IT professionals show surprising resilience in their outlook, with 81% expressing job security despite declining salaries. According to InformationWeek’s 2025 U.S. Tech and IT Salary report, the median salary for IT roles fell from $145,000 in 2023 to $115,000 in 2024, but is expected to rise to $125,000 in 2025. The report, which surveyed 400 technology professionals, notes that while nearly a third are dissatisfied with their compensation, factors such as work-life balance and flexible schedules are also key motivators for job satisfaction.
Demand for technology talent remains high as skills gaps widen, with only 7% of technology hiring managers expressing confidence in their ability to fill in-demand roles, according to a report by Robert Half published on February 13, 2026. The survey, which included 2,000 U.S. hiring managers, revealed that nearly two-thirds believe hiring skilled professionals has become more complex over the past year, with over 75% noting that skills gaps are increasingly evident within their teams. AI and machine learning were identified as the areas with the most significant talent shortages, alongside IT operations, infrastructure, and cloud architecture.
Why do we care?
There’s a contradiction buried in this data that I think most MSPs will miss, and missing it leads directly to a bad decision.
The InformationWeek salary data shows a 20% drop in median IT compensation in a single year. That’s the kind of number that shows up in a budget meeting and gets used to justify holding salaries flat. But then you look at the Robert Half survey — 2,000 hiring managers, much stronger sample — and 7% of them are confident they can fill the roles they need. Seven percent. And the specific roles in shortage? AI, machine learning, cloud architecture, IT operations. That’s not the generalist helpdesk market. That’s the exact capability stack MSPs are trying to build and sell.
So here’s the trap: an MSP reads the salary compression data, decides it’s a buyer’s market for talent, holds compensation flat on their best cloud and security engineers — and then loses them to an enterprise that read the Robert Half data and paid the scarcity premium instead.
You’ve optimized your way into a delivery crisis. And in a rationalization cycle, execution risk is disqualifying. If you lose the scarce engineers who make your AI and cloud strategy credible, you don’t just lose talent — you lose renewal leverage.