The 30-Day AWS Cost Audit: A Step-by-Step Playbook

55 min read

A practical week-by-week guide to auditing and optimizing AWS costs. Week 1: Visibility. Week 2: Quick wins. Week 3: Right-sizing. Week 4: Commitments.

AWS cost-optimization FinOps audit playbook

The 30-Day AWS Cost Audit

A step-by-step playbook for reducing your cloud bill

4 Weeks
Structured Plan
30-50%
Typical Savings
Actionable
Daily Tasks

Most AWS cost optimization guides are either too vague ("look for waste") or too overwhelming ("here are 147 things to check"). This playbook is different: it's a structured 30-day plan with specific daily and weekly tasks.

Follow it step by step, and by day 30 you'll have visibility into your costs, eliminated obvious waste, right-sized your resources, and set up ongoing optimization processes.

The 4-Week Structure

WEEK 1

Visibility

Understand where your money goes

See
WEEK 2

Quick Wins

Low-risk, high-impact changes

Cut
WEEK 3

Right-Sizing

Optimize active resources

Optimize
WEEK 4

Commitments

Lock in discounts

Commit
1

Week 1: Visibility

Understand where your money goes before you try to cut it

Days 1-2

Set Up Cost Explorer & Billing Alerts

Tasks:

  • Enable Cost Explorer in AWS Console
  • Set up budget alerts at 50%, 80%, 100% of monthly target
  • Enable Cost Anomaly Detection
  • Export last 6 months of data to CSV

Why This Matters:

You can't manage what you don't measure. Most AWS users never look at Cost Explorer until something goes wrong.

Days 3-4

Analyze Cost Breakdown by Service

Tasks:

  • Identify top 5 services by cost
  • Compare month-over-month growth rates
  • Flag any unexpected services on the bill
  • Document findings in a spreadsheet

What to Look For:

  • • EC2 usually dominates (30-50%)
  • • RDS is often #2
  • • Watch for data transfer costs
  • • CloudWatch can be surprisingly high
Day 5

Inventory All Resources

Tasks:

  • List all EC2 instances with sizes and utilization
  • List all RDS instances
  • Note which are tagged vs untagged
  • Create a "resource owner" spreadsheet

Pro Tip:

Use AWS Resource Groups Tag Editor to get a complete inventory across all regions at once.

Days 6-7

Create Cost Dashboard & Baseline

Tasks:

  • Create a simple cost dashboard (spreadsheet or tool)
  • Document current monthly spend as baseline
  • Set a target for end of 30 days (e.g., 20% reduction)
  • Share baseline with stakeholders

Week 1 Deliverable:

A documented baseline of current costs, top services, and a target reduction percentage.

Week 1 Goal: Complete visibility into your AWS spend
2

Week 2: Quick Wins

Low-risk changes that deliver immediate savings

Days 8-9

Delete Orphaned Resources

Tasks:

  • Find and delete unattached EBS volumes
  • Release unassociated Elastic IPs
  • Remove empty S3 buckets
  • Delete idle load balancers (no healthy targets)

Expected Savings:

5-15% of total bill. These are "free money" optimizations—zero risk, immediate savings.

Day 10

Clean Up Old Snapshots

Tasks:

  • List all EBS snapshots older than 90 days
  • Identify snapshots of deleted volumes
  • Delete unnecessary snapshots (keep recent + monthly)
  • Set up AWS Backup with retention policies

Expected Savings:

$0.05/GB/month. If you have 1TB of old snapshots, that's $50/month.

Day 11

Upgrade GP2 to GP3 Storage

Tasks:

  • List all gp2 EBS volumes
  • Upgrade to gp3 (zero downtime)
  • Update Terraform/CloudFormation for new resources

Expected Savings:

20% reduction in EBS costs. $0.10/GB → $0.08/GB, plus better performance.

Days 12-13

Optimize CloudWatch Logs

Tasks:

  • Set retention policies on all log groups (7-30 days)
  • Find largest log groups by storage
  • Review Lambda function log levels
  • Consider exporting old logs to S3

Expected Savings:

50-80% reduction in CloudWatch costs. Often one of the biggest quick wins.

