Cloud costs out of control? How to take charge with FinOps

FinOps (Financial Operations) is a practice that helps businesses take control of cloud costs through collaboration between finance, technology, and business teams. The goal is not to cut costs blindly, but to maximise the business value of every krone spent in the cloud — with full visibility, data-driven decisions, and cross-functional accountability.

Yet many Norwegian businesses experience a different reality: unpredictable invoices, resources running around the clock without being used, and a steadily growing cloud bill that nobody fully understands. According to surveys from Flexera and Gartner, roughly half of IT leaders experience unexpected fluctuations in cloud costs — and Norwegian businesses are no exception.

The problem is rarely the cloud itself. The problem is a lack of governance. This is where FinOps enters the picture.

What is FinOps?

FinOps — short for Financial Operations — is an operational practice that brings finance, technology, and business together to make data-driven decisions about cloud spending. It is not about cutting costs blindly, but about maximizing the business value of every krone spent in the cloud.

The FinOps Foundation, which is part of The Linux Foundation, defines FinOps as "an evolving cloud financial management discipline and cultural practice that enables organizations to get maximum business value by helping engineering, finance, and business teams collaborate on data-driven spending decisions."

The core of FinOps is simple: everyone who uses cloud resources should understand what it costs, and everyone who pays for cloud resources should understand what they get in return.

The three phases of FinOps

The FinOps framework operates in three iterative phases. It is not a linear process — organizations move continuously between phases as maturity increases.

Phase 1: Inform (Visibility)

The first step is creating visibility. You cannot optimize what you do not measure. In this phase, the focus is on:

• Mapping all cloud spending and allocating costs to the right teams, projects, and services
• Establishing a consistent tagging strategy for all resources
• Setting up dashboards and reporting that provide real-time insight
• Understanding cost drivers and consumption patterns

In Azure, this means adopting Azure Cost Management, setting up cost budgets with alerts, and establishing a resource group structure that reflects the organizational structure.

Phase 2: Optimize

With visibility in place, you can begin optimizing. Here you identify and implement concrete measures to reduce unnecessary spending:

• Remove unused resources (orphaned disks, empty resource groups, inactive VMs)
• Right-size over-provisioned services
• Implement reserved instances and savings plans for predictable workloads
• Use Azure Spot VMs for fault-tolerant jobs
• Enable auto-scaling to match capacity with actual demand

Phase 3: Operate

The final phase is about making cost optimization a sustained practice — not a one-time project:

• Define policies and automated rules for cost management
• Establish regular reviews of cloud spending with cross-functional teams
• Set KPIs for unit cost (for example, cost per transaction or cost per user)
• Build cost awareness into development processes and architecture decisions

Common mistakes that blow the cloud budget

Before we look at the solutions in detail, it is useful to recognize the most common mistakes. We see these repeatedly in Norwegian businesses:

Over-provisioning is the most frequent culprit. Developers often choose oversized VM types "just in case," and nobody scales them down afterward. A Standard_D8s_v5 VM in Norway East costs approximately NOK 2,800 per month — while a Standard_D4s_v5 costs half and is often more than sufficient.

Unused resources are equally common. Test environments set up six months ago that were never shut down. Disks left behind after deleted VMs. Load balancers with no backends. These "ghost resources" typically account for 20–30 percent of the cloud bill.

Lack of tagging makes it impossible to understand who is using what. Without consistent tagging, you cannot allocate costs, and without cost allocation, nobody can be held accountable.

No reserved instances means you are paying the full pay-as-you-go price for workloads running 24/7/365. Azure Reserved Instances provide up to 72 percent discount for a three-year commitment.

Wrong storage tier is easy to overlook. Data that is never accessed but stored on premium SSD instead of Archive storage costs many times more than necessary.

Azure-specific cost optimization

Since most Norwegian businesses we work with use Microsoft Azure, and Azure has its own data center region in Norway (Norway East and Norway West), let us look more closely at the most important optimization opportunities:

Strategy	Description	Typical savings	Best suited for
Reserved Instances (RI)	1- or 3-year commitment on VMs	30–72%	Stable, predictable workloads
Azure Savings Plans	Flexible commitment across VM types	20–65%	Variable workloads with steady consumption
Azure Spot VMs	Unused capacity at steep discount	60–90%	Batch jobs, CI/CD, fault-tolerant tasks
Auto-scaling	Automatic scale up/down based on demand	20–40%	Web applications with variable traffic
Right-sizing	Downscale over-provisioned resources	15–40%	VMs, databases, App Service plans
Dev/test pricing	Reduced rates for non-production environments	20–55%	Development and test environments
Storage tiering	Move infrequently used data to cheaper tiers	50–80%	Archive data, logs, backups
Azure Hybrid Benefit	Use existing Windows/SQL licenses	40–85%	Businesses with Software Assurance

