What Nobody Tells You About Enterprise Cloud Migrations

Cloud migrations are consistently underestimated. Not in terms of effort (most teams know they're hard), but in terms of what kind of hard they are.

Here's what surprised me most.

The technical work is the easy part

Application inventory, lift-and-shift planning, network architecture — this is well-understood work with clear playbooks. Your cloud vendor will provide a migration factory methodology. Your architecture team will have opinions. This part will get done.

The hard part is the politics, the prioritization, and the people.

App owners don't trust what they didn't build

Every application owner believes their app is special. And in a lift-and-shift, many of them are right — their app does have quirky dependencies, vendor licensing constraints, or latency sensitivities that a standard approach won't handle.

Build time for deep-dive discovery with each app owner into your plan. Don't let the cloud team make assumptions on their behalf.

Your wave plan will change — plan for it

We built a 5-wave migration plan over 18 months. By wave 2, we had replanned waves 3-5 twice. Business priorities shift. Decommission timelines slip. Network dependencies cascade.

Use your wave plan as a communication artifact, not a contract. Communicate changes proactively and explain the tradeoffs.

Cost optimization is a second project

Most migration programs are scoped to move workloads. Cost optimization — rightsizing instances, moving to managed services, retiring unused resources — is a separate workstream that requires separate focus.

If you don't plan for it explicitly, you will finish your migration and wonder why your cloud bill doesn't reflect the promised savings.

Cloud migrations that succeed aren't just technically sound — they're organizationally managed well. That's where program management adds the most value.