Each cloud service with its growing instructions: what we do first (sow), what happens during the work (grow), and what you are left with (harvest).
Inventory servers, apps and data. Decide what moves as-is, what gets reshaped and what retires.
Move in small batches with a rollback plan for each. Test before switching traffic.
Choose provider and region with you. Plan accounts, networks and who has access.
Build environments from written configuration, so they can be rebuilt the same way.
Agree what is looked after, how problems are reported and how quickly they are picked up.
Routine patching, capacity checks and fixes, with a log of every change.
Work out what you cannot afford to lose and how far back you need to go.
Automated backups stored separately, then a real restore rehearsal.
Pick the handful of signals that really mean trouble for your service.
Set up checks and alerts, tuned so they are not ignored as noise.
Read-only look at your cloud bill and what is actually running.
Spot idle resources, oversized machines and storage nobody uses.
Review accounts, permissions, exposed services and secrets handling.
Close what should be closed, add multi-factor sign-in, tighten access.
Understand where your app struggles: scaling, releases, configuration.
Reshape the parts that hold it back, one piece at a time.