Before we get to the business model, there is the matter of getting the tools
A software purchase in the MSP ecosystem passes through multiple hands before it reaches the person using it — vendor, distributor, possibly a cloud marketplace, then the MSP, then the client. Each step generates a quote, a purchase order, an invoice, a payment, and a reconciliation. Quantity changes require the whole chain to flex. Contract dates rarely align. Errors generate their own sub-process of correction, credit notes, and follow-up emails that nobody planned for and everyone absorbs.
This repeats for every tool in the stack. There are many tools in the stack.
The client has no idea any of this is happening. They think they are paying a monthly fee for IT support.
Every significant vendor in the MSP stack is meanwhile having the same conversation with itself: we need the channel for reach, we resent the margin the channel takes, we will build direct capabilities and deny we are competing with the channel while restructuring the partner program in ways that favor larger partners without announcing it, and we will sponsor the conference where we tell the channel how valued they are.
The MSP is caught in the middle of all of them simultaneously.
This is before anything has gone wrong.
The managed services business model is this. Charge a flat monthly fee to monitor, manage, and support a client’s technology environment. Keep the systems running. Patch the vulnerabilities. Respond when things break. Be accountable for everything so the client doesn’t have to think about it. Bundle in Microsoft licensing. Add security tools. Throw in whatever else the stack needs.
Car dealers have low margins on new cars. They make it up on service — predictable work, standardized pricing. The service bay is where the money is.
Retailers have low margins on product. Once you leave the store, their complexity ends. Nobody calls Best Buy at 11pm because the television is behaving strangely.
The MSP has combined the margin profile of the retailer with the support complexity of a business that gets called at 11pm. Frequently. About things that are genuinely complex. By clients who cannot understand why this is taking so long.
The margin on a Microsoft license resale runs around 15% on a good day. Hardware is worse — 6% is not unusual. In most professional services businesses, a margin below 50% is considered a problem worth solving. In the MSP world it is considered a revenue stream.
The service margin is where the business actually lives. The service margin is under pressure from every direction simultaneously.
This is not a recipe for effortless wealth generation.
Private equity discovered that MSPs have recurring revenue and pattern-matched to software. Recurring revenue. Fragmented market. Obvious consolidation play. Buy the MSPs. Offshore the help desk. Extract the margin. Flip in five years to a larger PE firm that will do the same thing at a higher multiple.
The part they missed is that recurring revenue built on trust behaves differently from recurring revenue built on a contract. The client stays because someone they rely on is accountable for their environment. The moment that person gets replaced by a standardized offshore process, the trust migrates — slowly at first, then faster than the model predicted.
The PE firm that made the acquisition is not the firm that discovers this. They have already flipped. The next buyer paid a higher multiple for a business whose underlying dynamics have not changed.
The same thesis is meanwhile being applied to the vendors in the MSP stack. Sophos and others have found their own private equity owners, each on their own five year clock, each optimizing for exit. Which means the MSP is now sandwiched between PE-owned tools and PE-owned consolidators, nobody with a ten year interest in the outcome, everyone extracting margin from the same ecosystem simultaneously.
What could possibly go wrong.
The tier one help desk ticket — password reset, printer not working, how do I do this in Excel — is becoming rarer. Not because of any single tool. Because users got more capable. AI handles questions that used to generate tickets. The colleague who knows gets asked on Teams before the help desk gets called.
What remains is the hard stuff. The weird interaction between two systems nobody designed to work together. The incident that requires judgment, not a script.
The offshore help desk is optimized for the volume that is disappearing. It is not equipped for the complexity that remains.
The most immediate Microsoft problem is not strategic. It is operational and it arrives constantly.
Microsoft does announce things. The problem is the volume. Changes, updates, new programs, revised terms, and licensing adjustments arrive in a firehose of partner communications between which it is essentially impossible to distinguish the important from the routine until something lands on a client invoice and requires an explanation.
The MSP finds out about a pending price increase from another reseller. Asks their Microsoft partner rep. Gets directed to another group. Extended Service Term billing appears on invoices. Nobody at the MSP knew it was coming. The client wants to know what it means. The partner rep is unavailable.
Microsoft is not doing this to make the MSP’s life difficult. They are running a global business of enormous complexity. The MSP is a rounding error in that calculation.
Which is precisely the problem.
So. To summarize.
High responsibility service business. Thin margins. Help desk handling fewer simple tickets and more complex ones. PE running a software playbook on a trust business while also buying up the vendors in the stack. Primary vendor whose internal decisions land on the MSP’s desk without warning, explanation, or apology.
What could possibly go wrong.
The MSP Contrarian: Unstacking the Business ModelAn independent publication. Not affiliated with, sponsored by, or particularly popular with any vendor in the managed services ecosystem. Subscriptions are free. Opinions are not.