Practical cost examples — Norway East

To make this concrete, here are some real price comparisons for the Azure Norway East region (approximate prices in NOK, March 2026):

Example 1: Web application (VM)

• Standard_D4s_v5 (4 vCPU, 16 GB RAM), pay-as-you-go: ~NOK 1,400/month
• Same with 3-year RI: ~NOK 490/month (65% savings)
• Spot VM: ~NOK 210/month (85% savings, but may be interrupted)

Example 2: SQL Database

• Azure SQL Database, General Purpose, 8 vCores, pay-as-you-go: ~NOK 8,500/month
• Same with 3-year RI: ~NOK 3,400/month (60% savings)

Example 3: Storage (1 TB)

• Premium SSD: ~NOK 1,300/month
• Standard SSD: ~NOK 500/month
• Cool Blob Storage: ~NOK 100/month
• Archive Blob Storage: ~NOK 20/month

The differences are enormous. A business running ten VMs without reserved instances is potentially paying NOK 100,000 more per year than necessary.

The FinOps maturity model: Crawl, Walk, Run

The FinOps Foundation uses a maturity model with three stages. This helps organizations assess where they stand and what the next step should be:

Stage	Label	Characteristics	Typical actions
Crawl	Crawl stage	Limited visibility, reactive cost management, few or no tags	Establish basic tagging, set up cost reports, identify the largest cost items
Walk	Walk stage	Consistent tagging, regular reporting, some optimizations implemented	Implement RI/Savings Plans, automate resource shutdown, establish budgets with alerts
Run	Run stage	Proactive and automated cost management, FinOps integrated into culture and processes	Real-time anomaly detection, automated right-sizing, FinOps KPIs in management reporting

Most Norwegian businesses we encounter are at the Crawl or early Walk stage. That is perfectly fine — what matters is getting started and moving in the right direction.

Tagging strategy and cost allocation

An effective tagging strategy is the very foundation of FinOps. Without tags, cost data is just noise. Here is a minimum set of tags we recommend for all Azure resources:

• Environment: prod, staging, dev, test
• Owner: Responsible team or person
• Project: Project name or code
• CostCenter: Cost center for internal billing
• Application: Application name
• ManagedBy: Terraform, Bicep, manual, etc.

Use Azure Policy to require mandatory tags when creating resources. It is far easier to require tags from day one than to tag thousands of existing resources after the fact.

FinOps culture: Cross-functional collaboration

FinOps is not just a technical initiative — it is a cultural change. It requires collaboration between three key groups:

Finance contributes budget expertise, forecasts, and financial reporting. They need understandable cost data, not technical jargon.

Engineers and developers make the daily decisions that impact costs — choice of VM size, architecture decisions, and whether the test environment is shut down after working hours.

Business leadership sets priorities and decides what is "worth it." A cloud resource that costs NOK 50,000 per month but generates NOK 500,000 in revenue is a good investment.

The key is to push cost responsibility down to the teams that actually use the resources — what the FinOps Foundation calls "decentralized decision-making with centralized visibility." Everyone sees the costs. Those who consume make the decisions.

How we can help

At UNOS SOFTWARE AS, we have broad experience with cloud infrastructure and cost optimization for Norwegian businesses. Through our cloud and infrastructure service, we help with:

• FinOps assessment — we analyze your current cloud spending, identify optimization opportunities, and create a prioritized action plan
• Azure optimization — we implement reserved instances, auto-scaling, right-sizing, and storage tiering
• Tagging and governance — we establish tagging strategies and Azure Policy rules that ensure consistent cost management
• Ongoing consulting — through our technical consulting service, we assist with continuous FinOps practice and optimization

Your cloud bill does not have to be unpredictable. With the right strategy and tools, Norwegian businesses can cut cloud costs by 30–50 percent — without sacrificing performance or flexibility.

Ready to take control of your cloud costs? Get in touch for a no-obligation review of your cloud spending.

---

Sources and further reading

• FinOps Foundation (2025). "FinOps Framework." finops.org
• Flexera (2025). "State of the Cloud Report." flexera.com
• Gartner (2025). "Cloud Cost Optimization: Best Practices." gartner.com
• Microsoft (2026). "Azure Cost Management and Billing documentation." learn.microsoft.com
• Microsoft (2026). "Azure Pricing Calculator — Norway East." azure.microsoft.com
• Dataforeningen (2025). "Cloud maturity in Norwegian businesses." dataforeningen.no