“MSPs Must Evolve”: Larry Cobrin on Private Equity, Profitability & AI in 2026

Larry Cobrin from MSP CFO joins Monjur to break down how data discipline, private equity pressure, and practical AI adoption are reshaping the MSP industry.

This episode explores why MSP success is increasingly about metrics, profitability, and operational maturity—not hype—and how founders and PE-backed firms are using better data (and selective AI) to scale smarter, not harder.

See how Monjur takes legal off your plate: https://hubs.la/Q042mCxG0

Connect with us:
LinkedIn: https://hubs.la/Q042mDZk0
X (Twitter): https://hubs.la/Q042mGRL0

Transcript

SPEAKER_00 0:03

Welcome to Talk Tech with Rob Scott.

SPEAKER_02 0:07

Hello and welcome. I'm Rob Scott and I'm your host. And today we're joined by Larry from MSP CFO. Welcome, Larry. Thanks for having me. It's my pleasure. Why don't you take a minute and and tell everybody about MSP CFO and what you guys do?

SPEAKER_01 0:25

Right. So MSP CFO is a reporting platform, but we don't like to think of it simply as reporting, you know, it's a you know, whatever we get out of the GL package or whatnot. What we try to do is discern opportunities for them using information in the PSA, finding a client that's less profitable than the rest of their clients, that's using more time, that's mispriced, hopefully giving them an actionable item so they can drive the bottom line using information that's already in their PSA and working their existing book of business.

SPEAKER_02 0:53

So you sell your services to manage service providers and help them understand the financial data in their business better?

SPEAKER_01 1:02

If we're achieving our goals, that if that is correct, yes.

SPEAKER_02 1:05

But you also do some work for vendors, is that right?

SPEAKER_01 1:08

Uh we do, we have done work for vendors from time to time where uh because we have access to the full data, this is a little bit afield of what we normally do. Uh, we will look at our database uh with anonymous data and try to validate whether or not the vendors are achieving their goals. Uh you know, we've done work for a security company and showed that they've reduced the number of security tickets in companies that have them versus companies that don't, that sort of thing. Uh we don't do a lot of it, but we do uh we like to validate that everybody in the market is is, as I say, achieving their goals.

SPEAKER_02 1:42

What type of MSP or size or however you think about it is the ICP for MSP CFO?

SPEAKER_01 1:51

Sure. So it's really anything um uh so it's anything above the level where the owner really knows everything. So a small two, three, four-person shop usually doesn't need us because they have their hands on everything. And also their data is not quite good enough. I would say more from an OML standpoint. We work better with slightly more mature MSBs because they generally have better data. Uh the revenue breakpoint where people start to talk to us is probably a million plus, a million five plus. And we have clients up to the several hundred million. Uh obviously for them, they there's layers of management and not everybody has disability. So the median is in the five to six million dollar range, 2020 plus engineers, but we have plenty below and plenty above that.

SPEAKER_02 2:35

And and and as far as the data that you study and see, what does it tell you about the type of year that we've had in 2025? What what are your impressions of 2025 based on uh the feedback that you've received from your customers?

SPEAKER_01 2:52

Well, it it's interesting because to say 2025 as a year, uh the same way you would say 2017 as a year, is really difficult to do because there's been so many external issues. One of those external issues is the ever increasing role of private equity and consolidation. Uh, another issue is everyone's investing and trying to work harder using, I mean, as we talked about the conference here, AI tools. Um in general, I think things are broadly improving. But there are, as I say, there's external factors. The turnover of ownership, you know, you walk around our conference, there's a certain demographic that's very well represented, and that would be people who are considering uh, you know, exiting their business, and that's what turned it over. So we are seeing more professional management and such.

SPEAKER_02 3:42

So we're talking about uh a trend where founder-led founder CEOs are being replaced with executive CEOs.

SPEAKER_01 3:54

Either executive CEOs or the founder-led CEOs now run three MSPs instead of one. Uh the private equity, what we find is that they are uh it in addition to being demanding it from a performance standpoint, they're demanding from a data standpoint. They want to know many, many, many metrics. So we're seeing a lot more of that as well. Uh and we're trying to accommodate as best as we as best we can to meet their needs.

