Found Millions in Wasted Cloud Spend

Here's a Fortune 500 Story

Presented by

Want to appear here? Talk with us

TOGETHER WITH FINOUT
How Just Eat Takeaway and SiriusXM Scaled FinOps with Confidence

Scaling FinOps shouldn’t be this hard

As teams grow, cloud costs get harder to track and reporting takes more time. Here’s how others solved it:

  • Just Eat Takeaway.com grew from 40 to 700+ engineers with 20× higher FinOps adoption.

  • SiriusXM unified visibility across AWS and internal platforms, cutting reporting time by 50% and improving team efficiency by 2–3×.

Heading to re:Invent?
Come see how we make cloud cost management simple, not stressful.

 

COST OPTIMIZATION
How I Found Millions in Wasted AWS Spend

A cloud expert walked into a Fortune 500 company expecting to find the usual waste like forgotten servers and unused storage.What he found was much worse: $1.2 million burning away each year on costs that gave the company nothing in return.

Problem One: Following Rules Without Thinking

The company was running expensive database systems 24 hours a day, 7 days a week. But developers only used them from 9am to 6pm on weekdays.

The fix was simple: turn off systems when nobody uses them. Remove expensive backups from tools that don't need them. Match the technology to what the business actually needs. This alone saved $520,000 per year.

Problem Two: Labels That Look Good But Tell You Nothing

The company had labeled 87% of their cloud resources. Their finance team loved it. Everything was organized by department and cost center.

They found a machine learning system costing $180,000 per year that only 3 customers used and brought in $12,000 in revenue. They found a dashboard costing $95,000 per year that executives looked at once every three months.

Adding labels that showed which features cost what, and which data was actually being used, saved another $315,000 per year.

Problem Three: Locked Into Yesterday's Technology

The company had committed to paying for specific types of cloud services for the next few years. This gave them a discount and looked smart on paper.

The problem? Their commitments were for the old systems. They were paying $890,000 per year for technology they weren't using anymore. Switching to shorter commitments that could change as their technology changed saved $180,000 per year and let them move faster.

The companies that spend smart don't just hunt for unused servers - they make sure every dollar spent connects to real business value.

 

CLOUD PROVIDERS
Top 10 FinOps Updates of Cloud Hyperscalers

AWS

CUR 2.0 now includes hourly granularity for EC2 On-Demand Capacity Reservations and ML capacity blocks. Track reservation coverage at resource-level detail, calculate utilization, and identify unused capacity to reclaim and reduce EC2 spend.

SageMaker Unified Studio lets admins configure custom tags automatically applied to all project resources. Ensures consistent tagging on notebooks and endpoints for reliable cost allocation without manual overhead.

S3 Tables now support tags for attribute-based access control and cost allocation. Apply cost center or environment tags to integrate tables into FinOps pipelines for accurate chargeback tracking.

Microsoft Azure

Azure Ultra Disk flexible provisioning model is GA, decoupling capacity, IOPS and throughput with GiB granularity. Delivers up to 50% cost reductions for small disks and finer control to avoid over-provisioning mission-critical workloads.

Object Replication metrics are now GA for Blob storage. Monitor pending operations and bytes to troubleshoot delays, optimize replication policies, and avoid unnecessary storage costs.

Google Cloud

Autoclass now supports buckets with hierarchical namespace for automatic storage tiering. Extends lifecycle-based cost optimization to HNS buckets, reducing storage spend without manual rules.

GKE logging agent processes logs up to 2× faster using fewer node resources. Reduces observability overhead, frees node capacity, and lowers costs in high-throughput logging environments.

 

FINOPS EVENTS
The Hybrid FinOps Advantage: Watch

Discover how FinOps 2.0 strategies deliver comprehensive optimization across your entire technology portfolio.

You'll Learn How To:

Achieve total cost visibility across data centers, multi-cloud, SaaS, and AI infrastructure

Optimize the complete technology stack with unified intelligence and automation

Break down silos between FinOps, ITAM, procurement, and engineering teams

Speakers

Jeremy Chaplin, Gerhard Behr & Victor Garcia

 

RESEARCH
Cloud FinOps Market Forecast to 2032

The Cloud FinOps market helps companies control their cloud spending through smart tools and services. Think of it like a budget tracker for cloud services, but much more powerful. The market was worth $12.2 billion in 2024 and will grow to $27.7 billion by 2032. That's a growth rate of 10.8% each year.

  • Big tech companies like AWS, Microsoft, Google, IBM, and Oracle lead this space. They build tools that watch cloud costs in real time, find wasted money, and help teams make better choices about their cloud use.

