Guide

Cloud Migration Checklist for SMBs

A practical, engineer-written checklist for small and mid-sized businesses planning a move to the cloud — from assessment and planning through cutover and post-migration support.

Cloud migration goes wrong for predictable reasons: an incomplete inventory, an undefined budget, unclear downtime expectations, and identity or backup gaps discovered after go-live. None of these are technical surprises — they are planning gaps.

Use the checklist below to work through each phase in order. The first two phases (assessment and planning) determine most of the outcome, so treat them as real work rather than paperwork.

1. Assessment — know what you actually run

Most failed migrations fail here. Before a single workload moves, inventory everything and grade it for cloud readiness.

List every server, application, and database, with owners and business criticality
Document current licensing (Windows Server, SQL, line-of-business apps) and renewal dates
Map application dependencies — what talks to what, and on which ports
Baseline performance: CPU, RAM, storage IOPS, and peak usage windows
Measure internet bandwidth and identify sites that need a circuit upgrade first
Flag legacy apps that cannot move as-is and need replacement or re-hosting

2. Planning — goals, budget, and success criteria

A migration without a written definition of success turns into an open-ended project. Fix the scope early.

Write the business driver: cost, resilience, remote work, compliance, or an aging server refresh
Set a target monthly cloud spend and compare it against current hardware + maintenance costs
Agree on acceptable downtime per workload (RTO) and acceptable data loss (RPO)
Prioritize workloads: start with low-risk, high-value systems such as file shares and email
Assign an internal decision maker and a technical point of contact
Build a rollback plan for every phase before scheduling it

3. Choose the migration approach per workload

Not everything belongs in the same bucket. Grade each application against these five options.

Rehost (lift and shift) — fastest path for stable servers with no code changes
Replatform — move to managed databases or app services to cut admin overhead
Repurchase — swap a legacy app for a SaaS equivalent (accounting, CRM, ticketing)
Refactor — rebuild only when the business case clearly justifies the spend
Retire or retain — decommission unused systems and keep anything blocked by compliance on-premises

4. Security, identity, and compliance

Cloud is not secure by default. Identity and backup are the two areas SMBs most often get wrong.

Enforce multi-factor authentication on every administrative and user account
Apply least-privilege roles — no shared global admin accounts
Encrypt data at rest and in transit, and document where regulated data lives
Configure independent, tested backups — cloud replication is not a backup
Enable centralized logging, alerting, and endpoint protection from day one
Confirm HIPAA, PCI, or contractual requirements are met before go-live, not after

5. Migration runbook and cutover

Run the move like a project with a timed script, not an all-nighter improvised on a weekend.

Pilot with a small user group and a non-critical workload first
Pre-seed large data sets ahead of the cutover window to shorten downtime
Write a minute-by-minute cutover runbook with owners and go/no-go checkpoints
Schedule cutovers outside business hours and notify users in advance
Validate DNS, email routing, printing, VPN, and line-of-business app access immediately after cutover
Keep the source environment intact until validation is signed off

6. Post-migration — optimize and support

The first 90 days determine whether the migration saves money or quietly inflates your bill.

Right-size instances and storage tiers after two to four weeks of real usage
Set budget alerts and review cloud spend monthly
Decommission on-premises hardware and stop paying for unused licenses
Run a restore test from backup to prove recovery works
Document the new environment and train staff on changed workflows
Move to proactive monitoring and managed support so issues surface before users report them

Common mistakes to avoid

Moving everything at once instead of phasing by risk
Assuming the cloud provider backs up your data — most do not, by default
Skipping bandwidth checks and discovering the office circuit is the bottleneck
Leaving legacy admin accounts and old VPN rules in place after cutover
Never right-sizing after go-live, then paying twice what you budgeted

Want this checklist run for your business?

CoreLynk engineers handle the assessment and planning phases with you — no call center, direct access to the people doing the work.

Schedule a Free 30-Minute Assessment →Explore Cloud Services