FinOps in 2026: Controlling Cloud Costs

Cloud spending is out of control - and most organisations know it. According to Harness's FinOps in Focus 2025 report, enterprises will waste an estimated $44.5 billion on underutilised cloud infrastructure in 2025 alone - approximately 21% of total cloud spend. The culprit? A fundamental disconnect between those who provision resources and those who pay for them.

FinOps - the practice of bringing financial accountability to cloud spending - has emerged as the solution. But implementing it effectively requires more than installing a cost dashboard. Building on the strategic IT management approaches reshaping technology leadership, this guide provides a practical framework for making cloud costs a competitive advantage rather than a budget black hole.

Why Cloud Costs Spiral Out of Control

The cloud promised cost efficiency. Pay for what you use. Scale up and down as needed. No more capital expenditure on hardware that sits idle. The reality has proven more complex.

Developer velocity trumps cost awareness. Engineers are measured on delivery speed, not cost efficiency. When spinning up a new instance takes seconds, the default behaviour is to over-provision "just in case" rather than right-size for actual requirements.

Visibility comes too late. Most organisations discover cost overruns weeks after they occur, when the monthly bill arrives. By then, the waste has already happened. The Harness research found that enterprises take an average of 31 days to identify and eliminate cloud waste - effectively guaranteeing a full month of unnecessary spending before any action is taken.

Complexity obscures accountability. Modern cloud bills can run to thousands of line items across multiple services, regions, and accounts. Without clear attribution, no one owns the problem. According to the FinOps Foundation's 2024 research, reducing waste has become the top priority for practitioners - yet many organisations lack the visibility to act effectively.

Commitment-based discounts require expertise. Reserved instances and savings plans can reduce costs by 30-70%, but they require accurate forecasting. The Harness study found that 55% of developers say purchasing commitments are ultimately based on guesswork, and 58% of organisations do not use reserved instances or savings plans at all.

What FinOps Actually Means

The FinOps Foundation defines FinOps as "an operational framework and cultural practice which maximises the business value of cloud and technology, enables timely data-driven decision making, and creates financial accountability through collaboration between engineering, finance, and business teams."

The key word is "collaboration." FinOps is not a tool you buy or a team you hire. It is a cultural shift that brings together three traditionally siloed groups:

• Engineering - who provision and consume resources
• Finance - who pay the bills and forecast budgets
• Business - who define the value that technology should deliver

When these groups work together with shared visibility and aligned incentives, cloud spending transforms from an uncontrolled cost centre into a strategic lever.

The Crawl, Walk, Run Maturity Model

FinOps maturity develops in stages. Trying to implement advanced optimisation before establishing basics leads to frustration and failure.

Crawl phase. Establish visibility. Tag resources consistently. Attribute costs to teams and projects. Create basic reporting that shows who is spending what. At this stage, awareness itself drives behaviour change - teams that can see their costs often reduce them without any other intervention.

Walk phase. Optimise actively. Right-size over-provisioned resources. Implement reserved instance strategies. Set budgets and alerts. Automate routine cost-saving actions. Engineering teams begin incorporating cost into technical decisions.

Run phase. Make cost a first-class engineering concern. Integrate cost data into CI/CD pipelines. Automate commitment purchases based on usage patterns. Predict and prevent cost anomalies before they occur. Cost efficiency becomes part of the development culture, not an afterthought.

Most organisations are still crawling. According to the FinOps Foundation research, human approval remains central to most automation efforts, with full automation still rare.

The 2026 FinOps Imperative

Several trends make FinOps adoption urgent for 2026:

AI and ML costs are accelerating. For organisations spending over $100 million annually on cloud, 45% report that AI and ML costs are already impacting their practices. Smaller organisations expect this to become significant soon. As I explored in the 2026 IT trends analysis, AI adoption is accelerating - and so are the infrastructure costs that support it.

Kubernetes complexity compounds waste. Container orchestration enables efficiency in theory, but often creates waste in practice. Research indicates that 82% of Kubernetes workloads are over-provisioned, with 65% using less than half of their requested CPU and memory. The average production cluster runs at just 10% CPU utilisation.

Multi-cloud strategies multiply complexity. Organisations increasingly use multiple cloud providers for resilience, compliance, or best-of-breed services. Each provider has different pricing models, discount mechanisms, and billing structures. Without unified visibility, optimisation becomes nearly impossible.

Economic pressure demands accountability. The era of "grow at all costs" is over. Boards and executives expect technology investments to demonstrate clear returns. Cloud spending that cannot be tied to business outcomes faces increasing scrutiny.

A Practical Framework for FinOps Implementation

Moving from cloud cost chaos to controlled optimisation requires systematic effort. The following framework provides a roadmap for IT leaders.

