The Problem: A Bill Nobody Could Explain
A 60-person SaaS company came to us with a $31,000 monthly AWS bill and a CFO asking hard questions. The engineering team could not confidently explain what was driving 40% of the costs. Sound familiar?
This is one of the most common situations we encounter. AWS bills grow organically as teams move fast — and the default AWS console makes it genuinely difficult to pinpoint waste without dedicated tooling and expertise.
The Audit: Where We Started
In week one, we ran a full Cost Explorer breakdown across every service, region, and tag. The results were predictable to us but shocking to the client:
- EC2 instances: 23 instances running at under 10% average CPU utilisation. Three of them hadn't had a connection in over 45 days.
- EBS snapshots: 847 snapshots, many attached to volumes that no longer existed. $1,400/month in pure storage waste.
- RDS instances: Two dev/test databases running on db.r5.2xlarge in production configuration, 24/7. They were only used during business hours.
- Data transfer: $2,100/month in inter-region data transfer for a logging pipeline that could have been redesigned to stay in a single region.
- NAT Gateway: A single NAT Gateway processing 4TB/month of traffic that could have been routed through VPC endpoints for S3 and DynamoDB at near-zero cost.
The Fix: 30 Days of Precise Remediation
We did not recommend a wholesale migration or a months-long transformation. We targeted the highest-impact, lowest-risk changes first:
- Right-sized 18 EC2 instances to Graviton3 equivalents — average 30% cost reduction with 10-15% performance improvement.
- Deleted 800+ orphaned EBS snapshots after a 7-day hold period and client sign-off.
- Scheduled dev/test RDS instances to stop automatically at 7pm and restart at 7am weekdays.
- Added S3 and DynamoDB VPC endpoints, eliminating the NAT Gateway data transfer cost for those services.
- Redesigned the logging pipeline to use a single region with cross-region replication only for DR.
The Result
Month-end bill: $18,400. A saving of $12,600 — 40.6% — in 30 days. No application code changed. No migrations. No downtime.
The client moved onto a Silver FinOps retainer. We now review their bill monthly, catch anomalies before they compound, and handle ongoing rightsizing as their workload evolves.
The Takeaway
AWS cost waste is not a sign of incompetence — it is a sign of a team moving fast without dedicated FinOps attention. The typical AWS bill has 20-35% addressable waste hiding in plain sight. The question is whether you have the tooling and expertise to find it quickly.
If your AWS spend is over $5,000/month and you have not had a structured cost review in the last 90 days, you are almost certainly overpaying. Book a free 30-minute AWS Cost Review and we will show you exactly where your money is going.