1.4 The Consumption-Based Model
Key Takeaways
- The consumption-based model means you pay only for the cloud resources you actually use — no upfront commitments required.
- Benefits include no upfront costs, no wasted resources, pay-as-you-go flexibility, and the ability to stop paying for resources you no longer need.
- Azure meters resource usage by the second, hour, or by transactions depending on the service.
- Reserved Instances and Savings Plans offer significant discounts (up to 72%) in exchange for 1-year or 3-year commitments.
- The consumption-based model eliminates the risk of over-provisioning or under-provisioning capacity.
The Consumption-Based Model
Quick Answer: The consumption-based model means you pay only for what you use. No upfront costs, no wasted resources. When you stop using a resource, you stop paying for it. Azure meters usage by the second, hour, or transaction.
How the Consumption-Based Model Works
In traditional IT, you purchase hardware upfront (CapEx) based on estimated future demand. If you overestimate, you waste money on idle hardware. If you underestimate, you cannot serve your users. The consumption-based model eliminates this problem.
Key Principles
- No upfront costs — No need to purchase hardware or commit to long-term contracts (unless you choose to for discounts)
- No wasted resources — You scale down or deallocate when demand drops
- Pay for what you use — Billing is based on actual consumption
- Stop paying when you stop using — Deallocate a VM and you stop incurring compute charges
Azure Pricing Models
Azure offers several pricing models to suit different workload patterns:
Pay-As-You-Go (PAYG)
The default pricing model where you pay for resources per second, per hour, or per transaction with no upfront commitment.
- Best for: Unpredictable workloads, development/testing, short-term projects
- Flexibility: Maximum — scale up or down at any time
- Discount: None (full list price)
Reserved Instances (RIs)
Commit to using a specific resource (VM, database, etc.) for 1 year or 3 years in exchange for a significant discount.
- Best for: Steady-state, predictable workloads that run 24/7
- Flexibility: Low — committed to the reserved capacity
- Discount: Up to 72% off pay-as-you-go pricing
Azure Savings Plans
Commit to a consistent amount of compute spending (measured in $/hour) for 1 year or 3 years. More flexible than Reserved Instances because the commitment applies across VM families and regions.
- Best for: Organizations with consistent compute spending but changing VM needs
- Flexibility: Medium — committed to spending amount, flexible on specific resources
- Discount: Up to 65% off pay-as-you-go pricing
Spot VMs
Purchase unused Azure compute capacity at deep discounts (up to 90% off). Azure can evict your VM at any time when it needs the capacity back.
- Best for: Fault-tolerant workloads — batch processing, dev/test, CI/CD pipelines
- Flexibility: High — but Azure can evict your VM with 30 seconds notice
- Discount: Up to 90% off pay-as-you-go pricing
Pricing Model Comparison
| Model | Commitment | Discount | Best For | Risk |
|---|---|---|---|---|
| Pay-As-You-Go | None | 0% | Variable workloads | Highest cost per unit |
| Reserved Instances | 1 or 3 years | Up to 72% | Steady-state workloads | Committed even if unused |
| Savings Plans | 1 or 3 years | Up to 65% | Consistent compute spend | Committed to spend amount |
| Spot VMs | None | Up to 90% | Fault-tolerant workloads | VM can be evicted anytime |
How Azure Meters Usage
Azure measures resource usage through meters that track consumption:
| Resource Type | How It Is Metered | Example |
|---|---|---|
| Virtual Machines | Per second of compute time | A D2s_v3 VM running for 730 hours/month |
| Storage | Per GB stored per month + transactions | 100 GB of Blob storage at Hot tier |
| Networking | Per GB of outbound data transfer | 50 GB of data leaving Azure |
| Databases | Per DTU/vCore per hour + storage | Azure SQL with 2 vCores for 730 hours |
| Functions | Per execution + per GB-second of memory | 1 million executions at 128 MB memory |
Important Billing Details
- Inbound data (data INTO Azure) is free in most cases
- Outbound data (data LEAVING Azure) is metered and billed after a free allowance
- Stopped (deallocated) VMs do not incur compute charges but DO incur storage charges for their disks
- Deleted resources stop all charges immediately
On the Exam: Remember that deallocating a VM stops compute charges but not storage charges. To completely stop charges, you must delete the VM AND its associated disks.
Factors That Affect Azure Costs
Several factors influence the total cost of an Azure solution:
| Factor | Impact |
|---|---|
| Resource type | Different services have different pricing (VMs cost more than storage) |
| Resource size | Larger VMs, more storage = higher cost |
| Region | Prices vary by Azure region (US East may differ from West Europe) |
| Bandwidth | Outbound data transfer is metered |
| Reserved capacity | Commitments reduce costs significantly |
| Azure Hybrid Benefit | Use existing Windows Server or SQL Server licenses to save |
On the Exam: You do NOT need to calculate specific prices. You need to understand the FACTORS that affect cost and which pricing model is appropriate for a given scenario.
Which Azure pricing model offers the largest potential discount?
A company deallocates a Virtual Machine. Which charges CONTINUE?
Which statement about data transfer pricing in Azure is correct?