SPEAKER_02 4:18

One strategy, I got a buddy who runs a PE firm, and I told him, I want to own my firm and run it, but I want you to help me implement the rigors of a PE firm as if you owned it. Um, because I think that that's really part of how the PE firms extract value is they run a tighter ship. Their budgeting processes are tighter, their hiring strategies are uh more like other companies, you know, in the industry, in other industries. They they tend to think of things in a business way, whereas some MSPs um tend to get in their own way and um get to be very reactive as opposed to proactive and strategic. Um what are you seeing, aside from the differences that you've already articulated, between a PE run MSP and a non-PE run MSP, maybe in the similar OML levels? What are some of the key differences that you see?

SPEAKER_01 5:22

Sure. So it depends who's running them. If it's still the founder has stayed on after, we don't see a ton of difference, although they are much more focused on growth. Remember, a founder-run business really wants to take home money and make payroll. A PE run business, they need to make payroll, obviously, but they also have to make sure they get an and meaningful asset growth. So the differences in operation is that I think the PE run firms are much more focused on growing both the top and the bottom line. Whereas the founder-run business, while they're focused on that, it is it isn't as central. Um I think a lot of, as you say, the discipline comes in once the PE comes in. Yeah, I they're they're more capital constrained, so they have to make more harder decisions.

SPEAKER_02 6:11

Yeah, I definitely have seen that. I talked to Arlen Sorensen this week, who's the founder of HTG and Evolved Peer Groups, and he said that sales is the Achilles heel of our industry. Um, and I have seen that with my clients as well. Um it also is indicative of the fact that uh when I talk to Evergreen and and some of the PE firms, they tell me the first thing they do when they buy a company is install a sales leader.

SPEAKER_01 6:44

Yeah, no, we actually, our company, our former salesperson, is now running an Evergreen Lyra uh MSV. My buddy Solier. Soleil, yeah, she's she's fantastic. She's grown into the role and she's doing an outstanding job from what I understand.

SPEAKER_02 6:58

Yeah, I I saw her this week and she seems very happy and a little overwhelmed in a good way, and she it works for her. That's awesome.

SPEAKER_01 7:07

Well, uh, I'm sure that makes you very proud to see see your teammates grow and and if somebody's gonna leave your company, growing into something like that is it really maybe if that's the way you would want it to happen.

SPEAKER_02 7:19

Yeah, that's really exciting. That's that's awesome. As far as uh what's on the roadmap for MSP CFO, you guys have a great reputation in the channel, you've got a large group of MSP clients or partners. Uh, what's coming in 2026 for MSP CFO?

SPEAKER_01 7:38

So, what we're trying to do is on the one hand, we're trying to do more of the same, continue to provide the right information for people to process, getting it, getting new insights on what they want to know to be helpful for their business. So we're not going to change who we are. But at the same time, if I'm being honest, one of the challenges that we have with our product is some people find it difficult to digest and get the insights that a seasoned user would be able to get out of it. So we're trying to build more, we're trying to improve the UX, always trying to improve the UX so people can use the information better. And then also find ways to build in workflows and automation and where appropriate, include AI to do some more of the work that a user would otherwise have done. Um, we've gone through this experiment a few times with several things that we've built. AI is an accelerant for us in building new reports. It's super helpful in that respect. But we're also looking every time we we build a new process, we A-B test it. We say, does it make sense do we need AI or do we not need AI? Things that we thought AI would be useful for, we have decided we'd rather do it traditionally without, because why introduce the possibility of hallucination? Why introduce token expenditures just to say we had AI when you don't need it? But there are places we will. So I hope we're hoping that 2026 we'll see a lot more uh improvement of the product, consolidation of the information to really push out insights rather than have people find that in the sister.

SPEAKER_02 9:05

That's really cool. So we've been talking to a lot of founders who have exited and planning to exit. And um, as founders, we all at some point think about our perfect exit. Uh, what's your plan as far as preparing for and potentially exiting someday?

