Dive Brief:
- Enterprises plan to expand their use of FinOps cloud cost optimization practices to rein in SaaS and AI spending, according to a FinOps Foundation report published Thursday. The nonprofit organization surveyed more than 800 companies for its fifth annual report.
- Nearly two-thirds of respondents intend to leverage FinOps to manage SaaS spend in the coming year and nearly half are targeting software licensing fees. A majority of FinOps practitioners — 63% — already manage AI costs across cloud and on-prem ecosystems, up from 31% last year, the report found.
- “Generative AI services from AWS, Google and Microsoft are coming into the FinOps practice right now as part of their regular cloud bills,” FinOps Foundation Executive Director J.R. Storment said. “But there’s a different set of levers for optimizing AI spend. It’s not just CPUs and memory and storage costs, it’s now GPUs and tokens and inference and training.”
Dive Insight:
As the FinOps ecosystem expanded to encompass vendor cost management tools and unified billing specifications, the practices associated with cloud management spread to other areas of proliferating tech spend.
The FinOps Foundation, a project of The Linux Foundation, added SaaS and data center to its list of cost-control candidates last year. The organization aims to introduce support for software services in version 1.2 of FOCUS, the FinOps Open Cost and Usage Specification framework due later this year.
FinOps units are becoming a ubiquitous enterprise presence. Among Fortune 100 companies, 93 participate in FinOps Foundation programs, according to Storment.
“Companies understand they need to do FinOps,” Storment said. “The big ones — the mature ones — have gotten good at it.”
As enterprises tamed cloud spend, FinOps priorities shifted.
Cloud workload optimization and waste reduction remained the top concern, cited by half of respondents to this year’s survey. Over the next 12 months, respondents said they intend to scale FinOps efforts and prioritize governance over cost cutting.
“The longer you have done optimization, the more you realize you need governance and policy to further those efforts,” FinOps Foundation COO Steve Trask said. “Optimizations are easy to tackle, while policies, processes and automation are aligned to more maturity as a practice.”
AI has climbed the ladder in response to the elevated levels of spending in cloud and on-prem attached to the technology’s appetite for compute and data resources.
Nearly every organization is investing in multiple infrastructure areas for AI, the foundation’s survey found, with 69% leveraging SaaS and 30% building out data centers and private cloud deployments.
The mix complicates cost optimization efforts. Last year, AI technologies were the prime driver of unmanageable cloud costs for many organizations, according to FinOps software provider Tangoe.
“You have to look in different places, and in each place there are different billing models, pricing models and optimization models,” said Storment. “The cardinality of options just goes through the roof.”