  • North America uses these tools the most, holding 38% of the market. Europe comes next with 27%, and Asia-Pacific has 22%. Companies in these places spend a lot on cloud services and need better ways to track every dollar.

  • Most buyers want ready-made software solutions rather than just advice. About 68% of the market is software tools, while 32% is services like training and support. Public cloud tools are more popular than private ones, taking 61% of the market.

  • Large companies buy most of these tools because they run thousands of cloud systems and need tight control. They make up 67% of buyers. Smaller businesses account for 33% and are growing fast as simpler, cheaper tools become available.

Recent updates show the market moving fast. IBM added new tools for AI workloads and container cost tracking. Amazon Web Services added AI helpers to find cost savings automatically. These companies keep adding features to stay ahead.

 

📺️ VIDEO
FinOps Weekly Summit Talks Now Available

More and more FinOps Weekly Summit Sessions are getting released. We’ll complete the release next week but you can already check them.

 

WASTE MANAGEMENT
Finding Waste Before It Finds You

This is Part 2 of a guide about finding and stopping waste in cloud computing costs. It continues the story of the FinOps Lab, where Fox and Raccoon teach teams how to manage cloud spending better.

The Lab uses three ways to look at data. First, they study what happened in the past. Second, they figure out why costs changed. Third, they predict what might happen next with a range of possible outcomes.

The real work starts after you make a report. The Lab learned that good cloud cost management needs constant attention through what they call operational loops.A loop has five steps. First, you watch for problems. Second, you figure out what the problem means. Third, you decide what to do. Fourth, you take action. Fifth, you check if it worked.

The Lab runs loops at different speeds. Daily loops catch small problems fast. Weekly loops look at bigger patterns. Monthly loops check if the overall plan is working.

The Lab follows one rule: only add a tag if it changes how you make money decisions. If a tag doesn't help you understand costs better, it just adds confusion.

The FinOps Lab started as simple cost reports. Over time it became a way for the whole company to think about cloud spending.

 

CASE STUDIES
BMW Cloud FinOps Transformed by GenAI on AWS

BMW Group manages over 10,000 AWS cloud accounts as of 2025, up from 4,500 just two years ago. With that many accounts spread across 30 production sites in 15 countries, keeping track of cloud costs became a real challenge.

To solve this problem, BMW Group built CLEAI, a smart assistant powered by AI that helps more than 5,000 BMW employees understand and reduce their cloud spending. The tool speeds up cloud cost analysis and fixes by up to 50%.

CLEAI works through simple conversation. Instead of clicking through complex dashboards, users just ask questions in plain language. For example, if someone needs to set up a database service, CLEAI can write the complete setup code following BMW's standards, including proper labels and cost-saving settings.

CLEAI connects three BMW systems together. The first is CLEA, which provides the user interface. The second is CLEA API, which stores all the cost and account data. The third is ICCA, which runs the AI processing.

When someone asks a question, the system first figures out what kind of help they need. Is it a data question? A best practice question? A code writing request? Based on that, it picks the right processing path.

 

🎖️ MENTION OF HONOUR
GenAI Driving Shift to Performance and Cost Accountability

A new study from Wharton Business School shows how companies are changing the way they spend money on AI tools. 72% of big companies now track how much money they make back from their AI spending. This is very different from before when teams could just try things out without proving the value.

  • Companies are spending more and more on AI cloud services like GPUs and training systems. But now the money people and buyers want strict rules about budgets and smart buying plans.

  • The study found that AI waste happens in surprising ways. Bad prompts waste money. Workers who don't know how to use AI well waste money. AI agents that get stuck in loops and keep running waste money. Systems that pull too much data or run too many things at once waste money.

  • The problems causing extra costs are changing too. It's not the usual cloud storage or servers anymore. Now it's AI agents, data storage systems, and programs that keep repeating themselves that cause surprise bills.

  • 30% of AI budgets now go to building things inside the company. 11% of companies are taking money away from old IT systems and HR to pay for AI instead.

Here's the twist: the real problem isn't the AI models themselves. It's the people. Companies need to train workers better and set up good rules for using AI. That's where the real savings will come from. Companies are moving from testing AI to making it work efficiently and prove its worth.

 

PROFESSIONAL SPOTLIGHT
Phillip Karg

BMW Quality FinOps

Phillip is doing great things at BMW such as the case study we shared above. Had the pleasure to talk with him and love how he drives FinOps!

 

Learn Hands On FinOps

We’ve created Learn FinOps Weekly because we believe that hands on learning is the way to make real impact with IT Budgets.

Carefully curated with the FinOps Weekly Team to align with our actionable way of doing things.

Learn Actionable FinOps, together