Episode 77: 75 Days of Uptime

Read the Transcript

Erick and Rich ponder the tradeoffs of using an all-inclusive managed services platform versus going with best of breed as well as techniques for migrating managed service clients to a standardized services stack. Then they’re joined by Yoav Susz, general manager for the U.S. at Atera, to discuss Atera Autopilot, a fully autonomous AI-powered help desk tool now in general availability. And finally, one last thing: a look at Internet Roadtrip, the world’s slowest massive multiplayer online game.

Discussed in this episode:

In Defense of Seams

Atera Autopilot is (Almost Eerily) Live

Thousands of people have embarked on a virtual road trip via Google Street View

 

Transcript:

Rich: [00:00:00] And 3, 2, 1. Blast off, ladies and gentlemen. Welcome to another episode of the MSP Chat podcast. Your weekly visit with two talking heads, talking with you about the services, strategies, and success tips you need to make it big and managed services. My name is Rich Freeman. I’m Chief Analyst Channel Master, the organization responsible for this show.

I am joined by your other co-host, Erick Simpson. He’s our chief strategist at Channel mastered. Erick, how goes it?

Erick: It’s going pretty good. Rich. We’re almost through Q2, if you can believe that. Memorial day’s right around the corner as we record this. So wishing everybody a nice memorial day in advance.

By the time you guys see this, it’ll be behind us. But yeah, [00:01:00] looking forward to taking that extra day off this week coming up. How are you doing? Yeah,

Rich: yeah. Holiday weekend coming up and our audience will never know the difference, but I actually I’m taking the whole week off. We will still have an episode of the show for you.

A week from now, we’ll just have prerecorded or possibly actually we’ll record it and and edit it very quickly after my vacation. But yeah you’ll never know the difference folks. But I will know the difference ’cause looking forward to a break.

Erick: Yeah. And you’ve earned it, rich.

Alright, so this being our last podcast, before your break, why don’t you lead us into this week’s story of the week?

Rich: Yeah you know what? This is interesting, Erick. A few weeks back we were both at the Kaseya Connect Conference. During that conference Kaseya announced the latest version of Kaseya 365.

The ops skew, if you will, of Kaseya 365. On the exact same day, synchro announced their synchro XMM product. And that Friday I published a post [00:02:00] on channel holic, my blog that was titled Death to Scenes. And the idea basically was everybody likes to talk about seamless integrations, but the reality is integrations have scenes and what ca the commonality between what Kaseya and Synchro were announcing that week is that they really believe that what MSPs need and increasingly will want.

Is a single platform source code integrated solution that does most or all of what they need to do. And gosh, Erick, it, it maybe within two hours of that post going out, I was contacted by Enable, and they said, we’ve got a different point of view on this. We would love to talk with you about that.

And I thought to myself if if I’m going to give them a chance to dissent from what I heard from Kaseya and Synchro and wrote about it, I should probably give Ninja one a chance to chime into. And so this week I interviewed Mike Adler, who’s the Chief Product and Technology Officer at Enable, and, peter Breton, who’s [00:03:00] been on the podcast before, he is their vice president of product strategy at Ninja One, and I got their perspective on the matter, which is similar to between those two companies, but a little bit different than what Kaseya and synchro are saying. So first of all, I think both Enable and Ninja One would agree if you are a younger, smaller less mature MSP, getting everything in one package probably makes sense for you because it’s just easier to deploy and to manage and to get started with that way.

As you get larger and as you grow more mature though you may outgrow the capabilities of that platform, or you are likely in fact to develop some requirements, some customer specific requirements that these platforms can accommodate. And it might be an industry specific thing and some regulatory obligation there.

There just might be one or two or three and a growing number of functions that only. A third party product is [00:04:00] gonna be able to deal with. And that’s when you really need to be working with a vendor who fundamentally believes in making those integrations as seamless as they can be. Now the trade-off so there’s a trade-off basically between these all in platforms and and, the more best of breed kind of approach and the best of breed approach obviously is it’s most flexible.

