3 min read

What is Amazon AWS Cost Anomaly Detection

December 19, 2022

What is AWS Cost Anomaly Detection

In these days of rising business and energy costs we can ultimately expect to see the burden on AWS data centers passed on to customers sooner or later. So keeping on top of your AWS cloud spend has never been more important.

When you have multiple AWS accounts, keeping track of cloud costs can be tricky, especially if there are unexpected changes to billing that you might not pick up until the AWS bill arrives.

Fortunately there is an AWS service that will alert you when your cost profile changes and starts accumulating charges that are outside the pattern of normal historical usage. That service is called Amazon Cost Anomaly Detection.

Amazon Cost Anomaly Detection is a machine learning-based service that helps organizations identify unusual spikes or dips in their AWS costs. It's designed to help businesses optimize their AWS usage and reduce unnecessary spending.

Here's how it works: Amazon Cost Anomaly Detection uses machine learning algorithms to analyze historical cost and usage data for your AWS account. It looks for patterns in the data and uses this information to build a baseline of what your normal AWS costs should be. If it detects a significant deviation from this baseline, it will send you an alert.

For example, if you suddenly start using a lot more compute resources than usual, Amazon Cost Anomaly Detection may send you an alert to let you know that your costs are likely to increase. This can help you take corrective action, such as scaling back your resource usage or shutting down idle resources.

In addition to sending alerts, Amazon Cost Anomaly Detection also provides detailed cost and usage data that can help you understand the root cause of the anomaly. This data can be accessed through the AWS Management Console, or you can use the AWS Cost and Usage Report to export the data to a third-party BI tool for further analysis.

Up until recently you could base the trigger for alerts on dollar amounts, ie: send an alert when expected usage exceeds a dollar threshold. As of December 2022 you can now specify percentages too and you can combine them with dollar amounts. So you could ask Anomaly Detection to send an alert when a detected anomaly’s impact is 25% higher than normal or the anomaly is $50 over and/or 30% higher than expected.

Overall, Amazon Cost Anomaly Detection is a valuable tool for organizations that use AWS and want to optimize their costs. It can help you identify and fix problems before they become too expensive, and it provides the data and insights you need to make informed decisions about your AWS usage. Whether you're a small business owner, a developer, or a data analyst, Amazon Cost Anomaly Detection is a tool that can help you save money on your AWS bills.

Cost Anomaly Detection can be found in the AWS Cost Management console.

Another way to keep an eye on costs is to review exactly what you have running in your AWS account. When you connect your AWS account to hava.io you will get an automatically generated network infrastructure diagram for every VPC detected.

This means all the resources running anywhere in your AWS account will be visualised, including copies of databases that shouldn't be running but are, entire test, staging or dev environments that may or may not still be in use. All the resources will appear on a diagram that is clear and easy to understand. Outliers and unexpected resources are regularly uncovered when clients first connect their AWS accounts to Hava.


You can get started with a free 14 day trial by following the button below.


Topics: aws
Team Hava

Written by Team Hava

The Hava content team