Cloud Adoption Roadmap: 5 Phases and a 12-Month Plan

9 min read ยท 2026-11-17

A cloud adoption roadmap is the plan that moves your company from its current infrastructure to cloud services in ordered phases. Each phase has an owner, dependencies, a decision gate and a milestone you can check. The example in this guide runs over 12 months in five phases: assess your workloads, build the plan and business case, set up the foundations, migrate in waves, then optimize and hand over to run teams.

The plan below is a working example for a mid-sized IT estate. It lists who owns each phase and what has to be true before the next one starts. It also gives a simple rule for ordering workloads, so you can adapt it and present it as it is.

The roadmap at a glance

Goal: Move priority workloads to the cloud within 12 months, with clear owners, controlled risk and measurable results. Duration: 12 months

  1. Assess and inventory (Months 1-2)

    Know exactly what you run, what it depends on and why you are moving it.

    • List every application and workload with its business owner, users, data sensitivity and current hosting.
    • Map dependencies between applications, databases and integrations.
    • Run a skills check on IT, security and operations teams.
    • Write down the business goals behind the move: speed, resilience, cost control or ending a data center contract.

    Milestone: Workload inventory signed off by IT and business owners, with goals agreed by leadership.

  2. Plan and build the business case (Months 2-3)

    Turn the inventory into a decision for each workload and a budget leadership can approve.

    • Choose what happens to each workload, using the seven common options listed below.
    • Score workloads and group them into migration waves, using the scoring rule below.
    • Estimate costs for each wave and set success metrics.
    • Write the RACI and the skilling plan, then get stakeholder approvals.

    Milestone: Gate 1 decision. Leadership approves the wave plan, the budget and the owners.

  3. Build the foundations (Months 3-5)

    Prepare a secure, governed environment before any production workload lands.

    • Set up the landing zone: accounts or subscriptions, network, identity and access.
    • Define security baselines, logging, backup and tagging rules.
    • Put cost alerts and budget owners in place for each team.
    • Train the first team that will run migrations.

    Milestone: Gate 2 decision. Security signs off the landing zone and one test workload runs end to end.

  4. Migrate in waves (Months 5-11)

    Move workloads in small, ordered batches and learn from each one.

    • Wave 1 (months 5 to 6): two or three low-risk workloads, mostly rehost, to prove the process.
    • Wave 2 (months 7 to 9): business applications with known dependencies, moved together.
    • Wave 3 (months 9 to 11): critical or complex workloads that need refactoring, with rollback plans.
    • After each wave, hold a review and update the timeline for the next one.

    Milestone: Gate 3 decision after each wave. Go, adjust or pause the next wave based on the KPIs.

  5. Optimize and hand over (Months 11-12)

    Lower running costs and move the new environment into normal operations.

    • Right-size resources and remove what nobody uses.
    • Shut down retired servers and close old contracts.
    • Hand runbooks and on-call duties to operations teams.
    • Write the plan for months 13 to 24: cloud-native projects and remaining workloads.

    Milestone: Year-one review delivered to leadership, with KPIs and the next roadmap approved.

Cloud adoption vs cloud migration vs cloud strategy

People mix up these three terms in planning meetings. A cloud strategy explains why the company uses the cloud and which principles guide the choices. A cloud migration plan covers moving existing workloads. A cloud adoption roadmap is broader. It includes the migration, plus the foundations, skills, governance, operating model and new cloud-native work.

Microsoft's Cloud Adoption Framework makes the same split. It offers a cloud-native adoption plan for building new solutions and a migration adoption plan for moving existing workloads. Many companies need both on the same roadmap. In this example, the first year focuses on migration and the second year adds cloud-native projects. If your leadership only asks "when are we out of the data center?", you are presenting a migration plan. If they ask "how will we work differently?", they expect the full adoption roadmap.

How to order workloads into migration waves

The order in which workloads move sets most of the risk in the plan. Microsoft's guidance asks teams to keep a workload inventory with the owner, business criticality and dependencies of each workload. For each one, it also asks you to record the migration strategy, success metrics, assessment results, target architecture and cost estimates.

To build waves, score each workload from 1 to 3 on four criteria: business criticality, technical complexity, number of dependencies and expected cost of the move. Workloads with low totals go first, because they teach the team without putting the business at risk. Workloads that depend on each other move in the same wave, even if their scores differ. Anything marked "retain" stays on the roadmap with a review date, so nobody forgets it.

To decide what happens to each workload, most teams pick from a common set of seven options. Base each choice on the business driver behind that workload:

  • Retire: switch it off because nobody needs it anymore.
  • Rehost: move it as it is to cloud servers.
  • Refactor: make small changes so it runs better on cloud services.
  • Rearchitect: redesign it to use cloud-native services.
  • Replace: swap it for a SaaS product.
  • Rebuild: write it again from scratch as a cloud-native application.
  • Retain: keep it where it is for now, with a reason written down.

Who owns what: roles and responsibilities

Settle ownership before the first wave, because every gate needs someone who can say yes. Kion advises getting the CEO and CFO on board early and involving security experts such as the CISO. It also suggests that technical input goes beyond the director of architecture to roles like the principal engineer. Microsoft asks you to record team responsibilities and capture approvals from technical and business stakeholders. A simple RACI on one slide is usually enough.

