Table of Contents
Most businesses do not stay with an IT provider because everything is going well. They stay because changing providers sounds worse.
The fear is understandable. Your current provider may hold administrator passwords, manage the Microsoft 365 tenant, renew the domain, operate the backups, control the firewall, and know the one strange thing that keeps an important application running. Owners picture a hostile handoff, missing credentials, and three days without email.
That is not what a well-run transition looks like. Switching IT providers should be a controlled transfer of access, documentation, and responsibility. The real risk is not changing providers. It is discovering during the change that your business never had control of its own technology.
Here is the transition process we use and the checklist we would want any business to follow, whether it hires KaselTech or someone else.
Before telling the incumbent provider anything, establish which assets belong to your company and who can access them. Your business should be the owner or primary account holder wherever the platform allows it. The provider should have delegated administrative access, not personal ownership.
- -Your domain registrar and DNS records
- -Microsoft 365, Google Workspace, Azure, AWS, and other cloud tenants
- -Firewall, switches, wireless access points, and network-management portals
- -Backup platforms, storage locations, encryption keys, and recovery accounts
- -Endpoint security, monitoring, remote-support, and password-management tools
- -Internet, phone, software, and hardware vendor accounts
- -Website hosting, analytics, advertising, and social-media accounts
- -Source code, automation, documentation, and configuration files created for your company
Some services may legitimately be purchased through the outgoing provider. That is fine, but the contract should explain what transfers, what must be replaced, and how long data remains available after termination. A subscription ending is manageable. An unknown subscription ending without warning is not.
Do not cancel service until this inventory exists. Once notice is given, the transition clock starts and goodwill can become unpredictable.
A useful inventory is more than a spreadsheet of device names. It tells the incoming provider what exists, who depends on it, how it is administered, and what would happen if it stopped.
- -Users, endpoints, servers, and locations
- -Internet circuits, public IP addresses, firewalls, switches, Wi-Fi, and VPNs
- -Microsoft 365 or Google Workspace configuration and licensing
- -Line-of-business applications and their vendors
- -Administrative portals and named account owners
- -Backup jobs, retention policies, storage locations, and recent restore results
- -Security tools, alert destinations, and incident-response contacts
- -Domains, certificates, warranties, renewals, and contract dates
- -Open tickets, known issues, planned projects, and recurring maintenance
The incoming provider does not need a perfect encyclopedia before starting. It does need enough information to keep critical services running and to identify the gaps that require discovery work.
If the outgoing provider has good documentation, the handoff is faster. If it does not, that is not a reason to stay. It is evidence that the transition needs a deliberate discovery phase.
An IT transition temporarily puts two providers near the same systems. That makes identity and logging more important, not less.
Give each provider separate named accounts. Do not pass around one shared domain administrator password. Require multifactor authentication, limit each account to the systems it needs, and preserve logs showing who changed what. Agree in advance on the exact time the outgoing provider's access will be disabled.
This is not just our preference. CISA's joint guidance for MSPs and their customers recommends least-privilege access, strong authentication, monitoring, logging, and clear contractual responsibility between the provider and customer.
A clean access transition normally looks like this: 1. Create new named accounts for the incoming provider. 2. Verify MFA and emergency access before changing anything else. 3. Export or preserve the logs needed to investigate later problems. 4. Rotate shared passwords, API keys, service-account credentials, and recovery codes. 5. Remove the outgoing provider's access at the agreed cutoff. 6. Audit privileged groups and remote-access tools after the cutoff.
The goal is not to treat the former provider as hostile. The goal is to leave no ambiguous access behind.
Monitoring agents, endpoint protection, backup software, and remote-support tools should not be ripped out all at once. Every old control should remain until its replacement is installed, checked, and reporting correctly.
A safe sequence is:
First, establish visibility. The new provider documents the environment, confirms administrator access, and deploys its monitoring where appropriate.
