Some MSPs Will Ride the AI Flywheel. Others Will Be Crushed By It.
The crushed ones will be smaller firms without the structural advantages of scale. Plus: an answer for MSPs to the high cost of tools and labor featuring agentic AI and open source software.
Filled with hopeful spirit in the season of good tidings last month, I wrote a post noting that while giant MSPs backed by heaps of private equity funding (and in two cases covered prominently here, venture capital) will be a massive presence on the managed services landscape in coming years, smaller MSPs with the foresight to carve out time for AI skill-building can overcome at least some of what are likely to be sizeable competitive disadvantages.
A few short weeks later, in the bracing cold of a post-holiday January, I’m starting to wonder if that’s really true.
A lot of the credit/blame for that slightly gloomy change of heart goes to a frank, must-read LinkedIn post and subsequent interview on the podcast I co-host in which Kevin Lancaster of BetterTracker and Channel Program warns MSPs in the clearest possible terms that “the AI $hit is about to get real” for reasons that go beyond money and people. In the age of AI, he observes, large MSPs have inescapable structural advantages that compound over time.
Beginning with data, the precious commodity that makes AI work. Big MSPs, by virtue of having more clients with more users, SaaS apps, and devices, have more of it than smaller ones, which translates to more operational insights.
“You’re correlating tickets, you’re correlating technician efficiency, you’re correlating your stack, you’re correlating your spend,” Lancaster (pictured) says on the podcast. Convert those insights into operational enhancements with help from AI, he adds, and the result will be a procession of sometimes modest efficiency gains that cumulatively yield better margins and lower prices, over and over and over again.
“That’s that flywheel effect that every business looks to get,” Lancaster observes.
In the case of managed service businesses, it also generates added profit to invest in more and even better AI tooling. And yes, sometimes that tooling will come from one of the many, many, many, many, many AI-native MSP automation startups I wrote about last year, which sell to small MSPs as well as large ones. But in other cases, it will come from dev teams right inside the house. Companies like New Charter and XTIUM, for example, are building highly customized AI solutions for service delivery, support, sales, and more. Titan and Shield Technology Partners are doing the same with the help of world-class AI engineers.
As Lancaster observes on LinkedIn, moreover, really big MSP rollups looking to jumpstart AI development may start buying some of those AI-native automation vendors I wrote about last year. Indeed, while I can’t name names, I can tell you that one of those vendors has received—and declined—multiple offers from MSP giants already.
Bids like that are motivated in part by the knowledge that internal AI innovation can inspire external AI services of the kind businesses will spend nearly $589 billion on this year and over $761 billion next year, according to Gartner. Mega MSP Integris, for example, is using AI to automate “chores” like closing its books every month, notes CEO Glenn Mathis.
“If that’s a problem for us, it’s a problem for a lot of different companies,” he says. “If we can solve it for ourselves, we can figure out how to productize it and sell it to our customers.”
Which ultimately produces an even bigger flywheel with even more dangerous implications for smaller MSPs in which more customers lead to more data leads to better AI leads to higher profits, lower prices, and more AI R&D, all of which results in even more data, even better AI, and even greater profits.
“It’s going to put tremendous pressure on that bottom half of the marketplace to compete,” Lancaster says.
Get moving
Which is not to say they can’t compete at all, just that they’ll need a different way to do so. Lancaster has a prediction for what many of them will choose.
“They won’t compete on size,” he writes, “they’ll compete on focus.” Meaning, among other things, that they’ll lean hard into cultivating industry and functional expertise that businesses need, can’t easily find, and will pay high premiums for.
“There’s always going to be room for niche and verticalized specialization, no doubt about that,” Lancaster says.
For everyone else, he continues, speed and maturity will be critical. Or rather, speed to maturity, which Lancaster defines in a second LinkedIn post that lays out a six-level AI maturity model for MSPs. Right now, he estimates, something like 70% of MSPs are somewhere in the first two levels, using consumer LLMs to boost employee productivity and maybe dabbling in AI-assisted operational enhancement. Getting past that into the more sophisticated tiers where the MSP rollups increasingly live is both absolutely critical and frustratingly time-consuming.
“You have to spend nights and weekends educating yourselves and really getting up to speed to progress along with the rest of the industry,” Lancaster says. “It’s going to take a while.”
Which is a problem, he adds, because time is not a luxury smaller, low-maturity MSPs enjoy at a time when the giant MSP flywheel is gaining momentum.
“Get your butts moving on this stuff,” Lancaster counsels.