It allows an MSP to handle, to accommodate more of these customer specific requirements and build the stack maybe that they’re they’re most comfortable with. The thing about a platform is that it, it’s very cost effective. It’s very efficient. It creates room for higher profits.

But at the same time, it’s difficult for any one vendor to be really good at everything. If you as an MSP are doing 10, 20, 30 different things, it’s gonna be hard to find one product from one vendor that is really gonna meet your needs in, in all of those categories and keep up [00:05:00] with your evolving requirements in all those categories.

And so this is another area where that that ability to integrate outside solutions is going to be key. And there’s a similar related issue that came up particularly in the conversation with Enable that I think is a good kind of illustration of what both these companies was trying to tell me.

And this was in the area of ai. So you think about AI and that is an a technology a corner of the industry that gets turned upside down more or less every week. There’s something new and exciting out there that you as an MSP might potentially wanna take advantage of, that your customers might potentially wanna take.

Advantage of coming up all the time. And so the ability to go out to a third party vendor that does this new AI thing, whatever it is, and integrate that product with your stack is gonna help you meet that customer demand, that end user demand faster than if you have to wait [00:06:00] for your one big platform vendor.

To work that into their product. And then the benjas and enables of the world will say, even when they do, they’re gonna have trouble being as good at that thing as the third party product. So I’m not here to endorse one point of view over the other, but I do want to complicate the picture a little bit and get your perspective on it as well.

Again, there are scenarios, there are MSPs who really are gonna be better off going with that all-inclusive platform and collecting the efficiencies and the profitability, productivity advantages that come with that. And then there are legitimate reasons why that might actually be counterproductive for some of the folks in our audience.

Erick: Yes, rich, and you used the word talk about how complicated the discussion is and I think it, it’s definitely accurate. To argue that there are different types of MSPs, right? I [00:07:00] think the solution that you ultimately land on or, plant your flag on dictates what your outcome, what your future desired outcome is.

For instance, I could argue Rich if I wanted to, get back into the fray and start another MSP practice with just the sole objective of building enough equity value as quickly as possible in order to exit then maybe a all encompassing platform might do the trick for me because I can cookie cutter and just basically focus on a specific type of business, right?

So I might be. Restricting myself to, a typical type of business or size of business or whatnot. Now the other option would be, oh, I start with that, and then I start adding in like additional third party services or solutions [00:08:00] that integrate with my basic platform. And that allows me to become a little bit more verticalized, a little bit more value add strategic to different types of clients that way or, we’ve talked about on the show before as well.

I want to do a, just a pure best of breed. Like I want to pick and choose from all of what’s on the, in the buffet line and create my own meal, which doesn’t really you know, I, so I sacrifice a few things if I do it that way, right? Some of that is integration, some of that is reporting and things like that.

So I think in each one of these different options that I’ve just laid out, there are benefits, but there are also some sacrifices that you make. Now what you just talked about was, if I’m reading the tea leaves, is if there is a way maybe you didn’t say this, maybe I’m just, imagining like how, what’s the best of all worlds, like no matter where I am, where I start and how I [00:09:00] mature and how I ultimately grow, is there an overarch?

’cause the key thing here is this what the promise of AI brings us and how can we access it in the most effective and fast and possible. Is there a vendor rich that has an AI platform that can overlay and integrate into any best of breed product, service, solution, or platform? Is that the holy grail?

Because if it is, when that’s solved, then you know, we’ve solved. The the challenge of the MSP being at whatever level of maturity or growth or strategic value that they wanna deliver, no matter what they’re using.

Rich: Possibly that would solve the AI piece of the equation.

But you would still have say the security piece of the equation. It, you still run into that situation where, you know and this is where you were talking about Erick, there is no perfect answer. There are trade-offs [00:10:00] either way. And what the MSP has to decide is am I okay with.

Enjoying the efficiencies of the platform, but giving up some of that, time to market advantage and specialized functionality advantage, or am I do I need to, am I willing to give up those efficiencies that come with the platform so that I can plug in, what I feel like is the product that my customer needs.