SPEAKER_01 9:23

So I was joking with someone earlier. I mean, my my my challenge is, and I think you and I can share this. We're of a certain age that I don't really want to work for somebody else though. And I'm too young to like necessarily be done. And so I want to keep doing what I'm doing. Uh, my goal is to continue to build a team such that my role is less needed, my presence is less needed. The joke that I make is I want somebody to join our company and ask who's that old guy walking around on the corner. That's the dream. Uh but in all honesty, you know, while we are nowhere near considering a sale, we get unsolicited requests all the time. If any of them made sense, we would absolutely consider it. But I really enjoy what I'm doing. This is the best job I've ever had. So I am not looking forward to trying to figure out how to fill seven days a week.

SPEAKER_02 10:12

I want to keep doing this. I'm in the same boat, by the way. I don't know what I would do. Uh, this is what I've been doing, you know. So if I suddenly uh, you know, was on a transition to retirement, and I'm like you, I'm too young to retire, I'm too old to go work for somebody. So in some ways we're stuck.

SPEAKER_01 10:33

Listen, if somebody there are people that talk about they would never sell. That would be dishonest. If someone offered me a ridiculous amount of money, um, I talked to somebody once and they said I would only sell to somebody who was bad at math. The number would have to be so ridiculous, but who am I to say I wouldn't take a number like that? Yeah. No one's offering it to me, so I I would get to continue to enjoy what I'm doing.

SPEAKER_02 10:53

Yeah, that's awesome. And I enjoy what I'm doing, and I think that's so important. And I've talked to so many founders who have sold and found emptiness and despair and depression and heartache, and um the grass isn't always at so much greener on the other side. Right.

SPEAKER_01 11:09

So it's always a question of balance. You know, we all do want to work hard and we all do want to have some time to goof off and enjoy our family and friends, but you need that balance. If you're working all the time and you don't have time with it, that's inappropriate. If all you're doing is playing golf and traveling, that can get tiring also. Yeah.

SPEAKER_02 11:28

I I know as leaders, we're all sort of sort of in this mix of like, what's the future of work gonna be? How do we need to guide our employees into this new era? I know for us, uh, we're fully bought into the agentic enterprise. Uh, you know, my theory is if I can't help my employees get good at AI, I have no hope helping my customers because my employees, at least trust me, work for me and have to do what I say. Right. My customers don't. I work for them. So I better get damn good at helping the people who uh have all the incentive to do what I ask them to do. It is, I can't help them get efficient with AI. I'm not gonna achieve my goals. And I'm just curious how you think about leading in this new world where AI is coming, but it's creating so much uncertainty. What kind of conversations are you having internally with your team about AI and and how you want them to use it and what your expectations are?

SPEAKER_01 12:27

Sure. So in my opinion, AI is we are just at the birth of AI. You know, somebody who's an expert is six months ahead of somebody who's starting from scratch. So what we tell what I tell our team is use AI wherever it's appropriate, never trust it. Use it wherever you can, use an LLM, try to find ways to use it, try to play and build automations using N8N or any other tools that are out there. Uh, because I want to unlock their mind on how we could bring that into what we do. So I give them a lot of leeway to be wrong. And with that, hopefully they would use it more and more. And we are seeing the using the foundational models as sounding boards, as writing tools. But what we're trying to do is really unlock the possibilities of what it can do. We're at a stage right now, I think, we have to, I'm less focused on the new chat GPT-5 or 6 or whatever comes next. I'm more focused on the use cases. So to understand the use cases, you just have to play with it to figure out what it could do. And listen, it gets harder than this because in addition to my employees and wanting to have them be present and part of the new agentic world. I have three daughters between 17 and 21 who are going to enter the workforce. I have no idea what kind of jobs are gonna be there for them. So I I also encourage them to use it more, understand it, to play with it. Because, you know, anybody tells you know what the next three years is gonna look like, I feel that they're they're being they're overstating their confidence.

