Global enterprises could be set to waste billions on their cloud infrastructure spend this year, due to their failure to create a culture of cost awareness amongst developers, new research shows.

A study of 700 engineering leaders and developers across the UK and the US, released by Harness, found a major disconnect between developers and financial operators when it comes to cloud costs, which could cost companies billions.

Over half (52%) of engineering leaders say the disconnect between FinOps and development teams is leading to wasted spend on cloud infrastructure costs.

Developers have limited insights into cloud waste, the report says, with less than half saying they have access to real time data on idle cloud resources (43%). About two in five (39%) say they have unused or orphaned resources, and a third say they have over or under-provisioned workloads. All of this contributes to a picture of imbalance.

Without a clear and balanced picture of cloud resource requirements, 55% of developers say purchasing commitments are ultimately based on guesswork.

As a result, enterprises estimate that 21% of their cloud infrastructure spend is wasted on underutilised resources, equating to an estimated $44.5bn (£35.25bn) in 2025 alone.

“Cloud infrastructure spend is one of the biggest line items for modern enterprises, right behind salary. Leadership teams should ask themselves if they are comfortable relying on guesswork to manage and optimize this spend,” said John Bonney, CFO at Harness.

“With a lack of visibility into cloud costs and constantly shifting requirements for workloads, needless inefficiencies and over-commitments are eating up resources and holding firms back when they are under pressure to do more with less.”

By automating cloud cost management, organisations can dynamically manage spend, eliminating waste and realising instant savings. This intelligent approach also removes the burden of managing resource scheduling for engineers while avoiding unnecessary resource usage.

The study also found a lack of transparency across organisations when sharing cloud spend data. Developers are often involved in cost optimisation too late in the process, and are not given the right insights they need to optimise cloud spend.

Most organisations are failing to set up fundamental cloud cost optimisation best practices, with nearly three quarters (71%) of developers not carrying out spot orchestration, and just under two thirds (61%) not rightsizing instances.

Further, over half (58%) do not use reserved instances or saving plans, and 48% do not track and shut down idle resources.

Despite these fault margins, 62% of developers want more control over, and responsibility for, managing cloud infrastructure costs.

Only a third (32%) of developers have fully automated practices to enforce cost-savings, the study also found, despite it taking enterprises on average 31 days to spot cloud waste such as idle or unused cloud resources.

Recommended reading

• Cloud Spending Surged By 21% Last Quarter
• Comment | Navigating Today’s Cloud Security Challenges
• Cybersecurity and Cloud Still Top Priorities for Financial Firms

It takes a similar amount of time to spot overprovisioned cloud resources and flag that they need to be right-sized (25 days).

Developers do see change on the horizon, however, as 86% of respondents feel like AI can impact their ability to optimise costs in the next year.

“The reality is developers don’t often view cost optimization as a priority. This disconnect leads to overprovisioned resources, idle instances, and inefficient architectures that quietly drain budgets,” explains Ravi Yadalam, senior director of product management at Harness.

“Cloud efficiency isn’t just a numbers game—it’s about creating a culture of cost awareness where enterprises shift their FinOps practices left, so engineers understand how their architectural choices directly impact both performance and financial outcomes.

“Developers must become equal partners in their cloud cost management strategy, gaining full visibility into spend so they can identify cloud waste and automatically eliminate it.”