Day 14

Review & Document Week 2 Savings

Tasks:

  • Calculate total estimated monthly savings
  • Document all changes made
  • Update cost dashboard with projected new baseline
  • Share wins with stakeholders

Week 2 Deliverable:

A documented list of quick wins with estimated monthly savings. Target: 10-20% reduction.

Week 2 Goal: Eliminate obvious waste (10-20% savings)
3

Week 3: Right-Sizing

Optimize resources that are running but oversized

Days 15-16

Analyze EC2 Utilization

Tasks:

  • Check AWS Compute Optimizer recommendations
  • Review CloudWatch CPU metrics (look for <30%)
  • Check memory utilization (requires CloudWatch agent)
  • Create list of right-sizing candidates

Rule of Thumb:

If average CPU is below 30% consistently, the instance is likely oversized by at least one size class.

Days 17-18

Right-Size EC2 Instances

Tasks:

  • Start with dev/staging environments
  • Downsize oversized instances (one size at a time)
  • Monitor for 24-48 hours after each change
  • Update to latest generation (m4 → m6i)

Expected Savings:

20-40% reduction in EC2 costs. This is often the biggest single optimization.

Days 19-20

Right-Size RDS Databases

Tasks:

  • Review RDS CPU and connection metrics
  • Identify idle or over-provisioned databases
  • Consider Aurora Serverless for variable workloads
  • Schedule maintenance window for resizing

Caution:

RDS right-sizing requires brief downtime. Plan changes during maintenance windows.

Day 21

Review & Document Week 3 Savings

Tasks:

  • Document all instances resized
  • Calculate cost impact of changes
  • Note any changes that need to be reverted
  • Update baseline with new projected costs

Week 3 Deliverable:

A list of right-sized resources with before/after costs. Target: Additional 10-20% savings.

Week 3 Goal: Right-size active resources (10-20% additional savings)
4

Week 4: Commitments & Sustainability

Lock in discounts and set up ongoing optimization

Days 22-23

Evaluate Savings Plans / Reserved Instances

Tasks:

  • Review Cost Explorer RI/SP recommendations
  • Calculate stable baseline usage (60-70% of avg)
  • Compare Compute SP vs EC2 SP vs RIs
  • Get approval for commitment purchases

Recommendation:

Start with Compute Savings Plans for EC2/Lambda/Fargate. Use RDS Reserved Instances for databases. Don't over-commit!

Days 24-25

Purchase Commitments

Tasks:

  • Purchase Compute Savings Plans (start at 60% of baseline)
  • Purchase RDS Reserved Instances for stable databases
  • Document all purchases and expiration dates
  • Set calendar reminders for renewal reviews

Expected Savings:

30-50% additional savings on committed usage. This compounds with right-sizing.

Days 26-27

Set Up Ongoing Governance

Tasks:

  • Enable AWS Config rules for tag compliance
  • Set up weekly cost anomaly review process
  • Create monthly cost review meeting cadence
  • Document tagging standards and enforce via SCPs

Why This Matters:

Cost optimization isn't a one-time project. Without ongoing governance, waste creeps back within 3-6 months.

Days 28-30

Final Review & Report

Tasks:

  • Compare Day 30 costs to Day 1 baseline
  • Create executive summary of all changes
  • Document lessons learned
  • Present results to stakeholders
  • Plan next optimization cycle (quarterly)

Week 4 Deliverable:

A comprehensive report showing: baseline, changes made, savings achieved, and ongoing governance plan.

Week 4 Goal: Lock in discounts and establish ongoing processes

Expected Results After 30 Days

30-50%
Total Cost Reduction
100%
Cost Visibility
95%+
Tag Coverage
Monthly
Review Process

Breakdown of Savings by Week

Week 1
0% (visibility)
Week 2
10-20%
Week 3
20-35%
Week 4
30-50%

Start Your 30-Day Audit Today

Get a head start with our free AWS Cost Analyzer. Upload your Cost Explorer CSV and see exactly where your money is going—the perfect Day 1 activity.

Start My Free Analysis

Results in 5 minutes • Perfect for Day 1

The Bottom Line

AWS cost optimization isn't complicated—it's just systematic. This 30-day playbook breaks down the process into manageable daily tasks. Follow it step by step, and you'll achieve significant savings while building the processes to maintain them long-term.

30 Days
Complete transformation
30-50%
Typical savings
Sustainable
Ongoing processes