The MSP AI Journey: Five Levels of AI Maturity

A client asks, "What should we be doing with AI?" Your answer is the starting point for building a Managed Intelligence Practice: a dedicated practice that helps clients adopt, govern, and use AI to improve how their businesses run.

Your answer says a lot about your managed service provider (MSP) business today. You might help them buy licenses. You might explain what employees can safely use. Or you might sit down with their operations team and redesign a process that is holding the business back. Each answer represents a different level of service, responsibility, and opportunity.

We created the MSP AI Journey to make that progression easier to understand. It describes five levels:

1. No AI Practice
2. AI Governance Awareness
3. AI Governance Enforcement
4. AI Enablement
5. AI Transformation

This is a practical planning framework, not an industry certification or a revenue forecast. The goal is to find an honest starting point and make the next move useful to your clients.

Download the one-page MSP AI Journey worksheet (PDF) to use with your team or in a client planning conversation.

Level 1: No AI Practice

At this stage, AI is something clients ask you about. You respond to requests, but you do not yet have a defined service around it.

You may enable ChatGPT or Claude accounts, or resell Microsoft Copilot licenses, but you don't know what they are using it for.

You are here if:

• Client questions about AI lead to ad hoc answers.
• Most of your AI work is provisioning accounts or licenses on request.
• You have no repeatable AI assessment, governance, or advisory offering.

How you earn: Any revenue is primarily tied to licenses or setup work. You have not yet built a distinct AI service.

Your next move: Send clients an AI Assessment Survey. Ask which AI tools their employees use, who uses them, what information they enter, and where they see opportunities to save time. Have both business leaders and employees respond. Use the answers to establish a baseline and start a focused conversation about AI policy and priorities.

Level 2: AI Governance Awareness

Here, you begin bringing AI conversations to clients proactively. You establish a basic inventory of which AI tools are being used and by whom, then set expectations for responsible use.

The deliverable becomes more substantial than a license. You introduce an acceptable use policy, educate employees, and use governance tooling to improve visibility into AI activity.

You are here if:

• You help clients introduce and communicate AI Acceptable Use Policies.
• You proactively discuss AI risks and responsibilities with their teams.
• You use governance tools to see which AI tools employees access within their coverage.

A useful policy should answer everyday questions: Which tools are approved? What information can employees enter? Who approves a new use case? When should a person review the result?

How you earn: Marking up Copilot licenses gives you an initial revenue stream, usually a few dollars per seat per month. Assessment, policy rollout, and education can become additional paid services.

Your next move: Enforce the policy and work to eliminate shadow AI: AI tools used outside the organization's approval and oversight. Roll out controls that block unauthorized access within the managed environment, give employees approved alternatives, and establish a process for requesting new tools.

Level 3: AI Governance Enforcement

At Level 3, you take responsibility for putting the agreed rules into practice. You understand how client teams use AI, configure controls around that usage, and respond when activity falls outside the policy.

For example, a client may approve a particular AI service for work while restricting unapproved services on managed devices. Your job includes implementing the controls, testing them, and making the approved path clear to employees.

You are here if:

• You can explain which teams use which AI tools and for what purpose.
• You enforce agreed restrictions and test that unauthorized access is blocked within the managed scope.
• You manage access and data-handling controls, with documented coverage and exceptions.

The organization's boundary is now enforced through approved accounts, services, access rules, and data-handling controls across the managed environment. Governance becomes an ongoing practice, much like managing SaaS: review access, onboard and offboard users, assess new tools, handle exceptions, and update controls as the business changes.

How you earn: Package AI governance as an add-on to your managed service packages, combining governance-tool margin with an ongoing management fee. The client pays for a continuing service that includes reviewing activity, maintaining controls, and handling exceptions.

Your next move: Pick a team with a useful, repeatable task and help it use approved AI effectively. Assign an owner for training and advisory, then agree on what improvement you will measure.

