A survey of 350 senior IT leaders in the U.S finds more than three quarters (78%) admit they find it difficult to consistently demonstrate a return on investment (ROI) in cloud computing, with revenue growth (43%), operational efficiency/productivity (36%), and cost savings (35%) being the metric most widely tracked.

Conducted by Wakefield Research on behalf of CloudBolt Software, a provider of an IT automation platform, the survey also finds more than half of respondents (55%) have difficulty linking cloud spend directly to business outcomes. Nearly half (48%) blame organizational misalignment and operational silos, followed closely by inefficient resource management (44%) such as poor tagging of infrastructure and inconsistent accountability as being the primary root causes of that disconnect.

Additionally, the survey also finds two thirds (66%) report they have cloud environments that are mostly-to-fully automated. However, 58% conceded it takes weeks or months to detect and fully remediate cloud-cost waste opportunities.

Mark Zembal, chief marketing officer for CloudBolt, said that would suggest IT teams are relying on various forms of manual automation, such as using Jira tickets to manage IT service issues, rather than an IT automation platform that makes it possible to detect and resolve issues in minutes.

The survey also suggests these issues might be further aggravated as more cloud-native applications are deployed. A full 98% of respondents said Kubernetes is becoming a major driver of cloud spend, but 91% remain unable to effectively optimize their Kubernetes clusters.

Additionally, many organizations (42%) are also expecting the management of hybrid cloud computing environments to become a higher priority, with more funding (39%) being allocated toward achieving that goal over the next six to 12 months, the survey finds.

Finally, 40% of respondents are also looking to optimize cloud spending allocated to the running of artificial intelligence (AI) models, otherwise known as FinOps for AI.

Collectively, these shifts make it clear that existing approaches to managing cloud computing environments are not going to scale to the levels that will be required, noted Zembal. IT teams will need to move beyond existing formulaic workflows that were never intended to support highly dynamic application environments, he added.

It’s not clear to what degree IT teams will move to unify the management of cloud computing environments. Most IT organizations have teams of specialists that are specifically responsible for various computing platforms, including servers running in on-premises IT environments. The issue that arises is that each team of specialists winds up performing the same basic tasks. CloudBolt is making a case for a more automated approach that enables a centralized IT team to manage multiple cloud computing environments at scale in a way that also serves to reduce the total cost of IT. In many cases, the efforts closely align with a set of best FinOps practices that IT organizations are increasingly adopting to rein in the cost of cloud computing, noted Zembal.

Regardless of approach, the total cost of IT is only going to continue to increase as more application workloads are deployed. Most organizations can’t afford to hire and retain the small army of IT personnel required to manage those workloads so the only real option is going to be relying on more automation to make up for what will otherwise become a chronic shortage of cloud computing expertise.