SPEAKER_02 14:01

I told people when the cloud came out and Microsoft and Google and everyone was all in on the cloud. I used to say we're one big data breach away from this whole thing falling apart. And in some ways, there's so much negative stories about AI. The consultants are saying most AI uh initiatives will fail in corporate America. Uh, we have all this money chasing AI for AI's sake, and we don't have the education to know which bot, which model, which use case that is going to win for the customer. It's like, you know, we wouldn't point uh an Oracle database at something we could do with a spreadsheet, but that's where we are right now with AI. There's such a lack of understanding of what use cases can win, which ones to focus on, what to do first. It's hard, even for MSPs and technical people, it's so hard to know what to do next.

SPEAKER_01 15:01

Sir, 100%. And I agree with you. On the one side, there are people saying that AI is just cool without being useful. And there's people on the other side that said, I have 100% confidence that AI will take over the world. I think they're both wrong. But I don't think we've figured out yet exactly. And we will. I think people playing with it, they will either intelligently get there or stumble upon use cases for AI. And there's a I mean, there's so much craziness. It's a fun story to watch. I'm not terrified of it, but I do want to stay on top of it.

SPEAKER_02 15:37

Yeah, and and I I think about it from the perspective of um as we move forward into this new world, and I agree with you, the extremists are boat drawing. Right. I remember people in the cloud telling me the cloud is just the new buzzword. It just means somebody else's computer, right? Well, who nobody will ever use the cloud. Well, I helped the largest bank in the world get in the cloud. I helped the whole largest healthcare system get in the cloud. It's not just a buzzword, it was a new way to do things, it was a new way to consume uh IT services. And there's still a lot of long runway for the cloud. It's not dead by any means, but this middle, this middle is still unfolding. And we all who are here and present have an opportunity to define what that middle looks like.

SPEAKER_01 16:31

Right. So here's the way I think about it. I think this isn't exactly the late 90s with the internet, but it's very, very similar in my view. And I remember the benefit of being a certain age, living through all that and studying it, thinking most of these companies will go under. Maybe I didn't think it at the time I'm not that smart, but most of them did go under. And I remember at the time, 98 or 99, Jack Wells, the CEO of General Electric, stodgy company. They do financing, they make airplane engines, do all that stuff. Refrigerators. Less so now, but yes. Um He didn't say we're gonna become an internet company. You're not gonna become an internet company if your job is to make airplane engines and MRI machines. What he said is I want every department to save me $50,000 using the internet. And to me, that's where I want to be. I don't want to be the company that's we're AI and then we have this promise that if it doesn't get fulfilled, we've been a failure. I want us to do what we're doing, but do it just a little bit better. And do it better in every corner that we can, and eventually we'll stumble upon a truly interesting use case.

SPEAKER_02 17:41

Yeah, I agree. My coach Dan Martell wrote a best-selling book called Buy Back Your Time. And in that book, the basic premise is you audit your time, you figure out the things that are amenable to being delegated to lower cost resources, you hire that person, and you bought back some time. Right. I've been thinking about that case same concept and the concept of AI. It's like, let's look at what you do every day, and what are the things that you're doing now that we can easily automate with AI?

SPEAKER_01 18:12

Well, there's two things. One, you got to make sure that AI can do what you want it to do. Um and two, it's really about what you not what it can do, but what you don't want to do. So I have the benefit, and you have the benefit as well of being a CEO. And as I built the company from scratch and started adding people, I adore the team that I have. And one of the things I like most about them is they do the things that I really didn't want to do. I'm not a natural salesperson. So I I've I've had a few salespeople that take that off of my plate. I get involved sometimes, but somebody else manages that. I can be a little scattered sometimes, I'm if I'm being honest. So we have customer success people who are very diligent and follow-up requests from beginning to end. And I instead of me trying to tackle 15 different tasks, they take care of the ones that are in front of them. So you talk about having AI do things. I'm trying to find AI to do things. Not only that, it well, first of all, what it can do, which it's not everything yet. And second, what I don't want to do, I still enjoy some menial tasks. I like some, I balance my books every morning. It takes me five minutes, it's somewhat mindless, but I enjoy it.

SPEAKER_02 19:24

Yeah, and it's all about um going back to like Gallup and Strength Finders. If that's a function of something that you like to do, it energizes you, it keeps you grounded, it's great. It's really finding those use cases that drain people. Yes, that can also be that are a good fit.

