Transcript
Good morning, good afternoon, good evening, everyone, and thank you for taking the time to tune in to episode four of AI for the modern MSP. Quick recap, in episode three, we basically explored how service and security delivery must be redesigned to meet the client expectations, right, of 2028. This episode, though, is going to feed a little bit more into the actual P&L for that, right? Much more. This episode is going to ask questions, I'm going to ask Will, introduce him in a moment, questions that operators feel every day, but rarely have the language to answer, right? So if AI is supposed to make my team more productive, where does that actually show up? So before we get into it, everybody, as you can see, I am joined by Will Dowling, not Jim Peterson today. So thank you for joining me, Will. If you would like to introduce yourself. Yeah, sure, absolutely. So for those of you who don't know me, hi, my name is Will. I'm the product marketing manager for our Connectways PSA and the suite of projects attached to it. So AI is a big, big thing that I love using. I'm an AI enthusiast, and so I'm super excited to be here today with everybody. So yeah, Martin, let's kind of kick this off, brother. All right. And for those who, this is the first episode, just a quick intro on myself. I forgot to do that a moment ago. My name is Martin Keer. I'm also a product marketing manager here at Connectwise. So welcome to the series. So yeah, so let's just kick things off then. So the productivity promise is real, but it's not automatic, right? AI is already changing what's possible inside, not just MSP service teams, but pretty much every team out there, right? Whether it's IT or even non-IT related, right? We're all looking for the same things, either repeat tickets or repetitive tickets, fast resolutions, consistent resolutions, consistent outcomes. But realistically, kind of like how we were back in the days of automation, the early days are struggling to be found to be actually productive in terms of tying that to the actual outcomes, correct? So Will, I'm going to throw a question to you here to start things off. When MSP leaders talk about AI and productivity, what are they usually picturing and where does that picture fall short? So generally when MSP leaders start thinking, yes, AI, productivity, they expect a system that's just going to automatically solve most of the problems and make everyone be able to do more all the time, right? But that's really not what winds up happening because they go, hey, I'm going to just bolt on some AI here and it's just going to solve my problems, right? So there are definitely a lot of gaps whenever we're talking through that. Yeah. So again, it's kind of a case of history repeating, right? Well, we now have automate, like back in the day, obviously, we now have automation. I can install an agent profit, right? I remember that was like early, early, early days of MSP world, right? So if we think about how we as a industry kind of absorbed and adopted automation back in the day, and we think about now lessons learned from that and bringing it forward to AI, would you say the gap between the AI adoption and financial results is a technology problem or a decision-making problem or maybe a combo of both? So that gap between AI adoption and seeing those financial results really isn't as much of a technical issue anymore. I mean, like how many MSPs have really figured out what to actually do with the capacity that AI, when correctly and intentionally implemented, creates? Like that's really the question. How many have figured out, hey, they need to be rethinking their processes to work with AI, not just bolting AI onto existing processes. It's absolutely a decision issue, in my opinion. Yeah. And it's also one of those where sometimes where you may even feed a... And I was talking to an MSP recently, actually, regarding this, and it's like AI is supposed to be this ultra super smart colleague of mine, like a team member, like the synthetic coworker, I believe you referred to it once. They are actually having AI review their processes, like, hey, could this be improved? If so, what are the good, better, or best alternatives for this process? And do we as a team have the capacity and the ability and the adaptability to not conform, but maybe to take feedback from AI, right? Because obviously, one MSP doing one thing, you are typically not the only MSP doing that type of process. 100%. So, where do you believe a, excuse me, just putting you on the spot nonstop here for 30 minutes. Love it. Love it. Let's do this one. Where do you think a leader should look? Or a leader, so it could be, you know, owner, leader of a team, leader of like an AI council, maybe. Where do you think they should look first if they want honest, true evidence of how AI is impacting their business, would you say? I love that question. So, I'm going to answer your question with more questions. You ready for this? Please do. So, if you're a leader, and you're looking at like, you know, where is that honest evidence of AI's impact inside of your business, right? Look at like, hey, did you removing some of the drudge work from your top techs make them happier at their jobs? How much would have a training replacement for that person