Cloud Costs Are on the Rise, Yet Containing Costs Remains a Challenge
Cloud computing started making its way into the enterprise in the late aughts, following the launch of AWS in 2002, Amazon's Elastic Compute Cloud in 2006, Google's App Engine in 2008 and Microsoft Azure in 2010. At the time, it was widely accepted that cloud computing was cheaper than traditional computing.
But while cloud computing undoubtedly makes it easier to deploy and use a wide range of applications and tools, as well as facilitates distributed working, its costs have steadily increased as workloads and cloud deployments expanded. According to Statista, enterprise spending on cloud computing increased from $61.7 billion in 2009 to an estimated $276.05 billion in 2021. Whereas in the early years, the majority of the spend was on data center hardware and software, in recent years the balance has tipped towards increased spending on cloud infrastructure services.
'Careless Cloud Management' Today
Cloud computing is complex, and it's only getting more so as vendors add new technologies and tools to their offerings. However, these additions don't necessarily mean deployments are easier to manage or cheaper to install and run.
To get a true understanding of costs, leaders must first understand what they are investing in and what it might possibly bring to their business. The best way to do this is to break it down by the storage, network and compute resources associated with production and the cost of goods sold vs. the resources used for testing, development and experiments, said Hyoun Park, CEO and principal analyst at San Francisco-based Amalgam Insights.
Typically there will be a linear relationship between some portion of the cloud spend and revenues and profit, he continued.
"At the same time, cloud computing makes it almost trivially easy to grab terabytes of data, duplicate those sources, and run compute-intensive workloads on that data without getting any business gains at all," he said. "Careless cloud management has led to the need for Cloud FinOps or cost management efforts that are now vital for any data or compute-intensive business."
He added that cloud spend is definitely slowing down. However, he advised taking that slowdown with a grain of salt. Take the example of a cloud vendor whose year over year growth is 20% and appears to have stagnated. That 20% revenue growth, he said, is still faster than the vast majority of businesses are growing.
Related Article: How Hybrid Cloud Is Enabling the Digital Workplace
Reconsidering Cloud Investments
Asked about reports that some companies are reconsidering their cloud investment, Park said the reality is that the majority of companies are now hybrid cloud, meaning they are rationalizing their use of on-premises servers, private cloud and public cloud to fit their needs.
“Steady and predictable workloads will always be cheapest on bare metal servers, assuming that companies have personnel on staff to manage those servers,” he said. “Every company needs to figure out how cyclical, variable and large their cloud resources are and figure out where those workloads really should be.”
Amalgam Insights estimates that companies that have not already made a concerted effort to cut cloud costs can find an average of 30% savings through a combination of resource optimization, provider optimization and hybrid cloud architecting.
Even still, Park admitted cloud costs are hard to get under control because of the number of services that fall under the cloud umbrella, meaning the usage can be both granular and very large. He noted that the major public cloud companies each have over 200 different services ranging from databases to machine learning to standard computer and storage options. These products typically do not align perfectly with similar products provided by other cloud providers.
The nature of cloud computing means there can be sudden surges of billing activity that can take place before companies can adjust their billing plans or resource pools. To get costs under control requires a continuous lifecycle approach to cost management, said Park. It starts with getting visibility to all relevant resources and usage. With this visibility, businesses can identify which workloads are associated with key cost or profit centers or with mission critical processes.
"By optimizing across billing, product, technical, and contractual environments, companies can create a virtuous flywheel of continuous optimization across all of the tools that cloud FinOps has to reduce and optimize costs while rightsizing based on the needs of the company," he said. "The goal is never to cut cloud computing costs to the bone, but to create an environment that maximizes business success and customer experience."
Related Article: Why Edge Computing and 5G Development Should Happen Together
The Complex Math of Calculating Cloud Costs
Anay Nawathe, principal consultant with global technology research and advisory firm Stamford, Conn.-based ISG, also stressed the complex relationship between cloud costs and business benefits. The question is more how a business uses the cloud to meet its goals than simply the cost.
Cloud computing can offer significant benefits such as scalability, agility, flexibility and enable the move from CapEx to OpEx, which can help businesses respond more quickly to changing market conditions and customer needs, leading to monetary gains.
However, he said organizations must intentionally use cloud-native services and architectures to realize these scalability, agility and flexibility benefits. Effectively using these cloud-native capabilities does not inherently mean increased costs — in fact, it frequently reduces the cost of operating in the cloud.
He added that most organizations now understand that without some level of application rationalization, a large cloud migration is rarely a cost-savings opportunity, especially when rehosting (i.e. lifting-and-shifting) most of their applications.
How McDonald’s Drove Productivity Through an Elevated Employee Experience
In the new remote/hybrid workplace, work/life boundaries are blurred and workplace stress is a top driver of mental health needs.