SPEAKER_01 19:41

Right. One of the one of the simplest ones. I just I I always like to be present in meetings. And taking notes in meetings, I found to be distracting. I'm trying to formulate what they really mean by that. And while I'm thinking about what they said and what I'm going to put to paper, I'm missing a little what they're saying next. The note-taking apps that they have out there, we use one called Grain. There's 15 or 30 or more of them out there. Outstanding AI solution. Love it. That to me is something I hated doing. It does it better than I did. It actually doesn't even cost very much. And I've been pretty pleased at that.

SPEAKER_02 20:16

Yeah, another another use case that that I encourage people to to explore is connect an LLM to your CR web. Yes. And because of that, you're able to do so much reporting. But the one prompt that I really like is ask the LLM how your ICP has evolved over the last three years. That one prompt, once you connect to your CRN, you do that one prompt, and you'll be at your race to first value and you'll be hoof.

SPEAKER_01 20:48

Interesting. I should try that because we we use uh, I don't know, use we use HubSpot. Yeah, same as and they have they have an MCP that we've exposed. Yeah. So we talk to it through our LM.

SPEAKER_02 20:57

Yeah, so start with that prompt. I've got a revenue intelligence record that before the month or the quarter ends, I have all the numbers. I know NRR, which is a hard thing to calculate for a SaaS company. I know upgrade sales and contribution of new ARR. And I know new logo sales contribution of ARR. I don't wait for accounting. As a CEO of my company, I have all the financial close data before the day ends, before the end of the quarter, before the end of the month. I've got a full revenue intelligence report, goes out to my team the next morning, and then accounting validates that.

SPEAKER_01 21:38

Yeah, no, we we yes. There's a lot of neat stuff that could be doing. Yeah.

SPEAKER_02 21:42

So for me, the next evolution of AI is RAG and OS compliance. You can't use an open system, it doesn't listen. Uh so if you're gonna do anything for business, you need anything production facing needs to be in a RAG system. That's closed, not open. Uh, I think once you start with that, now you start getting hugely different outcomes. And then focusing again on those little things. It's the little bots that are going to be the most deployed. Yeah. It's the note taker bot. It's the ticket deflection bot.

SPEAKER_01 22:18

So one of the things I'm trying to build right now, I've just I build it might be out there. I'm trying to build it for fun. Again, I want to build it so I can unlock my mind as how these things work. Yeah. I want something to scan Outlook, Slack, messages, tickets that we have tickets for internally, calendar, so that for every meeting, I get a briefing, a brief review memo for every meeting going in. And I know it can be done and it's fun for me to tinker. It's like the guy that, you know, builds a a a car. We could have just gone and bought a used car. But I like building it. I like tinkering with it, and eventually I'll get it. Maybe it won't be as good as something that's mass-produced, but the fact that I built it is my my enjoyment.

SPEAKER_02 22:59

Yeah, I was just in a uh breakout session at the biggest conference in our industry, and AI is a huge topic. And one of the speakers came up to me and said, You're the most technical speaker on AI at the whole conference. And I have no technical background. I'm just a lawyer.

SPEAKER_01 23:16

We're all six months ahead of somebody who hasn't even started yet.

SPEAKER_02 23:19

But the thing that I'm loving about AI is it speaks English. Yes. That's getting better. That's what gives guys like me hope. I I don't speak in HTML or or or uh whatever the JSON and all these other languages.

SPEAKER_01 23:33

Well that's the interesting thing, is because um a year ago you had to be a a a a uh uh prompt engineer. I actually did a chart to show the salaries of prompt engineers. Uh that went up and now it's gone back down because it's so plain language, you don't have to really engineer it. And it I compare that in presentations that I do to HTML programmer. Super valuable 20 years ago. You don't need them anymore because they're website builders. You want something custom perhaps, but most common people for what they want can just talk to some gamma, whatever it is, yes, and build a website, build a presentation. So a prompt engineering used to happen, now it's all plain English, and it's so much more correct than it used to be. And the thing is, uh I know what AI, I have an idea of what I can't do right now. I don't know what it won't be able to do in six months. I think back to the original Gemini, and somebody, I mean, not to get too into weird, it has to draw an image of World War II Nazis. And it drew a multicultural image, which is literally the opposite of what that class of people were like. It was just dumb and it's it's so much less so than it used to be.

