The AI Revenue Gap: Why 48% of Your Clients Want AI But Only 13% of MSPs Sell It

Kaseya’s 2026 State of the MSP Report carried a number that should keep every owner awake at night. Forty-eight percent of MSPs say AI and automation is the number one client need heading into 2026, ahead of security and backup. Only 13 percent are generating meaningful revenue from it.^[1]

That is not a technology problem. That is a packaging problem.

The demand is real. The revenue is missing. And the gap between the two is where the MSPs who figure this out first will build their 2026 growth engine while the rest chase shrinking deal sizes with the same services brochure.

“The MSP market is maturing, and rising competition is forcing providers to rethink how they grow. The strongest MSPs are tightening their operations, prioritizing efficiency and using data to clearly prove their value to customers.”

— Dan Tomaszewski, Executive Vice President of Channel, Kaseya^[1]

Why the Gap Exists

Most MSPs are using AI internally, they just are not selling it. According to the same Kaseya report, 53% of MSPs already use AI to automate ticketing, patching, and monitoring. But more than half have automated only about a quarter of their workload, and almost none have turned those internal efficiencies into a billable service line.^[2]

There are three structural reasons for this.

First: AI gets treated as a cost cut, not a product. When a workflow automation saves three hours of QBR prep time, most owners pocket the savings as margin and move on. That is not wrong, margin matters. But it means the client never sees the capability, never perceives new value, and never gets offered a higher tier. The efficiency gain stays invisible.

Second: Most MSPs cannot name the deliverable. “AI-enhanced support” is not a service. It is a scoping nightmare. When the deliverable is undefined, the price is undefined, and the sales conversation goes nowhere. As Scopable’s Kristina Shkriabina puts it, “MSP AI services pricing in 2026 needs to start with scope, not excitement.”^[3]

Third: The market is reselling tools, not outcomes. Buying a Copilot license at $30 per user and marking it up 15% is not an AI services business. It is a low-margin pass-through that trains your client to see AI as a commodity. The real margin lives in readiness, rollout, governance, training, and ongoing support, the work that surrounds the tool.^[3]

What the 13% Are Doing Differently

The small group that is generating AI revenue has made a specific shift: they moved from reselling licenses to packaging outcomes.

Here is what that looks like in practice, based on what working MSPs and channel analysts are reporting right now.

1. They start internal, then productize. One Thread partner built an HR knowledge agent in Microsoft Copilot Studio, trained on their own employee handbook and benefits documents, and published it into Teams. Once it worked internally, they packaged it for clients. The pattern is consistent: solve your own problem first, prove it works, then sell it. As Thread’s Matt Linn writes, “This is how AI sell-through begins, not with a roadmap or a sales pitch, but with something useful, built out of need, that turns out to be repeatable.”^[4]

2. They name six distinct service buckets. Scopable identifies six AI service categories, each with different risk profiles and pricing: AI readiness assessment, M365 Copilot rollout, AI governance packaging, AI-assisted managed services, workflow automation projects, and ongoing AI support.^[3] Each bucket is scoped, priced, and sold separately. That is what makes it real.

3. They price the surrounding work, not the license. For a 50-seat client, a Copilot readiness review prices at $1,500 to $2,500. The pilot rollout with policy, training, and 30-day review lands at $3,500 to $5,000. Ongoing admin and adoption support runs $750 to $1,250 per month. The license cost is the client’s Microsoft relationship. The service margin is yours.^[3]

4. They specialize vertically. MSPs focused on healthcare, finance, legal, or construction are finding reusable patterns. One clinic’s compliance workflow becomes the template for every clinic. One law firm’s document intake automation becomes a product. As Thread co-founder Mark Alayev says, “If you nail a few verticals, you will find patterns, and repeatable revenue.”^[4]

The Context: Why This Matters Right Now

The urgency is not only about AI. It is about what is happening to the rest of the MSP revenue model.

Deal sizes are compressing fast. The share of customers spending $25,000 or more annually fell from 75% to 41% year over year.^[1] Seventy-one percent of MSPs say acquiring new customers is now their biggest challenge. And demonstrating value early in the sales process, something 19% of MSPs now struggle with, nearly double the prior year’s rate.

