TLDR:
- AI is a new type of sales opportunity where the customer already wants the product.
- Purely technical Copilot setup work (licensing, configuration) is declining in value as AI automates it and competitors copy it easily.
- The real recurring revenue sits in consulting-style services: data compliance, adoption training, and custom agent development.
- MSPs can build a repeatable, low-cost security or AI-readiness assessment using a free PowerShell script and existing Copilot tools, then charge for it rather than giving it away.
Why is now the right time for MSPs to sell AI services?
Most business owners already use AI in some form, which is a major advantage compared to the early cloud computing era: nobody needs convincing that AI is real or useful, so MSPs can skip the education phase and move straight to a business conversation.
This conversation should be framed around money, not technology: talk in terms of productivity and profitability uplift, not the mechanics of large language models. AI adoption isn’t a rip-and-replace exercise. Copilot layers on top of a client’s existing Microsoft 365 environment, using data and permissions MSPs already manage, which makes it low-risk to trial and easy to remove.
How do you know when a client is ready for AI services?
Look for a few clear signals:
- Executives already using AI inside Excel or PowerPoint
- Staff overloaded with email and documents
- Heavy meeting loads that need summarising
- Information scattered across mailboxes, Teams, and SharePoint
These clients are ready for a conversation about moving from free Copilot Chat to a paid licence.
What AI services should MSPs focus on?
Purely technical work is declining fast in value, because AI itself is increasingly good at doing that work, and it’s trivially easy for competitors to copy. The growth areas look less like a support ticket and more like a consulting engagement.
Three areas stand out:
1. Compliance and data governance. With the updated Australian Privacy Act affecting all Australian businesses, there may be an opportunity for compliance work. Microsoft Purview tools such as data loss prevention, information labelling, and insider risk monitoring all require sitting down with a client to work out what data matters, where it lives, and who should see it, a consulting engagement… not just a configuration task.
For more information on how to create a secure Microsoft environment for AI, read our blog post “Copilot Minus The Risk”
2. Adoption and training. Helping staff move from basic prompting to more effective techniques, and keeping clients current on a fast-moving field, is easily repeatable across multiple clients once built.
3. Agent and solution development. Building agents in Copilot Studio carries the steepest learning curve of any layer, but it’s the area with the biggest long-term payoff, particularly when a solution is built once for a vertical (for example, accountants dealing with the ATO) and resold across similar clients.
For MSPs who don’t want to build this capability in-house, partner with other MSPs or specialists who’ve already invested in the higher-skill layers, rather than trying to cover the whole stack alone.
"Where the real opportunity probably is for most MSPs today... is the data security, data compliance, and the data maintenance options."
How can MSPs build a repeatable, low-cost AI assessment product?
In our recent webinar, Robert Crane demonstrated a free PowerShell script that extracts a client’s Microsoft 365 Secure Score data into a JSON file, which can then be fed into your chosen AI service to automatically generate a security assessment report, complete with separate executive and technical versions for different audiences.
This is a repeatable process MSPs can turn into a standard onboarding step or paid assessment product, without writing custom software.
Don't give assessments away for free. A priced assessment gets taken seriously by the client; a free one gets ignored.
AI isn’t a new category MSPs need to build from scratch, it’s an extension of the compliance, security and advisory work most MSPs already do, just with a new toolset.
The MSPs who move fastest from technical configuration into consulting-style compliance and adoption work will be the ones building recurring, defensible revenue.
Based on the Manage Protect webinar “Building a Profitable AI Practice,” featuring Microsoft MVP Robert Crane.
