.png)
How It Works
A PPA is a bilateral agreement negotiated directly with AWS, typically through an AWS account team or an AWS partner. The customer commits to a defined spending level or usage volume over a contract period, and AWS in turn offers discounted rates on specific services. The agreement is formalized as a custom pricing addendum to the customer’s standard AWS contract. Terms, discount levels, and eligible services vary by deal and are not publicly disclosed. PPAs are distinct from standard discount mechanisms like Reserved Instances or Savings Plans because the pricing is bespoke, negotiated in private, and not accessible through self-service AWS tools.
Why It Matters for Cloud Cost
A well-negotiated PPA can reduce unit costs on high-volume AWS services that standard commitment-based discounts do not fully address. Without one, large-scale AWS customers often pay more per unit than their spend level would justify. The risk, however, is that PPA negotiations are time-consuming, require AWS relationship access that smaller teams lack, and typically lock the customer into spending targets that can become liabilities if usage declines. Failing to meet the committed spend threshold in a PPA can erode or void the negotiated rates entirely. Organizations that rely solely on PPAs for cost savings without also optimizing their compute commitments through Savings Plans and Reserved Instances often leave additional savings on the table.
Usage AI: Usage AI operates alongside any existing PPA a customer has in place, layering automated Savings Plans and Reserved Instance management on top to capture the compute-level savings that PPAs typically do not cover, helping customers reach 30 to 50% overall reduction in AWS spend.