Somewhere right now, a FinOps engineer is writing a Jira ticket that nobody will take action on.

Somewhere right now, a FinOps engineer is writing a Jira ticket that nobody will take action on.

In a world where margins are under pressure and every dollar of efficiency translates directly to runway, valuation, or competitive pricing, how you manage cloud costs is a strategic decision, not an IT one.

Consider two companies with the same cloud bill. One has a motivated, talented FinOps team working hard to review costs and action recommendations.

The other has a system, partly automated and partly process-driven, that optimizes costs continuously.

A year later, one has growing savings and a leaner cost structure. The other is still triaging dashboards, despite their best efforts.

The difference isn’t effort or intent. It’s whether FinOps is a system or an intention.

Where FinOps Is Today

FinOps as a discipline has matured enormously. The tooling is better, the frameworks exist, awareness is high. Most companies of any meaningful size now have some form of cloud cost visibility.

But visibility is not a strategic advantage anymore. Everyone can see their waste. The differentiator has shifted.

In 2026, the question isn’t “do you have a FinOps practice?” It’s “how much of your FinOps practice runs reliably without depending on individual heroics?”

Why Good People Aren’t Enough

Many companies have invested seriously in FinOps. They have smart, motivated people who genuinely care about cloud costs. And they’re still overspending.

The reason isn’t a people problem. It’s a model problem.

At scale, manual FinOps looks like this: your cloud environment generates hundreds of optimization opportunities every month across AWS, GCP, and Azure.

Each one needs to be triaged, prioritized, assigned, actioned, and verified.

The people who can action them (engineering teams) have a full sprint of other priorities.

The people who find them (the FinOps team) don’t own the infrastructure.

Recommendations pile up faster than they get resolved. Meanwhile new waste is being created continuously as new resources get spun up.

Even a world-class FinOps team operating manually is playing whack-a-mole. They might clear the backlog this month. But stop moving and the waste builds up again.

The bottleneck isn’t knowledge or motivation. It’s the absence of a system.

Not Everything Should Be Automated

At this point the instinctive pushback is usually: “we can’t just automate decisions about production infrastructure.”

That’s correct. And it’s not what operationalization means.

Think of it as a spectrum.

Fully automated: Low risk, high confidence actions with no human needed. For example, deleting unattached volumes in dev/test environments, stopping idle non-production instances, flagging unused elastic IPs.

These are well-understood, easily reversible, and the cost of occasionally getting one wrong is low.

Process-driven: A human decision is required, but with defined ownership, a clear workflow, and an SLA so the decision actually gets made. E.g. rightsizing a production database, modifying reserved instance coverage, making architectural changes to reduce data transfer costs.

These need judgment. But judgment shouldn’t mean indefinite deferral.

Ad hoc: This is human judgment with no supporting process. No defined owner, no timeline, no follow-up mechanism. This is where most companies live today. It’s also where recommendations go to die.

The goal is to push everything as much towards automation on the spectrum as is safely possible.

Automate what can be automated. Build reliable process around what can’t. Eliminate the ad hoc category entirely.

The human in the loop isn’t the problem. The missing process is.

From Cleanup Crew to Strategic Function

Ad hoc FinOps is a treadmill. You fix waste, it creeps back, you fix it again. You spend effort recapturing the same ground every cycle. Stop running and you go backwards.

New waste appears constantly. That doesn’t change with operationalization. What changes is how quickly it gets caught and acted on.

Instead of surfacing in a quarterly review, it gets caught immediately and there’s a system that ensures it gets resolved.

The waste itself doesn’t disappear, at least not completely. The lag does. The mental overhead does. The tapping people on the shoulder to ask them to shut down an overprovisioned dev cluster does.

The organizational benefit is significant.

When routine work runs itself, your FinOps team stops being a cleanup crew and starts being a strategic function.

They can make architecture decisions, influence how new systems get built, and catch cost problems at the design stage.

That’s a fundamentally different and more valuable use of the same people.

Prevention is worth more than remediation. You only get to prevention when remediation is handled by a system.

We’ve Seen This Before

This isn’t the first time the industry has faced this kind of inflection point.

Before CI/CD, software deployments were manual, periodic, and stressful. Companies had skilled, dedicated release engineers who worked carefully to manage deployments.

But the model had a ceiling. Deployments were slow, risky, and infrequent, not because the people were bad at their jobs, but because the process couldn’t scale.

The companies that embraced CI/CD and infrastructure as code early didn’t just move faster. They changed the nature of the work entirely.

Deployment stopped being an event and became a continuous background process. The engineers who used to manage releases manually shifted to higher-value work.

The companies that held on to manual processes found themselves structurally unable to keep up, regardless of how talented their teams were.

FinOps is at exactly that inflection point now.

The companies that operationalize early won’t just save more money. They’ll change the nature of how cloud costs are managed in their organizations.

And the gap between them and companies still running ad hoc FinOps will grow in exactly the same way.

What Operationalized FinOps Actually Looks Like

What operationalized FinOps looks like in the details will vary across companies. However, there are overarching themes, and it doesn’t have to be anything exotic.

The companies building this advantage have:

• Continuous detection across their entire cloud environment. Not a monthly report someone pulls manually, but automated scanning that surfaces waste as it appears.

• Clear ownership so every resource and every cost center has a named team accountable for acting on recommendations. A recommendation with no owner is just noise.

• Predefined action thresholds, the spectrum defined in advance rather than negotiated case by case. This is what makes process-driven actions actually happen: when everyone already knows who approves what and by when, decisions stop getting deferred.

• Automated execution for the low-risk, high-confidence actions. Not everything, but enough that routine work happens without anyone having to remember to do it.

• Feedback loops that track realized savings, not just potential savings, so the system knows what was actioned, what was saved, and what drifted back.

The result is cost optimization running the way deployments run in a mature engineering organization: continuously and reliably, without requiring a meeting.

The Question Worth Asking

FinOps maturity isn’t measured by how good your dashboards are or how experienced your team is.

It’s measured by what happens to a cost optimization recommendation from the moment it’s identified to the moment it’s resolved, and whether that process is reliable enough to happen even when other priorities get in the way.

Here’s a simple diagnostic. When your FinOps tool surfaces a recommendation today, what happens next? Who gets notified? By when does it need to be actioned? Who checks that it was done?

If the answer is “it depends” or “someone on the team looks at it eventually,” that’s not a system. That’s a hope.

Every company has a FinOps intention. The question is whether you have a FinOps system.

Intentions don’t scale. Systems do.

Every company’s FinOps maturity looks different. I work with companies to audit their current setup, build a practical roadmap to operationalize, and implement the tooling to make it stick. If you’re ready to stop recapturing the same ground, let’s talk.

Book a call

Original post available in Jill’s Substack