.png)
How It Works
AWS Cost Explorer analyzes your historical on-demand usage, typically over a 7, 30, or 60-day lookback window, and calculates the Savings Plan commitment level that would have covered that usage. It then surfaces a recommended hourly commitment amount, a projected savings figure, and an estimated utilization rate. You can filter recommendations by Savings Plan type (Compute or EC2 Instance), term length (1-year or 3-year), and payment option (no upfront, partial upfront, or all upfront). Acting on a recommendation means navigating to the AWS Savings Plans purchase console and manually buying the suggested plan.
Why It Matters for Cloud Cost
Savings Plans can reduce EC2, Fargate, and Lambda costs by up to 66% versus on-demand pricing for Compute Savings Plans, or up to 72% for EC2 Instance Savings Plans. Cost Explorer Recommendations make that opportunity visible without requiring manual analysis. The limitation is that they are backward-looking: recommendations reflect past usage patterns and are refreshed roughly every 72 hours. Usage patterns that shift quickly, seasonal workloads, or fast-growing infrastructure can make a recommendation stale before you act on it. There is also no automation built in. Reviewing, deciding, and purchasing still falls to your team, and any commitment you purchase is yours to manage if utilization drops.
Usage AI’s CoPilot mode surfaces commitment recommendations with projected savings for customer review before any purchase is executed, while Autopilot goes further by purchasing and adjusting Savings Plan commitments daily without requiring human approval.