The pandemic forced businesses to accelerate their digital transformations, which, in turn, sped adoption of the cloud. According to a 2021 survey by O’Reilly, cloud adoption is rising steadily across industries — 90% of organizations now use cloud computing, up from 88% in 2020.
At its best, the cloud, nebulous though the term might be, promises to ease the transition from offline, manual processes to managed automation. For example, with the cloud, sales commissions previously tracked by spreadsheets can be automatically calculated and paid out on time. In a more futuristic scenario, cloud-based technologies can enable one to monitor and fine-tune an entire building’s HVAC from afar.
But the cloud has a major downside, and it concerns cost: More than a third of businesses have cloud budget overruns of up to 40%, according to a recent poll by observability software vendor Pepperdata.
This challenge has given rise to a cottage industry of companies developing software purported to optimize cloud costs on the fly. These cloud cost optimization vendors range from young startups to ripe acquisition targets, advertising their use of technologies like AI to detect which cloud services customers are using and reduce spending where possible.
“You can’t really separate out cost management from the rest of cloud management,” Gartner analyst Craig Lowery told TechCrunch over email. “About five years ago, cost management was very important because it was the most painful and least understood aspect of cloud adoption. Today, cost management is much better understood, and security and compliance have become just as, if not more important. [It’s] considered table stakes for a cloud management solution.”
Several years ago, the market, while nascent, was consolidating as incumbents in adjacent sectors saw the opportunities presented by cloud cost optimization. Microsoft in 2017 acquired Cloudyn, which provided tools to analyze and forecast cloud spending. Then, in 2019, Apptio snatched up cloud spending management vendor Cloudability, while VMWare and NetApp bought CloudHealth and Spot (formerly Spotinst), respectively, within the span of a few years.
The consolidation isn’t necessarily over, as evidenced by Intel’s $650 million purchase of Granulate this April. But the dust is beginning to settle.
“The consolidation phase is done, as most of the startups have been acquired and cloud management companies have moved on to solve broader problems,” Lowery continued. “The differentiations in these companies’ cloud management products come from the way in which all the cloud management functions work together, not something specific to cost management. They all do cost management pretty much the same way, though there are some small differences in feature and functionality.”
Unsurprisingly, Yuri Frayman, the CEO of cloud optimization startup Cast AI, disputed the notion that his company isn’t sufficiently differentiated. Cast AI is one of the few vendors that can both show cloud costs and reduce them automatically, he claimed, adding that it has capabilities to optimize cloud-native apps across providers, including Amazon Web Services, Google Cloud and Microsoft Azure.
“We started Cast AI out of personal frustration. We had just sold our previous startup, Zenedge, to Oracle, and one of the very interesting aspects of that business was the enormous growth of our cloud bill,” Frayman told TechCrunch via email. “Cast AI is now managing many thousands of clusters right now for customers large and small in adtech, e-commerce, fintech and more. More than a thousand organizations use Cast AI. Our customers’ apps range from just 100 CPUs to 211,000 CPUs.”
Hakim Weatherspoon similarly espoused the benefits of Exotanium, a cloud cost optimization company — and Cast AI rival — that he helped launch in 2019. Born out of research at Cornell University, Exotanium claims its technology can “dynamically” optimize cloud resources based on usage.
“Exotanium continuously adjusts your infrastructure based on the need to be as efficient as possible,” Weatherspoon said. “Other solutions require code changes or an operator to set policies. Many will point out where the issue is but will not tell you how to fix it, or even if some tell you how to fix it, you still need a team to spend time and effort to complete this change. Exotanium does this dynamically.”
Frayman largely attributes the growth in the market for cloud cost optimization to the nature of the cloud — specifically how the cloud can make consuming compute resources easier without necessarily reducing expenses. With a potential economic downturn on the horizon, businesses are also looking for budgets to slash, and the cloud is one of them, particularly for firms that boosted cloud spend during the pandemic.
According to a report by Anodot, for 59% of the organizations spending over $2 million a month on cloud usage, it can take days to detect an anomalous surge. During especially heavy usage periods, nearly 50% of companies see cloud costs spike from 10% to 19%.
Predictably, the majority of companies responding to a May poll from NetApp’s Spot (not exactly an unbiased source, granted) said that they plan to focus on cloud cost management in 2022.
“The current economic climate is accelerating the need for tools … that can help better understand and then eliminate unnecessary spending within enterprises’ cloud budgets — and do so without impacting application availability,” Webb Brown, the co-founder and CEO of Kubecost, which develops software to optimize Kubernetes cluster usage said. “Innovation is still the focus, but companies and developers are really starting to focus on saving costs.”
Major cloud service providers including Google and Microsoft offer first-party tools to “rightsize” spending. But some customers are reluctant to trust a vendor, given that they stand to benefit from recommending costlier plans.
“The pandemic forced a lot of companies to move faster to the cloud and develop cloud-native alternatives in a hurry,” Frayman said. “Suddenly, software-as-a-service solutions were in, and data centers were out. With that move, two important trends started in 2022: The realization that cloud costs were ballooning, and that the economy, market and inflation are not going in the right direction.”
Weatherspoon also blames long-term service commitments for driving enterprises to cost-optimization solutions. Multiyear contracts, the norm in the cloud industry, can reduce costs substantially but exact the toll of flexibility.
“Essentially owning [cloud] machines for three years at a time is counterintuitive and counter to [one] of the cloud’s biggest value propositions: the ability to pay as you go,” Weatherspoon said. “The pandemic and work-from-home only accelerated this.”
As for what the future might hold, Frayman and Weatherspoon predict the cloud cost optimization segment will expand as the demand for budgeting tools climbs. For one, the adoption of multicloud — using services from multiple cloud vendors — is expected to rise, bringing new cloud spend management challenges.
Gartner projects that worldwide spending on cloud services will grow 20.4% in 2022 to total $494.7 billion.
“Cost is the great consequence of an autonomous infrastructure,” Frayman said. “You will see more powerful observability, more automation and the addition of more cost variables across clouds.”
Weatherspoon added: “Most enterprises will be keeping some form of an on-premises data center and use the cloud for bursting capabilities. The next frontier that is starting but is immature is around multicloud and the ability to use the best available resources for a given job based on comparing the availability across all the major providers. Other trends that are interesting include dynamic and continuous resizing of instances based off need and changing this as the job goes through different cycles, as well as the use of Kubernetes and optimizing Kubernetes clusters across on-premises and the multiple clouds.”
Comment