The €50K AWS Cost Optimization Checklist

27 Quick Wins to Cut Your Cloud Bill in Half

By Carlos Infantes Lead SRE, United Nations | Ex-Amazon Infrastructure Engineer

Introduction: The €200K Wake-Up Call

I’ve reviewed hundreds of AWS architectures over 25 years. The pattern is always the same: startups post-Series A scaling fast, AWS bills growing faster.

Last quarter, I audited 7 Series A-B companies. Average AWS waste: €28,400 per year. Total across all seven: €198,800 annually - money literally evaporating into overprovisioned instances, forgotten volumes, and misconfigured storage.

The tragedy? 80% of these savings took less than 2 hours of work.

This checklist represents the collective €5M+ I’ve helped clients save. It’s organized by implementation effort and risk, so you can start with the no-brainer wins and build momentum.

Use this in 30 minutes. See results in 24 hours.

How to Use This Checklist

  1. Print or save this PDF - Keep it visible during implementation
  2. Start with Category 1 - Biggest impact, lowest risk
  3. Check boxes as you implement - Track your progress
  4. Calculate your savings - Most items show immediate cost reduction
  5. Share your results - Tag @thewisecto on LinkedIn or email carlos@thewisecto.com

Priority System:

  • 🟢 Green = Week 1 (No-Risk, Immediate Wins)
  • 🟡 Yellow = Week 2 (Low-Risk, High Impact)
  • 🟠 Orange = Month 1 (Requires Planning)

CATEGORY 1: Compute Quick Wins

Potential Savings: €10,000 - €20,000/year

🟢 Week 1: No-Risk Wins

  • 1. Right-size EC2 instances using AWS Compute Optimizer

    • Savings: €5K-10K/year
    • Time: 30 minutes
    • Navigate to Compute Optimizer → Review recommendations → Apply during maintenance window
    • Common finding: t3.large running at 12% CPU → downgrade to t3.small (60% cost reduction)

  • 2. Identify and stop zombie EC2 instances

    • Savings: €2K-5K/year
    • Time: 15 minutes
    • Filter instances by CPU <5% for 7+ days → Verify with team → Stop or terminate
    • Often found: Dev/staging instances running 24/7 unnecessarily

  • 3. Enable Auto Scaling for variable workloads

    • Savings: €3K-8K/year
    • Time: 1 hour
    • Set target tracking policies (CPU 60-70%) → Start conservative → Monitor savings
    • Real example: Client reduced instance hours 40% without performance impact

🟡 Week 2: Low-Risk, High Impact

  • 4. Purchase Reserved Instances for predictable workloads

    • Savings: €8K-15K/year
    • Time: 1 hour analysis
    • Review 30-day utilization → Commit to 1-year Standard RIs for >75% utilized instances
    • Tip: Start with 50% coverage, expand after confidence builds

  • 5. Migrate to Graviton instances (ARM-based)

    • Savings: €4K-12K/year
    • Time: 2-4 hours testing
    • Test compatibility → Create Graviton AMIs → Gradual rollout
    • Benefit: 20-40% cost reduction for same performance

  • 6. Switch Lambda to ARM64 architecture

    • Savings: €500-2K/year
    • Time: 30 minutes
    • Update function configuration → Test → Deploy
    • Bonus: Often sees performance improvement

🟠 Month 1: Requires Planning

  • 7. Review ECS/EKS cluster utilization and consolidate
    • Savings: €5K-15K/year
    • Time: 4-8 hours
    • Analyze node utilization → Reduce node count → Optimize pod sizing

CATEGORY 2: Storage & Database

Potential Savings: €5,000 - €15,000/year

🟢 Week 1: No-Risk Wins

  • 8. Implement S3 Lifecycle policies

    • Savings: €3K-8K/year
    • Time: 30 minutes
    • Transition to S3 IA after 30 days → S3 Glacier after 90 days → Delete after 365 days
    • Common use case: Logs, backups, archived media

  • 9. Delete unused EBS volumes

    • Savings: €1K-3K/year
    • Time: 20 minutes
    • Filter volumes “available” state → Verify unattached >7 days → Delete
    • Caution: Create snapshot first if unsure