At the same time, the talent gap is widening. The share of MSPs reporting difficulty hiring skilled technicians doubled from 9% to 16%. You cannot scale by adding people faster than the market can supply them.^[1]

Cybersecurity and backup remain reliable, 71% of MSPs reported year-over-year cybersecurity revenue growth. But the MSPs building the strongest forward position are using AI to do three things simultaneously: protect margins on existing services, create a new premium offering, and add clients without linear headcount growth.^[2]

How to Start Closing the Gap: A 90-Day Framework

This does not require a new hire or a six-figure platform investment. It requires naming what you are already doing, packaging it, and offering it.

Days 1 through 30: Audit and name. Identify three internal processes where you are already using AI or automation. Ticket triage. Patch management. QBR preparation. Write down what it saves in hours. That is your starting product.

Days 31 through 60: Package one offering. Choose the highest-visibility use case. Define the deliverable, the timeline, and the price. For most MSPs starting out, an AI readiness assessment at $2,500 to $4,500 per environment is the right entry point. It leads to everything else.^[3]

Days 61 through 90: Sell it to two existing clients. Do not start with new prospects. Offer the assessment to existing clients who trust you. Document the findings. Turn the deliverable into a case study. Then price the remediation and governance retainer that follows.

The goal is not to become an AI company. The goal is to make the AI you are already using visible, billable, and repeatable.

What AI Will Not Replace

Every credible analyst and practitioner in this space says the same thing: trust does not automate. Clients need accountability, a name, and a person answering at 2 a.m. Judgment does not automate. Novel decisions with business context require humans.^[2]

Human-in-the-loop is not a transition state. It is the part that becomes more valuable as routine work disappears. The technician who used to spend six hours preparing a QBR now spends three, and the extra three go toward strategic conversation with the client. That is the pitch. That is what the client pays for.

AI-native does not mean staffless. It means your people operate at a higher level because software absorbed the repetitive layer beneath them.

Frequently Asked Questions

Should I resell AI licenses as part of managed services?
License resale alone is a low-margin pass-through that trains clients to see AI as a commodity. If you resell, wrap it with readiness assessment, rollout support, governance, and adoption services. The margin is in the work around the license, not the license itself. ^[3]

What is the right first AI service to offer?
An AI readiness assessment. For most SMB clients, price it at $2,500 to $4,500 per environment. It inventories current AI usage, data exposure, identity controls, policy gaps, and use cases. It naturally leads to governance, rollout, and retainer engagements. ^[3]

How do I price AI services without hourly billing?
Separate delivery cost, tool cost, risk buffer, and target margin. Price per deliverable, not per hour. A Copilot rollout with policy, training, and 30-day review for 50 seats should land at $3,500 to $5,000 as a project, plus $750 to $1,250 monthly for ongoing support. ^[3]

Is AI revenue a real opportunity or just vendor hype?
The demand data is real: 48% of MSPs rank AI as the top client need. The revenue gap is equally real: only 13% are monetizing it. That gap exists because most MSPs have not scoped, packaged, and priced the work. The opportunity is in closing that operational gap, not in the technology itself. ^[1]

Sources and Further Reading

Kaseya. “AI Emerges as the Key to Scaling MSP Operations as Growth Gets Harder.” Press Release, April 14, 2026. https://www.kaseya.com/press-release/ai-emerges-as-the-key-to-scaling-msp-operations-as-growth-gets-harder
Shkriabina, Kristina. “The Future of AI in MSP Business in 2026.” Flamingo, June 10, 2026. https://www.flamingo.run/blog/future-of-ai-in-msp-business
Kristina Shkriabina. “MSP AI Services Pricing: What to Charge in 2026.” Scopable, 2026. https://scopable.io/blog/msp-ai-services-pricing-2026
Linn, Matt. “How MSPs Turn AI Services into Recurring Revenue.” Thread Service Magic Blog, March 11, 2026. https://www.getthread.com/service-magic-blog/how-msps-turn-ai-services-into-recurring-revenue
Australian Cyber Security Magazine. “Kaseya report points to tighter IT budgets and tougher growth for MSPs.” 2026. https://australiancybersecuritymagazine.com.au/kaseya-report-points-to-tighter-it-budgets-and-tougher-growth-for-msps

The AI Revenue Gap: Why 48% of Your Clients Want AI But Only 13% of MSPs Sell It