Get the whole story from the man himself
Beyond reading those LinkedIn posts I mentioned before, the best place to get Lancaster’s thoughts on the large MSP flywheel is directly from him on a recent episode of MSP Chat, the podcast I co-host. Then, when you’re finished with that one, I encourage you to stick around for the very latest episode, in which Dave Sobel of The Business of Tech podcast discusses the acquisition I wrote about recently. Lots more good stuff where that came from here.
Flamingo’s unorthodox plan to turn MSP systems of record into loss leaders
So if spending more on AI-powered automation is how big MSPs will goose margins in the years ahead, what can smaller MSPs with less data, fewer developers, and smaller budgets do to achieve the same objective?
How about spend less on software? That’s been Kaseya’s suggestion, certainly, since it began rolling out its Kaseya 365 product family just under two years ago, and fellow Miami-based software maker Flamingo has the same thought. Except that if Kaseya 365 is cheap, Flamingo’s OpenFrame suite is cheaper.
As in free.
Founded just over a year ago, Flamingo is a fledgling maker of MSP line-of-business software designed to increase margins by decreasing the top two expenses for smaller MSPs in particular: labor (which accounts for as much as 80% of an MSP’s costs) and what founder and CEO Michael Assraf (pictured) calls the “vendor tax” imposed by most managed service tool stacks. Flamingo’s goal is to banish that tax via open source software.
Its plan for achieving that goal has three components, beginning with OpenMSP, a no-fee community currently helping over 5,000 members collaboratively find, deploy, and implement open source management, ops, and security solutions. Of which there are way more than I, at least, had any clue. MSPs with sufficient skills and patience can integrate a suite of them on their own. MSPs without either can apply for beta access to OpenFrame, Flamingo’s pre-integrated suite.
Though that offering, like everything in the OpenMSP product directory, is free to download and use, Flamingo is not a charity. Backed by pre-seed funding from multiple venture capital firms, the company plans to make money on a cloud-based edition of OpenFrame equipped with Flamingo’s answer to the average MSP’s biggest expense, a pair of AI agents named Fae and Mingo that automate triage, scripting, and ticketing otherwise done by people. Pricing for that solution, which is scheduled to go GA at the end of the quarter, is still TBD.
Charging MSPs for that “system of action” software but not for the “system of record” software it automates is only one way Flamingo stands apart from competitors like ConnectWise, Kaseya, Halo, and NinjaOne. The company’s adamant resistance to cross-vendor integrations is another.
“We understand that might not be a great fit for every MSP,” Assraf says of Flamingo’s closed-door strategy, “but as time goes by, I hope we will prove to them that our stack is better.”
Initial evidence suggests many MSPs are willing to give Flamingo a chance to make that case. Some 150 MSPs are using the OpenFrame beta today and another 1,200 are waiting in line for a chance to give the software a test drive.
Also worth noting
CrowdStrike and Nord Security announced an SMB-focused strategic partnership combining the former’s Falcon platform with the latter’s secure access solutions.
CrowdStrike also announced a definitive agreement to acquire browser runtime security vendor Seraphic Security.
WatchGuard introduced Open MDR, which extends the company’s 24/7 detection and response service to third-party tools in addition to WatchGuard’s own technologies.
Acronis introduced Archival Storage, a long-term, encrypted, compliance-ready repository for infrequently accessed data.
Trend Micro introduced ÆSIR, a new AI-assisted research platform designed to identify and remediate zero-day vulnerabilities in AI infrastructure.
JumpCloud equipped its platform with shadow AI detection, zero trust for AI agents, and other AI security features.
D&H Distributing added the entire Fortinet security suite to the D&H Advanced Solutions+ ecosystem.
Armis, which will soon be part of ServiceNow, launched a new partner program aimed at expanding its global partner reach.
Former CISA chief Jen Easterly is now the current CEO of security community and conference host RSAC.
Pia named longtime exec James Allen its first chief community officer.
ScalePad, which you’ve read about here before, completed its acquisition of Produce8, which you’ve also read about here before.
Speaking of companies you’ve read about here before, Cork Cyber joined the Pax8 Marketplace.
Sound familiar? Nearly 40% of AI time savings are lost to rework like correcting errors, rewriting content, and verifying outputs, according to a Workday study with plenty more to say.
Google Cloud unveiled Gemini Enterprise for Customer Experience to unify shopping and customer service through agentic AI agents.
Dell announced updates to its PowerStore products designed to enhance intelligence, density, and security.
Commvault launched Unified Data Vault, a cloud-native data protection solution targeted at giving cloud developers a unified, secure vault for S3 data.
Wait, TikTok has a channel? It does now, for vendor partners anyway.
The Alliance of Channel Women opened nominations for its 2026 awards program to honor individual and corporate contributions to advancing women in technology channel roles.
Channelscaler named a new CFO, SVP of marketing, and VP of customer success.