  • 10. Delete old EBS snapshots

    • Savings: €500-2K/year
    • Time: 15 minutes
    • Identify snapshots >90 days → Verify no AMI dependencies → Delete
    • Automate with AWS Backup retention policies

🟡 Week 2: Low-Risk, High Impact

  • 11. Right-size RDS instances

    • Savings: €3K-10K/year
    • Time: 1 hour
    • Review CloudWatch CPU/Memory → Downsize overprovisioned instances
    • Example: db.r5.2xlarge at 25% CPU → db.r5.xlarge (50% savings)

  • 12. Enable S3 Intelligent-Tiering for variable access patterns

    • Savings: €2K-5K/year
    • Time: 30 minutes
    • Apply to buckets with unpredictable access → Automatic cost optimization
    • Works well: Media assets, user uploads, analytics data

  • 13. Convert GP2 EBS volumes to GP3

    • Savings: €1K-3K/year
    • Time: 1 hour
    • GP3 offers 20% cost reduction with same baseline performance
    • Modify volume type during maintenance window

🟠 Month 1: Requires Planning

  • 14. Migrate to Aurora Serverless v2 for variable database loads
    • Savings: €5K-12K/year
    • Time: Planning + migration (8-16 hours)
    • Best for: Dev/staging, variable production workloads
    • Scales to zero during idle periods

CATEGORY 3: Network & Data Transfer

Potential Savings: €3,000 - €10,000/year

🟢 Week 1: No-Risk Wins

  • 15. Implement VPC Endpoints for S3 and DynamoDB

    • Savings: €2K-5K/year
    • Time: 30 minutes
    • Eliminates NAT Gateway charges for AWS service traffic
    • Quick win for S3-heavy workloads

  • 16. Delete unused Elastic Load Balancers

    • Savings: €500-2K/year
    • Time: 15 minutes
    • Each idle ALB costs ~€20/month → Identify unused → Delete
    • Check target group health first

🟡 Week 2: Low-Risk, High Impact

  • 17. Enable CloudFront for static content

    • Savings: €1K-4K/year
    • Time: 2 hours setup
    • Reduces S3 data transfer costs → Improves performance globally
    • Configure cache TTLs appropriately

  • 18. Review and optimize cross-region data transfer

    • Savings: €2K-6K/year
    • Time: 1-2 hours analysis
    • Identify unnecessary replication → Consolidate regions where possible
    • Data transfer is one of the most expensive AWS costs

🟠 Month 1: Requires Planning

  • 19. Evaluate AWS Direct Connect for high-volume transfers

    • Savings: €3K-8K/year (for >5TB/month)
    • Time: Planning + setup (1-2 weeks)
    • Break-even typically at 5TB/month data transfer
    • Consider Snowball for one-time large migrations

  • 20. Consolidate multi-region deployments

    • Savings: €5K-15K/year
    • Time: Architecture review (4-8 hours)
    • Question: Do you really need US + EU + APAC?
    • Often single-region + CloudFront serves 95% of use cases

CATEGORY 4: Quick Configuration Wins

Potential Savings: €2,000 - €5,000/year

🟢 Week 1: No-Risk Wins

  • 21. Reduce CloudWatch log retention

    • Savings: €500-2K/year
    • Time: 15 minutes
    • Set retention to 7-30 days instead of “Never expire”
    • Export to S3 for long-term archive if needed

  • 22. Disable detailed monitoring on EC2 instances

    • Savings: €300-1K/year
    • Time: 10 minutes
    • Basic monitoring (5-min intervals) sufficient for most workloads
    • Keep detailed only for critical production instances

  • 23. Review and delete unused NAT Gateways

    • Savings: €1K-3K/year
    • Time: 20 minutes
    • Each NAT Gateway costs ~€35/month + data processing
    • Replace with VPC Endpoints where possible

