AI technology isn’t the issue for MSPs

Last month, I spoke at the Flow event in Nashville, an MSP conference hosted by Rewst focused entirely on AI and automation. Over several days I had the opportunity to talk with MSP leaders who are far ahead of the average MSP in adopting automation.

I left with one overwhelming observation.

AI technology isn’t our biggest challenge. Our business model is

For more than a decade, the MSP industry has been built around support, security, and professional services. Growth came naturally as our customers added employees, bought more licenses, and paid higher per-seat prices.

That growth engine has stalled.

Automation is reducing the cost of delivering our core services while competitive pressure will eventually reduce pricing. Our traditional business model is becoming a melting ice cube.

At the same time, customers are beginning to shift budgets toward AI and automation. The opportunity is obvious.

The question is not, “How do we use AI?”

The question is, “How do we build a scalable, profitable business around it?”

Most of the conversations I had were centered on technology. Which AI platform? Which agent? Which automation framework?

Those are important questions.

But they’re secondary.

The real question is what role an MSP should play in an AI-first world.

Some believe MSPs will become business consultants who analyze every customer’s workflows and build custom AI agents for each department.

I don’t.

Our industry has struggled for years to scale a truly strategic vCIO offering. Becoming high-end business analysts requires even more business expertise. And if every engagement becomes custom development, we’ve simply gone back to the days of project work and hourly billing. The very model we spent years trying to escape.

We’ve already learned that lesson.

If we’re going to win this transition, I believe we need a different set of principles.

• Build repeatable services, not custom engagements.
• Create solutions that are at least 85% reusable.
• Focus on services that every customer can consume (not just your top 5%).
• Build around common roles, business processes, and measurable outcomes.

Today, much of the AI work happening in our industry violates every one of those principles. It’s highly customized, requires scarce talent, and only makes economic sense for a handful of customers.

I’m not saying we shouldn’t do that work. We absolutely should. That’s how we’ll learn.

But learning is different from building a business model.

If our long-term strategy is simply to “AI our customers,” we’ll end up with a services business that doesn’t scale.

If we instead build repeatable, reusable AI-powered services that every customer can buy, we’ll create the next generation of successful MSPs.

Technology isn’t the issue.

Designing the right commercial model is.

I’ll be sharing much more about this over the coming months. It’s where I’m spending most of my time because I believe it’s the single most important strategic challenge facing our industry today.

The winners of the AI era won’t be the MSPs with the best AI. They’ll be the MSPs with the best business model for delivering it. Let’s go, people!

Related: Business owners are asking AI to find their next IT company. Are you coming up?