Every year, the same argument surfaces at MSP leadership off-sites. If you want to keep your best people, you have to pay them more. If you don’t, your competitors will. The talent war is real, and the only way to win it is to keep bidding.
The Service Leadership numbers say otherwise.
ConnectWise’s benchmarking division just published its 2026 Annual IT Solution Provider Compensation Report — and it contains a finding that contradicts the conventional wisdom so cleanly that it should probably be printed and taped to the wall of every MSP sales meeting. The most profitable managed service providers don’t pay top dollar. They pay less than their peers, experience more employee turnover, and still come out ahead.
Read that again. The best-performing MSPs lose more people and pay them less.
That’s not a typo. That’s a business model.
The Numbers
In 2025, 16% of IT services firm salary increases exceeded 6%. In 2026, MSP and VAR leaders expect only 8.1% of salary increases to hit that threshold. That’s a significant compression — and it’s not being driven by the bottom half of the market.
The stingiest operators, it turns out, are the most profitable ones. And the PE-backed firms are the most churn-prone in the entire study: 54.6% of PE-backed MSPs lost more than 10% of their workforce in 2025. For reference, among bottom-half performers, only 22.8% saw that level of attrition.
Top-quartile MSPs? 38.6% lost more than 10% of their workforce — well above what you’d expect from a company trying to retain talent through competitive compensation.
Meanwhile, small firms — those in the $1M to $2M revenue band — gave top-level raises to more than 30% of employees on average. They’re fighting the hardest to keep people. They’re also, on average, the least profitable.
Draw your own conclusions.
What the Data Is Actually Saying
Here’s what’s happening — and I’ve watched this pattern play out in enough partner orgs to have some conviction about it.
The belief that you need to outpay competitors to retain talent assumes that talent is the primary variable in your business model. High-performance MSPs have systematically removed that assumption from their operating model. They’ve built documented processes, scalable tooling, and repeatable delivery frameworks that reduce the blast radius of any individual departure. When someone leaves, they don’t leave with the job in their head. They leave and someone else runs the same playbook.
This isn’t callousness. It’s architecture.
The smaller, less profitable firms are doing the opposite. They’re compensating for the absence of systems by paying more to keep the people who are the system. That works until it doesn’t — and when it breaks, it breaks badly.
The report also surfaces something worth sitting with: the compression in headcount growth. In 2025, 36.4% of IT solution providers decreased headcount, while 43.5% increased. In 2024, those numbers were nearly reversed — 51.2% were adding people, only 35.7% were cutting. Service Leadership suggests that AI may be filling the gap where human headcount used to go.
That’s the inflection point. Not AI as a product to sell. AI as a labor substitute that changes the unit economics of service delivery.
The PE Model Is a Lab Experiment You Should Study
The 54.6% churn figure at PE-backed MSPs is striking, but it’s not a failure mode. It’s a deliberate operating thesis.
PE firms buy MSPs with the explicit intent to standardize, scale, and flip. They’re not trying to win the retention war — they’re trying to build an org that doesn’t need to. High churn at PE-backed shops is a signal that the rollup playbook is working, not that it’s broken. They onboard fast, systematize fast, and they’ve already accepted that the workforce is going to cycle.
“PE firms have organized their operations to accelerate hiring, training, and onboarding processes to address these trends,” Service Leadership noted. That’s a quote that independent MSP owners should print out and keep near their desks.
The PE playbook is a forcing function toward process dependency. If you’re not PE-backed and not thinking about your business the same way, you’re operating at a structural disadvantage relative to the firms that will eventually compete with you for the same customers.
The Dead Playbook
The playbook that’s dying is the one where talent retention through compensation is your primary competitive defense. It costs money you should be investing in systems, it creates dependencies on individuals that compound over time, and — critically — it doesn’t actually differentiate you.
If your customer relationships live in your best account manager’s head, you don’t have customer relationships. You have a liability that walks out the door every day and you hope comes back in the morning.
The MSPs winning this decade are the ones who’ve treated AI adoption, process documentation, and tooling investment as headcount substitutes rather than headcount supplements. The number of people in a firm is increasingly a lagging indicator of success, not a leading one.
The compensation data confirms what we’ve been seeing operationally for three years. The best shops have cracked the code on building businesses that can absorb turnover without flinching, because turnover was never the real variable.
What To Do With This
If you’re running an MSP and still fighting the talent war with your checkbook, the data is not on your side.
That doesn’t mean you should start underpaying people or engineering unnecessary churn. It means the investment priority should shift. Every dollar you’re spending trying to outbid a competitor for an experienced engineer is a dollar you’re not spending on the documentation, tooling, and onboarding infrastructure that would make that engineer’s departure a recoverable event.
Build the system. Pay fairly. Let the system carry the weight.
The profitable MSPs already figured this out. The Service Leadership data just confirmed it publicly.