🟡 Week 2: Low-Risk, High Impact

  • 24. Enable AWS Budgets and Cost Anomaly Detection

    • Savings: Prevents future overruns
    • Time: 30 minutes
    • Set budget alerts at 50%, 80%, 100% of monthly target
    • Anomaly detection catches unusual spikes early

  • 25. Implement comprehensive resource tagging

    • Savings: Enables future optimization
    • Time: 2 hours + enforcement
    • Required tags: Environment, Owner, CostCenter, Project
    • Use Tag Policies to enforce compliance

  • 26. Review IAM policies and reduce over-permissions

    • Savings: Security benefit + prevents costly mistakes
    • Time: 2-3 hours
    • Use IAM Access Analyzer → Apply least privilege
    • Prevents accidental expensive resource launches

🟠 Month 1: Requires Planning

  • 27. Set up AWS Cost Optimization Hub
    • Savings: Continuous optimization
    • Time: 1-2 hours setup
    • Centralized recommendations across all services
    • Review monthly for new opportunities

Implementation Roadmap

Week 1 Priority (No-Risk Wins)

Target Savings: €5,000 - €10,000/year

Complete items: #1, #2, #3, #8, #9, #10, #15, #16, #21, #22, #23

Time Investment: 3-4 hours Expected Result: Immediate bill reduction visible next month

Week 2 Priority (Low-Risk, High Impact)

Additional Savings: €8,000 - €15,000/year

Complete items: #4, #5, #6, #11, #12, #13, #17, #18, #24, #25

Time Investment: 6-8 hours Expected Result: Compounding savings, better visibility

Month 1 (Planned Changes)

Additional Savings: €10,000 - €20,000/year

Complete items: #7, #14, #19, #20, #26, #27

Time Investment: 15-20 hours Expected Result: Optimized architecture, ongoing cost controls

Total Potential Savings

Conservative Estimate: €23,000 - €45,000/year Aggressive Implementation: €35,000 - €60,000/year Time Investment: 25-35 hours over 4 weeks

ROI: Approximately €1,000-€2,000 saved per hour invested.

Common Mistakes to Avoid

❌ Don’t: Make changes in production without testing ✅ Do: Test in dev/staging first, then gradual production rollout

❌ Don’t: Delete resources without verification and backups ✅ Do: Create snapshots, document changes, get team confirmation

❌ Don’t: Optimize everything at once ✅ Do: Start with Category 1, build momentum, learn as you go

❌ Don’t: Focus only on cost ✅ Do: Balance cost, performance, and reliability

❌ Don’t: Implement and forget ✅ Do: Review monthly, adjust, continuous optimization

Next Steps

Share Your Results

Implemented this checklist? I want to hear about it!

  • Email: carlos@thewisecto.com
  • LinkedIn: linkedin.com/in/carlosinfantes
  • Tag: @thewisecto with your savings

Need Help?

Stuck on implementation? Book a 1-hour architecture review:

  • Review your AWS setup together
  • Identify your top 5 cost optimization opportunities
  • Get specific recommendations for your infrastructure
  • €120/hour - No commitments, no packages

Book now: thewisecto.com/mentoring

Weekly AWS Tips Newsletter

Want weekly AWS optimization tips delivered to your inbox?

Join 500+ cloud engineers getting:

  • 1 actionable AWS tip per week
  • Cost optimization case studies
  • Terraform patterns and best practices
  • SRE techniques from UN-scale infrastructure

Subscribe free: thewisecto.com/resources

About Carlos Infantes

Lead Site Reliability Engineer, United Nations 25+ years building infrastructure at scale. Former Amazon engineer. Helped 20+ Series A-B startups scale from 5 to 50 engineers without breaking the bank.

Credentials:

  • AWS Solutions Architect Professional
  • AWS Security Specialty
  • Managing infrastructure for 193 countries
  • €5M+ in verified client cost savings

The Wise CTO: CTO mentoring and advisory for Series A-B startups.

thewisecto.com | carlos@thewisecto.com

Version 1.0 - January 2025 © 2025 The Wise CTO. Free to share with attribution.

Bonus: Found this valuable? Share it with your team and tag @thewisecto on LinkedIn. I read every mention and often share additional tips in the comments.