ConnectWise CEO Manny Rivelo: AI Agents Shift Ticket Resolution and Labor Costs for MSPs

[00:00:01] Dave Sobel here, reporting from the floor of ChannelCon. I'm joined today by Manny Rivelo. He's the CEO of ConnectWise. And you're coming off a time of a lot renouncements. Three weeks ago, ConnectWise shipped what it's calling the ConnectWise platform, right? Rolling together a rebuild that folds PSA, RMM, Screen Connect, SIM, like all things together into a set of AI agents that you're using under a strategy called predictive IT. It's really a big change in the way ConnectWise goes to the market.

[00:00:29] Manny, it's the first time on the show. Welcome. Pleasure to be here, David. Look forward to it. So the platform went generally available on July 16th. Is your word? Oh yeah. It actually went earlier in July. But yes, it's been available for roughly a month now. Okay. And we had beta customers powering that using the platform. Fair. So what I want to get is, so assume an MSP owner that hasn't been tracking all the announcements. Right. What do I get? You get an integrated, let's call it a system of record,

[00:00:58] which means it's a PSA, RMM, with a series of other tools integrated into the platform, natively integrated into the platform. So what that means, it's not just common UX and UI and common authentication. It's all the services are integrated inside, common ticketing, common generic services inside the platform. And then on top of that, a system of agentic AI capability to do certain workflows. And we're building on that over time,

[00:01:28] adding more agentic capabilities into the platform. So now that converts a system of record, you hold a SaaS application with database into a system of action to take care of actions for NMSP. Is there a particular headline thing that people can implement right now that's the bit that you're talking about? Yeah, most of it is really around service delivery, ticket triage, ticket dispatch, ticket summarization, ticket recommendation around resolution, all of that kind of stuff.

[00:01:57] is available today. Co-pilot capabilities, so you could pull it and say, tell you what my most important tickets to work on. All that kind of capability is in the platform natively. And what you'll see is us adding more agentic capability around remediation here in the coming months. So you'll start to see a lot of RMM type remediations, working with bots, working with RPA to really remediate at MSPs,

[00:02:27] tickets. So I want to get a little bit of your take on the market. You've kind of come out and said that the old MSP model is out of runway. I like that. It's a big bold statement, but service leadership, your own benchmarking organization, just reported the best profitability runs the industry's had on record, right? Six straight years of 19% EBITDA. So how do you reconcile those two things? But if you look at the service leadership data, inside of it,

[00:02:54] if you look at best-in-class MSPs in general, the investments that they've been applying into, let's call it automation, agentic capability, has outstripped everybody else and their wages, they have a 14% increase on service wages. So what you're starting to see even in a service leadership data is those that are applying automation, agentic capability, are pulling away from everybody else.

[00:03:24] Everybody else is still producing per employee the same revenue as those that are applying AI technology. So yes, you're going to see over the coming year or years, that accelerate and people will be able to get more productivity, more revenue, better service delivery by using agentic capability in addition to the stack that they kind of has. So help me understand what specific portions of the model you think are the most problems, the runway is running out of, right? Where's the bit that people need to focus

[00:03:54] in their MSP business if they're not doing? It's free up the mundane work. So if you look at the equation for an MSP is, start an MSP, you build a set of services, you acquire customers. The more customers you acquire, the more tickets you end up receiving. The more tickets you're receiving, the more labor you have to hire to solve the tickets. A lot of that labor cost, those level one, level two tickets, reset my password, Wi-Fi doesn't work,

[00:04:23] our printer's broken, the list goes on and on. A lot of that can be automated through agentic capabilities and resolve. You free up that capacity, which is the bulk of your capacity. And what we're seeing when that gets deployed, MSPs are seeing anywhere between 30 to 60% ticket resolution, right? Repeat tickets go away because the agents are taking care of them. So you're starting to see that capacity regain and then applied toward revenue generation work, more higher level work inside an organization.

