Nearly one-third of managed service providers have put AI tools into operation, but only 12% can identify concrete productivity improvements, according to a new industry survey. The 2026 State of the MSP Business Report, released recently by Nerdio, draws from responses by more than 150 MSP decision-makers who oversee Microsoft 365 and Azure for their clients. The findings reveal a widening gap between AI deployment and tangible business outcomes, with most providers still working to turn experimental tools into measurable advantages.
The data shows that 32% of MSPs have integrated AI into their operations, while another 46% are either testing or preparing to roll out AI within the coming year. Despite this momentum, measurable gains remain rare. Only 14% of providers currently run an AI service line that generates revenue, and just 17% report that clients are inquiring about Microsoft Copilot or similar AI products—questions they aren't yet equipped to answer. Governance presents another obstacle: a mere 14% of MSPs consistently enforce a written AI governance framework across all customers using Copilot or other AI platforms. Meanwhile, six in 10 MSPs say the complexity of managing Microsoft environments is outpacing their capacity to grow. Over the past year, 29% expanded their customer roster without adding proportional engineering staff, compared with 17% who increased headcount in line with growth. Burnout has taken a toll as well, with 29% of MSPs losing technicians to workload pressure in the last 12 months. Concern about burnout rises sharply with workload intensity—from 34% among MSPs supporting fewer than 200 users per engineer to 54% among those managing more than 200.
"AI didn't create the divide in the MSP industry. It's making the one that was already there easier to see," said Will Ominsky, VP and GM of MSP Business at Nerdio. The real differentiator, Ominsky added, isn't whether an MSP has AI tools at its disposal—it's whether the provider has built an operational foundation strong enough to deploy them reliably across complex, multi-customer environments. He noted that MSPs that create a more uniform, repeatable service delivery model will be in the strongest position to transform AI from a trial into a quantifiable business edge.
The report identifies how a provider handles M365 and Azure across its customer base as the strongest predictor of outcomes—more influential than company size, staff count, or the number of tools in use. Providers are grouped into three categories: those relying on native Microsoft tools and manual workflows, those using community and open-source solutions, and those operating commercial multi-tenant management platforms. MSPs using commercial platforms can onboard a new user while accessing three or fewer separate portals 52% of the time, compared with 22% for those using native tools. When asked for compliance evidence, 77% of commercial-platform MSPs can deliver an audit-ready M365 assessment the same day, versus 53% of native-tool MSPs. The report suggests that adopting AI tools in isolation is unlikely to produce meaningful productivity shifts without first addressing underlying operational inefficiencies. It recommends that providers assess whether their current Microsoft 365 and Azure management processes enable consistent onboarding, reporting, and governance before scaling AI rollouts.
For MSPs, the path forward centers on establishing clear productivity benchmarks—such as technician time savings and service delivery efficiency—to gauge whether AI investments are yielding returns. With only 14% of surveyed providers running revenue-generating AI service lines, those that can show tangible improvements may be better positioned to convert internal AI adoption into customer-facing offerings. The report concludes that MSPs capable of documenting concrete results will have a clearer advantage as the industry moves from experimentation to execution. For providers navigating this transition, the ability to measure and communicate value may prove just as important as the technology itself. Yet the real test will be whether organizations can redesign their workflows to support sustained innovation, rather than simply layering new tools onto strained operations.