Kaseya Won’t Tollgate You
Unless you’re a vendor accessing its data, that is. Plus: Kaseya’s latest unified API progress and ConnectWise’s take on an MSP margin mystery and a debate about MSP rollups.
Deloitte created a minor stir on the socials recently in response to a research piece about “tollgating,” an increasingly common practice in which vendors charge users for access to their own data.
As the article itself observes, some of this isn’t exactly new. Hyperscalers, among others, have long charged data egress fees, for example, to discourage migration to competing platforms. AI, however, has added a new motivation to practices like that and extended it beyond hyperscalers to pretty much every software maker with an MCP server: Feeding data to end users and their hungry, autonomous agents costs money that someone has to cover.
That’s made data tollgating an urge impossible to resist for the enterprise app makers Deloitte mostly wrote about. Is Kaseya feeling the same temptation? After all, it’s got some 3 billion gigabytes of data on its servers. It’s not hard to imagine the company charging usage-based fees to cover some or all of what hosting and distributing that costs it.
And it will, eventually. Just not to MSPs.
That’s one of many facts shared by Kaseya CTO Pratik Wadher (pictured) during an illuminating conversation last week inspired by a recent press release about the new generation of unified APIs the company has been building out since Wadher stepped into his current job last September.
Everyone in Kaseya’s ecosystem benefits from the company’s huge trove of data, very much including Kaseya, for reasons I discussed two months ago. As Wadher observes, however, storing and serving those bits and bytes isn’t cheap.
“There’s a tremendous cost to it,” Wadher says.
As far as MSPs and corporate IT users go, Kaseya intends to foot that bill indefinitely, and it has no intention of charging egress fees either. Indeed, data portability (as opposed to the data gravity Kaseya itself enjoys) is one of the chief advantages of doing business with Kaseya, according to Wadher.
“Customers own their data,” he says, and they can take it wherever they want whenever they want without paying for the privilege.
Vendors, eventually, will be a slightly different story.
Part of the logic behind the “API-first” strategy Rania Succar described very early in her very first U.S. keynote as Kaseya CEO is that it allows the company to embed itself in an industry’s worth of third-party platforms, and the more such platforms, the better. That’s why one of the core attributes of the company’s new API platform is that anyone who wants to use it can do so more or less as they wish.
“It’s an open API,” Wadher says. “We’re being fair to everyone.” But not for free. Or at least not forever.
“The platform is open, but at the same time we want to make sure that we can monetize it, because there’s a cost to it,” Wadher says, adding that Kaseya needs to recoup at least some of that bill.
“We are going to charge for this,” Wadher says. “We’re working through that with our partners to figure out how.”
The goal, however, is not to discourage vendors from using the API. Kaseya has no problem with responsible third parties accessing its data for their own purposes, even if those third parties are system of action vendors and their purposes include building competing products. Such companies pose little long-term threat, according to Wadher, because they offer stand-alone products with no hope of replicating the deep integration and end-to-end capabilities offered by Kaseya’s platform.
“If you look at cyber resilience, cybersecurity, IT service delivery, there is probably no other company like Kaseya that will give you the benefit of all those products in one place,” he says. “That is our unique selling point.”
Three reasons Kaseya thinks MSPs will love open APIs
Vendors aren’t the only beneficiaries of Kaseya’s API openness. In fact, they’re not even the primary beneficiaries. MSPs are, or will be, according to Wadher, for at least three reasons.
1. Open APIs will enable MSPs to use the AI-powered interface of their choice. Remember when Succar told me she had big plans for connecting AI front ends to Kaseya’s back end via MCP last August? I certainly haven’t forgotten, and a large part of why I booked time with Wadher is that the recent press release I mentioned before quotes him on how the possibilities Succar hinted at are coming to life:
“With Kaseya Intelligence, we are building the first truly open, agentic IT platform—one that meets technicians inside the tools they already use, including Claude and Microsoft Copilot, and gives them the ability to move from real-time insights to fully automated remediation without ever switching context. That is not a feature; it is a fundamental shift in how IT gets done.”
It’s not a theoretical, someday shift either. Some of Kaseya’s larger, more sophisticated partners are using systems like Claude and Copilot in homemade AI interfaces, integrations, and workflows today.
“They’re using all different kinds of tools,” Wadher says. “They’re actually moving data out of our products into a custom area so that they can do some analysis.” Other MSPs you’ve read about here have similar ambitions, and Kaseya blesses them all.
“Our goal is really to meet the technicians where they are,” Wadher says. “If they prefer to jump, that’s fine. They can go to the different tools and do more complex queries or more complex workflows.”
