AWS re:Invent 2025 - What's new in AWS cost Optimization (COP202)

AWS re:Invent 2025 - What's New in AWS Cost Optimization (COP202)

Introduction

• This session was presented by Rick Oaks, the product management lead for AWS optimization tools.
• The session covered new features and capabilities in AWS cost optimization, including:
  • EBS optimization and automation
  • Idle NAT gateway detection
  • A new cost efficiency metric in the Cost Optimization Hub
  • Updates to the Savings Plan purchase experience

EBS Optimization and Automation

• AWS has launched new EBS optimization capabilities in Compute Optimizer:
  • Automatic snapshot and deletion of unattached EBS volumes
  • Automated upgrades of EBS volumes (e.g. GP2 to GP3, IO1 to IO2)
• These optimizations can be applied immediately or through automated policies:
  • Policies can be scoped to entire accounts, organizations, or specific subsets
  • Policies can be configured with criteria like tags, regions, and estimated savings
  • Automation includes safety nets like rollback capabilities and health checks
• The goal is to make it easy for teams to implement "low-hanging fruit" optimizations without significant effort.

Idle NAT Gateway Detection

• Compute Optimizer now includes recommendations to detect idle NAT gateways:
  • Analyzes 30 days of usage data and checks for NAT gateways not referenced in route tables
  • Identifies NAT gateways that are likely unused backup/failover resources
• Provides visibility into these idle resources so teams can evaluate if they are still needed.

Cost Efficiency Metric in Cost Optimization Hub

• A new "cost efficiency" metric has been added to the Cost Optimization Hub:
  • Combines potential savings from all optimization recommendations
  • Divides total potential savings by total optimizable spend
  • Provides a percentage that represents the efficiency of the environment
• This metric allows teams to:
  • Set targets and track efficiency over time
  • Identify high and low efficiency areas for optimization focus
  • Demonstrate the business impact of optimization efforts to leadership

Savings Plan Purchases

• AWS has launched a new "Database Savings Plan" offering:
  • Provides up to 35% discount on a 1-year commitment for database services like RDS, Aurora, DynamoDB
  • Can be purchased through the Savings Plan Recommendations and Analyzer tools
  • Automatically applies to eligible usage without requiring architectural changes
• The tools allow analysis of historical usage to determine the optimal Savings Plan purchase amount:
  • Identifies the maximum savings waterline based on past usage patterns
  • Provides flexibility to customize the purchase amount
  • Tracks coverage and utilization of the Savings Plan over time

Customer Story: Delta Airlines

• Bonnie Fernall from Delta Airlines shared their journey with Compute Optimizer and cost optimization:
  • Established a central dashboard to provide visibility and ownership of optimization opportunities
  • Implemented automation for low-risk optimizations like unattached EBS volume cleanup and EBS volume upgrades
  • Leveraged the new cost efficiency metric to track progress and drive alignment with leadership
  • Took a phased approach, starting with quick wins before tackling more complex optimizations like EC2 right-sizing
  • Emphasized communication and change management to build trust and adoption across the organization
  • Achieved 27% monthly cost savings through their optimization efforts

Key Takeaways

• AWS continues to invest in cost optimization tools and capabilities to help customers save money:
  • New automation features make it easier to implement "low-hanging fruit" optimizations
  • Metrics like cost efficiency provide a holistic view of optimization performance
  • Expanded Savings Plan options, including for database services, offer more flexibility
• Successful cost optimization requires a combination of tooling, processes, and organizational change management:
  • Visibility, ownership, and communication are critical to driving adoption and results
  • Automation can help scale optimization efforts, but should be implemented thoughtfully
  • A phased, iterative approach allows teams to build trust and momentum over time