cost you? Look at like, hey, are you CSAT scores going up? How much is that netting you purely in retention, not in having to go out and get new customers? And what's your team able to do now, that they weren't able to do in the past, because you automated part of their day, like AI is not going to just automatically improve margins, it's not going to, it creates the conditions for that margin improvement. So when we're talking about like future state, for MSPs, you know, we're looking at MSPs in 2028, potentially, who, you know, they move past like activity metrics and measure productivity in terms of outcomes. You're looking at, you know, capacity as more of a strategic, you know, resource, and you can really articulate this financial impact down clearly to like, your owners, your clients, etc. And you've really kind of closed that gap, whenever you're looking at AI adoption versus, you know, profit and loss through what you're doing with that additional time, additional funding, additional capacity, right? So, traditionally, right, MSPs have always looked at, you know, cost per ticket average, or average time, technician utilization rates, you know, these are the metrics that we all know and love. And what's the word I'm looking for, KPIs, right, like the KPIs that we've all become accustomed to, which have then fed into QBR, or maybe not quarterly business reviews, I know some companies, they don't do them quarterly. So they're like technical business reviews. Those conversation highlights or topics now are definitely going to be different as we move forward. It's not just a case of, hey, I closed 1000 tickets for you this month, pay me, right? It's not that anymore. So if cost per ticket becomes less meaningful, what do you see replacing those traditional KPIs that we have come to love and understand over the past few years? Look, man, let's face it, like most of the metrics that we're using today, are designed for a reactive task driven service desk when you think about it. So you know, if you're looking at measuring capacity versus measuring output, that's a completely different conversation, right? Capacity is going to replace your old metrics, you know, your old metrics of utilization and cost per ticket, assume, you know, human resolution speeds and human resolution time. Capacity is going to really be able to help you measure what you can do now that you couldn't have done with the bandwidth you had already before. So like, you know, what's the highest value work your team can be doing? Is it retention centric work? Is it revenue generating work? Capacity is really that planning variable that we need to look at in the future. And output, like, that's a retrospective, man, if we, you know, that's what we've already done. If we know that we can really start to look at like, hey, what can we absorb? Where can we reinvest some of this funding that we're seeing the time and savings from and where can we use that to really grow? So that's, that's as far as what I'm looking at going, hey, what replaces our cost per ticket? Well, it's that capacity. Yeah. Yeah. And when I think of, you know, what product, like what productivity actually looks like for the service team, it's going to be totally different. And I remember a phrase that I heard many years ago, when people were, I'm not going to say scared, but they were a little afraid, I guess, a softer word for scared, I guess, you know, automation isn't there to replace you. It's there to empower you. Well, now I think AI is just that, oops, virtual background is, my hand is here. I generally believe, I genuinely believe that, you know, AI is exactly that. It is not here to replace you. It is here to empower you, but like times a thousand, right? The possibilities are genuinely endless. So what would you say the risk, like some of the risks there, you know, like for optimizing for metrics that were designed for a model AI is replacing? If you're optimizing, just going, hey, all right, we're going to optimize everything, right? Well, guess what? You're not reporting on things like business effectiveness or your team's effectiveness and impact. Impact's big, man. Like, you know, if we're talking impact, like productivity isn't just about activity anymore. That's where a lot of people are going, oh yeah, we got all this extra stuff done, right. But what was the impact of that? Like if you're looking at your most productive guys and they're not making a major impact into your business, into your customer's happiness, et cetera, then they're not being used to their habit or their highest and best use. Like they're just being used to continue to do what they're doing. They're just doing more of that. But if you're using it for the highest and best use, like what's the impact that you want them to make? Like that's what really needs to be measured when we're looking at that. And if we're not optimizing for metrics that weren't designed for that, then we've got a problem. Like we can't just measure what's easy, right? We've got to look at, you know, those easy metrics. Let's face it. They don't matter anymore. That's what we're taking. It's low hanging fruit at that point. Like we're measuring easy metrics. Guess what? It's just a nice thing to look at on a dashboard. I hate to say it, but