When we audit a new client's cloud bill, we almost always find 20-40% waste. The five patterns below are the ones we find most often.
1. Idle and Orphaned Resources
Every cloud account we've audited has them: EC2 instances running for a project that ended six months ago, unattached EBS volumes, load balancers pointing to nothing. One client had $8,400 per month in orphaned storage volumes alone.
How to find it
- AWS Trusted Advisor and Azure Advisor both flag idle resources for free.
- Tag everything with an owner and expiration date.
- Set up a weekly report of resources with zero traffic over the last 14 days.
2. Over-Provisioned Compute
Most workloads run at 15-25% utilization on average. A client running 40 production VMs reduced their compute spend by 31% by systematically right-sizing based on actual 30-day utilization data.
The best time to right-size was three years ago. The second best time is during your next quarterly review.
3. Wrong Storage Tiers
The price difference between S3 Standard and Glacier Deep Archive is about 23x. One client with 180TB cut their storage bill by 67% with lifecycle policies.
4. Egress and Data Transfer Surprises
We once found a client spending $4,200 per month on cross-AZ traffic between two microservices that didn't need to be in different availability zones.
5. Committed Use Discounts Left on the Table
Reserved Instances and Savings Plans can reduce compute costs by 30-72%. Analyze your 30-day baseline usage — commit for that baseline, pay on-demand for the rest. A client saved $47,000 annually on a 1-year Savings Plan.
Making Optimization Continuous
Cloud cost optimization is not a one-time project. We build cost monitoring into every managed cloud engagement. If your cloud bill has been creeping up, get in touch for a no-obligation cost audit.