A cloud migration assessment is the discovery-and-analysis step that produces the facts a migration plan needs. It inventories servers, applications, dependencies, data platforms, and licenses, then evaluates each for cloud readiness, right-sizing, and cost - yielding a total-cost-of-ownership projection, a recommended strategy per workload, and a sequenced roadmap.
Key Takeaways
-
An assessment inventories five things: servers and infrastructure, applications, dependencies, data platforms, and licensing.
-
Dependency mapping is the highest-value output - it shows which workloads must move together.
-
Microsoft's Azure Migrate is the first-party tool for discovery, assessment, and dependency analysis.
-
The deliverables matter most: right-sizing recommendations, a TCO projection, a per-workload strategy, and a roadmap.
-
Skipping assessment is the most common reason migrations overrun on cost and time.
What is a cloud migration assessment?
A cloud migration assessment is a structured discovery exercise that answers three questions before you move anything: what do we have, how does it interconnect, and what is the best way to move each piece. It replaces assumptions with evidence - a validated inventory, a dependency map, and performance data - so the migration plan rests on facts rather than tribal knowledge or out-of-date diagrams.
It is the foundation of any credible Azure migration and modernization effort and the first thing a disciplined program does. In Microsoft's Cloud Adoption Framework, assessment sits at the start of the migration journey, and in practice it is what makes the rest of the plan possible. If you are building the wider program, our guide to how to plan an Azure migration shows where assessment fits among goals, landing zone, and waves, and the broader stages of an Azure migration place it in the full sequence from discovery to optimization.
What do you inventory and evaluate?
An assessment captures five categories. Each contributes a distinct input to the migration decision.
Servers and infrastructure
Inventory every physical and virtual server: operating systems, CPU, memory, storage, and - critically - actual performance data over time, not nameplate specs. Performance data is what enables right-sizing, so the cloud target is matched to real usage rather than over-provisioned to match an aging on-premises box. Capture network configuration and any appliances (load balancers, firewalls) that workloads depend on.
Applications
Catalog the applications running on that infrastructure: technology stack, version and support status, integration points, owners, and business criticality. This is where end-of-support software, undocumented systems, and orphaned applications surface - each of which changes the plan. Business criticality drives sequencing; technical condition drives the strategy choice later.
Dependencies
Dependencies are the single most valuable thing an assessment uncovers. Applications rarely run in isolation - they call databases, services, and other applications, sometimes across the estate. Dependency mapping reveals these connections so workloads that talk to each other are grouped into the same migration wave. Migrating one half of a tightly coupled pair is a classic cause of outages, and only a dependency map prevents it.
Data platforms
Databases, data warehouses, and analytics workloads need their own evaluation: size, schema complexity, performance characteristics, and integration with reporting. Azure targets typically include Azure SQL, Azure Synapse, and Microsoft Fabric, and the assessment determines which fits each data estate and what migration approach preserves data integrity and reporting continuity. Where the data estate is large or messy, it is worth assessing alongside a data engineering and warehousing lens so the move sets up clean, well-modeled data rather than carrying problems forward.
Licensing and cost
Finally, capture the licensing position - operating systems, databases, and middleware - and current run costs. This feeds the total-cost-of-ownership comparison and surfaces opportunities such as bringing eligible licenses to Azure. It also keeps the business case honest: Azure consumption is billed by Microsoft and varies by service and usage, so the assessment models expected spend rather than assuming a flat saving.
What tools run an assessment?
Microsoft's Azure Migrate is the first-party hub for discovery and assessment. It supports agentless and agent-based discovery of servers, performance-based right-sizing, and dependency analysis that visualizes how machines and processes communicate. Microsoft also provides targeted assessments for SQL and web applications. Azure Migrate is the common backbone, often complemented by partner tooling and manual analysis for application-level nuance that automated discovery cannot infer - and Microsoft updates these capabilities over time, so specifics should be confirmed against current documentation.
Want an assessment that ends in a costed plan, not a spreadsheet? Centric runs discovery and assessment as the front of its Azure migration and modernization engagement - so you leave with right-sizing, TCO, and a per-workload strategy you can act on.
What does an assessment produce?
Tools gather data; the assessment's value is in the deliverables that turn data into decisions:
-
Validated inventory and dependency map - the authoritative picture of the estate and how it interconnects.
-
Right-sizing recommendations - cloud targets matched to real performance data, avoiding the over-provisioning that makes a naive lift-and-shift expensive.
-
Total-cost-of-ownership projection - current-state versus target-state cost, including expected Azure consumption, to ground the business case.
-
Migration strategy per workload - a recommended approach for each application from the 6 Rs (rehost, replatform, refactor, rearchitect, rebuild, replace). Choosing well is its own discipline; our framework for lift-and-shift vs. refactor vs. rebuild explains how workload characteristics map to each option.
-
Sequenced roadmap - workloads grouped into waves by dependency and risk, with a low-risk pilot first and milestones throughout.
Together these are the difference between an assessment that informs a decision and one that just describes the estate.
How long does an assessment take?
It scales with the size and complexity of the environment. Discovery typically runs for a period long enough to capture representative performance and dependency data, after which analysis and deliverable production follow. Centric's published guidance on overall migration timelines - a few weeks to several months depending on environment size, workload complexity, and approach - applies to the program as a whole; the assessment is an early, time-boxed portion of that, and it is the part most worth doing thoroughly.
How does Centric run an Azure migration assessment?
As a Microsoft partner, Centric runs assessment as the opening of its *Assess and Plan* phase: discovering infrastructure, applications, and data platforms; using Microsoft tooling such as Azure Migrate for dependency analysis and right-sizing; and translating the findings into a per-workload strategy and a sequenced roadmap with milestones. The output is built to be acted on - a costed plan, not a static audit. For a closer look at the method and what clients receive, see how Centric runs migration assessments.
Done well, the assessment de-risks everything that follows: the business case is grounded in real cost data, the wave plan reflects real dependencies, and each workload moves the right way.
Considering a move to Azure? Start with the facts. Centric's Azure migration and modernization team can run a full assessment and hand you a right-sized, costed, sequenced plan.
Frequently Asked Questions
What is a cloud migration assessment?
A structured discovery-and-analysis step that inventories servers, applications, dependencies, data platforms, and licenses, evaluates each for readiness, right-sizing, and cost, and produces a TCO projection, a per-workload strategy, and a sequenced roadmap.
What do you inventory in a migration assessment?
Five categories: servers and infrastructure (with real performance data), applications (stack, support status, criticality), dependencies between workloads, data platforms (databases, warehouses, analytics), and licensing and current run cost.
What is Azure Migrate?
Microsoft's first-party hub for migration discovery and assessment. It supports server discovery, performance-based right-sizing, and dependency analysis, plus targeted assessments for SQL and web apps. Microsoft updates its capabilities over time.
What is dependency mapping?
The process of identifying how applications and servers communicate, so interdependent workloads are migrated together. It is the highest-value assessment output, because moving half of a coupled pair is a common cause of outages.
Do I need an assessment before migrating?
For anything beyond a trivial estate, yes. Skipping assessment is the most common reason migrations overrun on cost and time, because the plan ends up built on assumptions rather than validated inventory, dependency, and performance data.