if everyone's learning more efficiently, more effectively, and what's really going to differentiate you from your competitors at that point is things like your retention, your service quality, and frankly, how well you're able to displace those other guys and grow because of the extra things you're able to do with your team's best use to make you more impact to things like the happiness. So if you're not looking at those, you're really just measuring how many tasks AI is replacing, which doesn't really give you any useful information. So you've really got to review like what capacity is recovering and what's that impact and how are you using that impact? So I know, obviously, we had a quick discussion about, you know, this particular episode before we were on air, but something that just dawned on me right now is I believe that this shift to AI, like, and I know this is mostly about profit and like lead episode and like the But one thing that improves that is, as always, communication. I truly see that this as an inflection point in the industry, where MSPs are like, hey, hey, client, these metrics that you have gauged, like you have gauged our success rate on have been A, B, C, and D. Well, with AI, those metrics are going to be X, Y, and Z. Oh, but we're going to work with you on redefining what those are. Yeah. I know this question wasn't in the show notes, but I'm going to throw it at you. Here we go. Love it. Do you believe or do you see potential where your metrics for one client may be potential? Different for others, depending on what that client is wanting or needing to do in their own business journey. You know, it's not like a one size fits all, like, hey, here are the 10 KPIs that we go for for all of our clients. I can see that being some overlap, obviously, of course, but then, you know, I think we call them in the technical world, we call these the snowflakes, right? I can really see those becoming more apparent, but I do believe AI will help us maintain that drift a little bit. What are your thoughts on that? Yeah, I mean, honestly, I agree. AI is going to take a lot of that stuff that we've just been like, oh, yeah, no, these are our 15 metrics that we think make us successful as MSPs. Now, those are out the window, guys. It is what you're going to be doing to tailor your service to your customers and using AI to be able to help achieve that. And that's, again, that's the impact. That is where you're looking at going, cool, this person needs these things, right? This is, you know, what this is worth to us in contract value. Great. How do we achieve that? Look at that from an AI perspective and go, cool, where can we use AI to achieve these specific impacts? At that point, you've got a way to be able to track that and you can actually use those metrics custom tailored for that customer to differentiate yourself from every other guy out there who's still just measuring the same 10 KPIs across all of their customers. This is where you start to differentiate and this is where you start to make a bigger impact to the market. This is also where you make sure that things like your retention are continuing to stay up because you don't want to over automate stuff. If you're sitting there and it feels like they're just talking to an AI, you've done too much automation and too much AI. You need to look at where your people, the human element, can really start to shine and really start to drive relationships because someone might be mad at you, but if you have a great relationship with them and your techs have a great relationship with them, they're much less likely to turn just because something happened that you couldn't really control. This is really where you start to build relationships and start to differentiate what you're doing from the other guys. Yeah. And one thing that I've always believed in is one of the pillars of productivity is consistency. And I truly, we're talking in the future now, MSPs of 2028, this is where everything we say right now is absolutely correct, of course. I can honestly see a lot more MSPs being much more vertically aware. Right now we have a lot of MSPs out there that do go across multiple verticals because we want clients. But I think there'll be more intention, intentionality when it comes to, hey, you know what? We are really, really good in this particular vertical or maybe these two verticals. We want profit. Let's try and be consistent. These two verticals are our bread and butter. That has to be the majority of our services. And these other clients who don't fall into that vertical, well, obviously, I mean, I'm not saying drop them, but what I am saying is you'll definitely find yourself honing in on. Yeah. Again, again, that's what I think. Yeah. I think you're right on that. I think the rise of the specialist MSP is really what we're going to see as we move towards 2028, et cetera. At one size fits all, MSP isn't going to be as successful because if everyone's automating the same stuff with AI, what makes them different? What makes you want to stay with them? So we really need to make sure that we're looking at that, differentiating ourselves between, hey, is there a market that we're really, really good at serving? Cool. Specialize into those markets, into whichever markets we're