[00:04:52] That frees up that labor cost. And that labor cost is one of the most expensive costs for MSPs. And also burnout for MSPs and then finding talent is pretty difficult. You said in a couple of those numbers, I'm going to, the ones that I pulled out there, 45% off handling time, 30 to 40% fewer tickets, five to 12 points of margin. Yeah. But are those dual models or are those customer models? Those are? Which came out of production time? Those came out of production data

[00:05:22] that we're noticing as they've been deploying as they've been deploying agentic capabilities. We've also now put into the platform the capability where, think about it as an odometer, as the agent is doing its task, it's reporting what it's doing and giving you your savings back, right? Your capacity back. And we're not telling MSPs what to do with that capacity. What we want to say is you have the capability of free up hours. That allows you to do a lot more. You can now get much closer, right, to your end customers

[00:05:51] servicing better with higher level work versus resetting a password. Okay. The customer in there you named in the GA release lands about half the model, right? Is that the Starport is the one named specifically there. Is that a realistic number or is Starport like an early mover on that kind of half the model? They're a realistic number. Okay. The MSPs and this is why we've created this dashboard is they all measure productivity differently.

[00:06:21] Like some folks we heard, well, I have ticket resolution. I no longer need to hire another technician, right? Others were, well, I used to have a dispatcher. I used to have two dispatchers. I lost the dispatcher now that AI agent is doing the dispatching and triage, right? So I don't need to replace it. So it was quantified differently and that's why we're trying to get to one set of metrics that over time we can build that out. Doing the same thing with the SLI data, right? We're starting to pull

[00:06:49] how much money are you spending in hard cogs, meaning tools, AI, et cetera, and then what are your outcomes? And we're starting to see the correlation there also. So you have more data with time, but yeah, it's very clear that a lot of this repetitive mundane work is going to disappear and offering you the capacity to get much closer. One of the things I do here at MSP Radio is track what MSPs are actually saying in the communities where they're the most honest.

[00:07:20] And right now, the conversation is dominated by one theme. The people selling AI as the solution don't understand what it's actually like to run an MSP practice. This is why Pax8 makes the most sense to me. They're the only marketplace in the channel where MSPs can cut through the AI noise and turn it into a business, a guided path to repeatable monetization, a curated catalog of agentic solutions to address specific small business challenges,

[00:07:49] and a unified marketplace that makes provisioning, governance, and operations simple instead of risky. If you're looking for fewer moving parts and a stronger business, that's PAX 8. Get started at pax8.com. That's P-A-X, the number eight, dot com. The other thing that I really thought was interesting is you said that the larger MSPs are benefiting the most here, right? Yes. Best in class MSPs. Okay. Not always larger. Not always larger. Okay. So how,

[00:08:19] if the audience, a lot of the MSP market does skew smaller, right? So like there's a long tail of that. Like what does that do for the ones that are in that struggling end of that? How can they leverage this together? Yeah, the larger MSPs because of their size can invest in the development of some of this agenda capability themselves. And it leads to an interesting conversation. It's the build versus buy this capability so they can maybe build some

[00:08:48] of this capability. The reality is that's a tough value proposition because, and there's confusion in the market. There's this concept of core versus context. The core to an MSP is the service delivery, the intimacy, understanding their customers, meeting them right where they need to in their backyard, their neighborhood, right? And getting that intimacy. The context is sometimes the tool sets they need to run the business. AI is interesting because AI gets viewed as core. It may not be core.

[00:09:17] It's only core if it helps you with the core services, giving you intimacy to that customer and differentiated intimacy to that customer or service delivery that's differentiated. If not, it's actually context and you need to use it, right? So what we're doing is we're integrating that into the full stack and that allows any MSP of any size to not be in the software building business but leveraging the software to get the outcome for themselves and for the end

[00:09:47] customers. So the smaller MSP who might be more nimble can then deploy the technology, realize that it's context but positioning it to deliver the core which is the services, right? So that's the way we look at the market and that levels the playing field. So don't chase the bigger MSPs and try to create a software development house. You don't need to do that. We and the ecosystem are doing that for you. Okay. I want to spend the second half of our discussion on the economics bits, right? So where does

[00:10:16] cost a partner and who actually pays for it? I'm going to try and be really precise so I want you to push back on me anywhere I get it wrong, right? Because I want to make sure I get it right. My understanding is there's three meters here per user for the co-pilots, per ticket for high volume ticketing and consumption pricing for the complex work. Yeah. Have I got it right? Yeah, you did. So if you, the co-pilot capability, I'm going to change it just a little bit. The co-pilot capability is me, let's say as a technician

