To Reserve or Not to Reserve

Introduction

I get asked a lot about whether an organization should reserve capacity for their Azure App Services or continue paying as you go. They are concerned that they might commit to much more than they need and financially might not work out for them after 1 or 3 years. let’s break this down thoroughly so you get the full picture of Azure App Service Reservations, how they work, and why you’d want to use them

What is an Azure App Service Reservation?

Azure App Service is a PaaS (Platform as a Service) that lets you host web apps, APIs, and mobile backends without worrying about infrastructure. You pay for the underlying App Service Plan, which defines the compute resources (VM instances, cores, memory, scaling, etc.).

Normally, App Service Plans are billed on a pay-as-you-go (PAYG) basis (hourly). However, with App Service Reservations, you can pre-commit to using App Service capacity for 1 year or 3 years in exchange for a discounted price.

So:

PAYG = Flexible, but higher cost.
Reservation = Commitment, but much lower cost.

How do Reservations Work?

You choose the scope of the reservation:
1. Single subscription → applies to that subscription only.
2. Shared scope (billing scope) → applies to all subscriptions in the same billing account.
3. Management group scope → covers subscriptions under a specific management group.
You select the reservation details:
1. Resource: App Service Plan
2. Region (e.g., East US, West Europe, etc.)
3. SKU (pricing tier, e.g., Premium v3 P1v3, P2v3, etc.)
4. Term: 1 year or 3 years
5. Quantity: Number of instances (cores) you want to cover with the reservation
Azure automatically applies the reservation benefit:
1. When you run workloads in App Service that match the region + SKU + resource type of the reservation, the PAYG charges are replaced by prepaid reserved capacity charges.
2. If you run more instances than reserved, you pay PAYG for the excess.

How Does it Save Money?

Discounts are typically 35%–55% off PAYG rates (depending on SKU and term).
Best savings are with 3-year commitments.
Example:
Let’s say you’re running a P1v3 App Service Plan (2 cores) in East US.
- PAYG = ~$146/month per core
- 1-year reserved = ~$100/month per core
- 3-year reserved = ~$75/month per core
  That’s a 50%+ savings if you commit for 3 years.

Key Benefits

Cost savings: Predictable workloads (always-on apps, production systems) save the most.
Flexibility:
- You can exchange reservations (e.g., switch region or SKU) if business needs change.
- You can cancel reservations (with a 12% early termination fee).
Automatic application: No need to assign a reservation to a specific App Service Plan; Azure will apply it wherever it finds a matching workload.
Budget predictability: You prepay, so you avoid variable PAYG spikes.
Scoped optimization: You can apply it at shared scope to cover multiple subscriptions.

Limitations & Things to Know

Applies only to App Service Plans, not to other PaaS (Functions, Logic Apps, etc. unless hosted on App Service Plan).
You must pick the right SKU and region. Reservations are SKU/region-specific.
If you use fewer instances than reserved, you still pay for the full reservation (no refunds for underuse).
Reservations don’t cover networking, storage, or additional services, only compute cost of App Service Plan.
Billing is either monthly payments (spreading the cost) or all upfront.

When Should You Use Reservations?

Use App Service Reservations when:
You have steady, predictable workloads that will run continuously (production apps, corporate websites, APIs).
You want to cut costs long-term and can commit for 1 or 3 years.
Your workloads don’t frequently change regions/SKUs.
You’re managing multiple subscriptions and want to optimize enterprise-wide costs.

Avoid reservations if:

Your workload is short-term, experimental, or highly variable.
You’re not sure which region/SKU you’ll need long-term.

Summary

At the end, it is a calculated process to find out if it will help your organization, keeping in mind the following items:

Save up to 55% over PAYG rates.
Lock in predictable costs for budgeting.
Enterprise flexibility with exchange/cancel options.
Applies automatically across subscriptions for maximum coverage.
Perfect for always-on apps where compute demand won’t disappear.

Reach out to our experts at The Training Boss, we can help you identify the need and the best short term and long term solution for your business.