working on and go from there. I see James commented saying, hey, James, good to see you, buddy. Thanks for joining us. But yeah, I think as these business owners and tech managers and tech leads all start having these conversations with their clients, they're going to realize, hey, we're not only just good in this particular vertical, we actually kind of enjoy this vertical a lot more than we thought we might. Right? Specialist MSPs are the thing of the future, in my opinion. Yeah. SMSPs? Oh, we're adding more. Sorry. Sorry. Time out. I apologize. I'm not going to that term. Okay. I am an SMSSP. Anyway, so when we think of how people are using AI today, right, and getting the efficiency gains today, I think we can agree that the majority of MSPs out there are obviously looking at the low-hanging fruit, the easy stuff, right? Ticket triage, I say easy stuff, it's all relative, ticket triage, categorization, tickets, just looking at tickets, right? So documentation and knowledge capture is also a big one, right? Because as we know, the two things, like you and I, we've been techs, and the two things we hated the most were, well, hate is a strong word, disliked timesheets and documentation, right? Yeah, exactly. So now if you can have AI fulfill that, or at least work with you to make that easier and frictionless, that adoption is going to just accelerate completely. So what would you say, obviously AI is not new-new, it's new, where would you say MSPs are actually seeing the time savings right now and improving what they can do with their capacity in practice today, from what you've even heard from partners already? I mean, honestly, you called out a lot of it. The gains right now are really in the low-hanging fruit, we're still early in the game. So a lot of people are taking and going, okay, what's that stuff that our team still has to do? But it's not driving the business forwards or really making, again, an impact. You know, that's the triage, that's your ticket categorization, your documentation, anything that's not making a actual impact to move your business forward, that's where you need to be automating, that's where you need to be using AI to go, cool, let's just get this off of our team's plates, make sure that we've got some level of human oversight on it so that it's not just going off the rails. And as we're doing that, make sure that we're building into things like time resolution and even bigger things like predictive resolution and escalation reduction, because we're able to do things like reduce our escalations, guess what, we're making our customers happier, we're making our customers able to know that they can depend on us. And really, the next step on that's kind of embedding AI at every friction point, not just bolting AI on, but embedding it in your processes and looking at it from a holistic perspective. Not just going, here's what our process is, we're just going to stick an AI on it. No, like that's not really something we need to be doing. We need to be making sure that that is really integrated. So welcome back, Martin. Thank you. Sorry about that. Technical difficulties. Wow. Not just technical, but verbal too. Verbal, technical, all the difficulties. All the difficulties this morning. May I just say though, while I was gone, phenomenal answer. Thanks buddy. So yeah, I was going to say, let's dig into that topic of what's everyone doing they shouldn't still be doing? You know? That's one of those topics that I'm just like, look, if you're having to solve the same issue more than twice, why? Why are you doing that? Make sure your solutions are capturing that and it's automating the resolution, right? Yeah, absolutely. And again, you know, like we've, I hate to keep harping on about this and repeating myself, but we've definitely been here before with automation, right? It's one of those where like, we've had automation to help us. Now we have AI along with automation to help us. So there is very little room to, I mean, at the end of the day, I've always believed that, you know, clients, customers, respectfully, they will always have their issues, right? That they need to report. Otherwise we'd all be out of business. Exactly. The main factor there though, is let's just make sure that the issues that they bring up in three months are different. I can handle different issues. That's no problem. That's what I'm there for, right? As an MSP to help them through their issues and be the hero or whatever. But if they're the same, so I would also, yeah, I would like, when it comes to like things that they shouldn't be doing, there's a difference between being productive and being busy. I agree. And I think right now, and we can absolutely include ourselves in this one, you know, in the early days, you know, everybody's doing something with AI and that's fantastic. You're learning what works, how you work best with AI. I have realized, you know, and the more you use it, the more you, you're better to, you're, you can, you can better articulate how you approach using AI yourself. So one thing that I would encourage. Oh, I think Martin froze again. Are you there? All right, good. There you go. Yep. All right. Poor Wi-Fi this morning. I apologize. You should call an MSP. I really should. I really should call an MSP. Maybe