Presented by

Want to appear here? Talk with us

Cloud environments are becoming more shared and AI-driven, but many cost allocation models were built for a simpler world.

When that foundation breaks, FinOps stops being actionable.

Join Nicole Boyd from Capgemini and Elly Rauch from Finout to learn how leading teams move from basic cost allocation to TCO that supports real decision-making at scale.

Tuesday, February 10, 12 PM EST | 9 AM PST

👉 Join us to start the year with a clearer path to decision ready TCO

Azure Savings Plans confuse smart people because they work differently than normal discounts. Shannon from Azure Musings explains what actually happens: A Savings Plan replaces your expensive pay-as-you-go compute charges with cheaper hourly pricing. You commit to spending a fixed amount per hour for one or three years, and Azure automatically applies better rates to your eligible compute up to that commitment amount.

The savings do not appear as money subtracted from your bill. The savings are the higher costs you never get charged in the first place. Here is how the hourly window works:

Azure processes your Savings Plan benefit every single hour. Each hour stands alone. If you use $100 of eligible compute in an hour and your commitment is $80, Azure prices $80 worth at the cheaper Savings Plan rate and charges the remaining $20 at regular pay-as-you-go rates. If you only use $60 of compute in an hour but your commitment is $80, you lose that extra $20 of benefit. It does not roll over to the next hour.

Your monthly compute bill is $100,000 at pay-as-you-go rates. That is about $137 per hour if you run 24/7. Savings Plan pricing is typically 30 percent cheaper for the same compute. So that $137 per hour becomes about $96 per hour under the plan. You buy a Savings Plan with a $96 per hour commitment. Over a month, that costs you about $70,000. Your avoided cost is $30,000 per month. You never see that $30,000 as a refund or credit. You just stop getting charged the expensive way.

Most teams should start with partial coverage, not full coverage. Pick an hourly commitment you know you will use most hours. Let the rest stay pay-as-you-go. Good coverage is not 100 percent. Good coverage is a commitment that matches your boring, predictable compute so you stop overpaying for usage you know will be there every month.

Learn how to manage your Azure Savings Plans in our Azure FinOps Course

Read More

CUDOS (v5.7.1) and CORA (v0.0.11) updates fix reporting for Shield Advanced and multi-period savings, improving data reliability for financial stakeholders.

Amazon Bedrock now supports a 1-hour prompt caching TTL for Anthropic models, significantly cutting inference spend for repeated or agentic workflows.

RDS Blue/Green Deployments reduced switchover downtime to ~5 seconds, lowering the operational risk and cost of maintenance windows.

Broadened regional availability of C8i (Intel Xeon 6) and Graviton4 (C8gn/R8g) instances offers up to 40% price-performance gains for compute and database modernization.

Read All AWS Updates

Vertex AI Search introduces flexible pricing, allowing teams to switch between consumption-based and monthly subscription models to stabilize budget forecasting.

Cloud SQL Fast Clone (GA) reduces storage spend and provisioning time for MySQL and PostgreSQL by replacing full volume copies with near-instant, cheaper clones for dev/test environments.

The N4A (Axion Arm) family is now GA, providing a high-efficiency compute option for teams looking to optimize cost-per-core on Arm-compatible workloads.

Read All GCP Updates

No Updates (again …)

Read All Azure Updates

Don’t just find savings. Get them fixed.

Topics we’ll explore:

Date: February 26th, 6PM CET / 12PM EST

FinOps expert Frank Contrepois shares how to bridge the gap between engineering and the boardroom by focusing on context-driven data and strategic constraints.

Listen on Apple Podcast

Listen on Spotify

Here are six automation strategies that can cut AWS bills by 30-50% without needing constant human oversight.

Automated EC2 Rightsizing: The fix is to set up a Lambda function that checks CloudWatch metrics every week. It looks at CPU and memory usage over 14 days. When it finds instances using less than 20% CPU on average and less than 40% at peak times, it sends an alert. This simple automation can save 15-30% on EC2 costs.

S3 Storage Lifecycle Policies: The solution is to set up automatic lifecycle rules for every S3 bucket. Move files to cheaper storage tiers based on age: Intelligent Tiering at 30 days, Glacier at 90 days, Deep Archive at 180 days. A monthly Lambda function can check that every bucket has these policies in place.This approach saves 30-50% on S3 storage costs.

