Cloud migration assessment tools: what they measure and which to use

Which assessment tools do AWS, Azure and Google Cloud offer?

Each provider has a free or low-cost assessment toolset aimed at moving workloads to its own platform. Use the one for your target cloud first, then add third-party tools where you need a neutral or multi-cloud view.

AWS: AWS Transform and Migration Evaluator

AWS has changed its tooling recently. The AWS Application Discovery Service documentation now says the service is no longer open to new customers and recommends AWS Transform for similar capabilities. Migration Hub has also closed to new customers.

According to the AWS Transform user guide, the service collects and analyses on-premises server inventory from multiple sources, groups applications, identifies dependencies and builds prioritised migration waves. It can also design a landing zone and translate source network configuration into AWS VPCs. AWS states there is no additional charge for using it.

Migration Evaluator focuses on the business case. It uses an agentless collector or uploaded inventory exports to find over-provisioned servers, compare licence options (bring your own licence versus licence included) and project AWS costs. It is requested through AWS and usually delivered with AWS or a partner. For more on the AWS side, see our AWS migration assessment guide.

Azure Migrate discovery and assessment

Azure Migrate uses a lightweight appliance to discover VMware, Hyper-V and physical servers, SQL Server instances and web apps, then produces readiness, sizing and cost assessments and a business case. Microsoft lists Azure Migrate as a free service; partner tools used inside it may charge. Its dependency analysis captures TCP connection data without agents for VMware, Hyper-V, physical servers and servers in other public clouds. Microsoft recommends the agentless option; agent-based analysis only exists in the classic view, which is scheduled for deprecation by the end of 2026.

Google Cloud Migration Center

Google Cloud Migration Center scans on-premises environments for servers and SQL Server, MySQL and PostgreSQL databases, or accepts manual uploads. It generates rapid cost estimates and TCO reports, identifies application and network dependencies, and gives technical fit suggestions for each workload.

Third-party and manual tools

Independent tools such as Flexera One and Device42 provide discovery and dependency mapping that is not tied to one cloud. They help when you are choosing between providers, when your estate spans several data centres, or when you need a CMDB that outlives the migration. Many teams also start with simple exports, for example a VMware inventory from vCenter or RVTools, and a spreadsheet of application owners. Manual data is quick to gather but goes stale quickly and misses dependencies.

How do you choose an assessment tool?

Pick the native tool for your likely target cloud, then check it against your estate. If you have not chosen a cloud yet, run a neutral tool or two native tools side by side.

• Coverage: does it support your hypervisors, physical servers, operating system versions and databases?
• Collection method: agentless collectors are faster to deploy; agents give deeper process-level data. Check what your security team will approve.
• Data location: where is collected data stored, and does that fit your GDPR and internal policies? Inventory data often includes hostnames, IPs and user names.
• Dependency depth: network connections only, or also application and process detail?
• Cost model quality: does it support reserved pricing, licence reuse and your negotiated discounts?
• Exports: can you export results to your planning tools, or are you locked into one console?

How to run an assessment with these tools

An assessment of a few hundred servers usually takes weeks rather than days, mostly because of the data collection window. The steps:

1. Define scope and goals. Data-centre exit date, cost target, compliance limits and which systems are out of scope.
2. Deploy collectors. Install the appliance or agents, open the needed firewall rules and confirm data is arriving.
3. Collect over a full business cycle. Include month-end, payroll or seasonal peaks so sizing reflects real demand.
4. Validate with owners. Interview application owners to confirm dependencies, downtime tolerance and business priority.
5. Choose a strategy per application. Use the 7 Rs of migration: retire, retain, rehost, relocate, repurchase, replatform or refactor.
6. Build the business case. Right-sized cloud cost, migration effort, double-running period and licence impact.
7. Plan waves. Group dependent systems, start with a low-risk pilot and schedule the rest.

Common limitations of assessment tools

Tools save weeks of manual work, but they have blind spots. Plan for these:

• Short collection windows miss peaks and give undersized recommendations.
• Like-for-like cost estimates assume a rehost. Replatforming to managed databases or containers changes the cost picture.
• Hidden dependencies such as batch jobs that run quarterly, file shares or hard-coded IP addresses can be missed.
• Licensing rules for Microsoft, Oracle and other vendors need human review; a tool's licence view is a starting point.
• Data egress and support costs are often left out of tool estimates. Add them before you present the business case, and use cost optimisation practices from the first wave.

Example: what an assessment phase delivers

In Opsio's SilverRail Technologies migration, the assessment phase was a full infrastructure and needs analysis carried out with SilverRail's architects. It identified the key pain points, set the target architecture on AWS, and produced a business case with ROI calculations, a risk and feasibility report and project costs, plus a roadmap for the later phases. The next phase then tested those findings with a pilot migration of selected workloads before the main cutovers.

How Opsio helps

Opsio runs cloud migration assessments for AWS, Azure and Google Cloud, combining native tools with owner interviews to produce a costed plan and migration waves. See our cloud migration services.

Frequently asked questions

What is the best cloud migration assessment tool?

The best tool is usually the native one for your target cloud: AWS Transform and Migration Evaluator for AWS, Azure Migrate for Azure, and Migration Center for Google Cloud. Add a third-party tool if you need a neutral comparison across clouds.

Are cloud migration assessment tools free?

The native ones mostly are. Microsoft lists Azure Migrate as free and AWS says AWS Transform has no additional charge. Third-party discovery platforms are licensed commercially, and you still need staff time for interviews and validation.

How long does a cloud migration assessment take?

For a few hundred servers, plan for several weeks. Most of that time is performance data collection over a full business cycle, so start the collectors early.

What replaced AWS Application Discovery Service?

AWS states that Application Discovery Service is no longer open to new customers and recommends AWS Transform, which provides discovery, dependency analysis and wave planning.

Do assessment tools find all application dependencies?

No. They detect network connections between servers during the collection window. Infrequent jobs, shared files and business process links need to be confirmed with application owners.