Cloud Migration Checklist: 50 Things to Check Before You Migrate

Cloud migrations fail for predictable reasons: insufficient assessment, underestimated complexity, missed dependencies, and cost surprises. This 50-item checklist prevents all of them.

Phase 1: Assessment & Discovery (1–12)

1. Complete a full inventory of all servers, databases, and applications
2. Collect 14–30 days of actual CPU, memory, disk and network performance data
3. Map all application dependencies before you move anything
4. Identify all external dependencies — third-party APIs, licensing servers
5. Identify applications that cannot move to cloud (regulatory or technical)
6. Classify all data by sensitivity (public, internal, confidential, regulated)
7. Inventory all software licences and determine cloud eligibility (BYOL)
8. Document all custom integrations and APIs that need re-pointing
9. Identify all scheduled jobs, backup jobs, and maintenance windows
10. Identify workloads requiring specific hardware (GPU, FPGA, high-memory)
11. Run Azure Migrate, AWS Migration Evaluator, or equivalent assessment tool
12. Upload your assessment data to TCOIQ for multi-cloud TCO comparison

Phase 2: Cost & Business Case (13–20)

13. Calculate total current infrastructure TCO including hardware, data centre, staff
14. Get cloud pricing quotes for all workloads — not just compute
15. Factor in data egress costs for data-intensive workloads
16. Include one-time migration costs: tooling, parallel running, testing, training
17. Model reserved instance savings for baseline workloads
18. Calculate break-even timeline and present to stakeholders
19. Get budget approval with a 20% contingency buffer
20. Document expected ongoing optimisation savings

Phase 3: Security & Compliance (21–28)

21. Review all compliance requirements (GDPR, PCI-DSS, PDPA, SOC2)
22. Confirm target cloud has required compliance certifications
23. Design cloud network architecture (VPCs, subnets, security groups)
24. Plan identity and access management — SSO, MFA, least-privilege
25. Design encryption strategy for data at rest and in transit
26. Plan secrets management — never store credentials in code
27. Define security baseline policies enforced from day one
28. Plan cloud security monitoring and alerting strategy

Phase 4: Architecture & Design (29–36)

29. Decide migration strategy: Rehost, Replatform, Refactor, Retire, or Retain
30. Design target state architecture — don't copy on-premises patterns to cloud
31. Plan for high availability across at least two availability zones
32. Design disaster recovery — define RTO and RPO per workload class
33. Plan database migration — homogeneous or heterogeneous engine change
34. Design networking connectivity — VPN or Direct Connect/ExpressRoute
35. Plan DNS migration strategy (critical path item on go-live day)
36. Design tagging and resource organisation before creating anything

Phase 5–7: Execution, Cost Management & Post-Migration (37–50)

37. Define migration waves — start with simplest, non-critical workloads
38. Run a pilot migration before committing the full programme
39. Set up landing zone infrastructure before migrating any workloads
40. Test all workloads in cloud environment before production cutover
41. Define and test rollback procedures for every workload
42. Plan parallel running period with clear go/no-go criteria
43. Implement mandatory resource tagging from day one
44. Set up cloud cost budgets and alerts
45. Schedule non-production environments to stop outside business hours
46. Plan first reserved instance purchase for 30–60 days post-migration
47. Run cost optimisation review at 30, 60, and 90 days
48. Decommission all source infrastructure on schedule
49. Conduct post-migration retrospective
50. Share learnings with the broader team

TCOIQ automates items 11 and 12 — upload your discovery data and get a comprehensive multi-cloud TCO report in minutes.