Episode #945 of the MSPi PrimeCast
AI is creating new opportunities for MSPs, but Matt Peters, CEO of Fixify, believes getting value from it requires more than simply adopting the latest technology.
During his conversation with Joey Pinz, Peters drew on his experience across engineering, cybersecurity, managed detection and response, and IT service delivery to explain what happens when technology meets the realities of customer environments. His perspective is especially relevant as MSPs determine where AI fits within their operations—and what it will actually take to make it work.
Here are eight AI realities from Peters’ conversation that MSPs should understand.
1. AI Needs a Business Outcome
Peters believes MSPs should begin with a basic question when evaluating an AI solution: What numbers are going to move?
That could mean the percentage of Tier 1 tickets covered, improvements in time to first response, or the amount of staff time reclaimed. The technology itself may be exciting, but Peters argues that MSPs ultimately need to understand what will change and what it will cost to achieve that change.
MSP Action: Establish the specific operational metrics an AI initiative is expected to improve before committing resources to it.
2. Understand the Work Before Automating It
Peters recommends examining the existing ticket base to understand where the work actually occurs. That includes diagnostic effort, resolution time, ticket types, and how work is distributed across the team.
If an MSP wants AI to give technicians time back, it first needs to know what is consuming that time.
MSP Action: Analyze ticket data and technician workload to identify where automation could create the most meaningful capacity.
3. Different Types of Work Require Different Solutions
Variance is a recurring theme in Peters’ approach to operations.
He describes low-variance work as a natural fit for traditional automation. AI can be effective with medium-variance work when it has sufficient context. High-variance work—such as complex migrations, data analysis, and business transformation—is where human expertise remains particularly valuable.
MSP Action: Categorize recurring work by its level of variance instead of assuming every process is equally suited to AI.
4. AI Is Only as Useful as the Context Around It
Peters rejects the idea that AI can simply be dropped into an MSP environment and immediately solve problems.
He points to functional knowledge bases, APIs, access controls, customer information, and proper configuration as important pieces of making AI useful. In his experience, missing context is a significant reason AI automation can fail.
MSP Action: Evaluate the quality of your documentation, knowledge base, integrations, and access before expecting AI to handle more complex work.
5. Deployment May Be Harder Than the AI
One of Peters’ key observations from building Fixify is that the difficult part isn’t necessarily the AI itself. It is deployment, learning, and providing enough context for the technology to be effective.
He also cautions MSPs to understand what a vendor will require from their own team. A solution can stall if implementation becomes another responsibility assigned to an employee who already has a full-time job.
MSP Action: Before selecting an AI platform, determine who will implement and maintain it, how much time that will require, and what assistance the vendor actually provides.
6. The Customer’s Perspective Still Comes First
Long before his work with Fixify, Peters learned an important lesson while developing technology products: good software alone doesn’t guarantee success.
Technology becomes useful when it is deployed and solves an actual customer problem. Peters says the customer’s perspective is ultimately the perspective that matters.
That lesson carries directly into AI. An impressive capability means little if it doesn’t improve the service an MSP can provide its customers.
MSP Action: Evaluate AI not just by what it can automate, but by whether the resulting service experience is better for the customer.
7. Scaling Is About More Than Increasing Throughput
Peters learned while helping scale services at Mandiant that organizations often focus first on making things faster or automating more. He argues that variance can be the more important challenge.
Processes that depend on undocumented knowledge or change dramatically from one situation to another become difficult to scale. Understanding those variations and creating repeatable approaches can make increased throughput possible.
MSP Action: Identify inconsistent and undocumented processes before attempting to scale them with technology.
8. AI Can Change the Economics of MSP Growth
Perhaps the biggest opportunity Peters sees for MSPs is breaking the linear relationship between additional work and additional headcount.
If automation and AI can handle appropriate Tier 1 and some Tier 2 work, MSP professionals can spend more time understanding customer environments, completing projects, handling complex issues, and providing consultative services.
Peters doesn’t view that as a reason to eliminate people. He believes AI can augment them and allow MSPs to place talented employees where human expertise provides greater value.
MSP Action: Measure AI’s potential not only in labor saved, but in the additional higher-value work your existing team could take on.
What This Means for MSPs
Peters’ perspective on AI is neither that it will solve everything nor that MSPs should dismiss its potential. His argument is much more practical.
AI can transform an MSP business, but only when leaders understand the work they’re trying to improve, provide the necessary context, measure meaningful outcomes, and recognize where people continue to deliver the greatest value.
For MSPs, the opportunity isn’t simply to automate more. It’s to become more deliberate about what should be automated, what should be assisted by AI, and where talented people can make the biggest difference.
Catch the full conversation on MSPi PrimeCast Episode #945 and connect with Matt at https://www.linkedin.com/in/matt-peters