The MSP Race to the Bottom Is Over. Here's What Comes Next.

Let’s start with the number that should be keeping you up at night.

MSP project margins fell to roughly 13% in Q4 2024, down from 23% a year earlier (ConnectWise Service Leadership Index). That’s not a dip. That’s a floor falling out. In the same quarter, 18% of MSPs ran at a loss — up from 14% the prior quarter. Worldwide managed-services revenue growth slowed to about 1%.

So the work is getting cheaper, more firms are underwater, and the market that bailed everyone out for fifteen years has stopped growing. Read those three facts together and the conclusion is hard to dodge: the traditional outsourced-IT model is being commoditized in real time, and AI is pouring gasoline on it.

This post is the big-picture case. The supporting math — the AI execution gap, the margin arithmetic, the hidden tax on custom apps — gets its own treatment elsewhere. Here we’re going to name the trap, then name the way out.

Competing on price is a death spiral, not a strategy

Here’s the uncomfortable truth most MSP leaders already feel but haven’t said out loud. When your core offer is “we’ll run your IT cheaper and more reliably than you can,” you’ve built a business on a promise that AI is systematically dismantling. Ticket triage, patch management, monitoring, first-line support — the exact tasks that justified the labor model are the ones automation eats first.

And once your differentiator is price, the next quote is always lower than yours. There’s always a firm willing to bid the work at a margin you can’t survive on. You can win that fight for a while. You cannot win it forever.

When the thing you sell can be bought cheaper next quarter, you don’t have a moat. You have a countdown.

The math makes the trap explicit. Recurring managed-services margins run around 46% (Service Leadership Index) — roughly three and a half times the 13% you’re now getting on project work. The market already prices the difference. Recurring revenue typically commands somewhere in the range of 1.5–2x the valuation multiple of one-off project revenue, and private MSP M&A is clearing at a median of about 8.9x EBITDA (Aventis Advisors). The premium isn’t for revenue. It’s for revenue your client can’t easily walk away from.

So the strategic question isn’t “how do we cut cost to defend the commodity?” It’s “how do we stop selling the commodity at all?"

"We’ll do your AI strategy” is the right instinct and an incomplete play

Smart MSPs already see where this goes. That’s why so many are repositioning as AI advisors and strategy partners. About 43% of channel firms planned to sell AI-related software and services in 2024 (CompTIA). The instinct is correct: move up the value chain, sell judgment instead of hours.

But there’s a gap, and it’s a big one. Gartner expects at least 30% of generative-AI projects to be abandoned after proof-of-concept by the end of 2025. Advice is easy to sell and hard to keep. A strategy deck is a one-off project — exactly the low-margin, easily-replaced work you’re trying to escape. The slides get delivered, the engagement ends, and your client is back in the market next year.

The money isn’t in advising on AI. It’s in operationalizing it. In being the firm that doesn’t just tell the client what to do but actually makes the work move, every day, across their real systems — and keeps making it move long after the strategy engagement is forgotten.

That’s the difference between a consultant your client thanks and a partner your client can’t function without.

Build with AI. Execute with Kinetic.

This is the pivot. Stop defending a commodity. Start owning a solutions practice your customers can’t rip out.

The principle is simple:

AI advises. Humans decide. Workflows execute.

AI is brilliant at recommending, detecting, forecasting, and classifying. It is not where accountability lives, and it is not where the work actually happens. The work happens in orchestration — across the PSA, the RMM, identity, M365, finance, and the dozen other systems your client already runs. That execution layer is the durable, high-margin, sticky thing. And it’s exactly what gets left out of the strategy deck.

Kinetic is that layer. You build the intelligence and the strategy with AI; you execute and operationalize with Kinetic. Concretely, that means:

• Multi-tenant by default, so one solution serves your whole book instead of being rebuilt per client.
• Configuration-driven delivery — you onboard clients in days, not weeks, and industry studies put config-driven work at roughly 2–3x faster than traditional development.
• 100+ connectors to PSA, RMM, AD, M365, and any REST API — an experience and orchestration layer on top of the tools your clients already own. No rip-and-replace.
• IL5-certified, federal-grade governance and audit trails, so “AI made a change in production” is something you can prove, permission, and reverse.

This is the same posture that’s already working for MSPs like Dataprise and Advanced: stop reselling someone else’s roadmap, and start owning the layer where work gets done.

The part nobody puts in the proposal: the tax on what you build

A quick warning, because it’s the trap inside the pivot. If your move “up the stack” is a pile of bespoke custom apps, you’ve just traded one margin problem for a worse one. The industry rule of thumb is that software maintenance runs about 15–20% of build cost every year — forever. Build ten custom solutions across your client base and you’ve quietly hired a permanent team just to keep them from breaking.

Configuration-driven, multi-tenant delivery is how you escape that. You build a solution once and configure it across the book, instead of maintaining a snowflake per client. The custom-app maintenance tax deserves its own post — but you should price it into every “let’s build something custom” conversation starting now.

The choice in front of you

You’re at a fork, and the market is going to force the decision whether you make it deliberately or not.

Path one: defend the commodity. Compete on price. Squeeze labor. Hope AI plateaus before it eats the rest of your margin. The data says how that ends.

Path two: convert one-off project work into high-margin, recurring, sticky solutions. Build the intelligence with AI. Execute and operationalize with Kinetic. Become the layer your clients can’t pull out without breaking how their business runs.

The gap between those two paths is not theoretical. For an illustrative $50M, 200-client MSP, the modeled difference is a roughly $22M three-year profit gap, about +$4.5M in annual recurring revenue, around +$14M in enterprise value, and 60% less build effort to get there.

Those are illustrative numbers. Yours will be different — which is exactly why you should run them.

Run your numbers and see your own 3-year profit gap. Then take a look at how Kinetic works for MSPs.

The race to the bottom is over. It was always going to end this way. The only open question is whether you spend the next three years defending a shrinking commodity — or building the thing that comes next.