And yeah. And that I think is what both Enable and n Ninja would tell you they’re trying to do is to allow you to use as much of their platform as you wish to without sacrificing too much efficiency when you integrate something in from somebody else.

Erick: Yeah, I think that’s the, the ultimate goal for any platform is to allow interconnectivity.

No matter what you’re using. And so maybe, maybe I’m chasing a unicorn here with this kind of AI idea bit, but I’m thinking, [00:11:00] rich, maybe that fills the gaps and allows you to plug in your security and use how much of whatever platform you want, but then allow you to integrate with other third parties in a much more maybe efficient way or a different way, or not waiting for the platform vendor themselves to evolve that way.

And I’m thinking specifically from a reporting perspective, from a single pane of glass perspective and overlay that allows that fills in that gap. I know that, today as we speak, in 2025 we’ve got this movement from these large platform vendors to either build a.

As much functionality within their own platform. They’ll build it, buy it, acquire it, integrate it or create API connections into it to provide that single pane of glass experience. I think what you’re talking about is the AI readiness and leading, lead industry leading focus of in, of building [00:12:00] the AI component of these platforms to enable the partners to do potentially what they can’t with somebody else.

Rich: Yeah, I mean that AI story is a good example of what, from a certain point of view could be a disadvantage or a handicap that comes with going with an inclusive, all-inclusive kind of platform. Let’s move on to your tip of the week, Erick. ’cause it actually relates to what we were just talking about here in terms of unified service stacks and potentially migrating from whatever you’re using today to something else.

Erick: Yeah. Uncanny how that happened today rich very timely. So yeah, what I’m talking about here and it’s just the idea of consolidating and migrating either legacy stacks or legacy clients into your new stack or bundle of services. I was just hosting a webinar this morning and we were talking about, how to have the cybersecurity [00:13:00] conversation with reticent or resistant customers, if you will. And, I poll the audience and I ask them how many of you have reticent customers that don’t wanna move forward? And a lot of them, I, if they were being true, I think everybody would raise their hand that, that they don’t want to either move into your next required set of services or portfolio of services because they just, either from a budget perspective, from a strategic perspective, from a, just a, a CD or F customer perspective, they’re just not going to move forward that way.

So how do you get all of your clients migrated over into kind of maybe a re-imagining of your stack? I know a lot of partners, rich, that are still using legacy. Solutions, from 15 years ago or 20 years ago for a specific component of their stack and think, oh, I’m just gonna continue using that even though I’m adding new things that are meeting the needs of today’s business [00:14:00] by owners and mitigating risk and things like that.

From a cyber security perspective, I’m thinking if you reimagined your entire stack today from scratch, if you were starting from scratch, a blank whiteboard, what would that stack be today? I’ll bet you, rich, it would be pretty different. Maybe not. In all respects, there would be significant differences in what’s in that stack.

So if you’re gonna reimagine your stack today, or you’re going to standardize and require all of your clients to subscribe to this stack, that means that you have to upgrade existing legacy clients. Your new agreement, your scope of work needs to include. Your agreement needs to include this package that every new client must subscribe to in order to be your client.

So how do we do that? Just a couple of thoughts. Couple of quick thoughts. Number one, I’ve said this several times maybe on this program as well, but if you’ve done everything that you can possibly [00:15:00] do Rich, to get a client to move in the direction you want them to move in, and they just are not going to move in that direction, then, but you have other clients that are, so let’s say your a and B clients, most of ’em take your new subscription.

Then you’ve got these CD and F customers that just are not gonna budge for whatever reason, right? There may be all kinds of reasons for that. First of all, the revenue that you will realize from getting everybody trued up that will true up into that new. Service relationship with you that makes it easier for your team to deliver service more efficiently.

You’re sleeping better at night. Your clients are sleeping there at night because you’ve got best, you’ve got the upgraded stack and security that as part of it, that revenue will make it less painful for you to exit those other customers that become risks for you now, and not just from a cybersecurity perspective, rich, but from a service efficiency and profitability perspective as well.

