Cloud migration projects fail for predictable reasons: insufficient assessment, unrealistic timelines, and treating migration as a purely technical exercise. This checklist distills lessons from dozens of migrations into a practical, phased approach.
Phase 1: Assessment (Weeks 1-4)
Inventory your environment: - [ ] Catalog all servers, databases, and applications - [ ] Map dependencies between services (which talks to which) - [ ] Document current resource utilization (CPU, memory, storage, network) - [ ] Identify compliance requirements (data residency, industry regulations) - [ ] List all integrations with external systems
Classify workloads by migration strategy: - Rehost (lift-and-shift) — Move as-is to cloud VMs. Fastest, lowest risk. - Re-platform — Adopt managed services (e.g., self-managed DB to RDS). Moderate effort, good ROI. - Refactor — Rearchitect for cloud-native (containers, serverless). Highest effort, highest long-term benefit. - Retire — Decommission workloads that are no longer needed. - Retain — Keep on-premises (regulatory, latency, or cost reasons).
Financial modeling: - [ ] Calculate current total cost of ownership (include facilities, power, staff) - [ ] Estimate cloud costs for each workload using provider calculators - [ ] Factor in migration costs (tooling, temporary dual-running, staff time) - [ ] Define the break-even timeline
Phase 2: Planning (Weeks 4-8)
Architecture design: - [ ] Choose target cloud provider(s) and regions - [ ] Design network topology (VPC, subnets, peering, VPN/Direct Connect) - [ ] Plan identity and access management (SSO, IAM roles, service accounts) - [ ] Define security baselines (encryption, firewall rules, monitoring) - [ ] Select managed services for databases, queues, storage
Migration waves: - [ ] Group workloads into migration waves (start with low-risk, non-critical) - [ ] Define success criteria for each wave - [ ] Plan rollback procedures for each workload - [ ] Schedule migration windows to minimize business impact - [ ] Assign ownership for each workload migration
Phase 3: Foundation (Weeks 8-12)
Build the landing zone: - [ ] Provision accounts/projects with proper organizational structure - [ ] Deploy networking (VPC, subnets, NAT, DNS, VPN) - [ ] Configure identity federation (SSO, MFA) - [ ] Set up monitoring and logging (CloudWatch, Datadog, ELK) - [ ] Implement IaC (Terraform/Pulumi) for all foundation resources - [ ] Deploy CI/CD pipelines for infrastructure changes
Phase 4: Migration Execution (Weeks 12-24+)
For each wave: - [ ] Provision target infrastructure - [ ] Migrate data (with integrity validation) - [ ] Deploy application components - [ ] Run parallel testing (old and new environments) - [ ] Validate performance benchmarks - [ ] Switch traffic (DNS, load balancer, or blue-green) - [ ] Monitor closely for 48-72 hours post-migration - [ ] Decommission source infrastructure only after validation period
Phase 5: Optimization (Ongoing)
- Right-size instances based on actual utilization data
- Implement autoscaling for variable workloads
- Purchase reserved instances or savings plans for stable workloads
- Set up cost monitoring and budget alerts
- Schedule quarterly optimization reviews
Common Pitfalls to Avoid
- 1Migrating everything at once — Wave-based migration reduces risk
- 2Ignoring data transfer costs — Egress fees add up, especially for multi-cloud
- 3Skipping the rollback plan — Every migration needs a tested way back
- 4Underestimating DNS TTL — Lower TTLs before migration day
- 5Forgetting compliance — Data residency requirements can block entire migrations
Conclusion
A successful cloud migration is a business transformation, not just a technical project. The checklist above provides structure, but the key success factor is cross-functional alignment: IT, security, finance, and business stakeholders all need to understand the plan, the timeline, and their role in the process.