Three years ago, “FinOps” was a buzzword that most IT leaders could safely ignore. Today, it’s become critical infrastructure. I’m watching companies spend 30-40% more on cloud services than they need to, and the problem is getting worse as cloud adoption accelerates.

The fundamental issue is simple: cloud makes it too easy to spend money, and traditional IT budget management wasn’t designed for this consumption-based model. Someone needs to own this problem, and that someone is IT.

Why Cloud Costs Spiral Out of Control

In the old world of on-premises infrastructure, spending was constrained by physical reality. You had a fixed number of servers, a fixed amount of rack space, and procurement processes that created natural friction. If someone wanted new infrastructure, they had to submit a request, wait for approval, and go through purchasing.

Cloud demolished all those constraints. Now any developer with AWS credentials can spin up a $10,000-per-month database instance in under a minute. No approval needed. No procurement process. No visibility until the bill arrives thirty days later.

I’ve seen this pattern dozens of times: a developer spins up resources for testing, forgets to shut them down, and suddenly the company is paying $3,000 per month for an idle environment. Multiply that across hundreds of developers and dozens of projects, and you’ve got a serious cost problem.

The other issue is architectural inefficiency. When you’re buying physical servers, you think hard about sizing because you can’t easily change it later. In the cloud, you can just pick bigger instance types and call it done. I routinely see applications running on instances that are 3-4x larger than needed, burning thousands of dollars monthly because nobody bothered to right-size them.

Why Finance Can’t Own This

You’d think cloud cost management would fall under Finance’s purview—after all, it’s about managing costs. But Finance doesn’t have the technical knowledge to understand what’s being purchased or whether it’s necessary.

When Finance sees a $50,000 AWS bill, they can tell you it’s higher than last month, but they can’t tell you why or what to do about it. They don’t know the difference between EC2 and RDS. They can’t evaluate whether you need that expensive instance type or if a cheaper alternative would work. They can’t identify zombie resources or architectural inefficiencies.

Finance can track and report on costs, but they can’t optimize them. That requires technical understanding of how cloud services work, how applications are architected, and what trade-offs are available.

Why IT Needs to Own It

IT is the only organization with both the technical knowledge to understand cloud infrastructure and the authority to make changes. You understand what resources are being used, why they’re being used, and whether cheaper alternatives exist. You can identify waste, enforce standards, and implement automation to prevent future overruns.

More importantly, IT owns the relationship with cloud providers. You’re negotiating enterprise agreements, managing reserved instances, and making architectural decisions that drive long-term costs. Finance can’t do any of that effectively.

But here’s the key insight: owning cloud cost management isn’t just about cutting costs. It’s about enabling the business to spend money efficiently. When IT does FinOps right, teams can move faster because they understand the cost implications of their decisions upfront rather than getting surprised by bills later.

What FinOps Actually Looks Like in Practice

Real FinOps isn’t about cutting cloud spending to zero. It’s about creating visibility, accountability, and optimization workflows that make cloud spending predictable and aligned with business value.

First, you need tagging standards and enforcement. Every cloud resource should be tagged with the project, team, environment, and cost center that owns it. This sounds obvious, but most organizations have terrible tagging hygiene. Without good tags, you can’t attribute costs accurately, which means you can’t hold teams accountable.

Enforcing tagging requires automation. Don’t rely on developers to remember to tag resources—build it into your infrastructure-as-code templates and CI/CD pipelines. Auto-delete untagged resources after a grace period. Make good tagging the path of least resistance.

Second, you need regular cost reviews with engineering teams. I run monthly FinOps meetings with each product team where we review their cloud spending, identify anomalies, and discuss optimization opportunities. These don’t need to be long—thirty minutes is usually enough.

The key is making teams aware of their spending and giving them the tools to optimize it. Most developers have no idea how much their applications cost to run. Once they see the numbers, they’re usually eager to reduce waste.

The Low-Hanging Fruit

Every organization I’ve worked with has the same easy wins available:

Right-sizing instances. Most workloads are over-provisioned by 30-50%. Use cloud provider tools or third-party monitoring to identify underutilized instances and downsize them. This alone typically saves 15-20% of compute costs.

Shutting down non-production environments outside business hours. Dev and test environments don’t need to run 24/7. Shut them down nights and weekends. That’s an immediate 70% cost reduction for those resources.

Cleaning up zombie resources. Every organization has resources that were spun up for a project that ended months ago but never got deleted. Idle load balancers, unused storage volumes, orphaned snapshots—it all adds up. I’ve found $20,000+ per month in zombie costs at multiple companies.

Buying reserved instances or savings plans for steady-state workloads. If you’ve got workloads that run 24/7 with predictable capacity, you should be buying reserved capacity at a 30-40% discount. The ROI is immediate and guaranteed.

These optimizations don’t require sophisticated tools or deep expertise. They just require someone to take ownership and actually do the work.

The Tools You Actually Need

You don’t need expensive third-party FinOps platforms to get started. The cloud providers themselves offer surprisingly good cost management tools—AWS Cost Explorer, Azure Cost Management, GCP’s cost tools. They’re free and provide most of what you need for basic FinOps.

What you do need is someone dedicated to using these tools regularly. FinOps can’t be something IT leadership does in spare time between other priorities. It needs dedicated focus, even if it’s just 25% of one person’s role initially.

As you mature, you might want third-party tools that provide better multi-cloud visibility, more sophisticated anomaly detection, or automated optimization recommendations. But start with the basics and add tooling as you identify specific gaps.

The Cultural Challenge

The hardest part of FinOps isn’t technical—it’s cultural. You’re asking teams to care about costs when they’re focused on shipping features. You’re introducing friction (tagging requirements, approval processes) into workflows that were designed for speed.

This is where executive support matters. If the CFO and CTO are aligned that cloud cost management is a priority, teams will take it seriously. If it’s just IT complaining about costs, it’ll get ignored.

Frame FinOps as enabling business goals, not blocking them. When teams understand their cloud spending, they can make better trade-offs. They can invest in performance where it matters and cut costs where it doesn’t. That’s better for everyone than arbitrary across-the-board cuts when costs get out of control.

Making It Stick

FinOps isn’t a one-time project—it’s an ongoing operational discipline. You need regular cadences for cost reviews, continuous optimization of resources, and constant reinforcement of best practices.

The organizations getting this right treat FinOps like they treat security or operational excellence. It’s built into standard processes, measured consistently, and owned by specific people who are accountable for outcomes.

If you’re in IT leadership and you’re not already doing FinOps, start now. Your cloud costs are only going to grow as the organization adopts more cloud services. The longer you wait to get control of this, the harder and more expensive the cleanup will be.

And if you’re already doing some FinOps but it’s not getting traction, push harder on the accountability piece. Make teams own their costs. Give them visibility and tools, but also make them responsible for the outcomes. That’s when FinOps moves from IT overhead to business value.