Bundall, Gold Coast QLD 07 5610 1930 [email protected] Serving businesses across Queensland

Cloud Migration Checklist: 12 Practical Steps to Avoid Costly Mistakes

By Creative Technologies team
Published 17 November 2026 | Last reviewed: November 2026

A cloud migration checklist turns a risky move into a staged one: audit your applications, data and licences, test backups and rollback, set up identity and multi-factor authentication, check bandwidth, train staff, then verify everything after cut-over before you decommission anything.

This guide is for business owners and managers planning a move to Microsoft 365, Microsoft Azure or AWS. Work through this cloud migration checklist before you book a cut-over date. Our cloud services page shows how we run cloud platforms across Queensland.

What should you fix before a cloud migration, not after?

Fix anything that is cheaper to sort out while the old system still runs: messy data, unknown dependencies, untested backups, loose access and missing licences. After cut-over, the same fixes compete with live users, live data and a support queue.

Most migration pain is discovered during the move rather than created by it. The application only one person understands. The backup that reported success for a year and was never once restored. Microsoft’s guidance on building a migration plan with Azure Migrate makes the same point: know your estate and its dependencies first, then clear blockers such as network limits and training.

A cloud migration checklist gives that work an order, so the cheap fixes happen in the planning week, not the cut-over weekend.

The cloud migration checklist: 12 steps in four stages

Work through the stages in order. Stages 1 and 2 decide whether the move is safe to schedule; stages 3 and 4 decide whether it stays that way. Tick each step off, and treat this cloud migration checklist as the record of the plan.

Stage 1: Audit what you have

1. Inventory every application, server and data store. List what runs the business: line-of-business applications, file shares, databases, email, print services. Note each item’s owner. Discovery tools automate the collection, but Microsoft’s workload assessment guidance warns they can miss undocumented dependencies, so owners must confirm the list.

2. Map dependencies and integrations. Record what each application talks to: the database behind it, the finance system it feeds, the supplier portal it uses. If an application calls a database all day, they move together or stay connected until both are across. Undocumented integrations are the classic cause of a Monday outage after a cut-over. Our business systems integration work starts from this map.

3. Clean up data and retire what you no longer need. Do not migrate duplicate files and the mailboxes of people who left years ago. Archive what you must keep, delete the rest, and retire applications nobody uses. A smaller estate moves faster, costs less to store and fits a shorter cut-over window.

4. Audit licences and subscriptions. Count licences against the people who use them, and check what your agreements entitle you to in the cloud. The surprises here are dull and expensive: extra seats to buy, editions not covered, software paid for twice. Finding them now is cheap; finding them mid-project is a budget variation.

Stage 2: Protect the move

5. Back up everything, then test a restore. A backup job that reports success has proved nothing until a restore works. Before any stage moves, back up the systems and test-restore a sample. Decide how much data you can afford to lose and how fast each system must be back. Our data backup and disaster recovery service is built around tested restores.

6. Write a rollback plan with clear triggers. Write down how you return to the old system, who makes the call and what triggers it: failed sign-ins, a database that will not sync, an application that will not open. Agree a deadline too. If the stage is not working by then, roll back rather than troubleshooting on.

7. Set up identity, access and MFA before day one. Create accounts, groups and permissions ahead of the move, with multi-factor authentication on from the start, not in the tidy-up afterwards. MFA is one of the Australian Signals Directorate’s Essential Eight strategies, and accounts without it sit exposed during the busiest week of the project. Our cyber security team treats identity setup as part of the migration.

Stage 3: Prepare the ground

8. Check your network and bandwidth. Your internet connection sets the pace for moving data. Measure it before you schedule anything, and plan bulk transfers for quiet hours or over several nights. Microsoft lists network bandwidth and latency among the blockers to settle before a plan is finalised: replication speed decides whether your cut-over window is realistic or wishful.

9. Move in stages and book cut-overs around the business. Start with a pilot group, learn from it, then move the rest in waves. Schedule cut-overs outside your busiest periods and tell staff the dates early. A staged move gives you clean points to pause or roll back, which a weekend marathon never does.

10. Train staff before go-live, not after. People need to know how to sign in, where their files live and what to do when something looks different. Short pre-move sessions and first-week guides prevent most support tickets and workarounds. Training belongs in the migration budget, not in someone’s spare time.