Phase 1: Establish Visibility (Months 1-2)

Before optimising anything, you must understand what you have.

Implement consistent tagging. Define a tagging taxonomy that maps resources to business dimensions - cost centres, products, environments, teams. Enforce tagging through policy, not just guidelines. Untagged resources should generate alerts.

Consolidate billing data. Aggregate costs across accounts, regions, and - if multi-cloud - providers. Native cloud tools provide account-level visibility, but cross-cloud views require third-party solutions or custom development.

Create attribution reports. Build dashboards that answer: What are we spending? Where is it going? Who is responsible? These reports should update at least daily - monthly billing cycles are too slow for effective management.

Identify quick wins. Even basic visibility reveals obvious waste - idle resources, oversized instances, forgotten development environments. Document these for Phase 2.

Phase 2: Optimise Actively (Months 3-6)

With visibility established, begin systematic optimisation.

Right-size over-provisioned resources. The Harness research found that 61% of organisations do not right-size instances. Start with the largest workloads - a 10% reduction on your biggest spend has more impact than eliminating tiny inefficiencies.

Implement commitment strategies. Reserved instances and savings plans require forecasting, but the savings are substantial. Begin with stable, predictable workloads where commitment risk is low. Build confidence before tackling variable workloads.

Automate routine savings. Schedule non-production environments to shut down outside working hours. Automatically delete orphaned resources. Set up alerts for cost anomalies. The goal is to make cost efficiency the default, not a manual effort.

Establish governance. Create policies for resource provisioning - required tags, size limits, approval workflows for large deployments. Prevention is more effective than remediation.

Phase 3: Embed in Culture (Months 6-12)

Sustainable FinOps requires cultural change, not just tooling.

Give engineers cost visibility. The Harness research found that fewer than half of developers have real-time visibility into cloud waste. Integrate cost data into the tools engineers already use - IDEs, dashboards, CI/CD pipelines.

Align incentives. If engineers are measured only on delivery speed, they will optimise for speed. Include cost efficiency in team metrics and performance discussions. Celebrate cost savings alongside feature delivery.

Create feedback loops. When teams reduce costs, show them the impact. When costs spike, investigate promptly and share learnings. Make cost management visible and valued, not hidden and punitive.

Invest in skills. FinOps is a discipline that requires training. The FinOps Foundation offers certifications and resources. Build internal expertise rather than relying solely on external consultants.

Common FinOps Mistakes to Avoid

Treating it as a one-time project. Cloud environments change constantly. New services, new workloads, new pricing models. FinOps is an ongoing practice, not a cleanup exercise.

Focusing only on cost reduction. FinOps is about value, not just savings. Sometimes spending more on cloud is the right decision - if it enables faster delivery, better reliability, or new capabilities. The goal is informed decision-making, not arbitrary cuts.

Centralising all responsibility. A central FinOps team can provide expertise and tooling, but cannot manage costs across a large organisation alone. Engineering teams must own their costs, with central support and governance.

Ignoring the 52% disconnect. The Harness research found that 52% of engineering leaders cite the disconnect between FinOps and development teams as a root cause of waste. Tools alone cannot fix this - it requires deliberate effort to bridge organisational silos.

Waiting for perfect data. Tagging will never be 100% complete. Forecasts will never be perfectly accurate. Start with what you have and improve iteratively. Perfect is the enemy of good enough.

Quick Reference: FinOps Health Check

Use these questions to assess your organisation's FinOps maturity:

Visibility

• Can you attribute 80%+ of cloud costs to specific teams or products?
• Do cost reports update at least daily?
• Can engineers see their team's cloud costs without asking finance?

Optimisation

• Do you actively right-size resources based on utilisation data?
• Are you using commitment discounts (reserved instances, savings plans)?
• Do non-production environments shut down automatically outside working hours?

Culture

• Is cost efficiency part of engineering team metrics?
• Do developers consider cost when making architectural decisions?
• Is there a defined owner for cloud cost management?

If you answered "no" to more than three questions, FinOps should be a 2026 priority.

Building Cloud Cost Competence

The organisations that master FinOps in 2026 will gain significant advantages - not just in reduced spending, but in agility and strategic clarity. When you understand what cloud resources actually cost and what value they deliver, technology investment decisions become easier and more defensible. For a practical, tactical guide to implementing these principles on AWS specifically, see my piece on how I cut our AWS bill by 30%.

My IT management services help organisations implement FinOps practices - from initial visibility and tagging strategies through cultural change and ongoing optimisation. Whether you are starting from scratch or looking to advance from crawl to run, a structured approach accelerates results.

Get in touch to discuss how FinOps can transform cloud spending from a budget concern into a strategic advantage for your organisation.