What can you actually commit to?
Every provider sells a discount in exchange for a forward commitment. The shape of the commitment, and the flexibility you keep, differs by cloud.
| Cloud | Instruments | Discount range (indicative) | Flexibility note |
|---|---|---|---|
| AWS | Savings Plans, Reserved Instances | up to about 72 percent | Savings Plans trade specificity for flexibility across compute |
| Azure | Reservations, Azure Savings Plan | up to about 72 percent | Reservations are exchangeable; Savings Plan applies across compute |
| GCP | Committed Use Discounts (spend based and resource based) | up to about 70 percent | Sustained use discounts also apply automatically |
| OCI | Universal Credits (annual flex or pay as you go) | tiered by commitment | Support Rewards offset Oracle support fees |
Discount ranges are indicative and depend on term, region, instrument, and payment option. Always verify against the provider's current pricing for your exact footprint.
What layers on top of the instruments?
Above the commitment instruments sit multi year enterprise agreements that trade a spend commitment for a discount tier. They share one dangerous trait: use it or lose it.
| Agreement | What it trades | The catch |
|---|---|---|
| AWS Enterprise Discount Program (EDP) | Multi year spend commitment for a discount tier | Shortfall on the commitment is owed |
| Azure MACC | Consumption commitment for enterprise pricing | Shortfall clause: unspent commitment is still owed |
| GCP enterprise agreement | Committed spend for tiered discounts | Use it or lose it on the commitment |
| Oracle Universal Credits | Annual credit pool for OCI services | Unused annual flex credits can lapse |
Where does negotiating power come from?
Discounts are not won by asking. They are won by changing the provider's expected value of the deal.
A defensible, bottom up forecast of consumption. Without it you either over commit and pay a shortfall, or under commit and leave discount on the table.
Knowing what comparable buyers actually pay turns a list price conversation into a market price conversation.
Renewal windows, quarter ends, and the provider's own targets create moments where the same ask lands very differently.
The credible option of placing a workload on another cloud, or keeping it on private infrastructure, is the single strongest lever in the room.
Sizing a commitment to a forecast.
Coverage strategy is risk adjusted, not discount maximised. The goal is to cover the spend you are confident you will use, and flex the rest.
| Layer | Share of footprint | Instrument | Why |
|---|---|---|---|
| Stable base load | about 60 percent | 3 year commitments | Highest confidence, deepest discount |
| Predictable growth | about 20 percent | 1 year or flexible plans | Some confidence, keep exchange rights |
| Variable and spiky | about 20 percent | On demand and spot | Uncertain, avoid carrying unused commitment |
// shares are indicative; the right split is set by the confidence interval on your own forecast
Related reading
Commitments sit inside a wider program. See the cloud cost optimization guide for the full picture, the Azure cost optimization guide for the MACC in detail, and our white paper, The Azure MACC Negotiation Guide. When you are ready to act, our cloud commitment negotiation service runs the process for you.
People also ask
Frequently asked questions
Get the full negotiation guide
A practical, buyer side playbook for sizing and negotiating commitments across AWS, Azure, GCP, and OCI, with the leverage checklist and the drawdown math.
The Cloud Spend Navigator
One practical cloud cost lever, with the numbers, every two weeks. Written for buyers, never for vendors.
Put a defensible number on your cloud spend.
No provider in the room, no published price list. Tell us your footprint and we will scope the savings against your billing data — we reduce your cloud spend or we reimburse our service fee.
The Cloud Spend Navigator: what changed in cloud pricing, commitments, and FinOps — no vendor spin.