31% of IT leaders waste half their cloud spend

Nearly half of IT decision-makers believe their organization sees more than a quarter of its cloud spend wasted, with around one in three (31%) saying that the waste exceeds 50%, according to a report from VMware.

Analysts and other cloud specialists see cloud cost overruns being a significant issue in the enterprise today, but they place the blame on developer and IT culture. By changing how the organization thinks of cloud investments — along with adopting a far more cynical view of cloud contract language — much of the cloud spend problem can be reduced, many industry observers contend.

“A majority of enterprises agree that organizational silos complicate cloud management, making it difficult to maintain visibility, control, and governance in the public cloud,” according to according to VMware’s report, which drew on a global survey of 1,800 senior IT decision-makers.

But IDC Research Director Rob Tiffany also sees enterprise vendor relationships playing a nontrivial role in cloud spending problems. 

“AWS and Google are not traditional IT players, so Microsoft is trying to get [enterprises] hooked” into Microsoft cloud environments, and some enterprise CIOs think, “‘I’ll just sign and then I don’t have to think about it,’” Tiffany said.

Cloud cost woes and cures

Enterprises need to more aggressively integrate FinOps tools into their operations, making sure to provide access to just about anything that touches any cloud, Tiffany said, pointing to HP’s purchase of Morpheus Data as one of the several cloud cost management tools that should be considered. 

CIOs need to be “looking for waste and idle virtual machines. It might turn out that you are only using 3% of the VM or users are not even hitting these things,” Tiffany said. “Not using FinOps software enough to be the eyes and ears of wasted and unused [cloud resources] is something that a lot of companies” are doing.

But finding a cloud utilization problem won’t help if the contract doesn’t enable chargebacks for such discoveries, Tiffany said. 

Even if such chargebacks are not supported, Tiffany encourages enterprise CIOs to “go through accounts payable” records for any cloud purchases, including shadow purchases. “If you find things that are wasteful, it could be used as leverage with another cloud,” he said.

A problem IDC has observed is enterprises purchasing FinOps tools, but giving those tools insufficient access. 

“I want my AI agent to log into our AWS dashboard and find duplication, unused services and to make sure it can do it across all of our clouds,” Tiffany said. “Make sure that APIs are exposed for every last service.”

Still, Tiffany sees corporate culture — and specifically IT and developer cultures — making cloud spend analysis difficult.

Culture transformation needed

“CIOs are typically not rewarded for nitpicky things like that, including maintaining old systems,” Tiffany said. “They get love for doing new, innovative things. [Tracking cloud spend duplication] generally doesn’t advance their careers.”

Today, though, agentic AI may make a difference. Historically, CIOs have been hesitant to distract their overburdened tech staffs by chasing small-cost problems. But assigning such chasing tasks to a virtual agent can be a key difference.

“In the past, CIOs have had pushback. ‘I don’t have the manpower for that.’ But with agentic AI, that changes the game. CIOs can now advocate to HR and CEOs to incentivize going after these inefficiencies and maybe change their bonus structures,” Tiffany said.

Matt Kimball, VP and principal analyst with Moor Insights & Strategy, encourages CIOs to consider all cloud forms, including SaaS contracts that are often overlooked. “Those big contracts with Salesforce, are they included?” he asked. 

Moor has found many CIO clients “who have signed up for overprovisioned instances, spending way too much money on [cloud resources] that they will never use,” Kimball said, offering an example of how a one-time cloud addition quietly morphs into an ongoing yearly charge. 

“Your Oracle rep says you need this module that nobody ever touches. And now it’s in the contract every year. Before you know it, it’s a recurring fee,” Kimball said. “Remember that rationalizing your cloud spend is an evergreen exercise.”

Mark Troller, CIO of expense management firm Tangoe, encourages CIOs to not overly fixate on the waste percentages reported by VMware.  

“I don’t get hung up on whether cloud waste is 30% or 50%. The real issue is definition. If you only count unused compute or storage, 50% feels overstated,” Troller said. “But when you include the indirect inefficiencies I see every day like duplicate SaaS licenses, shadow IT spend, idle environments, or poorly architected workloads driving up network costs, the true waste is likely higher.”

Roman Rylko, CTO at IT consulting firm Pynest, also sees cloud cut overruns being common.

“Many CIOs have a completely justified feeling that half of their cloud costs are wasted. For example, they deployed an environment for a project and forgot to turn it off, or reserved extra capacity that sits idle,” Rylko said. “What often just irritates is the opaque billing structures from providers. Without constant monitoring and a FinOps approach, it is almost impossible to understand where exactly the money is spent.”

Rylko cited a recent client “where the developer environments worked 24/7, although the team used them only during business hours. An automated schedule to shut down environments in the evening and restart them in the morning gave savings of a double-digit percentage already in the first month.”

Rylko argued that such anecdotes are why he sees “the issue of cloud optimization is not only finances, but also development culture. If engineers do not understand that each of their VMs or containers costs money, no CFO reports will help. Effective strategies include making engineering teams accountable for spend and ensuring they understand the cost impact of their resources.”