Level 4: AI Enablement

At this stage, your conversations expand from governing AI use to helping people work better with it. This is where your clients truly start to experience the superpower of AI in their daily work.

You work with client teams to identify practical use cases, provide training, and help those habits stick. Examples include inbox triage, meeting summaries, and sales follow-up drafts. The value comes from fitting AI into daily work, setting expectations for review, and checking whether the result is useful.

You are here if:

• You run AI training that is specific to a team's work.
• You have deployed practical use cases and review adoption and results.
• Clients pay you for AI advisory and ongoing guidance.

This is where an MSP can develop a virtual Chief AI Officer (vCAIO) offering: helping a client prioritize opportunities, assign ownership, and review progress. The title should come with a defined service and regular decisions to make together.

How you earn: Training fees, implementation projects, and advisory retainers compensate you for expertise and ongoing involvement.

Your next move: Map a client workflow from beginning to end. Look for repeated handoffs, duplicate entry, delays, and growing backlogs. Choose one bottleneck where a change could improve the whole process.

Level 5: AI Transformation

At Level 5, you help change how the client's business operates. Your work may span several teams and systems, combining automation, AI agents, and custom applications around a business outcome.

Consider a tax firm. At Level 4, you might help its team use AI to email clients and chase missing files. At Level 5, you are building full agentic workflows that perform tax return preparation itself: organize source documents, extract and reconcile information, populate draft returns in the firm's tax software, and flag exceptions for a qualified preparer to review. You are helping the firm run a core line-of-business process using AI, with professional review and sign-off built into delivery. At this level, you can save your clients thousands of hours every year.

Before building, read our post on how to evaluate a use case for a client AI solution.

You are here if:

• You map workflows and identify bottlenecks with business owners.
• You deliver and maintain automations, agents, or applications that address those bottlenecks.
• You coordinate a program of improvements with defined owners and measurable outcomes.

Success measures might include turnaround time, rework, backlog, or capacity. Establish a baseline before the build and account for the effort still needed to review outputs and resolve exceptions.

How you earn: Discovery and implementation fees can grow into recurring revenue for operating, supporting, and improving the workflows. The roadmap's ambition is an AI practice that could become comparable in revenue to your IT practice.

Your next move: Make the first successful workflow dependable, document the measured outcome, and use what you learn to expand into related workflows or teams.

How to find your current level

Open the MSP AI Journey PDF with your team and tick the signs that are true today. Your starting level is the highest row where all three signs are present. Ask for evidence behind each check mark: a policy in use, a tested control, a training program, or a live workflow.

Then assess two things separately: your internal adoption and your client adoption. Where is your own MSP on the journey? Where are the clients you serve? Working through the stages internally gives your team a chance to build the muscle and knowledge once before delivering the same services to clients. Your own operations become a place to learn how to govern AI, train teams, run workflows, and support them over time.

Apply the client assessment client by client. One client might already be at Level 5 while another is at Level 2. Record those differences instead of assigning your entire client base the level of its most advanced project.

Recommended steps to build your Managed AI Practice

Turn the roadmap into an operating plan for your practice:

1. Document each client's maturity level in your CRM or PSA. Add a field in ConnectWise, Halo, or Autotask for the current level, along with the next milestone and review date. Give account managers a shared view of where each client stands.
2. Assign a dedicated leader to run the Managed AI Practice. Give that person ownership of the roadmap, delivery standards, team development, and client outcomes, with protected time and authority to move the work forward.
3. Take a few trusted, innovative clients to Level 5 first. Choose clients willing to collaborate, share process knowledge, and work through early iterations. Deliver a meaningful business outcome, measure it, and develop case studies with their permission. Use those engagements to refine the playbook before scaling across the client base.
4. Organize delivery around a repeatable playbook toward Level 5. Define the assessment, governance rollout, training, workflow discovery, and implementation steps. Give each stage clear deliverables and criteria for moving forward so the next engagement builds on what you have already learned.

Get a few clients all the way there. Build the evidence, delivery experience, and repeatable process around those successes. Then scale.