AWS Cost Anomaly Detection now supports percentage-based thresholds
Starting today, customers of AWS Cost Anomaly Detection will be able to define percentage-based thresholds when configuring their alerting preferences. AWS Cost Anomaly Detection is a cost management service that leverages advanced machine learning to identify anomalous spend and root causes, so customers can quickly take action to avoid runaway spend and bill shocks. The percentage-based alerting configuration enables customers to capture the anomalous spend dynamically rather than having to estimate an absolute dollar amount.
Starting today, customers of AWS Cost Anomaly Detection will be able to define percentage-based thresholds when configuring their alerting preferences. AWS Cost Anomaly Detection is a cost management service that leverages advanced machine learning to identify anomalous spend and root causes, so customers can quickly take action to avoid runaway spend and bill shocks. The percentage-based alerting configuration enables customers to capture the anomalous spend dynamically rather than having to estimate an absolute dollar amount.
To setup AWS Cost Anomaly Detection, customers configure cost monitors and define when and how to receive alerts. Up until today, customers have only had the option to specify a fixed-dollar threshold (difference between actual and expected spend) for their alerts. With percentage-based thresholds customers will have even more flexibility and choice when to be alerted based on their needs. Customers can now set rules like “send me an alert when the anomaly’s impact is 25% higher than the expected amount” or “send me an alert when the anomaly’s impact is over $40 AND/OR the impact is 30% higher than the expected amount”.
Percentage-based thresholds for AWS Cost Anomaly Detection is available in all AWS commercial regions, including China. You can get started with this new features using the AWS Cost Anomaly Detection console or programmatically via the public APIs at no additional cost. To get started with percentage based thresholds in AWS Cost Anomaly Detection, please visit our documentation.