they can have AI resolve my issue before I even realized there was one. So where was I going with that? Oh yeah. So while I can't call out all of the things that MSP should not be doing, what I can call out is one thing that I believe that they should be doing. And that is having a unified front on how that company uses AI. Because right now an MSP, let's just say simple numbers, has 10 teams. Each of them have 10 specific tools relevant for that team. That's a hundred tools. AI tools are thousands upon thousands upon thousands, let alone the ones that you can even have Claude Code write for yourself, right? So it's one of those where form an AI council in your MSP. And these would be from members of each of the departments and all get on the same page with how you're going to be using AI. Because if you are, and back to the P&L, if you're spending a ton of money on all of these different tools, you're coming into your P&L there. Absolutely. I 100% agree. Yeah, I think that's something I strongly believe in. No, I agree. And I think that's, you make a few good points there that I want to just kind of dial in on. First off, you mentioned productivity without purpose, man. Productive versus busy. That's a trap. Like, great. So we did a whole bunch more tickets. But what do we actually accomplish with that? That's really where we need to look at that. And as we're doing that, some of the hardest things to get ROI out of are the workflows that just have bad data backing them. Like I harp on this constantly, man. Get your data in order. If your input data is bad, your output data is going to be just as bad. It's just going to be just as bad faster than before. So it's going to be bad. It's going to look pretty. Yeah. But it will still be pretty bad. And when we're looking at like productivity without purpose, again, this isn't about tasks anymore. It's, you know, look at like, hey, how many human touch points per resolution are we tracking? It's about that resolution quality more than about time at this point. And, you know, it's it really comes down to like, there's that version of AI productivity that just it looks really impressive on the dashboard, but it really doesn't do much for the business. So, you know, if we're looking at that, as we're doing that, we need to look and approach AI implementation as, you know, fundamentally different than a lot of other tools like you mentioned it. So, you know, if I've got 100 AI tools, I'm paying for all of those, am I getting that return on investment? When you look at outcomes, when you look at that, again, intent, where are we going? What is our goal? What is our long term goal? Where do we need to take this defined outcome, right, and go, hey, this is where we want to be. Define that outcome first. And then that tooling section decision, that needs to be your second choice. Because you can sit and put as many tools as you want, but if you don't know what the outcome you're trying to get to is, guess what? You're not, you're just going to have tools, you have a bunch of tools with no defined path forward for them, there's everyone's going to have them, and you're going to have lots of people who are less busy, but not driving that impact. So things like you mentioned AI councils, things like steering committees, being able to go, hey, from a business perspective, from an operating principle perspective, where do we go from this? What do we want to be in four years? And where do we want to be? How do we get there? And are there tools that can solve multiples of these outcomes that we want to get to? Instead of going, hey, I'm going to bolt this on here, this here, this here, this here, this here, no, it's, hey, is there a set of tools or a couple of tools that we can use and let them talk to each other, you know, maybe a unified platform that has AI in it that we can start to really use to drive those outcomes that we want for our business, rather than just going, hey, we need a tool to fix this problem. Hey, we need a tool to fix this problem, because that's really what I've seen a lot of people doing is they're in the race, because AI is moving so fast right now, it really is. And I think that is actually just, I mean, on the timing, that was a wonderful answer to the question. That is an awesome note to leave it on. But you're right, you know, like, AI is moving so fast. And just when you think you figured it out, something comes up next week. And if you're too busy this week, to get to your AI tooling and playing with AI, you feel like you've missed out already, but you already feel behind in just a week. It's it's a radically changing landscape. And we just hope that these live series that we're putting on for you out there really does help you stay in touch, stay grounded, and we will keep having these for you. We really do appreciate the interaction here and being able to speak to you. And Will, I know it was short notice for you for joining me today, but I super appreciate it. I really do. I'd love to have you on next month as well, if you're available. Absolutely. Yeah, I'm in. Awesome. All right, guys. Well, thank you so much for joining today. We really do appreciate your time. And yeah, until the next one, we'll see you next time. Take care, everybody. Have a great one. Bye.
