Cloud costs now No. 2 expense at midsize IT companies behind labor

Midsize IT companies now spend huge percentages of their revenues on cloud services, driven by a growing number of AI workloads running in the cloud, according to a new survey.

While a third of IT companies say they spend between 5% and 8% of annual revenues on cloud services, another 29% spend over 13%, according to a survey of 100 CFOs at SaaS and other IT companies with up to 1,000 employees. On average, the companies represented spend 10% of revenues on the cloud.

AI and machine learning workloads make up 22% of the cloud costs at the organizations represented, according to the survey, commissioned by FinOps vendor Cloud Capital. Cloud costs are now the second largest expense at many IT and SaaS companies, trailing only payroll and other employee-related spending.

The AI tradeoff

For now, IT companies and SaaS providers are swallowing these cloud costs in anticipation of projected efficiency or productivity increases coming from AI, says Spencer Pingry, cofounder and CTO at Cloud Capital.

“With the AI workloads in particular, as we see the cloud spend go up, a lot of people are dealing with the increase for now, because they think there’s a certain set of payoffs that will be realized,” he says. “But once we realize those payoffs, I don’t think cloud spending is going to cut back; it’s going to increase instead.”

The question for their CIO customers then becomes, at what point do these IT and SaaS vendors begin passing on these costs to them?

“As cloud spend rises, price increases are one common lever to protect margins,” Pingry says. “But in a market where competition to land and retain customers is intense, raising prices can backfire by increasing churn and making it easier for competitors to win switchers.”

The Cloud Capital survey shows midsize IT vendor CFOs and their CIO partners struggling to contain cloud spending, with significant cost volatility from month to month. Three-quarters of IT org CFOs report cloud spending forecasts varying between 5% and 10% of company revenues each month, Pingry notes.

Costs of AI workloads are harder to predict than traditional SaaS infrastructure, Pingry adds, and organizations running major AI workloads are more likely to report margin declines tied to cloud spending than those with moderate AI exposure.

“Training spikes, usage-driven inference, and experimentation noise introduce non-linear patterns that break the forecasting assumptions finance relies on,” says a report from Cloud Capital. “The challenge will intensify as AI’s share of cloud spend continues scaling.”

A huge increase

While Cloud Capital doesn’t have historical data on cloud spending as a percentage of revenue — the company plans to track it going forward — the survey’s numbers represent a huge increase from historical trends.

Gartner in 2024 calculated that the entire IT budget at midsize companies across several industries took up just 3% of revenues, with software publishing and internet services firms spending just over 6% of revenue on IT.

The survey’s average of companies using 10% of revenues on cloud services seems high, but there’s little doubt that cloud spending is increasing, says Ed Frederici, CTO at Appfire. The software vendor currently spends about 1.6% of its revenues on cloud computing, he says.

Still, cloud compute and storage costs are rising, he says. “When you’re using consumption-based pricing, you can go from a few dollars a day to $1,000, so we’re very diligent about keeping an eye on it,” he says.

While Appfire doesn’t have a lot of AI-heavy cloud workloads, Frederici sees how cloud costs can ramp up quickly as AI compute and storage needs feed off each other.

“The more AI knowledge you want to have, the more storage you’re going to have,” he says. “It’s kind of this non-virtuous cycle — the more capable you make your AI, the more you drive up your two most expensive costs.”

Too much flexibility

Other IT leaders see the survey’s results as more realistic. At many organizations, developers have too much power to spin up new cloud resources for short-term projects, says Rick Clark, global head of cloud advisory at digital transformation provider UST.

“Cloud spend climbs toward 10% of revenue when consumption is disconnected from business value and when we confuse developer flexibility with productivity,” he says. “In the data center era, no one would have allowed developers to rack and install hardware, yet in the cloud era, we gave them near-infinite infrastructure choices with no economic guidance.”

Cloud services in themselves aren’t inherently too expensive, but many organizations shoot themselves in the foot through unintentional consumption, Clark adds. “Costs rise when the system is built without a clear understanding of the value it is meant to deliver,” he adds.

In addition, the fear of missing out on AI is leading to skyrocketing cloud spending, he says.

“No CxO wants to explain to the board why another company used AI to leap ahead,” Clark adds. “This has created a no-holds-barred spending spree on training, inference, and data movement, often layered on top of architectures that were already economically incoherent.”

IT leaders, working with CFOs, can take several steps to control cloud spending, experts say. While some experts point to IT teams using FinOps practices, Clark says they’re just part of the solution.

“FinOps is helpful for finding obvious waste such as unused resources, but that is a one-time Band-Aid,” he says. “The larger issue is that FinOps often pushes cost responsibility onto developers who have no insight into revenue, margin, or the economic goals of the workload they are building. Cost is a business decision, not a technical decision.”

IT leaders need to work closely with CFOs to set goals for cloud spending that match the needs of the IT team and the goals of the business, says Cloud Capital’s Pingry. CFOs need input on cloud budgets as they eat up a greater portion of revenue, and a good relationship between the CIO and CFO can lead to a better understanding of both departments’ needs, he adds.

“The advice I’d give to any IT leader is to just work on the relationship with finance, because it’s often seen as like a negative, with finance coming to look for ways to cut things, but they just want things to be on a plan, and as long as that plan is being hit, then finance is very happy,” Pingry says.

Appfire’s Frederici recommends that IT leaders regularly check their teams’ cloud use and be diligent about cutting off cloud resources that are no longer needed. IT teams can also set up alerts to tell them when cloud usage or spending is deviating from norms, he says.

Appfire focuses on cloud optimization and regularly shuts down cloud resources that are no longer needed, he says.

“Something that very common is your developers, your QA folks, have the ability to spin up any non-production environment they want,” he says. “They spin it up for an experiment, and then they forget to exist and spin up another one for the next experiment. All of a sudden, you have really extensive sprawl that’s sitting there.”