How to Future-Proof Your Employee Experience Strategy in 2023
A framework to navigate through economic uncertainty
Challenges to Efficiency in 2023: Your Employees Need the Digital Workplace of the Future
The era of asking employees to do more with less is upon us
The Essential Role of Communicators in Fostering Wellbeing in the Digital Workplace
Join us for practical insights on how digital communicators can support employees to thrive in the digital workplace
Addressing Employee Needs and Wants with a Digital Workplace
The workplace is getting more and more digital – both in how we work and where we work
Maintaining a Human-Centered Approach During Digital Transformation
When it comes to digital transformation - people drive change, not technology
Some organizations with under-utilized data centers and fast-approaching renewal dates have found limited short-term cost savings from rehosting migrations, but even then, these cost savings will quickly diminish without a quick shift to cloud-native services. And yet not every workload is appropriate for public cloud hosting. For example, workloads that prioritize performance or data governance over agility and scalability can be better suited for a private cloud or on-premises deployments, said Nawathe.
“Rather than completely moving out of public cloud, most organizations will employ a hybrid cloud strategy where workloads will be hosted where they best belong,” said Nawathe. "While some organizations that have begun to repatriate select public cloud workloads back into their own data centers or colocation facilities, most organizations can find significant cost optimization from ramping up their cloud cost management (i.e., FinOps) capabilities."
Organizations must also understand that legacy IT volumes should not be the basis of committing to cloud capacity volumes without optimization first, he continued. "We have seen many of our clients commit to and pay for capacity far in excess of their actual usage, often for a few percentage points of improved volume discount. This can play out to a hyperscale charge of twice what the business would otherwise pay if the business rightsized its contractual commitment."
Related Article: A Cost-Cutting Cloud Optimization Strategy for Your Skyrocketing Cloud Bills
Build the Right Cloud Model
The ROI on cloud use is clear as long as companies focus on the right governance model, set of policies and technology integrations, said Sandeep Chellingi, head of cloud innovation services at Edison, NJ.-based Orion Innovation.
He said cloud-first solutions can reduce data center operation costs and achieve a total cost of ownership of between three-to-five years. To keep costs down, enterprises moving to the cloud must build a sustainable business model by applying the right policies and governance guardrails to implement best cloud cost optimization practices.
In this respect, he advised companies to calculate the total cost of running in data centers compared to the cloud, and factor in the latest cost-optimization practices, which can save companies millions of dollars. "Controlling cloud costs is a growing challenge. Enterprises need to adopt FinOps and governance best practices early on in the cloud journey to keep spending under control. We are seeing common challenges like orphaned resources, cloud auto-scaling, idle cloud resources, backup/snapshot management, improper provisioning, and improper cloud spend visibility and accountability, all contributing to increasing costs.”
Lack of Visibility Adds to the Challenges (and Gives Rise to FinOps)
For Clinton Ford, DataOps champion at Palo Alto, Calif.-based Unravel Data, the two biggest contributing factors to rising costs in cloud computing are the modern enterprise’s need to conduct real-time analytics on increasingly large and complex data sets and the lack of visibility into what it costs to run these workloads.
He points out that public cloud services like AWS and GCP have made it much easier to spin resources up and down as needed, but organizations frequently over-provision cloud resources. Industry analysts estimate that at least 30% of cloud spend is “wasted” each year – some $17.6 billion. Data management is the fastest-growing segment of cloud spending, he added, and for modern data pipelines in the cloud, the percentage of waste is significantly higher, estimated at closer to 50%.
Ford shared some of the key findings from a recent survey the company conducted of data team leaders on their most pressing challenges and priorities for the coming year. Two-thirds of surveyed data teams said cloud spending is now a KPI of high strategic importance. Almost half (44%) of respondents also stated that they believe they are leaving money on the table when it comes to their public cloud utilization. Alarmingly, almost a quarter of respondents (23%) said they were unable to even estimate what percentage of their cloud resources went unused.
Unfortunately getting at this information is anything but easy. Modern data pipelines and data apps are enormously complex, interdependent, and the sheer size and scale of data workloads only amplifies the challenge of identifying cost savings opportunities.
In order for data engineers to take action on cost optimization, they need to be empowered with the granular-level usage details that help them to make informed and defensible choices — and do so without sacrificing what they really care about — reliability and performance. It’s why interest in the practice of FinOps has grown so dramatically over the past year.
The unifying principle of FinOps is that by bringing finance, engineering and business teams together to make better decisions around cost and performance, they will then act in a more responsible manner, provided of course they have access to the right data to inform their decision making process.Citing data from the 2023 State of FinOps report by the FinOps Foundation, the biggest challenge facing organizations trying to establish a FinOps culture is “getting engineers to take action” on cost optimization. Yet the authors also note that with so many data projects on their backlog and nearly unlimited cloud resources at their disposal, it is understandable that data engineers naturally prioritize new data pipeline creation and timely data delivery over cost optimization.
About the Author
David is a European-based journalist of 35 years who has spent the last 15 following the development of workplace technologies, from the early days of document management, enterprise content management and content services. Now, with the development of new remote and hybrid work models, he covers the evolution of technologies that enable collaboration, communications and work and has recently spent a great deal of time exploring the far reaches of AI, generative AI and General AI.