Effective FinOps requires more than understanding the roots of cloud spend

When corporations turned to cloud computing in earnest in the early 2000s, the idea was to save money. Rather than running on-premises data centers with spare capacity to meet surges in demand, businesses paid large-scale global cloud providers to access the on-demand, low-unit-cost computing their giant data centers could offer.

Yet twenty years later, while cloud has grown to constitute close to half of enterprise IT budgets , companies estimate that an astonishing one-third of that spend may be wasted. Client after client I speak with is grappling with rationalizing cloud spend.

In this blog, I’ll take you through the key structural features of cloud-based computing to better prepare you to get to grips with the spending they entail.

Elasticity as a double-edged sword

Cloud computing is part of - and has enabled - a much broader set of changes in corporate IT activity this century, which we need to consider together to understand rising costs.

First, it allowed for the development of Software as a Service products (SaaS) delivered over the internet and charged for on a subscription and/or consumption basis, with costs rising and falling according to usage. Second, it facilitated agile, or iterative, software development processes because processing power can be scaled up or down as configurations change.

SaaS and agile development contrast with a previous longstanding model, in which the business requested a product and finance agreed on a budget with engineering based on the expected incremental revenue of the product. IT spending, in other words, was once a delicate negotiation between the different stakeholders of a business and their interests. In contrast, engineers' day-to-day decisions largely determine cloud spend today.

The greater flexibility afforded by cloud has undeniably led to better products and more rapid development. But, with finance teams largely cut out of the loop, it has also led to rising costs. The elasticity offered by cloud computing, in other words, is useful but also carries significant risks. Businesses need to update decision-making processes to take the risks of elasticity into account, along with external factors engineers don’t look into, like rising energy costs and the impact of interest rate fluctuations on purchase decisions.

The misconception

In principle, there is no reason why cloud computing has to involve more computing power consumption just because engineers have more decision-making power. Agile systems can still be built with efficiency in mind. But for two reasons, they are often not.

The first is cultural. With engineers in charge, their preferences - for optimal design and outcomes - are generally prioritized over cost. There is also a focus on building new and exciting things, rather than cleaning up existing parts of a stack, meaning redundant and orphaned cloud services remain hidden within systems.

The second is visibility. Few cloud providers offer automated alerts at an individual application level unless engineering teams build this in. Look at the quarterly cost statement provided to an enterprise client by a hyperscaler, and you’ll be presented with reams of hard-to-parse data. The dispersed nature of the information makes it harder to get a consolidated and actionable view point.

In other words, it is hard for a finance team and business team to discern precisely what cloud services they are paying for. This problem is compounded as corporations embrace multi-cloud approaches, with services and data deluges from many hyperscalers.

Reimposing accountability

So how do you reimpose financial planning on a cloud model without killing the flexibility it provides?

The first element is cultural change - getting engineers to be increasingly aware of the implications of their decisions. It would be inefficient to put every engineer through a cost training course, but senior technicians, who oversee architectures, processes, and systems, need to be engaged in the FinOps process from day one and cascade principles down through their teams. In addition, embedded governance should be enabled right from the architecture level decision-making.

The second is transparency. Hyperscalers charge on per resource per second basis. Corporate data teams need to tame the endlessly complex systems that agile development techniques have created and throw light upon exactly what value they deliver in terms of revenue and at what cost. This should not be a one-off point-in-time audit of a company’s systems because they will change rapidly. Instead, processes must be developed to ensure that any use of an IT resource will immediately be visible.

The third key plank in effective FinOps is adopting unit-based economics. Once cloud usage is transparent, the cloud engineers engaged in FinOps should always have the pulse on the gross profit margin by quantifying the cost to produce and cost to serve. They should further create automations to ensure efficient cloud consumption. Artificial intelligence can also analyze fully transparent data, allowing for predictive and proactive system changes and associated impact to cost.

The ultimate aim should be gaining a picture of the unit of revenue produced for any unit of cost input so that cloud spending can be traded off against any other cost center in the company.

Doing it right or making every move count in your journey

At the EY organization, we help clients embed cost transparency at every stage of the cloud journey, whether the enterprise is new to cloud or seeking to optimize its existing cloud spend. We also joined the FinOps foundation, an organization at the forefront of evolving FinOps best practices.

EY teams are working with clients to help deliver rapid improvement assessments with dashboards that map to value, helping exit Data Centers and migrate to cloud cost-neutral, implementing controls, and helping leverage the best with multi-cloud professionals. EY teams are investing in accelerators and developing specific services that factor Risk, Sustainability, and Compliance that addresses specific industry challenges and strategizing over a long-term horizon for ROI.

Cutting back on cloud spending, or not moving to cloud at all, would be a short-sighted response that slows down innovation and hurts long-term competitiveness. Instead, we suggest EY clients think strategically and apply financial management at every step of the cloud journey, from architecture design to choosing the optimal tech stack, leveraging insights on consumption at every step to ensure we maximize return on investments.?

The views reflected in this article are the views of the author and do not necessarily reflect the views of the global EY organization or its member firms.