2. It will make life easier for all of an MSP’s employees. As in both the flesh-and-blood kind and the agentic ones.
“The API framework is designed for both human and non-human actors,” Wadher says, meaning it will eventually allow both technicians and their agents to ask the same questions, get the same answers, and act on them in response.
3. It will enable MSPs to share agents and automations with peers. “The whole goal here is that whatever an MSP does, because it’s an open platform, they can make that available to the rest of the MSPs,” Wadher explains. “The platform actually gets better over time because everyone else is using it and contributing to it.”
The wait until that process gets underway won’t be long either, according to Wadher. “The goal is really to start surfacing the API platform and to start having customers start using it by the end of this year,” he says. Like perhaps November, when the APAC edition of Kaseya’s Connect conference takes place in Sydney?
“The platform will be the big story,” Wadher predicts.
Kaseya Assist takes command
As you just read, big, sophisticated MSPs are building AI front ends into Kaseya’s back-end data lake. For the moment, most smaller MSPs lack the know-how to do the same. But presumably as LLMs and LLM-powered technologies like Copilot grow more capable, they’ll start constructing custom AI interfaces too, right?
“I see it happening the other way around,” Wadher says. Large MSPs are building AI interfaces because they have no alternative right now, he notes. Most of them would happily let a vendor like Kaseya bear the burdens of funding software development and paying down tech debt if they could. And smaller MSPs aren’t eager to take on those costs either.
“They just want to run a business,” Wadher says. “They don’t want to get encumbered with building other solutions.”
The vehicle Kaseya will eventually use to free them from that encumbrance is Assist, the company’s AI assistant. Today, it’s basically a standard-issue chatbot offering quick answers to basic questions about individual solutions. Over time, as the unified API comes more fully to fruition, Assist will become a centralized source of deeper insights from across the Kaseya portfolio.
“If you’re looking at an endpoint, for example, you’ll be able to see if that endpoint is secure. You’ll be able to see if it is being backed up,” Wadher says. “You’ll be able to look at all the tickets against that particular endpoint.” And without investing in a do-it-yourself project.
Service Leadership takes on the service margin mystery
There’s a mystery embedded in Service Leadership’s recently published 2026 Annual IT Solution Provider Industry Profitability Report.
It’s not why the industry’s top quartile MSPs grew EBITDA 25% last year, versus 10% for their median peers and 5% for the lowest quarter. That’s a long-standing pattern for which Service Leadership has a long-standing explanation: Top-ranked MSPs sell end-to-end packages of standardized services across all accounts in a targeted client base while consistently employing rigorous, disciplined operational processes. What’s different in last year’s data is the advantage top performers enjoyed in service margins, and the degree to which that advantage appears to be mounting.
Service margins are my shorthand for a metric Service Leadership calls service multiple of wages. To calculate it, you total up all the money an IT provider makes delivering services and divide it by all the money they spend on salary, bonuses, benefits, and beyond for everyone on their service delivery team. Higher SMWs correspond to higher profits and growing numbers suggest growing profits.
That top-quartile MSPs do better than others on SMW is nothing new. Service Leadership has long advised clients to target 2.5x on that metric, and best-in-class clients have been averaging 2.7 for a while. The mystery is why that number is rising so fast right now.
“In the most recent three quarters we’ve seen that marching up to 2.8-ish, 2.9-ish, and in Q1 it was 3.01,” says Peter Kujawa (pictured), ConnectWise’s EVP and general manager of Service Leadership and IT Nation. “The only other time we’ve seen a 3 on that was at the bottom of the Great Recession when wages tanked.”
Except that everyone’s SMW presumably rose that time. This one is much more limited in scope. “The median is just barely going up,” Kujawa says. “The bottom quartile is flat.”
Service Leadership’s new report, as thorough as it is, offers no explanation for either why service margins are increasing sharply for top-tier MSPs or why they’re not rising for anyone else. But Kujawa has a theory.
The last three quarters correspond very roughly with the period of time in which MSPs have been adopting service desk automation software in meaningful numbers, he notes, and it’s reasonable to think best-in-class MSPs were among the earliest and most effective adopters. The upshot, Kujawa believes, is that top-quartile providers are growing measurably more productive and hiring more slowly as a result.
“What the data is showing us is that the best in class are starting to be more disciplined in their backfilling,” he says, adding that the bottom-line impact is just beginning.
“I think personally in a year, the best in class will be at somewhere between 3.25 and 3.4 on a service multiple of wages,” Kujawa predicts.
