Why

Centrally purchased RIs, Savings Plans, and CUDs benefit workloads across multiple BUs, but the purchase sits with one team. If commitment savings stay in the purchasing account, consuming teams see inflated on-demand costs. Redistribution creates accurate unit costs and the right incentive structure — BUs that right-size into committed instance types benefit from the discount.

What

Define a commitment savings redistribution policy and implement it in the billing pipeline. Choose from three models depending on your organisation’s financial structure.

How

Choose a Redistribution Model

Model	How It Works	Best When
Proportional by covered usage	Savings distributed by each BU’s share of covered hours	BUs own their P&L
Centralised budget	RI/SP cost stays in central platform cost centre. BUs see on-demand equivalent pricing	Cloud is a central service
Amortise and redistribute	Spread upfront/monthly cost across benefiting BUs by covered hours. Each BU sees true blended rate	Maximum transparency needed

Commitment Redistribution Example
═══════════════════════════════════════════════════════════════

  RI Pool: 20× m5.xlarge — $14,400/month
  On-demand equivalent:     $21,600/month
  Total savings:             $7,200/month

  BU              Covered Hours    %       Savings Credit
  ──────────────  ─────────────   ─────   ──────────
  Payments        8,640            60%     $4,320
  Analytics       4,320            30%     $2,160
  Dev/Test        1,440            10%       $720

Map Commitments to Consuming Resources

Requires logic in the billing pipeline to match commitment discounts to consuming resources/accounts.

Provider	Native Support
AWS	Cost Categories with Split Charge Rules support proportional redistribution. Amortised cost view in Cost Explorer shows per-account blended rates.
Azure	Cost Allocation Rules can redistribute reservation costs. FOCUS export includes `EffectiveCost` (amortised) alongside `BilledCost`.
GCP	CUDs/SUDs apply at project level. Use BigQuery billing export to calculate `on_demand_equivalent - actual_cost` per project for savings attribution.

Implement in Billing Pipeline

For cloud-native: configure Cost Categories (AWS) or Cost Allocation Rules (Azure) with proportional split by covered usage.

For custom: build a scheduled job that calculates on_demand_equivalent - actual_billed per commitment, allocates the delta proportionally, and writes adjusted cost records to the billing data store.

Reconcile and Document

Ensure the sum of redistributed savings equals total commitment savings (zero-sum). Document the policy, including which commitments are in scope, the redistribution formula, and the reconciliation process.

Deliverable Checklist

Redistribution model chosen and agreed with Finance
Commitment-to-consumer mapping logic built
Cloud-native rules or custom pipeline operational
Reconciliation confirms zero-sum (no leakage)
Policy documented: scope, formula, reconciliation process
Showback reports reflect redistributed savings