Then, protect recovery. Verify that backups are completing and perform at least one test restore before altering the old backup platform.
Next, move security controls. Install and validate the replacement endpoint and email-security tools before removing the old agents. Watch carefully for conflicting products during the overlap.
Then, move support. Give employees a clear support number, email address, and go-live date. Tell them what will look different and what will not.
Finally, remove old access and tools. Uninstall abandoned agents, close provider accounts, rotate credentials, and confirm that alerts now reach the right team.
This order keeps the business observable and recoverable throughout the change.
"The backups are working" is not a handoff artifact. A successful restore is.
- -The most recent successful backup for each critical system
- -The retention period and earliest available recovery point
- -Where backup data is physically or logically stored
- -Who owns the storage account and encryption keys
- -How quickly a file, mailbox, server, or application can be restored
- -Whether a recent sample restore has completed successfully
Keep an independent copy of essential documentation and recovery information outside the systems being transitioned. CISA's risk considerations for MSP customers similarly emphasize offsite backups, activity logs, limited provider privileges, and continuity planning.
If nobody can demonstrate a restore, treat the backup as unverified. Do not discover that distinction during an outage.
The exact timing depends on the environment, but assigning work to phases keeps a transition from becoming an open-ended scavenger hunt.
Days 1-3: control and communication. Confirm decision-makers, critical systems, vendors, support contacts, ownership, and the final date with the outgoing provider.
Days 4-10: access and discovery. Transfer documentation, establish named administrator accounts, inventory systems, preserve logs, and identify missing credentials or unknown dependencies.
Days 11-20: tool deployment and validation. Deploy monitoring, support, backup, and security tools in a controlled order. Test alerts and restore data. Begin resolving the highest-risk findings.
Days 21-30: cutoff and cleanup. Remove old access, rotate secrets, uninstall abandoned agents, reconcile licenses, review open issues, and give leadership a concise environment summary and improvement plan.
Complex migrations can take longer, but basic accountability should not. At the end of the first month, you should know what you own, what remains at risk, and what the new provider is doing next.
Professional providers understand that clients change direction. They may enforce the notice period and invoice correctly, but they transfer client-owned information without theatrics.
- -Refusing to provide administrative access to a tenant owned by your business
- -Claiming that basic network documentation is the provider's proprietary information
- -Withholding passwords or backups to force payment outside the contract
- -Removing security or backup tools before the agreed transition date
- -Using shared accounts that make activity impossible to attribute
- -Being unable to explain which licenses and data will disappear at termination
- -Insisting that your domain, cloud tenant, or core vendor accounts must remain in their name
The best time to prevent these problems is when signing the original agreement. CISA provides an SMB vendor-assessment resource that includes a managed-service-provider vetting use case. Ownership, security responsibility, incident notification, data return, and transition assistance belong in the contract before the relationship starts.
- -A current inventory of users, devices, infrastructure, cloud services, and vendors
- -Documented administrator and emergency-access procedures
- -Verified monitoring, endpoint protection, patching, and backups
- -A list of unresolved risks ranked by urgency and business impact
- -A clear support process employees can actually use
- -A 90-day stabilization plan and a longer-term technology roadmap
- -Confirmation that the former provider's access and tooling have been removed
You should not need to understand every technical detail. You should be able to see who is accountable, what has been verified, what remains unknown, and what happens next.
A good transition is not a dramatic weekend cutover. It is a measured handoff in which access is transferred, protections overlap, backups are verified, and responsibility becomes clearer each week.
If dissatisfaction with your current provider has been outweighed by fear of the move, start with ownership and documentation. You can learn a great deal before giving notice, and none of that work is wasted if you decide to stay.
KaselTech handles the outgoing-provider coordination, account recovery, documentation, tool deployment, and first-month stabilization for managed IT transitions. During August, onboarding and transition are included for qualifying new managed IT agreements, with no long-term contract. Start with the free technology review or talk through the handoff with us.