A cynic might note here that ConnectWise, which just launched a new platform with built-in AI functionality, has good reason to speculate without evidence that AI functionality is making MSPs more profitable and productive. ConnectWise, however, hopes to provide that evidence soon when it publishes the first results of its “Automation Index,” a new measure of AI readiness, adoption, and financial impact based on six inputs, including SMW. Watch for it in Q3.
Automation security matters too
Mythos, Fable, GPT-5.6, and models like them are generating a lot of fear, uncertainty, and doubt in the security world right now. Arve Kjoelen, Barracuda’s CISO, brings reason- and experience-based insight to the topic on the latest episode of MSP Chat, the podcast I co-host. Check it out here, and check out our many, many other episodes here.
An answer to the loosely integrated versus tightly integrated MSP rollup debate
Service Leadership attempted to resolve another mystery in its latest report and did, but not in a way that’s likely to satisfy any of the parties most directly invested in it.
Those would be the private equity firms and other investors buying up MSPs in big numbers. Broadly speaking, they have two philosophies about portfolio building. One (The 20 is a good example here) holds that your best bet is to align everyone around a single brand, identical tool stack, and shared operating processes. The other (think New Charter) believes in letting acquired companies retain their brand and culture while sharing resources in areas like sales, marketing, and finance. Service Leadership calls platforms in the first camp fully integrated and those in the latter one semi-integrated.
Which approach is better? Literally billions of dollars are riding on the answer. And having crunched the numbers, Service Leadership can now say the correct answer is … both.
“The [revenue] growth rates are in fact faster for the semi-integrated folks, but the service gross margin and EBITDA are better for the fully integrated folks,” Kujawa says.
And indeed, per Service Leadership’s latest data, semi-integrated platforms grew revenue 26.3% last year, versus 20.6% for fully integrated ones. On the flip side: fully integrated platforms recorded 19.1% gross margins versus 16.8% for semi-integrated ones and came out ahead on EBITDA as well, at 19.5% versus 18.6%.
Those numbers reflect a trade-off between disruption and efficiency, Kujawa believes. Semi-integrated platforms do less onboarding after acquisitions, resulting in less churn and more time for revenue building. Fully integrated rollups, by contrast, collect the longer-term benefits of standardization and economies of scale.
Which outcome is better? That’s a little like asking whether a Roth IRA or a traditional one is better, Kujawa argues. “It depends on what your tax bracket is and what you’re trying to achieve.” Same goes for building MSP platforms.
“It depends on what you’re trying to achieve,” Kujawa says. “Both models can be successful if properly executed is our conclusion.”
One last note for any MSP reading this and thinking I don’t care what works for private equity because I have no intention of ever working with it. Like it or not, you will have to compete with PE-backed providers, and the data says they’ll have an important edge on you.
“Generally speaking, private equity companies outperform the non-private equity companies on EBITDA by about three to five points,” Kujawa says.
Over on The Business of Tech
Host Dave Sobel recently paid automated visits to a bunch of MSP websites to see how many of them reference AI services.
“Across roughly twenty-nine hundred websites, sixty-eight percent have no AI mention anywhere on the site. Even among the larger, more visible MSPs — the ones you’d expect to be out front — it’s about half. The AI gold rush you keep hearing about has not reached most MSP marketing.”
Also worth noting
Barracuda has acquired IAM vendor Evo Security.
Infoblox has agreed to acquire network intelligence and observability software maker Kentik.
Accenture and Google Cloud are collaborating on pre-built, industry-specific agentic AI solutions for mid-market organizations.
Attribute, from DoiT, is designed to map AI token, GPU, and shared infrastructure usage to specific customers, features, and agents automatically.
MSP360’s new public API lets users connect their RMM directly to external dashboards, PSA platforms, and business intelligence tools.
More evidence that AI security matters: ESET’s analysis of nearly 900,000 AI skills turned up tens of thousands of suspicious and thousands of outright malicious ones.
Blackpoint Cyber has launched an AI SOC agent.
Keeper Security’s Endpoint Privilege Manager product now has agentic AI governance capabilities.
Cynet’s cyber platform for SMBs is now available on the Pax8 Marketplace.
Optimize365’s M365 security management platform is now available on the Microsoft Marketplace.
Optimize365 also has a new alliance agreement in place with CyberFOX.
Tia Scripting, the latest addition to TeamViewer’s DEX platform, helps IT teams to generate device automation scripts from natural-language prompts.
Michelle Graff is the new SVP of global partners and channel at Veeam.
Jeff Spridgeon is the new CEO at AI-powered email security vendor Trustifi.
Chris Stoddard is the new CRO at Jitterbit.