Reserved Instances, Savings Plans, and Committed Use: The Cloud Discount Playbook Most Businesses Ignore
5/19/20263 min read
Public cloud providers — AWS, Microsoft Azure, and Google Cloud Platform — offer powerful discount programs that can reduce compute costs by 30 to 72 percent compared to standard on-demand pricing. These programs are well-documented in provider documentation and are actively discussed in cloud architecture communities. They are also consistently underutilized by mid-market organizations, which continue paying on-demand rates for workloads that have been running predictably for months or years.
The reason for this gap is not ignorance. Most IT leaders are aware that reserved instances and savings plans exist. The gap exists because optimizing cloud spend requires someone to own it — to analyze usage patterns, model commitment scenarios, and execute the purchasing decisions. In organizations without a dedicated FinOps function, that ownership falls through the cracks between IT, finance, and operations.
This post is a practical guide to the three main cloud discount mechanisms — what they are, how they work, and how to evaluate which combination is right for your environment.
On-demand pricing is what cloud providers charge customers who have not made any commitment. It is the most expensive way to run a predictable workload — and it is what the majority of mid-market businesses are paying for the majority of their cloud compute.
Mechanism 1: Reserved instances
Reserved instances — called Reserved Instances on AWS, Reserved VM Instances on Azure, and Committed Use Discounts on GCP — are capacity commitments that exchange flexibility for price. You commit to using a specific instance type in a specific region for one or three years. In exchange, the provider discounts the hourly rate by 30 to 72 percent depending on the commitment term and payment structure.
The discount math: an AWS m5.xlarge instance running on-demand costs approximately $0.192 per hour — $1,676 per year. The same instance on a three-year reserved commitment with all-upfront payment costs $0.073 per hour — $639 per year. That is a 62 percent reduction for a single instance. On a 50-instance deployment, the annual savings exceed $50,000.
The risk: reserved instances are most effective for workloads with stable, predictable usage profiles. If a workload is decommissioned before the reservation term expires, the remaining commitment still bills. Reservation strategy requires workload stability analysis before purchase.
Mechanism 2: Savings plans
AWS Savings Plans and Azure's equivalent offer a more flexible alternative to reserved instances. Rather than committing to a specific instance type, you commit to a minimum dollar amount of compute spend per hour — and receive discounts of 20 to 66 percent on all usage that falls within that commitment, regardless of instance type, size, or region.
Savings Plans solve the primary limitation of reserved instances: instance type flexibility. If your workload mix evolves — you migrate from one instance family to another, or your compute needs shift between regions — your savings plan discount applies automatically. For organizations whose infrastructure is actively evolving, savings plans typically deliver better outcomes than instance-specific reservations.
Mechanism 3: Enterprise discount programs
Beyond the self-service discount mechanisms available to all customers, AWS, Azure, and GCP all operate enterprise discount programs that require negotiation with a dedicated account team. AWS Enterprise Discount Program, Azure Monetary Commitment, and GCP's Private Pricing Agreements offer 15 to 40 percent reductions from standard pricing in exchange for annual spend commitments.
These programs are available to organizations spending approximately $100,000 or more per year on cloud services — a threshold that most mid-market businesses with meaningful cloud deployments have already crossed. Access requires engaging your provider's enterprise sales team directly. Most mid-market organizations have never had this conversation. The account rep who handles your support tickets is not the person who can offer enterprise pricing. You need to ask for an enterprise account team engagement specifically.
How to build the right combination
The optimal discount strategy for most mid-market organizations is a layered approach: reserved instances or savings plans for the stable, high-utilization base layer of compute; on-demand pricing for variable and burst workloads; and an enterprise discount negotiation for organizations meeting the spend threshold.
Building this strategy requires a usage analysis — minimum 90 days of consumption data — to identify which workloads have the stable utilization profiles that make commitment purchasing appropriate. Running reservation coverage reports in your cloud provider's Cost Explorer or equivalent tool is the starting point. From there, the coverage gaps reveal where on-demand pricing is being paid unnecessarily.
Sigma Technology Consulting conducts cloud discount strategy analysis as part of our cloud cost audit service. If your organization is spending more than $50,000 per month on cloud compute and has not evaluated reservation coverage in the past 12 months, contact us at sigmatechconsult.com. The savings in most cases cover the cost of the engagement within the first billing cycle.
Sigma Technology Consulting, Inc.
25 Years of Experience, Vetting & Procuring Technology Vendors
Contact Us
Support
© 2026. All rights reserved.