Put one name per line, not a department. When a wave slips, the program lead needs to know exactly who to call. It also helps to name a backup owner for each line, so work continues when someone is away.

  • Executive sponsor (CIO or CTO): accountable for the roadmap, approves each gate.
  • Cloud program lead: responsible for the timeline, dependencies and quarterly updates.
  • Platform or infrastructure team: responsible for the landing zone and migrations.
  • Security and compliance: responsible for baselines, consulted on every wave, signs off Gate 2.
  • Finance: consulted on the business case, responsible for budget tracking and cost reviews.
  • Business owners: consulted on wave timing and responsible for user testing.
  • HR or learning team: responsible for the skilling plan with IT managers.

Skills, change and governance

Skills are a real dependency, so treat them like one on the timeline. Microsoft asks you to document a skilling plan with the training each role needs, learning paths and timelines. Put training before each phase that needs it: landing zone skills before month 3, migration tooling before Wave 1, cost management before the optimization phase. If a team has never run cloud workloads, plan time for people to practice in a sandbox.

Governance sets the rules that keep the new environment under control: who can create resources, how resources are tagged, which data can live where, and how costs are checked. Name the people who own these rules before the first production workload is deployed. Write the rules during the foundations phase, not after Wave 2, when exceptions have already piled up. For regulated data, add compliance checks to the Gate 2 and Gate 3 criteria so nobody can skip them.

Risks to plan for and how to answer them

Each risk below needs an owner and a planned response on the roadmap, not just a line in a risk register.

  • Hidden dependencies break an application after the move. Response: dependency mapping in phase 1 and moving linked systems in the same wave.
  • Costs grow faster than planned. Response: budget alerts from day one and a cost review after each wave.
  • Security gaps in the new environment. Response: security sign-off at Gate 2 and baselines applied by default.
  • Teams lack the skills to run migrations. Response: skilling plan scheduled before each phase and a small first wave.
  • Business users resist the change or schedules clash. Response: business owners choose wave dates and test before cutover.
  • Downtime during cutover. Response: rollback plan and agreed maintenance windows for every critical workload.

KPIs that show the roadmap is working

AWS, in its cloud adoption explainer for small and medium businesses, organizes the work around four focus areas: prioritizing activities, assessing capabilities, executing to deliver business outcomes and measuring outcomes. Measurement is the one that gets dropped once migrations start. Pick five or six metrics and report them at every gate.

Take a baseline for each KPI during phase 1. Without a baseline, you cannot show leadership what changed.

  • Workloads migrated against plan, by wave.
  • Actual cloud spend against the business case.
  • Incidents and downtime caused by migrations.
  • Time to provision a new environment, before and after.
  • Share of resources that follow tagging and security rules.
  • Number of people trained against the skilling plan.

Cloud adoption roadmap template for leadership

Executives do not need the full workload list. They need a one-page view: five phases on a 12-month timeline, three gate decisions, the owner of each phase, the top three risks and the KPIs. Use one row per phase, show milestones as markers and draw arrows only for dependencies that block a date.

Keep one version of the plan that everyone works from, holding all decisions, estimates and strategies, so nobody presents an old copy. Kion suggests a regular cadence, such as quarterly, so the team revisits the plan in a structured way instead of reopening it after every problem. Put that review on the calendar from day one. Each quarter, update the dates, move workloads between waves when needed, record the gate decisions and present the new version to the same audience.

Common mistakes to avoid

  • Starting migrations before the landing zone is approved. Set Gate 2 as a hard dependency for Wave 1.
  • Moving the most critical system first to show quick results. Start with low-scoring workloads and build confidence.
  • Treating the roadmap as an IT document only. Involve finance, security and business owners from phase 1.
  • Leaving "retain" workloads off the plan. Give each one a reason and a review date.
  • Skipping baselines. Measure costs, incidents and provisioning time before anything moves.
  • Writing the plan once and never updating it. Review it every quarter and record what changed.

Frequently asked questions

What are the phases of a cloud adoption roadmap?

GitLab describes four steps: assessment, planning, implementation and optimization. In practice, it helps to split implementation into a foundations phase (landing zone, security, governance) and a migration phase run in waves. That gives you five phases with clear gates between them, as in the plan above.

How long does cloud adoption take?

It depends on the number of workloads, their complexity and the team's skills. This guide uses 12 months as an example horizon for priority workloads, with a second plan for remaining systems and cloud-native projects. Your own timeline should come from the inventory and the wave plan. Update the dates after each wave instead of fixing them on day one.

What should a cloud adoption roadmap template include?

It should list phases, dates, owners, dependencies, gate decisions and milestones. Add the migration strategy and wave for each workload, the main risks with their responses, and the KPIs you will report. Keep a one-page version for leadership and a detailed version for the program team.

How do you present a cloud roadmap to leadership?

Show one page with the phases on a timeline, the three gate decisions, the budget and the top risks. Explain the business goal first, then the plan. End with the decision you need from them, such as approving Gate 1 or the budget for the next wave.

Generate this roadmap with AI