What Causes AWS Costs to Rise When Usage Stays About the Same?

Have you ever looked at your AWS bill and thought:

“We’re not really using more than before. So why is the bill higher?”

I’ve always found this interesting because a workload can look pretty stable on the surface, while different parts of the AWS bill are changing underneath.

If I were digging into it, I wouldn’t assume compute was the problem right away. I’d probably look at a few other areas too:

• Data transfer: Cross-AZ, cross-Region, or internet traffic can add up even when overall application usage hasn't changed much.
• Storage and requests: Your stored data may keep growing, and services such as S3 can also add request and retrieval charges.
• Supporting services: NAT Gateways, load balancers, CloudWatch logs, snapshots, and replication are easy to overlook when looking at the main workload.
• Pricing and configuration: A change in instance type, capacity setup, or Savings Plan/Reserved Instance coverage can affect the bill even when usage looks similar.

That's what makes these increases tricky. You can look at CPU, instance-hours, or traffic and still not immediately see where the extra spend is coming from.

So I’m curious has this happened to you?

What looked normal in your AWS usage, but turned out to be responsible for the higher bill?

And how did you eventually figure it out?

Was it data transfer, storage, requests, supporting infrastructure, pricing coverage or something you weren't expecting at all?

I'd be interested to hear what you found and how you tracked it down.