Cloud migration is the process of moving data, applications and operations from on-premises or legacy systems to a cloud environment, or from one cloud to another. A cloud migration strategy is the plan that decides, application by application, what moves, how it moves and in what order.
TL;DR
- Cloud migration moves data, applications and operations to public, private or hybrid cloud, or between clouds.
- AWS groups the ways to move an application into six strategies, the 6 Rs, and each application gets its own.
- A migration runs in five phases: assessment, planning, preparation, execution and optimization.
- Start with non-critical systems, test at every stage and keep a backup and recovery plan.
- The main risks are data loss, security gaps and business disruption, and a detailed plan reduces all three.
- Specialists help most with re-architecture, compliance, cost optimization and post-migration support.
What is cloud migration?
Cloud migration is the transfer of data, applications and services from traditional physical servers to cloud-based infrastructure. The target can be a public, private or hybrid cloud, and moving between cloud platforms is called cloud-to-cloud migration. Typical resources moved include data, applications and operations.
The aim is to modernize operations, streamline workflows and respond faster to market demands. It works only with careful planning, because every application has its own dependencies and risks. If the real question is whether to rebuild first, the rewrite vs refactor framework helps you decide, and the legacy modernization guide covers the wider picture.
Which cloud migration strategy fits each application?
There are six common strategies, known as the 6 Rs: rehost, replatform, repurchase, refactor, retire and retain. Rehost is the fastest and refactor the deepest. Choose per application, not once for the whole estate, based on its business value, its dependencies and the effort you can afford.
| Strategy | What it means | Typical fit |
|---|---|---|
| Rehost | Move the application as it is ("lift and shift") | Stable systems that need to leave the data center quickly |
| Replatform | Move with small changes, such as a managed database | Applications that gain from managed services without a rewrite |
| Repurchase | Replace with a SaaS product | Commodity functions such as email or CRM |
| Refactor | Re-architect for cloud-native services | Core systems that need scale and faster releases |
| Retire | Switch off what is no longer used | Duplicated or obsolete applications found in the inventory |
| Retain | Keep on premises for now | Systems with compliance, latency or cost constraints |
What are the phases of a cloud migration?
A cloud migration typically runs in five phases: assessment, planning, migration preparation, execution and optimization. Each phase has a clear output that the next one depends on. Timelines vary with the complexity of the environment, from weeks for small applications to months for large enterprise systems.
- Assessment: Audit your current environment, identify which applications are suitable for migration and map the dependencies between systems and any potential risks.
- Planning: Build a migration roadmap with goals, success criteria, timelines and cost estimates, and prioritize critical systems for early migration.
- Migration preparation: Choose your cloud provider and migration strategy, and set up a project team with the technical skills to oversee the process.
- Migration execution: Begin with non-critical systems and move data, applications and services in small increments, testing at each stage to keep disruption minimal.
- Optimization: After the move, monitor performance continuously and tune workloads for cost efficiency and security. The FinOps optimisation service covers the cost side.
What should a cloud migration checklist include?
A good checklist covers eight points: goals, inventory, provider, costs and timelines, risks, security, testing, and backup and recovery. Work through them before the first system moves. Each point turns a vague intention into a decision someone owns, which is what keeps a migration from stalling.
- Define goals and objectives: Set clear, measurable KPIs, such as cost reduction or improved application performance.
- Inventory your current environment: Document all applications, systems and data, along with their dependencies.
- Choose a cloud service provider: Evaluate providers on scalability, security and compliance features.
- Estimate costs and timelines: Estimate migration costs and the savings you expect afterwards, and set realistic timelines.
- Conduct a risk assessment: Identify risks such as data loss or downtime and prepare contingency plans.
- Establish security measures: Apply encryption, identity management and security protocols during and after the move.
- Plan for testing: Reserve time for post-migration testing to confirm that systems work as expected.
- Create a backup and recovery plan: Make sure a robust disaster recovery plan is in place before cutover.
What benefits can a cloud migration deliver?
The main benefits are scalability, cost efficiency, agility, security, disaster recovery and collaboration. They are not automatic, since they depend on the strategy chosen for each workload. A rehosted application gains mostly hosting flexibility, while a refactored one can use autoscaling, containers and serverless services.
Cloud platforms adjust resources to demand, so an e-commerce shop can absorb holiday traffic without overprovisioning. The original version of this article stated that companies often see a 20-40% reduction in IT operating expenses after migration. It named no study, so read it as an indication and measure your own baseline first.
Leading providers also offer encryption, identity management, continuous threat monitoring and compliance with standards such as GDPR, HIPAA and ISO 27001. Multi-region replication supports disaster recovery, and tools such as Google Workspace and Microsoft 365 enable real-time collaboration across locations.
Which risks stall a cloud migration?
The main risks are data loss, security vulnerabilities and extended business disruption. They usually come from missing dependency maps, untested cutovers or unclear ownership. A detailed plan with security controls, staged testing and a recovery strategy reduces all three, and it should be agreed before any production system moves.
The original post claimed that 80% of companies attempting do-it-yourself cloud migrations experience critical outages. It cited no source, so it is kept here only as the original author's statement and flagged for editorial review. What is certain is that skipping the assessment phase and the backup plan makes outages more likely.
How does working with specialists change a cloud migration plan?
Specialists add technical depth, risk mitigation, cost optimization, compliance knowledge and support after go-live. They are most useful when applications need re-architecting or when regulation applies. WWG supports migrations on platforms such as AWS, Azure and Google Cloud, from secure data transfer to performance tuning.
On risk, WWG runs pre-migration assessments and builds redundancy and failover into the plan to protect business continuity. On cost, it applies auto-scaling, rightsizing and reserved instances so you do not pay for unused resources. On compliance, it checks that the migration meets requirements such as GDPR, HIPAA or SOC 2 and runs regular security audits.
The work does not end at cutover: WWG continues monitoring, optimization and support as your needs change. See the cloud migration service and the DevOps migration service for how this is organized.
Planning a move to the cloud? Talk to a WWG engineer about your applications and the migration path that suits them.
Sources
FAQ
Frequently Asked Questions
A cloud migration takes from a few weeks for a small application to several months for a large enterprise system. The timeline depends on the complexity of your environment and the scope of the move. A thorough assessment up front is the best way to turn that range into a realistic estimate.
Yes, a phased approach is common and usually safer than a single cutover. Many businesses migrate non-critical systems first, refine the process on them and only then move essential operations. Each stage gives the team a chance to test, fix problems and adjust the plan before the stakes rise.
Work with a provider that offers encryption, identity and access management and regular security audits. Apply the same controls to data in transit and at rest, and run compliance checks against the regulations that apply to your industry. Agree these measures in the plan before any data moves.
After the migration, monitor performance, optimize for cost efficiency and keep security measures up to date against evolving threats. Treat optimization as an ongoing activity rather than a final phase, because usage patterns change over time. Review cost and performance reports regularly and adjust resources before waste builds up.
