April 4, 2026 · Tim Fraser, Cloud Operations Lead

AWS Costs Explained for Non-Technical Managers

Your finance team wants to know why the AWS bill went up. Your team says something about NAT gateways and data transfer. You need to give an answer that makes sense to the board, not just to the engineers.

Here's how AWS billing works in plain English, and what questions to ask your team.

The big four: where AWS money goes

For most organisations, 80-90% of the AWS bill comes from four services:

1. EC2 (Compute) — your servers

This is the cost of running virtual servers. Think of it like renting office space — you pay by the hour, and bigger rooms cost more.

• More servers running (someone launched new ones)
• Bigger server sizes (someone upgraded an existing one)
• Servers running 24/7 that are only needed during business hours
• Old test/staging servers nobody turned off

2. RDS (Databases) — your data storage

Managed database servers. These tend to be expensive because they run 24/7 and are often oversized "just in case."

• Multi-AZ deployments (a backup copy running in another data centre — good for production, wasteful for dev/test)
• Oversized instances — a database sized for peak load that only hits peak for 2 hours a day
• Development databases running the same configuration as production

3. Data Transfer — moving data around

This is the sneaky one. AWS charges for data leaving its network (egress), and even for data moving between availability zones within the same region.

• Applications chatting across availability zones (cross-AZ traffic)
• Large file downloads by customers
• Backup replication between regions
• Logs being shipped to external monitoring services

4. S3 (Storage) — your files

Object storage for files, backups, logs, and media. Usually cheap, but it adds up when nobody cleans up old data.

• Log files accumulating without lifecycle policies (auto-deletion)
• Old backups never cleaned up
• Data stored in expensive storage classes when it could be archived

Reading the AWS bill

AWS Cost Explorer (Billing → Cost Explorer in the console) shows spending by service, by month, and by trend. The most useful view is Group by: Service over the last 3 months. This immediately shows which service is responsible for any spike.

But if you don't want to learn the console, that's exactly the kind of question you can delegate — either to your team, or to a tool that speaks plain English.

How plainfra helps managers with cost visibility

plainfra connects to your AWS account (read-only) and answers questions in plain English. No console, no jargon, no learning curve.

> "What's driving our AWS costs this month? Give me the top 5."

You get a clear breakdown — dollars, percentages, and trend arrows. If NAT Gateway costs jumped 47%, plainfra tells you that and suggests why.

> "Are we paying for resources nobody is using?"

plainfra scans for idle instances, unattached volumes, and unused load balancers — and gives you dollar amounts for each one.

The weekly health report takes this further. Every Monday, plainfra emails you a PDF with cost trends, new anomalies, and changes since last week. You walk into standup with a clear picture — no console required.

When plainfra flags a cost issue, you can turn it into a ticket for your team to investigate. The finding has all the detail they need: which resource, how much it costs, and what the recommended action is.