Stage 4: Prove it worked

11. Test against a written post-move checklist. Test what the business does: sign in, open and save files, print, run the line-of-business application, send and receive email, process one order. Microsoft recommends functional, integration, security and performance testing after a migration, with time in the plan for fixes. Keep the old system running until testing passes, and record results against your cloud migration checklist.

12. Review security settings and costs in the first month. Cloud environments drift. Permissions spread, test resources keep running and bills creep up. Book a review to check access, confirm backups run, remove whatever the migration left behind and compare the first bill with the plan. Decommission the old servers only after that review.

Which mistakes make a cloud move expensive?

These mistakes show up in most migrations, and a cloud migration checklist catches each one while fixing it is still cheap.

Mistake What it causes Checklist step
Skipping the dependency map Its database stays behind, so the application fails or runs on old data Step 2
Migrating years of uncleaned data You pay to move duplicates and ex-staff files, and the cut-over stretches Step 3
Backups never test-restored A failed stage becomes data loss instead of a delay Step 5
One big-bang cut-over for everything One fault affects the whole business, with no clean point to roll back to Step 9

Each mistake is cheap in the planning week and expensive in the cut-over weekend, and staged moves with tested backups suit most businesses.

What does a cloud migration cost?

There is no honest fixed figure, and a provider who quotes one before discovery is guessing. A cloud migration checklist surfaces the costs that usually surprise people while there is still time to plan for them.

The usual surprises are licence gaps found at step 4, running old and new systems side by side, staff training hours, and replacing an application that is not cloud-ready. None are reasons to stay put. Find them in the audit, not on the invoice.

What if an application should not move at all?

Some should not. Unsupported applications, software tied to hardware on site, or systems needing an instant response from on-site machines may be better kept on site, replaced first, or moved later.

Assessment tools sort workloads into ready, ready with minor changes, and not ready. A hybrid setup, where some systems stay on site and connect securely to the cloud, is a normal end state for many businesses. Your cloud migration checklist needs a stop option per application.

How long does a cloud migration take?

It depends on the number of users, the volume of data and the complexity of your applications, which is why honest timeframes come after discovery. Anyone who promises a date before the audit in your cloud migration checklist is finished is working from hope. A staged plan starts with a pilot, tests each wave before the next begins, and leaves a buffer for post-move fixes. A small, tidy environment can move within weeks; a legacy-heavy one should expect preparation to take longer than the move itself.

What to do next

  1. Work through this cloud migration checklist and mark every step done, booked or not applicable.
  2. Fix the gaps in stage order. Do not lock in a cut-over date until stages 1 and 2 are complete.
  3. If the dependency map or licence position is guesswork, get it reviewed. Our IT consulting team plans and delivers migrations, so the advice and the work sit with one team.
  4. Book the first stage as a pilot, and judge the rest of the plan by how it goes.

Frequently asked questions

What should be on a cloud migration checklist?

A cloud migration checklist should cover an inventory of applications and data, dependency mapping, data cleanup, a licence audit, tested backups, a written rollback plan, identity and MFA setup, network checks, staged cut-over dates, staff training and post-move testing and reviews.

Do we still need backups once our data is in Microsoft 365?

Yes. Microsoft keeps the Microsoft 365 service running, but protecting your data against accidental deletion, ransomware or a compromised account is your responsibility. A separate backup gives you an independent copy you can restore from, held apart from your everyday environment.

How long does a small business cloud migration usually take?

It depends on the number of users, the volume of data and how tangled the applications are. A staged plan with realistic timeframes should come out of the discovery work before any dates are promised, and a small pilot move usually comes first.

What is a rollback plan, and why does a migration need one?

A rollback plan sets out how you return to the old system if a migration stage fails, who makes the call and what triggers it. It matters because problems found during cut-over are time-critical, and deciding the way back mid-incident wastes the maintenance window.

Sources

Sources checked November 2026.

Talk to us about your move

If working through this cloud migration checklist raised more questions than it answered, it has done its job. Talk to our Bundall team on 07 5610 1930, or send us a message about your systems and timeframes, and we will help you plan the next step.


Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top