SPEAKER_02 24:41

The evolution um of the tools and the perception in the marketplace, six months is the new year. Uh the cycles are so fast. For us, when we started Monger in 2022, uh we added about 40 customers in that first year. In the first three months of our AI platform, we uh have 141 AI legal assistants that accelerate in 12 weeks. In 12 weeks. Everything's accelerated. So it's like for us, it's like uh a 5x of our cloud native growth, which was still north of 200% every year. Right. So it's not like we took AI and suddenly started growing. We're growing over 200%. And now with AI, I tell my team, I have a 10x vision. And I think that there are opportunities for MSPs customers so that MSPs can go in and help them with AI and create those kinds of transformational outcomes. And that's when managed services become strategic.

SPEAKER_01 25:53

I I hope so. I hope our industry survives as well. But the problem is, is you have to be on it and be fast because what you were able to build so quickly, in 18 months somebody can build in three weeks. My benefit, and this is why what they don't have is they had they can build it technically, but they might not have the knowledge.

SPEAKER_02 26:12

I have two things that put me in a competitive advantage. One, I've got 900 customers that trust me. Right. That pay me every month. That's number one. Number two, I have the legal domain expertise. I'm not gonna have the best tool. I don't need the best tool. As long as my tool satisfies my customers, right? My tool is the right tool. They may build one that is better than mine objectively, but as long as I'm satisfying my clients' needs, my clients aren't gonna leave me over a tool because it's so much more than a tool. It's a relationship, it's a legal relationship. I'm just trying to take a profession that is very antiquated and slow and build on what I already did with a SaaS platform, now an AI platform that's dynamically updated. So instead of dynamically updating legal contracts for the cloud, now I'm dynamically updating knowledge bases for AI. It's really not that big of a leap for us, right?

SPEAKER_01 27:13

The core for us is still the legal agreements and still legal services. We're trying to do the similar thing, understanding people's finances and how everything's going in the market. The interesting thing with using the AI is that uh it can do everything it does, but one of the nice things is one person you go and look at one company, another person go and live in another company. And that takes time. With AI, I can independently send hundreds of agents into every company to take a look at them, put everything back in a singular document, and then see what's going on across everybody. I think that's where it starts to get real fun.

SPEAKER_02 27:47

That's where we go, and we you we talk about OML and evolve. Right. I've created a maturity level for legal. Level one, legal protection and compliance. Do your documents protect you legally of the services that you're offered? Level two, standardization and efficiency. How do you update them? Uh, how do you deliver them? Uh is it seamless for your customers to sign? Have we reduced the friction from the sales process? That's level two. Level three is intelligence. That's what you get when you have all the data, yes. And so um, I think we're all on a journey, and I'm grateful to be on it with people like you. And uh uh if someone wanted to get in touch with uh MSP CFO, how would they do that?

SPEAKER_01 28:40

Uh we have our website www.mspcfo.com. We have a contact us button at the bottom. Uh we're happy to talk to anybody. We're a small company. They want to talk to me directly, just drop my name in there and I'll I'll get back to them. Um yeah. And we're again, we advance our product. We do everything to interactions with our community. Our community is smarter than we are, so we like the interaction.

SPEAKER_02 29:05

Coming from a very smart guy like you that says a lot, ladies and gentlemen, there you have it. Larry from MSB CFO. Thank you for being here. Thanks for having me.

SPEAKER_00 29:14

You've been listening to Talk Tech with Rob Scott, brought to you by Monger. Monger is the first mover in providing contracts as a service solution specifically designed for IT-managed service providers. There's task-enabled legal solutions based on industry-leading templates that are customized for each client and periodically updated to ensure that FSPs always have the latest protections and are legally compliant. For more information, visit LiveZerve.com.com.