Cost Anomaly Detection: AWS Cost Anomaly Detection can catch these problems within 24 hours. Set it to alert when daily costs jump by $100 or more. Route those alerts to a Lambda function that automatically tags suspicious resources for review. For extra protection, the Lambda can stop newly launched instances that trigger cost spikes, with safeguards to protect production systems.

Spot Instance Automation: Spot Instances cost 70% less than regular On-Demand instances. But managing them manually is too much work for most teams. The answer is Auto Scaling Groups with a mixed approach: 20% On-Demand instances for baseline capacity, 80% Spot instances for cost savings. The savings are huge: 50-70% for workloads that fit the model.

Reserved Instance Monitoring: Set up a Lambda function that runs monthly to check Reserved Instance utilization through the Cost Explorer API. If utilization drops below 70%, it sends an alert to review the portfolio.

Tagging Enforcement: Use EventBridge to trigger a Lambda function whenever someone creates a new resource. The Lambda checks for required tags like Environment, Team, CostCenter, and Project. If tags are missing, it stops the resource and sends an alert.

The fastest way to see results is to start with S3 lifecycle policies in week one, then add EC2 rightsizing in week two. These two changes deliver the quickest return. Then add tagging enforcement, cost anomaly detection, Reserved Instance monitoring, and Spot instance strategies over the following four weeks.

Build the automation once and let it run.

Read More

A cloud infrastructure expert found $50,000 in waste across two company audits and says up to 42% of cloud spending is pure waste.

A SaaS company was paying $21,000 per month for AWS. Everything worked great. No outages. Good uptime. But the audit found massive waste just in their EC2 service:

Old snapshots from 2019-2021 that nobody touched: $960 per month. Servers running at 12-15% capacity but sized for full power: $329 per month wasted on unused capacity. Reserved Instances for a dead project while active servers paid full price: $1,500 per month lost. Stopped servers still paying for storage: $34 per month. Orphaned storage volumes sitting unused for 8-14 months: $244 per month

Total found: $3,787 per month or $45,444 per year. That's 18% of their entire infrastructure budget. The kicker: Their monitoring showed everything was green. The waste was invisible in their dashboards.

A mid-size company spending $48,000 per year on AWS asked for a quick cost review. The obvious waste came fast: 26 unused Elastic IPs costing $1,092 per year. Services deleted three years ago but still paying for IP addresses. Then came forgotten Lightsail instances with zero activity for months. Database backups from deleted projects. Total: $400 per month or $4,800 per year. But the real problem was security: 126 IAM users when only 20 were actually active. Some passwords were 10 years old. Some employees had been gone for 8 years but still had access. Multi-factor authentication was optional.

The hardest part wasn't finding the waste. That took one hour. Getting approval to delete things took one week.

Read More

Dan Watson just shared something that most FinOps teams miss completely. You can have the best cloud cost tools in the world, but they won't matter if your team is afraid to speak up. Watson calls this psychological safety, and it's the foundation that makes everything else work. Here's what it means in practice.

When teams feel safe, engineers admit they over-provisioned resources instead of hiding them. Finance people challenge bad forecasts instead of staying quiet to avoid conflict. Product managers raise cost concerns early instead of waiting until budgets explode.

When teams don't feel safe, the opposite happens. People hide problems, avoid hard conversations, and stick with wasteful practices because change feels risky.

Loss aversion makes people afraid to try cost savings that might fail.

Status quo bias keeps teams doing things the old way to stay safe.

These biases get worse when people fear blame or ridicule.

So how do you build this safety in your FinOps practice?

Stop treating cost optimization like a witch hunt for mistakes.

Frame it as learning and improvement instead.

Have leaders admit their own blind spots first.

When the boss says "I don't understand this either," everyone else feels permission to ask questions.

Run blameless reviews after cost incidents.

Focus on what broke in the system, not who messed up.

Bring finance, engineering, and product people together regularly.

Trust builds when people work side by side, not when they throw reports over the wall.

Reward the people who surface problems early, even when the news is bad.

Read More

FinOps & SAM Expert Author

One of the greatest contributors to FinOps Weekly Blog. Richa has been doing amazing content and it’s a pleasure to have her in the blog team!

Connect with the Professional of the Week

February 9th Kickoff. Enrollment closing

Alfonso is creating a masterpiece.

The biggest educational we’ve done till date.

Learn all you need to optimize resources in Azure.

199€ Super Early Bird – Ends December 21st (CLOSED)

299€ Early Bird: Opens January 2nd – Ends Today

⏩ 499€ Launch Offer: Opens Today

699€ Starting Offer: Opens XXX

999€ Final Price