Why Most FinOps Teams Are Still Stuck—And How to Break the Loop

The 2025 CloudBolt Industry Insights (CII) report revealed that 58% of organizations take weeks or longer to remediate cloud-cost waste—even when the signal is clear.

That’s the real tension: teams may have visibility, but the follow-through still lags. Dashboards are improving, but the loop between insight and action remains painfully slow. And the speed of cloud waits for no one.

At FinOps X 2025, CloudBolt Chief Product & Technology Officer Kyle Campos argued that what many call “mature” FinOps—cost visibility, tagging hygiene, anomaly alerts—isn’t the finish line. It’s the ceiling. And it’s holding teams back from what comes next: automation that closes the loop. 

FinOps Today: A Broken Feedback Loop 

In most organizations today, the optimization loop plays out like this: a change in cloud usage triggers a signal. That signal shows up in a report or dashboard—maybe flagged by an alert. Someone sees it. Then comes evaluation: what changed, is it expected, is it risky? A ticket gets filed. Days pass. Prioritization becomes a negotiation. Engineering may or may not have bandwidth. Eventually—maybe—a fix is deployed. Sound familiar? 

It’s a slow, inconsistent process that leans heavily on people to keep things moving. Kyle called it a kind of “report factory”—where FinOps teams spend more time chasing approvals and monitoring ticket queues than driving meaningful change. Everyone’s busy, but no one’s in control. 

This creates what Kyle called a “two-speed reality”: insights move faster than teams can respond. Signals pile up, backlogs grow, and action—if it happens at all—lags by weeks or months. By then, the infrastructure has already changed. 

“You start to get this pile-up… and you’re not able to get through that leg fast enough,” Kyle said. “And now you start to feel overwhelmed with the amount of things on your plate. Suddenly everything feels urgent, and prioritization turns into a bottleneck.” 

And the friction doesn’t stop there. Because FinOps often lacks the authority to act directly, the insight gets kicked downstream to Engineering—and the whole loop begins again. 

Engineering receives the signal—but often re-evaluates it through a different lens. Is it still valid? Does it align with sprint priorities? Is it worth addressing now? What looked like a clear opportunity to FinOps becomes just another ticket in a crowded backlog. The process repeats: more evaluation, more prioritization, maybe—eventually—action. 

It’s a second loop, built on the same manual friction as the first. It adds delays, complexity, and risk that the signal degrades before anything gets done. 

When FinOps depends on people to manually move insights forward, the loop doesn’t close. It fragments. And over time, the model becomes unworkable—not just inefficient, but fundamentally incapable of scaling with cloud complexity.  

This isn’t just a process problem. It’s a structural one. But it’s also solvable. Continuous optimization offers a fundamentally different approach—one built for the speed, scale, and sprawl of modern cloud environments.

From Manual to Continuous: Shifting the Operating Model 

Continuous optimization isn’t just about speed—it’s about changing how decisions get made. Instead of manually reviewing reports or firing off tickets one by one, the system becomes the operator. Signals stream in, and automation does the rest. 

Machine learning monitors conditions in real time. Evaluation happens as data arrives. Policies—configured by FinOps and aligned to business context—determine what matters. What’s risky? What’s noise? What thresholds reflect your org’s goals? 

Even gray areas can be codified. Maybe certain teams must be looped in before action, or approvals are only valid during change windows. These concerns can be built into policies. You’re not hoping someone sees a ticket—you’ve designed the system to know what to do. 

When a signal meets the right criteria, it’s automatically queued for action. No begging. No chasing. Just a clear, declarative response that happens by design. 

The difference this makes isn’t marginal—it’s exponential. A 31-day lead time might feel acceptable on paper, but compare that to 3 hours. Over time, the financial delta is massive. And it doesn’t stop at cost savings. Faster action means better resilience, fewer incidents, and more freedom for Engineering to focus on innovation—not cleanup. 

The result? No more backlog pressure. No more JIRA escalations. Just continuous, policy-driven momentum—built for the way cloud works today. 

Breaking Through the Ceiling 

The shift we need isn’t just about speed. It’s about acknowledging that the old operating model is broken. Manual optimization can’t keep up with the complexity and velocity cloud. The need has evolved. The scope has expanded. But most teams are still stuck in loops that don’t close. 

That’s why continuous optimization isn’t a nice-to-have. It’s becoming the only viable path forward. Moving from signal to action in hours—not weeks—improves cost efficiency, resilience, and engineering velocity.

The opportunity now is to stop treating visibility as the destination—and start embedding intelligence into the flow of action itself. When automation is driven by policies rooted in your real-world priorities, the right thing happens by default—not after weeks of hallway escalations. 

We don’t have to keep begging. We don’t have to file more tickets. The future of FinOps won’t be won through hustle. It will be won through systems that adapt in real time, codify what matters, and act at the speed of cloud. 

CloudBolt was built to help teams do exactly that. With policy-driven automation, real-time evaluation, and support for hybrid, public, and Kubernetes environments, our platform helps close the loop—continuously, intelligently, and without the overhead. 

The ceiling isn’t your data. It’s the execution gap. And it’s still wide open.