AI business topics

Deep thoughts on MSPs in the AI age

by

A lot of MSPs are embracing AI as “internal automation” without rethinking their service catalog, value story, staffing model, or long‑term client strategy. That’s how you end up with a more efficient version of the same commoditized business instead of an AI‑era MSP and it might also be how you end up being replaced by AI that your clients are implementing.

Stop Building AI Toys and Start Designing Your AI‑Era MSP

Most of todays MSPs fit into the very standardized form that was pushed to them as the path to success by a host of vendors.

  • Revenue mix:
    • 60–80% MRR from infrastructure care‑and‑feeding (servers, endpoints, patching, backups, 365 admin),
    • plus project work (migrations, refreshes, occasional security work).
  • Headcount:
    • Mostly Tier 1/2 handling reactive tickets and routine changes,
    • a few senior engineers/architects,
    • minimal dedicated roles for security, data, or business consulting.
  • Value story:
    • “We keep things up, patched, backed up, and secure; here’s your ticket count and uptime report.”

It’s not that this model doesn’t produce revenue and profits today, it’s that this model won’t appeal to anyone in a few short years. This particular train is about to come into station, which means now is the time to restructure and get on the next profit train.

AI agents will erase a huge amount of low‑complexity, high‑volume work: password resets, simple “how do I?” questions, basic reporting, and even most troubleshooting. Almost everything runs in the browser now so there’s little to install or configure anymore. If your revenue model assumes a steady stream of tickets and manual work, the tech is now lined up directly against your P&L.

Some MSPs are drifting into building tools, bots, and micro‑SaaS without committing to being a product company (support, roadmap, GTM, legal, security, etc.). That half‑step is a distraction that rarely becomes a real profit center and often weakens the core services business. If your big AI strategy is ‘tickets went down 30%,’ you don’t have an AI strategy, you have a temporary margin bump and a future pricing problem. Everyone will get more efficient at doing the basics. All of the major tool vendors are building in AI that will make that a certainty. Not only that but business software is building in AI that will not only automate your client’s work but will also make the software self-healing with or without AI assistance but not your assistance.

How does the MSP evolve?

Flip your revenue mix from infrastructure‑heavy to advisory‑heavy

Infrastructure and endpoint care remain essential, but they’re increasingly delivered through, AI‑driven stacks (M365 + Defender + Intune + Copilots, etc.) and become the “platform fee,” not your main offering. Software is just something you provide, not spend your time supporting, because it’s largely self-supporting.

Adopt a modern “stack”

  • Identity‑first: Entra, conditional access, Zero Trust principles.
  • Security: All the Defenders, MDO, MDR/XDR, security baselines, compliance and shadow IT controls.
  • Training: Microsoft 365, Copilot and other AI’s, Security, Data Management and Privacy.
  • Data Management: Control of where data is (DLP), how agents access it (Purview) and aggressive archiving
  • AI: Your clients industry tools brought under governance.

This means leaving perhaps what are today your most profitable service lines behind and reinvesting in your business with new tools, perhaps new people with these areas of expertise or retrain those who are onboard with your vision. Obviously you don’t want to do that on day one, but you do want to stop investing in the old and begin your pivot today and with haste. You need to be repositioned by 2028.

Adopt new service lines

Managed AI services

You can start this today. Businesses are clamoring for someone to take charge and you need to show authority over AI and the vast and rapid level of change. Change and uncertainty is good for business. This is your chance to shine.

Start with:

  • AI readiness assessments (data, licensing, risk, culture).
  • AI acceptable use and ethics policies, governance, and Shadow IT control.
  • Copilot deployment, prompt and use‑case training, tuning usage and security over time.
  • “AI optimization” retainers: ongoing review of prompts, automations, and integrations to keep business outcomes improving.

All of this means that AI governance becomes a first‑class service: defining acceptable use, data boundaries, app approvals, and monitoring

Managed security and compliance with AI in the loop

This is where the next generation MSP will focus the majority of their efforts. If your helpdesk was your profit center today, then this will be your profit center tomorrow.

Fractional CIO/CISO with AI focus

The best MSP are doing this today. If you’re still providing quarterly business reviews, then you’re already behind the wave. You need to move up the stack on this one and hire people that really understand how business works and how technology supports its success.

  • Strategy sessions focused on automation roadmap, AI use cases, vendor consolidation, and ROI.
  • Budget development
  • Process modernization
  • Tie AI outcomes to business metrics: time saved, error reduction, new capabilities.

Your reports then become focused risk posture, AI adoption, and workflow improvements, not just closed tickets and security incidents avoided.

How does this change your head count?

Dramatically is the one word answer. Essentially MSPs are going to flip upside down.

  • Less emphasis on Tier 1/2 reactive roles over time as AI and automation handle the basics.
  • More emphasis on:
    • Security engineers and analysts.
    • Data/automation specialists (Power Platform, Copilot Studio, scripting).
    • Client‑facing advisors who can speak “business outcomes,” not just tools.

AI doesn’t automatically reduce headcount; it changes what your people should be doing to be worth their seat cost.

However in reducing your helpdesk staff, who I have long called the front door of your business, be sure that you don’t let go of customer service as your top priority. MSPs are customer service businesses. It’ll still be true that people like to do business with people they like. So make yourself likeable, white-glove, polite, helpful, friendly and not robotic. A client should never feel like they are stuck in your process. They should feel taken care of.

Rebranding your MSP

As is typical of our industry, new acronyms are being invented daily to describe this new MSP. I’d try to avoid that trap. Things are too in-flux right now to adopt any particular new pigeon hole. Instead, communicate your new value story relentlessly.

Change your messaging everywhere

  • Update proposals, client facing decks, and onboarding materials to clearly explain how AI and automation are part of the service—and why that increases, not decreases, your value.
  • Adopt a new elevator pitch
  • Update your website
  • Update your MRR bundle now to include, AI readiness assessment, governance policy, deployment, training, monthly optimization, and reporting 

The traditional MSP model has meet its limits but that doesn’t mean that your business has to meet its end.

Related past blog articles:

How Third Tier can help:

You may also like

Leave a comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.