Rate Optimization: Commitments & Discounts

Lesson 1 of 3

Reserved Instances & Savings Plans

Evaluate commitment discount options across AWS, Azure, and GCP, and select the right commitment strategy for each scenario type.

FinOpsDecode is an independent training product and is not affiliated with, endorsed by, or sponsored by the FinOps Foundation or any certification body.

Commitment Discounts: The Core Trade-off

Commitment-based discounts (Reserved Instances, Savings Plans, Committed Use Discounts) offer 40–70% savings over on-demand pricing in exchange for a 1 or 3-year commitment. The risk is commitment misuse: buying too many commitments results in unused capacity you still pay for; buying too few means continued on-demand pricing for workloads that could be discounted. The practitioner's job is to find the right coverage level—typically 70–80% of steady-state usage committed, with the remainder on-demand or spot for flexibility.

Commitment Types by Cloud Provider

AWS

  • Standard RIs: max discount, least flexible
  • Convertible RIs: can exchange instance type
  • Compute Savings Plans: apply across families
  • EC2 Instance Savings Plans: tied to instance family

Azure

  • Reserved VM Instances: 1 or 3-year term
  • Azure Savings Plan for Compute: flexible across regions
  • Dev/Test pricing for non-prod workloads
  • Hybrid Benefit for Windows/SQL licenses

GCP

  • Committed Use Discounts (CUDs): resource-based
  • Flexible CUDs: apply across machine families
  • Sustained Use Discounts: automatic, no commitment
  • Spot VMs: lowest cost, preemptible

Commitment Buying Process

  1. 1Establish a 30–90 day usage baseline to identify steady-state compute patterns.
  2. 2Calculate on-demand spend for always-on workloads.
  3. 3Model coverage scenarios at 60%, 70%, and 80% commitment levels.
  4. 4Start conservatively (60–70%) for first purchase to reduce risk.
  5. 5Review utilization monthly and adjust coverage through additional purchases or exchange.
FinOpsDecode Rule

Commit to what is stable and predictable. Keep variable and experimental workloads on-demand.

Practice this topic

Reinforce this lesson with scenario questions tagged Reserved Instances, Savings Plans, Rate Optimization.

Go to Practice