Legacy system migration remains one of the most challenging—and rewarding—digital transformation initiatives enterprises undertake. Done right, cloud migration delivers operational agility, cost optimisation, and innovation velocity. Done poorly, it becomes an expensive exercise in technical debt relocation.

After guiding dozens of enterprises through cloud migrations—from 20-year-old mainframe systems to mid-2000s monoliths—we’ve distilled our learnings into a strategic playbook.

Assess Before You Migrate: The 6R Framework

Amazon popularised the “6 Rs” of cloud migration, and they remain the gold standard for strategy formulation:

  1. Rehost (Lift-and-Shift): Move as-is to cloud infrastructure. Fastest path, minimal re-architecture, but limited cloud-native benefits. Best for: Time-sensitive migrations, applications with unclear usage patterns.
  2. Replatform (Lift-and-Reshape): Minor optimisations during migration—containerisation, managed database services, auto-scaling. Best for: Applications with clear cloud benefits but limited refactoring budget.
  3. Repurchase (Replace): Swap custom-built systems for SaaS alternatives. Best for: Commodity functions like CRM, email, collaboration tools.
  4. Refactor (Re-architect): Complete rebuild using cloud-native patterns—serverless, microservices, event-driven architecture. Highest benefit, highest cost. Best for: Core differentiating systems where agility is competitive advantage.
  5. Retire: Decommission applications no longer needed. Best for: Zombie systems consuming maintenance budget without delivering value.
  6. Retain: Keep on-premises for compliance, latency, or cost reasons. Best for: Systems with hard regulatory constraints or dependencies that make migration prohibitively expensive.

The Strategic Decision Matrix

Don’t apply one strategy to all systems. We use a 2×2 matrix plotting “Business Value” against “Technical Complexity” to guide approach:

  • High Value + Low Complexity = Refactor (maximise cloud benefits)
  • High Value + High Complexity = Replatform (balance benefit and risk)
  • Low Value + Low Complexity = Rehost (quick wins)
  • Low Value + High Complexity = Retire/Replace (avoid money pits)

Case Study: Financial Services Migration

A London-based fintech firm approached us with 15 years of accumulated systems—a classic distributed monolith with 50+ tightly-coupled services, shared databases, and synchronous communications creating cascading failures.

Our assessment revealed:

  • 12 services accounting for 80% of business value
  • 8 services effectively unused (< 100 requests/month)
  • 5 services replicating functionality available in modern SaaS tools

The strategy:

  • Refactor: 4 core services → serverless architecture (AWS Lambda, DynamoDB)
  • Replatform: 8 mid-value services → containerised on EKS
  • Replace: 5 services → SaaS equivalents (Auth0, SendGrid, Stripe)
  • Retire: 8 services → decommissioned
  • Retain: 2 services → on-premises (regulatory requirements)

Results after 9 months:

  • Infrastructure costs down 45%
  • Deployment frequency up 12x (weekly → multiple daily)
  • System reliability improved from 99.1% to 99.97%
  • Development velocity increased 3.5x (measured by feature delivery)

Critical Success Factors

  1. Executive Sponsorship: Cloud migration isn’t IT’s project—it’s business transformation. Secure C-suite commitment and cross-functional alignment.
  2. Skills & Culture: Cloud-native patterns require cloud-native mindsets. Invest in training and hire cloud expertise early.
  3. Data Strategy First: Most migration complexity lives in data. Map dependencies, design migration patterns, and test thoroughly before touching production.
  4. Incremental Migration: Big-bang migrations fail. Use the strangler fig pattern—route traffic progressively to cloud systems while legacy continues operating.
  5. Cost Governance: Cloud flexibility creates cost unpredictability. Implement FinOps practices, tagging strategies, and budgets from day one.

Your Migration Roadmap

Week 1-4: Discovery & Assessment

  • Application portfolio analysis
  • Dependency mapping
  • TCO modelling
  • Risk assessment

Week 5-8: Strategy & Planning

  • 6R classification
  • Architecture design
  • Migration wave planning
  • Team assembly

Month 3-6: Pilot Migration

  • Non-critical applications first
  • Validate architecture decisions
  • Build operational muscle
  • Refine runbooks

Month 7-18: Full Migration

  • Progressive waves by business value
  • Continuous optimisation
  • Decommission legacy infrastructure
  • Knowledge transfer

Month 19+: Optimisation & Innovation

  • Cost optimisation
  • Performance tuning
  • Cloud-native feature development
  • Continuous improvement

The cloud migration journey is less about technology and more about orchestrating organisational change. Start with clear strategy, measure relentlessly, and optimise continuously.