FinOps: master your cloud costs without sacrificing performance

"It was supposed to be cheaper”. This phrase resonates and often endures in the minds of IT decision-makers. Everyone legitimately believed that the cloud would bring flexibility, speed, and cost savings. And they weren’t wrong. On paper, it all looks perfect: pay-as-you-go, scalability, automation. But the reality is more nuanced. Resources pile up, projects accelerate, practices diverge. And the same thing happens to the bill.

The problem isn’t the cloud. It lies in how it’s used and managed. This is where FinOps comes in. Initiated by the FinOps Foundation, this discipline maximises the business value of the cloud. It doesn’t just reduce costs; it makes them understandable, manageable, and aligned with organisational priorities. The real question isn’t how much we spend, but why we spend it. Is a service expensive due to poor configuration or because it has become critical to the business?

FinOps isn’t a magic wand. It’s a continuous process that finally brings together IT, finance, and business teams. The shared objective: get the most out of the cloud without losing control.

So, how do you get started? Where do you begin? That’s what we’ll explore. FinOps is essential for regaining control over your cloud.

What does “FinOps” mean?

The term FinOps is a contraction of Finance and DevOps. It’s more than a semantic blend. FinOps refers to an operational approach that enables organisations to manage their cloud spending as accurately as possible, grounded in real-world usage.

It’s a continuous balancing act that reconciles technical performance, actual usage, and available budget. In short: spend in the right place, at the right time, for the right reasons.

Three pillars structure this approach:

• Visibility: having a clear, shared, and real-time view of consumption.
• Accountability: each team (technical, product, or business) actively manages its resources.
• Continuous optimisation: adjusting configurations, automating shutdowns, revisiting commitment models.

When properly implemented, FinOps aligns the company’s financial vision with the teams’ innovation dynamics. It doesn’t slow you down rather it channels the cloud’s power intelligently.

Why is FinOps becoming essential?

The cloud has changed the game. You pay for what you consume, not what you plan. This flexibility is valuable, but without clear governance, overspending is almost inevitable. A poorly managed cloud migration can cost more than a well-amortised on-premises infrastructure.

In many organisations, the distribution of roles complicates management. IT teams provide, business teams consume, and finance observes - often one-quarter after the event. Therefore, FinOps becomes a structuring skill for preventing bad habits becoming the norm.

Concrete examples abound:

• A pre-production cluster left running overnight to “save time in the morning”.
• Oversized instances operating at only 15% of their capacity.
• Managed services enabled for exploratory purposes and then never turned off because they slipped off the operational radar.

None of these scenarios are critical on their own but taken together, they add up.

What are the 5 key steps in a FinOps approach?

Adopting FinOps doesn’t require a radical transformation it requires a rigorous methodology.

1. Audit the current state: clarify before acting

Before optimising, you need to understand. The initial audit collects consumption data across all environments. This assessment quickly reveals blind spots: oversized resources, inactive instances, dormant storage.
A crucial practice for an effective audit is systematic resource tagging (environment, team, project, etc.). Unattributed costs are unmanaged costs. By enforcing mandatory tagging policies at resource creation (via Terraform, Helm, etc.) you restore clarity and accountability.

2. Structure a FinOps team: hybrid profiles

A FinOps process doesn’t rely on a single actor. It requires a dedicated or cross-functional team combining multiple profiles: cloud engineers, architects, finance professionals, and business stakeholders. Their shared role isn’t that of an air traffic control tower, rather a facilitation unit serving projects.

3. Define KPIs: manage by value

What isn’t measured can’t be managed. Define clear, actionable, and shared indicators:

• Cost per product or client: links technical spends to business value.
• Resource utilisation rate: identifies room for improvement.
• Proportion of optimised costs versus gross cost: measures the real impact of FinOps.

Our advice: measure value, not just spend.

An aggregated bill does not tell you much. Unit metrics tell more: cost per build, per order, per thousand requests. These measurements provide a common language connecting architecture, performance, and business objectives, enabling decisions that are not just based on intuition.

4. Establish review rituals: pace your decisions

Schedule regular sessions (weekly or monthly) to analyse KPIs and turn them into actions. Should capacity be reduced? Should a service be reconfigured? These rituals build the right habits. One concrete example: automating the shutdowns of non-critical environments (test, staging) during nights and weekends, reducing their bill by 20–40% from the first week.

5. Optimise usage and negotiate with providers

• Choose the right instance types: clouds offer various instances (on-demand, reserved, spot, etc.). Reviewing default instance types isn’t micro-optimisation, it can reduce costs by 30–60% in some cases. Best practice: size according to observed usage and let autoscaling handle variability.
• Use savings plans: reserved instances and savings plans are financial instruments, not technical levers. Gradual coverage, periodically reassessed (e.g., for stable workloads), protects better than chasing maximum discounts that dictate architecture.
• Manage storage efficiently: storage often consumes silently. Implement lifecycle policies to move rarely accessed data to cheaper tiers. Delete unattached volumes and snapshots.
• Negotiate with providers: supplier relationships are a strategic lever. For effective negotiation, rely on solid metrics: consumption history, usage forecasts.

Which tools support a FinOps approach?

Tool choice depends on your maturity and environment complexity.
 

• Native cloud provider tools: a good starting point.
  • AWS cost explorer: trend visualisation and custom reports.
  • Azure cost management: advanced budget tracking and recommendations.
  • Google cloud billing reports: detailed view and export to BigQuery.
• Specialised solutions: for multi-cloud visibility and advanced analytics.
  • Apptio Cloudability: centralised view, predictive analytics.
  • nOps: automated drift detection integrated into the DevOps cycle.
• Custom approach: some organisations build their own dashboards for maximum agility (via BigQuery, Grafana, Metabase) by collecting data through APIs.

Our advice: tools depending on governance. 

Efficiency comes less from the tool itself than from how well teams use and interpret it. A tool cannot replace the conversations it is meant to provoke. A minimal foundation (budgets by scope, anomaly alerts, enforced tagging policy) often delivers more progress than a dashboard factory.

FinOps case study: optimising Plecto’s cloud costs

Kaliop recently helped Plecto, a European scale-up specialising in team engagement, to reduce its cloud infrastructure costs by 50%. Plecto chose to migrate from a hyperscaler to Scaleway, moving toward a European cloud with a modernised, observable, and optimised architecture.

Discover the FinOps methodology applied for Plecto.

The limits of FinOps: clarity and vigilance

FinOps is a valuable approach, but it’s not a miracle solution.

• It doesn’t replace a well-thought-out architecture: no optimisation can compensate for poorly conceived infrastructure.
• Optimisation has limits: beyond a certain point, the effort to save a few percent costs more than the savings achieved.
• ROI is gradual: this is a practice to embed over time. True value is built through repetition and rigor.
• Preserve team autonomy: FinOps is a framework for accountability, not a control machine. Watching where every penny goes can stifle innovation.

FinOps: a key skill to adopt today

FinOps is neither a trend nor a one-click solution. It’s a structured, long-term methodology that transforms the relationship with the cloud. It empowers teams, creates a shared culture, and gives decision-makers the tools to secure technology investments without stifling innovation.

For CTOs, infrastructure managers, or architects, it’s a powerful lever for more detailed, more responsive, and strategic management.

Is your cloud bill opaque? Kaliop turns complexity into clarity. Contact our experts for an initial FinOps maturity assessment and discover how well-managed cloud spending can become your strongest ally in innovation.