FinOps in Focus
2025
Control Cloud costs from code to Production
Managing cloud costs has evolved from a financial imperative to a strategic engineering challenge. Beyond just reducing expenses, cloud cost optimization is now a catalyst for driving business agility, product velocity, and competitive differentiation.
Co-written with AWS
As a leading cloud provider, Amazon Web Services (AWS) has observed that customers who incorporate FinOps best practices achieve what they call the "Innovation Dividend"; uncovered savings that can be redirected to strategic initiatives. Based on the AWS Cloud Value Benchmark Study, organizations that implement FinOps best practices realize 27-40% cost savings compared to on-premises environments, unlocking significant opportunities for innovation.
While cloud credits and free tiers have made it easier to start building, Harness and AWS together are committed to helping customers optimize costs throughout their cloud journey, not just after launch. Through our partnership, we focus on providing builders with precise, actionable cost insights without causing decision paralysis.
Customers are balancing immediate experimentation needs with long-term cost efficiency, and this is why empowerment, training, and developer alignment are key to maintaining innovation velocity while ensuring efficient resource usage. Working with our partners like Harness, we strive to enable continuous optimization that unlocks innovation while making FinOps an integral part of the product development lifecycle, rather than an afterthought.”
Pranav Bhushan, Global Head of Cloud Economics & FinOps, AWS
According to Gartner's Forecast Analysis, the global cloud market is expected to reach $723B in 2025 and potentially $2-2.4T by 2030, highlighting the critical importance of cost optimization before, during, and at scale.
Which is why we surveyed 700 developers and engineering leaders in the US and UK to uncover how organizations are transitioning from traditional cost management to cost-aware development, fostering collaboration across teams, and leveraging automation and AI for continuous optimization in 2025.
Developers as
FinOps Partners
Developers are uniquely positioned to facilitate financial discipline with their deep understanding of internal systems. Which is why organizations that foster developers to partner with their FinOps peers achieve technical excellence and financial responsibility. By empowering development teams with cost visibility and ownership, companies unlock unprecedented efficiency. Unfortunately, most organizations have yet to eliminate internal siloes – leading to misaligned business objectives and wasted cloud spend.
52%
of engineering leaders say the disconnect between FinOps and developers is leading to wasted spend on cloud infrastructure costs.
Cloud spend, once managed by finance teams in isolation, is now embedded deeply within the software development lifecycle. Organizations are realizing that cloud efficiency isn’t just a numbers game—it’s about creating a culture of cost awareness. The reality today is that developers often view cost optimization as someone else's problem. This disconnect leads to over-provisioned resources, idle instances, and inefficient architectures that drain budgets.
55%
of developers ignore
cost management
And it doesn’t stop there. Only 35% cite ensuring cloud cost efficiency as a key measure of success. Though this is cause for concern, understanding the rationale is critical. Since developers aren’t being incentivized to control spend, they likely also aren’t being penalized for overspend either.
When we surveyed engineering leaders on which teams are accountable for cloud cost management, developers were often far from culpability:
Cloud operations team
Finance team
IT operations team
Infrastructure team
Application teams
Development team
Product teams
51%
45%
41%
29%
22%
13%
13%
Cloud operations team
51%
Finance team
45%
IT operations team
40%
Infrastructure team
29%
Application teams
22%
Development team
13%
Product teams
13%
Collaboration between finance, product, and engineering teams is essential to turn cloud cost management into a shared responsibility. Developers must become equal partners in the cloud financial strategy – a task that most are willing to take on.
62%
of developers want more control over, and responsibility for, managing cloud costs.
62%
This desire for increased responsibility represents a crucial opportunity for organizations to bridge the gap between development and financial governance. Rather than viewing cost management as a burden, developers are showing genuine interest in becoming financial stewards of their cloud resources.
Tool Gaps Drive Hidden Costs
The truth is, when cost optimization exists outside the development process, it becomes an afterthought rather than a design principle. This leads to systemic waste that compounds with each deployment. The reason being is that there are blindspots that aren’t being addressed.
29%
of developers say their cost optimization tools are fully integrated into the software development lifecycle (SDLC).
When we asked developers about their level of visibility into their cloud infrastructure expenses, it wasn’t surprising to learn that only:
43%
Have full visibility into real-time data on idle cloud resources
39%
Have full visibility into real-time data on unused or orphaned cloud resources
33%
Have full visibility into real-time data on over/under provisioned workloads
Without visibility into the impact of their spending or incentives to optimize, developers default to over-provisioning to ensure performance—treating cloud instances as an infinite resource rather than a finite business investment. Dev environments sit unused during off-hours, forgotten test instances run indefinitely, and oversized resources collect dust. By the time these overruns surface, expenses have already occurred and require significant refactoring to fix.