[00:10:46] or anybody sitting inside an MSP polling the system and asking it for data. No different than you would do a search, just an AI-agentic search, if you will. So that's based by user. It's for your personal productivity. Right. And you prompt it the way you want and you get the data you want. What you refer to as per ticket, it's not really per ticket, it's per workflow. Okay. A ticket may have five different workflows. Okay. Right? So we,

[00:11:15] when one of the workflows could be we see a ticket, we summarize, I mean, we triage the ticket, we dispatch the ticket, we put it, David, on your task to do. We go and reach into your documentation and say, here's, and in your history and say, this is how the ticket gets resolved. This is exactly what we should do. We summarize all of that. Okay. That's a workflow. Okay. That's a task that a human would do. Okay. Now we could do it

[00:11:44] through AI technology. In seconds, nanoseconds, right? Right. And deliver that consistently every time based on your specific data. So it's not a ticket because that once, let's say you get to that point and now you as a technician, you're a policy, I want to review it and you review it and you say, okay, I agree. Click. You can have another workflow, which is the remediation part. Right. Right. Which again, would be a human. Right. So what we're doing is trying to address that per workflow and over time

[00:12:14] integrating those workflows deeper and deeper. Now, when you get into other areas, other services that we will launch over time, you can get into remediating a very complex task, troubleshooting a very complex task. That's not, that's, that could resetting a password is a very simple thing. Changing your whole environment, your network infrastructure. So that will be consumption based. So yeah, but the two that are in play today are per workflow and per use. Okay.

[00:12:44] So what I'm getting at though is if we're charging per ticket per workflow for a product that promises 30 to 40% fewer tickets, your revenue goes down when it works, right? Am I off? Am I off? No. And so that's what tickets that you have to handle as an MSP. The agents are handling them in the background. The ticket volume doesn't go down per se. It's just what you work on. So if an MSP is working on 100 tickets on any given day, 30, 40, 50, 60% of those can be handled

[00:13:14] by agents. Okay. So you're still working on 40 tickets, but the total ticket volume is still 100. We still open 100 tickets. We still resolve 100 tickets. We still document it. We still send out bingo cards to make sure the end customer is happy on 100 tickets. All that's still happening. So it's the task to the MSP, the labor task to the MSP. From an end user perspective, there's still 100 tickets. Okay. But if I'm an MSP who's flat fee per seat

[00:13:43] until my renewal, you've just made a chunk of my cost variable and metered by doing workflows and tickets. So when the token bill spikes in like month four, is there a cap or do I eat that bit? Because there's no variability to my fixed fee engagement. No, because I understand. But our job is to deliver that workflow at a fraction of the cost of a human delivering that workflow

[00:14:13] with consistency and better outcomes. Okay. Right? So if, again, simplify it. It's $10 an hour for an employee to do something. Our job is to do that in pennies. In seconds. Right? Right? That's the delta. So what you're basically doing is offsetting the tasks that are mundane that take human capital that is too expensive and a technology can drive it faster for you. That's all. Okay. Here's what I'm hearing

[00:14:42] from MSPs on backup. They want control. Control over storage. Control over costs. Control over what happens when something breaks. Comet Backup gives you that. Bring your own storage. White label it for your clients and keep margins where they belong. With you. It's why Comet Backup keeps showing up when MSPs ask each other what actually works. See for yourself at CometBackup.com. The other thing I wanted to ask about is

[00:15:11] so your proof point in all this is that your security division asked you for 200 hires, right? And you deployed agents instead. Correct. Okay. Those 200 people didn't get hired. Right. So it was actually 400. Okay. When we pushed the 200 and you're absolutely correct. Okay. We deployed agent technology and we did not hire those and were able to become much more efficient. Okay. and now you're telling partners it isn't about headcount. So help me understand.

[00:15:40] Square that circle for me a little bit. Well, I couldn't have hired 400 people anyway, right? Okay. All right. I mean, you can't get your hands on that. It would have been inefficient. Facility, space, and all of that. So what we were able to do is not hire those resources. Okay. Take care of most of the level one, level two work through agent technology. We were able to go from a service level objective that we would handle your ticket in an hour to a service level agreement that we would not only handle your ticket but resolve your ticket

[00:16:10] in 15 minutes. Okay. And we have nine, five, seven nines of accuracy on that. We missed like one a month but we can't handle that for some reason. Right? So the agent technology. So now we've delivered a better service right? I had a better outcome altogether and I couldn't have been able to hire that talent. I would have been looking at real estate but am I going to find space for that many people?

