Cloud and DevOpsJanuary 28, 202615 min read

The Enterprise Cloud Migration Playbook: From Assessment to Cutover

Maria ChenCloud Practice Lead, CodeXcape
Planet Earth at night showing global network connectivity and cloud infrastructure

Cloud migration programs fail for predictable reasons: unclear scope, skipped discovery, “lift-and-shift” without refactoring debt, and no FinOps discipline after go-live. Successful programs treat migration as a product, with waves, metrics, and explicit rollback criteria.

CodeXcape has led cloud migrations for manufacturing, healthcare, fintech, and professional services clients across North America, Europe, and Asia-Pacific. This playbook is the structure we use whether the target is Azure, AWS, or a hybrid footprint.

Phase 1: Discover (weeks 1 to 4)

You cannot migrate what you have not inventoried. Run automated discovery (Azure Migrate, AWS Application Discovery Service, or third-party CMDB scans) supplemented by application owner interviews.

  • Application catalog: name, owner, users, revenue and criticality tier, tech stack, dependencies
  • Infrastructure: VMs, databases, storage, network flows, identity integrations
  • Compliance: data classification, residency requirements, industry regulations (HIPAA, PCI, SOX)
  • Operational metrics: current monthly cost, incident rate, RTO and RPO requirements
  • Technical debt flags: unsupported OS, end-of-life databases, single points of failure

The 6R decision framework

  • Rehost (lift-and-shift), fast but carries legacy inefficiency
  • Replatform, minor changes (e.g., move SQL Server to Azure SQL Managed Instance)
  • Refactor, rework for cloud-native (containers, serverless), highest long-term value
  • Repurchase, move to SaaS (Salesforce, Dynamics 365, Workday)
  • Retire, decommission unused apps (often 10 to 20% of portfolio)
  • Retain, keep on-prem temporarily due to latency or regulatory constraints

Phase 2: Design landing zone (weeks 3 to 8)

A landing zone is your cloud foundation: accounts and subscriptions, networking, identity, logging, and guardrails. Build it once, correctly, before migrating workloads ad hoc.

  • Hub-spoke or mesh network topology with centralized egress and DNS
  • Identity federation, Entra ID / Okta integrated with on-prem AD
  • Centralized logging (Log Analytics, CloudWatch, SIEM forwarders)
  • Policy-as-code, Azure Policy, AWS Config, OPA gates on non-compliant resources
  • Tagging standard, cost center, environment, owner, application ID (required for FinOps)
  • Backup and disaster recovery baseline, geo-redundant storage, tested restore runbooks

Phase 3: Migrate in waves (not big bang)

Sequence migrations from low risk to high risk. Wave 1 might be dev and test and internal tools. Wave 2: customer-facing but cloud-ready apps. Wave 3: core ERP, payment systems, or latency-sensitive workloads.

Each wave includes

  • Detailed runbook with step order, owners, and validation checks
  • Rollback plan, what triggers it, who approves, how long it takes
  • Parallel run period when possible (sync data, compare outputs)
  • Cutover window communicated to stakeholders with support escalation path
  • Post-migration hypercare, 2 to 4 weeks of elevated monitoring

Phase 4: Optimize and operate

Migration is not the finish line. Month 1 post-cutover: right-size over-provisioned VMs, enable autoscaling, purchase reserved instances for steady workloads, and eliminate orphaned resources. Establish a monthly FinOps review with application owners.

  • Typical quick wins: 20 to 35% cost reduction in first 90 days post-migration through right-sizing alone
  • Enable auto-shutdown on non-production environments nights and weekends
  • Move cold data to cheaper tiers (Blob Archive, S3 Glacier)
  • Introduce infrastructure-as-code so environments are reproducible and auditable

Common pitfalls to avoid

  • Migrating without fixing dependency maps, outages during cutover
  • Treating cloud as “someone else’s computer” with same legacy architecture
  • Skipping training, ops teams cannot support what they do not understand
  • No executive KPIs, program loses funding when benefits are anecdotal

A structured migration reduces risk and accelerates time-to-value. CodeXcape provides assessment workshops, landing zone implementation, wave execution, and managed cloud operations for enterprises globally. Email sales@codexcape.solutions to schedule a discovery session.

Share this article

Instagram does not support direct web sharing. We copy the link so you can paste it in your post or story.