An Azure reservation is a one or three year commitment to a specific resource, such as a virtual machine size in a region, that discounts that usage by roughly up to 72 percent against pay as you go. Its distinguishing feature for buyers is flexibility after purchase: reservations can be exchanged for others or refunded within published limits, which lowers the risk of committing to the wrong thing. The Azure savings plan for compute is the more flexible sibling, trading some discount depth for the ability to apply across instance types and regions like a spend rate commitment. The buyer's job is to cover only the steady floor that runs every hour, sized to a forecast you can defend, after rightsizing, so the commitment stays highly utilized.
Here is how reservations work, the exchange and refund rules, and how to choose between a reservation and the savings plan.
How does an Azure reservation work?
You commit to a quantity of a specific reservable resource, most commonly virtual machine instances of a given size in a region, for a one or three year term, paid upfront or monthly at no extra cost. Azure then applies the reservation discount to matching usage automatically, hour by hour, until the reserved quantity is consumed; usage beyond it bills at pay as you go. Reservation scope can be set to a single subscription or shared across an enrolment, and instance size flexibility lets a reservation apply across sizes within a family group, which raises utilization. The commitment is utilized only when matching resources actually run, so an oversized reservation strands spend just as a Savings Plan does on AWS.
What exchange and refund flexibility do reservations have?
Azure reservations can be exchanged for other reservations and refunded, within published limits and policies that have tightened over time, which is what makes them less risky than a pure use it or lose it commitment. Exchange lets you move a commitment to a different size, region, or term when the workload changes, and refund lets you exit, subject to limits. Because policies change, treat the current exchange and refund rules as indicative and verify them against Microsoft documentation before relying on them in a sizing decision. The practical point is that this flexibility lets you commit a little more confidently, but it is not a substitute for sizing to a real forecast.
Reservation or Azure savings plan: which should you buy?
| Dimension | Azure reservation | Azure savings plan for compute |
|---|---|---|
| What you commit to | A specific resource type in a scope | An hourly compute spend amount |
| Flexibility | Exchange and refund within limits | Applies across instance types and regions |
| Discount depth | Deeper on a stable, known workload | Slightly smaller for the added flexibility |
| Best for | Stable VMs that will not move | An evolving fleet whose mix shifts |
The common pattern is reservations on the stable, known core and an Azure savings plan on the flexible remainder. Discount figures are indicative and depend on resource, term, and region; verify against current Azure pricing.
How do you size reservation coverage?
Coverage is a risk decision. Look at the hourly run rate of the reservable resource over a representative period and find the floor that runs reliably every hour, then commit to that floor, not the average or the peak. Rightsize the virtual machines first so you are not reserving against oversized instances, and use instance size flexibility and shared scope to keep utilization high. Prefer one year terms where the forecast is less certain and three year terms only where demand is proven, accepting the deeper discount as payment for carrying more risk. Aim for high utilization rather than the highest coverage percentage.
A worked example
A scaling fintech was about to reserve three year capacity at its peak virtual machine count. The peak only ran during business hours, so a reservation at that level would have sat idle every night, stranding spend. Sizing instead to the genuine overnight floor after rightsizing, using shared scope and instance size flexibility, produced reservations that ran close to fully utilized, with the variable daytime demand left on pay as you go and an Azure savings plan covering the flexible remainder. The book stayed highly utilized as the fleet shifted, part of the program that left the company 41 percent lighter on cloud spend, verified against billing data and anonymised.
Frequently asked questions
How do Azure reservations work?
Can you cancel or change an Azure reservation?
Should I buy a reservation or the Azure savings plan?
Size your Azure commitments the right way
We help enterprises size and manage Azure reservations and savings plans to a forecast they can defend, with high utilization and no stranded spend, as an independent advisory that takes zero provider commissions and answers only to you. Our guarantee: we reduce your cloud spend or we reimburse our service fee, on a Fixed Fee or a no risk Gainshare basis. Download the Azure commitment kit, read the deeper Azure cost optimization guide, and compare the models in Azure savings plan versus reservations.
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.