Because if they’re not using [00:16:00] the new stuff, then you’re still having to keep up managing these old platforms and these old services and solutions that way. Tip one, once you get to that point, you are then going to message your reticent customers that they need to make a decision whether you have a good, better, best or just this upgrade package to accept that by this date in the future.

If not, you will automatically upgrade them to that new package of services and they will see that reflected on their next invoice. And the small percentage of folks that, complain, that’s your first indicator of who’s not a good fit for you in growing your business moving forward. And I’m not saying you give them a pass or you give them break.

You may work with them a little bit, give them a little bit of a discount maybe to get them to say to, to stay happy and if you want to keep them. But again, they are still preventing you from [00:17:00] spending more time with your a and b clients and bringing on new A clients and B clients because you’ve got a capacity math problem that you’re trying to solve with your technicians and engineers as well.

And then lastly, I guess the tip would be once you do. Everybody that’s going to move forward into this new package of services, this new service relationship with you if you will then demonstrate the value and benefit of them making that decision in your next strategic meeting with them and showing them how much better you’re delivering service against your SLA.

You’re increasing your csat, you are responding to tickets in a much more timely fashion or things like that, right? And able to spend more time with them strategically, rather on, on business growth outcomes rather than on technical outcomes.

Rich: I so often it seemed to go to the three S’s, the three S’s that characterize the most successful, fastest growing MSPs in the industry.

[00:18:00] And it’s scale, building scale into workflows and processes. It’s specialization by solution type or vertical industry, and it’s standardization. That’s the third s And so really a lot of what you’re talking about here is just acknowledging that to the degree you are not standardized in terms of your stack and you don’t have your customers all on pretty much the same stack.

You are holding yourself back from. Possible growth. You’re slowing down the growth of your business. And so I do endorse, what you’re talking about here in terms of addressing the issue. I think you make great suggestions in terms of how to do that in a way that doesn’t just blindside a customer and gives them, and use some room in there to think about the future of your relationship together.

But I especially like the last piece of that, to the degree that I’m underscoring standardization, something that is important for you as an MSP when you standardize and you collect the benefits of [00:19:00] standardization that helps you out. It doesn’t necessarily help out the customer.

And I do making sure that you call out to the customer that yes, you had to make some changes, but this benefited you as well as me. Here are some things that I’m doing better for you now because we made this change that I told you’ve gotta do by June 1st, or else a thing.

I think that’s a really important part of the process you’re describing.

Erick: Yeah. And I think that there’s also the opportunity rich potentially because you’re consolidating and you’re making different decisions and maybe some of the solutions in your stack weren’t quite as MSP friendly from a pricing and a consumption basis as what you’re moving towards now there may be and considering the efficiencies that you’re going to gain by making this transition.

There may be some cost savings that you might be able to telegraph to [00:20:00] some of your clients. Wouldn’t it be great to say, Hey, normally to make this migration we would have to charge you this much, but because we’re expecting such better efficiency and we’re reducing some of our internal costs and things like that increase is only gonna be this much, right?

So again, just looking at all the different levers and techniques that we can bring to bear to get clients to do the right thing and follow our guidance. And of course, understand that nothing’s free and the cost of doing business is going up. They’re raising their rates to their customers, that, the economy is impacting them as well.

And I think having that strategic conversation and building that much closer business outcome-based relationship, that true business partner relationship. Allows you to be candid about these things with clients in ways that we might not have felt comfortable being before because we are really all in this together.

Kind of reminds me of, not the same thing, but just that sense that we’re all in this together when we were all trying to get through the pandemic [00:21:00] together and working together with our clients and going the extra mile. Heck, I know there are a lot of MSPs out there, rich, that really laid out for their clients, right?

And they benefited from that because, the clients appreciated that. And here’s just another way to say, Hey we’re in this together. Let’s continue growing our relationship strategically and doing what’s right for each other so that we can continue allowing you to, grow your company and align our recommendations to help fuel that growth.