[00:16:40] So that, what we didn't, and the folks that are there that were doing level one work are now doing more advanced work, right? So we get to use the capacity. They become happier employees. It's the same we did to ourselves the same exact thing that we're recommending MSPs do to grow and scale their business. Applier talent where there's more value. Makes sense to me. So last couple of questions. You inherited ISO. Like you inherited the platform and you dismantled it.

[00:17:10] We evolved it. Okay, evolved it. So last November the company was talking about advancing it. In June the headline essentially says you've taken it apart. What should a partner read from those headlines and directions? What should they read into it? Look, the direction that the company had around ASIO was a good direction. It was MSPs have grown over time and have developed a tool set tool squall and it would be nice if we could consolidate that tool and make the tools better integrated

[00:17:40] and work better together. Give you a simple example. There's ticketing in an RMM. There's ticketing in a PSA. Why would you need two ticketing solutions? They do different things. One ticketing solution as a common service, right? That thesis is always correct. The challenge was architecturally we had to dismantle some of the bowels of the tool and rebuild those. So David Rasapur and the engineering team have come in and enhanced that and changed that.

[00:18:10] ASIO is still inside for lack of a better word, right? It's just been engineered for the right level of scale that we need. The fundamental differences then organically and through an acquisition we made in January, Zofric out of Toronto, we've now embedded AI into the platform. Not bolted it on, it put it back in the platform as a common shared service living with the data right in center. So now we can build new agents and they're not bolted on.

[00:18:40] They're inside that platform. Okay. Completely inside that platform giving us that efficiency and that scale. So we feel really good. So then how do we communicate that? We're going to say it's A0 Next Gen, it's A0 2.0, it's all of that. We just said, look, what it is is the ConnectWise platform and we renamed it the ConnectWise platform. Makes sense. I appreciate the transparency, sir. So last question there and I'm going to give you the chance to kind of argue your own side here. What's the thing MSPs that are listening to this that are doing right now that they should stop doing in the next 12 months

[00:19:10] and what happens to the ones that don't? I think the biggest thing that worries me about this industry, there's concern that you have to be big right to change and transform through this. Absolutely not. I mean, I think you first of all adopt, get curious on what's happening in the world. As I said today on stage here, if you have 12 employees, 50 employees or 100 employees, I guarantee you at least have that many agents because everybody's using it.

[00:19:40] The most common thing is we use Google and now Google gives you an agent response, right? That's much more that we're getting all acclimated and accustomed to that. So begin to adopt AI, CAS. Don't worry about if you have to build it. Really evaluate your core versus context framework. Say, what am I really good at and what do I want to deliver and how do I buy the tool set to deliver those services better and allow me to scale and become much more

[00:20:10] intimate with the partner base. So I think there's a lot of folks that I got to build this thing. When we get to our advisory council, we talked to them, we had a great meeting about 45 days ago and if you had asked me six months ago, some of the larger MSPs were saying we're going to build this stuff. Some of the larger MSPs are building some things that are core to their business but many of them said, well, we drove our token cost to 70k a month and we've seen no transformative

[00:20:40] transformation because we're building our science projects. So be careful of building science projects. Stop that. Start applying the solutions that are in the market and get the benefits. We're happy to show you where to start. We have hundreds and hundreds of partners that are already starting and growing every single month. So the opportunity is there and don't spend your time wisely like you have the best. Practical advice. May, thanks so much for joining me today. Really appreciate it. Pleasure, David.

[00:21:09] What would you fix in your business with the right playbook? Small Biz Thoughts members get a library of templates and operational resources, recorded member calls, and classes through IT service provider university. Operational education built specifically for independent MSPs. Start at smallbizthoughts.org. Interested in advertising? Head to mspradio.com slash engage. The Business of Tech is written

[00:21:39] and produced by me, Dave Sobel, under ethics guidelines posted at businessof.tech. Thanks for listening. I'll see you on the next episode. Produced by Picture This Video, part of the MSP Radio Network.