This silent drain becomes more severe when multiplied across different teams and projects. And without real-time insights integrated into their daily workflows, developers and engineering leaders can assume with near certainty that a significant portion of their budgets are wasted. 66% of developers estimate that at least 20% of their cloud infrastructure spend is wasted on underutilized resources. And we soon discovered why.
55%
of developers say it's too hard to understand their cloud spending commitments with the various cloud providers they use
These blind spots not only impact the bottom line but also create tension between teams as they struggle to retrospectively justify unexpected costs and identify responsibility for cloud waste. With these shadow workloads being seen as a perpetual tax of sorts, it becomes easier to grasp why 55% of developers say it's too hard to understand their cloud spending commitments with the various cloud providers they use. The same percentage of developers also disclosed that their cloud purchasing commitments are ultimately based on guesswork.
AI Powers Modern FinOps
Cloud pricing models now include a myriad of variables that shift constantly – think of instance types, regions, commitment levels, and usage patterns. And while already juggling technical requirements, security concerns, and delivery deadlines, developers are overwhelmed by the additional effort required to make cost-effective infrastructure decisions.
of developers say the complexity and constant cost changes make it hard to optimize cloud infrastructure spend
Though there is still a great deal of room for improvement, automation has made simplified . By implementing intelligent tools that continuously analyze usage patterns, recommend right-sizing opportunities, and automatically adjust resources based on actual demand, organizations can shield developers from this complexity.
68%
of developers don’t have fully automated cost savings practices
Though just about a third of developers have fully automated cost-saving tools, this provides an opportunity to understand which optimization strategies teams are prioritizing in their automation journey. By examining these initial steps, organizations can learn from early adopters and create more effective roadmaps for expanding their own cost optimization practices.
Developers were asked to identify which of the following methods they use to optimize cloud costs and they revealed:
Tracking and shutting down idle resources
Reserved instances or savings plans
Rightsizing instances
Spot orchestration
52%
42%
39%
29%
Tracking and shutting down idle resources
52%
Reserved instances or savings plans
42%
Rightsizing instances
39%
Spot orchestration
29%
Ironically enough, artificial intelligence is emerging as the next frontier in cloud cost optimization. While automated scripts can handle predictable events like shutting down weekend resources, AI brings sophisticated pattern recognition and predictive capabilities to the table. And while no one is questioning that AI offers powerful cost optimization capabilities, actual adoption and regular usage remains a critical area to monitor.
Developers were then asked about the role AI-driven recommendations play in their cloud cost optimization strategy. They informed us that:
We rely heavily on AI
Moderate AI and manual work
AI is not a key factor
We don't use AI for cloud spend
32%
47%
17%
4%
We rely heavily on AI
32%
Moderate AI and manual work
47%
AI is not a key factor
17%
We don't use AI for cloud spend
4%
As you can see, adoption is still warming up. But that’s honestly to be expected as there are many variables to analyze, especially for organizations in highly regulated industries. But slow adoption doesn’t seem to have a strong effect on the perceived value of AI in the coming year.
86%
of respondents feel like AI can impact their ability to optimize costs in 2025
Engineering leaders and developers are very optimistic about AI's potential to transform cloud cost optimization in 2025. This confidence stems from AI's proven ability to provide predictive insights, automatically adjust resources based on usage patterns, and identify optimization opportunities that human analysts might miss.
Cloud Cost Control in 2025
With cloud costs representing an ever-larger portion of technology budgets, organizations can no longer afford to treat cost optimization as a separate process from development. Integrating cost optimization tools directly into the software development lifecycle becomes not just beneficial, but critical for survival in 2025 and beyond.
When optimization tools are fully embedded in the SDLC, developers gain real-time visibility into the financial impact of their decisions, enabling them to make cost-effective choices during initial design rather than costly corrections after deployment. Which is what more than half of leaders admit is taking place today.
52%
of engineering leaders say their organization focuses on cloud cost optimization once applications/features are deployed to production versus during development.
Addressing cloud costs after deployment creates a costly game of catch-up that organizations can't afford to play. When cost optimization becomes a production-stage concern, teams face the painful choice between living with expensive architectural decisions or undertaking risky refactoring of live systems.
As the final takeaway, realize that controlling costs in 2025 requires more than just giving developers tools–it requires a cultural shift where developers feel ownership of their cloud spend. Without this cultural foundation, even the most sophisticated cost optimization tools will struggle to deliver lasting impact.
Your next steps forward
Before we get too ahead of ourselves, let’s recap our findings and highlight a few key takeaways. That way you have a clear summary of AI’s impact within software delivery. Ultimately, the integration of AI into development workflows requires careful consideration. While AI tools can significantly reduce cognitive load, they also introduce new complexities around tool selection, integration, and governance.
And the reality is, most organizations don’t have a real picture of these nuances prior to starting their journey. Companies must carefully balance the benefits of AI automation against the need for human oversight and understanding. From a desired outcomes perspective, the goal should be to use AI to augment developer capabilities rather than replace human judgment.
