to confirm that projected costs of prior migrations were reasonably accurate, but they will go a step further: they will endeavor to quantify the economic value of the cloud in terms of a favorable return on investment for the migration.
This process will not only involve measurement of costs incurred, but will also factor revenue impacts. For example, did the cloud allow new products and services to come to market more quickly than the legacy rack-and-stack model of commissioning new on-premise resources? If so, what was the value of that accelerated revenue realization? Can the value of beating competitors to new opportunities (and so market share) be measured? Did the elasticity of the cloud allow rapid expansion in existing markets to meet bursting demand that otherwise would have resulted in stock-outs? Did the geo-diversity of the cloud allow entire environments to be expanded to new continents in days rather than months to catch an emerging market opportunity? If so, can the revenue impacts be quantified?
These and similar questions should be examined to quantify the value of previous migrations and better inform the business cases of in-process or future migrations. Whenever possible, such financial projections should paint a holistic picture of the migration inclusive of revenue impacts – especially when cloud resources are projected to be more expensive when measured purely on a cost basis. Higher projected costs in no way guarantee a negative return on investment for a project!
Management Accounting for Business Decisions: Phase II
As organizations’ FinOps capabilities evolve to the stage where they can conduct solid unit and customer costing, they should keep looking forward to the next phase of maturity: Activity Based Management, which is a model of managerial cost accounting for cloud costs similar to that which has existed in manufacturing for decades.
Activity Based Management can benefit managerial activities ranging from budgeting and forecasting to operational enhancement for improved profitability. For example, instead of budgeting based on prior periods’ spend (scaled up by a growth factor?), or budgeting based on a percent of revenue, budgeting can be done by forecasting volume output and the directly-related activities that trigger the consumption of cloud services. And once accurate projections of cloud consuming activities are identified, the assumptions used to generate those projections can be challenged by looking for profit-improving innovations that will reduce them.
That analysis will in turn lead to challenging a variety of technical architectural decisions that not only might impact the activities associated with output, but might impact the ability to measure and trace costs to begin with.
Eventually, the integration between FinOps practitioners, managers and technical teams may become so thorough that many FinOps activities will actually be absorbed by the technical teams themselves. In this theoretical, highest stage of FinOps evolution, the standalone FinOps teams that had sprung up and served the organization so well will likely shrink into a reduced advisory and knowledge-sharing role.
Why should today’s organizations go to the trouble of creating and supporting a FinOps team? Many organizations without a FinOps team are feeling tremendous pain while trying to contain and report on costs. For these organizations, the answer is obvious.
Other organizations may not be feeling particular pain points even though they don’t yet have a FinOps team. Ironically, the outcome for them may be even worse, because by driving without headlights they have no way of knowing how out of control some of their cloud costs may have become or what opportunities they may be missing.
The FinOps Foundation has a very helpful model to help practitioners prioritize their activities: Crawl, walk, run! One can’t do one without first doing the other. Review each of the stages of evolution above, and consider the potential business value of each – particularly those at the higher end of the evolution. Note that each depends on the most basic fundamentals of FinOps being in place, and those fundamentals rely on the formation of a FinOps organization as the first step.
About the author:
Rich Hoyer is the Director of Customer FinOps at SADA.