In this comprehensive article, I will walk you through how to use Azure Migrate Assessment to evaluate your enterprise estate, optimize workload sizing, and build an airtight business case for your migration to Microsoft Azure.
Table of Contents
- Azure Migrate Assessment
- What is an Azure Migrate Assessment?
- The Key Benefits of Running an Assessment First
- Evaluating Your Workloads: The 4 Core Assessment Types
- The 4-Step Azure Migrate Assessment Workflow
- Deep-Dive: Understanding Performance-Based Rightsizing
- Licensing & Cost Optimization: Unlocking Enterprise Savings
- Best Practices for Enterprise Cloud Readiness
- Summary
Azure Migrate Assessment
What is an Azure Migrate Assessment?
At its core, Azure Migrate Assessment is an intelligence layer built into the central Azure Migrate portal. Instead of relying on static inventory spreadsheets or manual server surveys, Azure Migrate deploys a lightweight appliance into your local environment to continuously discover servers, databases, web applications, and file shares.

Once telemetry is gathered, the assessment engine evaluates your estate against Azure infrastructure targets. It provides answers to four critical questions every Chief Technology Officer (CTO) asks:
- Readiness: Are my current workloads technically compatible with Azure?
- Sizing: What specific Azure VM SKUs, SQL tiers, or App Service plans match my workload requirements without over-provisioning?
- Cost: What will my monthly Azure bill look like, taking into account Azure Hybrid Benefit and Reserved Instances?
- Dependencies: Which servers talk to each other, and which applications must be migrated together to avoid latency breakdown?
Architectural Insight: An assessment is a non-destructive, read-only exercise. It runs in parallel with your production workloads without impacting performance or modifying local configurations.
The Key Benefits of Running an Assessment First
- Eliminates Over-Provisioning: On-premises hypervisors are notoriously over-provisioned. A physical server might have 64 vCPUs allocated, but performance telemetry might reveal it operates at under 8% average CPU usage. Azure Migrate rightsizes your cloud footprint based on actual utilization rather than allocated capacity, frequently cutting projected cloud costs by 30% to 50%.
- Identifies Blockers Early: Obsolete operating systems, non-standard disk formats, unsupported database configurations, or unsupported kernel versions are surfaced during the assessment phase—allowing your team to remediate issues before scheduling migration waves.
- Maps Unregistered Dependencies: In legacy datacenters, documentation is rarely up to date. Automated dependency analysis reveals hidden database connections, background API calls, and shared network drives that would otherwise cause application outages post-migration.
- Generates Board-Ready Business Cases: Azure Migrate includes a built-in Business Case tool that pairs infrastructure costs with operational savings, carbon footprint reduction (sustainability metrics), and software licensing optimizations.
Evaluating Your Workloads: The 4 Core Assessment Types
| Assessment Type | Target On-Premises Workloads | Key Evaluation Output | Primary Cost Drivers Analyzed |
| Azure VM Assessment | VMware, Hyper-V, Physical Servers, AWS/GCP Instances | Azure IaaS VM SKU sizing, OS compatibility, disk storage types (Standard, Premium, Ultra). | vCPU count, RAM, IOPS, throughput, Azure Hybrid Benefit (AHUB). |
| Azure SQL Assessment | On-prem SQL Server instances, Failover Clusters, AGs | Modernization readiness for Azure SQL Database, Azure SQL Managed Instance, or Azure SQL VM. | Compute vCores, storage capacity, IOPS, feature parity checks. |
| Azure App Service Assessment | IIS Web Applications, ASP.NET, Java apps | Platform-as-a-Service (PaaS) compatibility for Azure App Service plans. | Port bindings, authentication providers, ISAPI filters, web config settings. |
| Azure Files & Storage | On-premises SMB and NFS file shares (Windows/Linux) | Azure Files SKU recommendations (Standard vs. Premium, redundant media types). | Total storage volume, active file throughput, churn rates, redundancy needs. |
The 4-Step Azure Migrate Assessment Workflow
To execute an accurate assessment, follow this structured, phased workflow across your enterprise infrastructure.
1. Deploy the Azure Migrate Appliance:1 – 2 Hours.
Download and deploy the lightweight appliance VM (OVA for VMware, VHD for Hyper-V, or script for physical servers) into your local environment. Connect the appliance to your Azure Migrate project using secure outbound HTTPS credentials.
2. Execute Agentless Continuous Discovery:Â 1- 30 Days.
Allow the appliance to discover inventory across hypervisors and OS guest layers. Enable agentless application discovery to identify running software, installed roles, and database instances across your enterprise servers.
3. Analyze Application Dependencies: Continuous Monitoring.
Enable agentless dependency analysis across critical server groups. The appliance captures incoming and outgoing TCP connections to map multi-tier application architectures and identify dependent server groups.
4. Configure & Run Assessment Reports: Interactive Analysis.
Create targeted assessment reports inside the Azure portal. Customize properties including target Azure region, sizing criteria (performance-based vs. as-on-premises), percentile utilization (P95/P99), comfort factors, and licensing options.
Deep-Dive: Understanding Performance-Based Rightsizing
The single most powerful configuration knob inside Azure Migrate Assessment is the Sizing Criterion. You can choose between two fundamental approaches:

1. As On-Premises Sizing
This setting takes your current local server allocations (e.g., 16 cores, 64 GB RAM) and finds the exact matching Azure VM size. While safe, this approach carries forward years of legacy inefficiency and over-provisioning into the cloud.
2. Performance-Based Sizing (Recommended)
Performance-based sizing analyzes historical performance telemetry collected by the appliance—including CPU usage, memory utilization, disk IOPS, disk throughput, and network adapter throughput.
When configuring a performance-based assessment, pay close attention to these three properties:
- Performance History: Set this buffer duration to at least 30 days. This captures full monthly business cycles, end-of-month financial processing, and weekly batch jobs.
- Percentile Utilization: Default to the 95th percentile (P95) or 99th percentile (P99). This filters out temporary anomalies or brief reboot spikes while ensuring your target Azure VM comfortably handles sustained peak operational loads.
- Comfort Factor: A safety multiplier applied on top of performance metrics (e.g., a Comfort Factor of 1.3x adds a 30% buffer to account for future organic business growth).
Licensing & Cost Optimization: Unlocking Enterprise Savings
An assessment is not merely a technical readiness check; it is a strategic financial model. To build an accurate financial plan for your CFO, your assessment configuration must incorporate Microsoft’s key cloud licensing mechanisms:

Azure Hybrid Benefit (AHUB)
If your organization maintains active Software Assurance (SA) on Windows Server and SQL Server core licenses on-premises, you can apply Azure Hybrid Benefit inside the assessment settings.
- Windows Server AHUB: Waives the base Windows OS licensing fee on Azure IaaS VMs, allowing you to pay the lower Linux compute rate.
- SQL Server AHUB: Allows you to swap on-premises SQL Server core licenses into Azure SQL Database, SQL Managed Instance, or SQL IaaS VMs, often delivering up to 55% savings on database workloads.
Reserved Instances (RIs) & Savings Plans
For predictable, steady-state workloads (such as core domain controllers, ERP servers, and production database instances), select 1-Year or 3-Year Reserved Instances inside your assessment settings. Combining 3-Year Reserved Instances with Azure Hybrid Benefit can yield total savings of up to 72% compared to standard pay-as-you-go pricing.
Best Practices for Enterprise Cloud Readiness
To ensure your Azure Migrate Assessment delivers actionable results, adhere to these recommendations:
- Grant Sufficient Appliance Permissions: Ensure the discovery appliance has read-only administrative access across vCenter Server or Hyper-V hosts. For deeper guest-level discovery (identifying installed applications and SQL instances), supply domain-level guest credentials.
- Filter Out Decommissioned Workloads: Before running your final report, tag and remove legacy VMs slated for retirement. Including dead servers skews capacity metrics and bloats projected cloud spend.
- Group Workloads by Application Boundaries: Rather than creating one massive assessment containing 2,000 servers, construct logical Assessment Groups based on application dependencies (e.g., “ERP System”, “Customer Portal”, “Data Warehouse”).
- Evaluate Modernization Targets (PaaS over IaaS): Look closely at the Azure SQL and App Service assessment recommendations. Rehosting database servers to Azure SQL Managed Instance eliminates OS management, backup orchestration, and manual patching overhead while reducing infrastructure compute costs.
Summary
Executing a thorough Azure Migrate Assessment transforms an enterprise cloud migration from an unpredictable risk into a phased, cost-optimized engineering discipline.
Ready to begin your cloud transformation journey?
- Set up an Azure Migrate Project in your target Azure tenant and region.
- Deploy the Azure Migrate discovery appliance into your local environment.
- Allow telemetry to collect for a minimum of 30 days to capture peak utilization cycles.
- Generate rightsized cost reports incorporating Azure Hybrid Benefit and Reserved Instance discounts.
You may also like the following articles:

I am Rajkishore, and I am a Microsoft Certified IT Consultant. I have over 14 years of experience in Microsoft Azure and AWS, with good experience in Azure Functions, Storage, Virtual Machines, Logic Apps, PowerShell Commands, CLI Commands, Machine Learning, AI, Azure Cognitive Services, DevOps, etc. Not only that, I do have good real-time experience in designing and developing cloud-native data integrations on Azure or AWS, etc. I hope you will learn from these practical Azure tutorials. Read more.