Rich: Folks, we’re gonna stay on theme here for this episode. We’ve been talking about managed services, tools and tool stacks. We’ve been talking a little bit about the role of ai in the world of managed services. We’re gonna keep doing that. And we’re gonna do it specifically in the context of what is arguably the most ambitious AI tool for MSPs that I am aware of right now.

It’s a tool called Autopilot. It is a fully autonomous agentic capability for the RMM solution from a terra. As we record this, it just came out officially. General availability [00:22:00] this week. And when Erick and I come back from the break, we’re gonna be joined by Joab Sues, he’s the general manager at a Terra for the US.

He’s gonna tell us all about autopilot, what it is, what it does the results that they’ve been seeing from this thing in in testing. And we’ll just give you an opportunity to think about the degree to which you already, you are comfortable turning over. Ticket triage resolution start to finish without any technician involvement whatsoever.

That is all coming your way right after the break.

And welcome back to part two of this episode of the MSP Chat podcast or spotlight interview segment where we are very pleased. Be joined by Yoav Suze. He’s the general manager for the US at Atera. Yo, welcome to the show. Pleased to be here, rich. Thanks so much for having me. I was telling you off the air when the idea [00:23:00] first came to me, we should have somebody from AERA on the show.

I, I reached out to folks and we worked on timing. We came up with with a, an episode of the show and a week for the conversation that turned out to be a good fit. And then lo and behold, that, that kismet that timing turned out to be absolutely perfect because this is the week, if I’m not mistaken, that you folks are introducing a very interesting new technology called Autopilot.

We’re gonna talk a lot about that and about what it says about agentic AI and managed services and where that’s headed. But before we do that, before we get started, tell folks a little bit about yourself and a little bit about Atero. Sure.

Yoav: So one, rich I thought you had this all planned out.

I don’t, I didn’t think that this was by chance. I thought this was all planned out, but in any case my, my name is Jo, originally born and raised in Tel Aviv, Israel spent 10 or so years in London. I’ve been living in New York for the last 11 years. Really joined a terror to lead the US operation about a year and a half ago.

Terra’s always operated in the [00:24:00] us but today we, we have about 60 to 70% of our customer base. Of our 13,000 customers are here in the. A year and a half ago we decided we needed boots on the ground. And as we got market and work with bigger and bigger companies, we needed to have a presence here.

And we’re about 15 people here in the New York office growing pretty rapidly. For those not familiar with the Terra, I would say that our claim to fame was with it. We were a all in one IT platform. So we basically provided everything from RMM to PSA to remote access, to network discovery, all in one place.

And that’s what made us very popular with ASPs over the years. But I’d say that our claim to fame over the last couple of years has been everything we’ve been doing in ai. For the last three years, we’ve been fortunate to be partnering with Microsoft and with open ai building kind of different products, which are serving.

Up until now we’ve been basically selling a product called Copilot, which is like a tool for technicians. So think about it as something which allows it technicians to just have a lot. More bandwidth, right? Whether that’s being able to write [00:25:00] power PowerShell scripts using AI or to discover os using AI or summarize tickets or create knowledge base articles, bunch of different things.

That’s been our claim to fame, I’d say over the last year and a half. And what we’re releasing tomorrow to the world officially is autopilot. And as you said, we’re gonna talk a lot more about that, but we could not be more excited.

Rich: Before we move on to autopilot, let’s talk just a little bit more about copilot because that, what that tool does is much more similar conceptually, I think to what ai MSPs and our audience think about when they think about AI in, in managed services, where there is a tool that is augmenting your capabilities.

It is assisting with triage and so on. So I tell folks a little bit about. The functions that copilot performs how your MSP partners tend to build it into the workflow and that the impact it it has had for those MSPs.

Yoav: Yeah, I think that the best way of thinking about copilot is like a suit of armor or like the Ironman suit, if you want to think [00:26:00] about it that way.

It’s basically augmenting the capabilities of the IT technicians and it’s doing that across a different parts of their workflow, right? So if you think about it, you can use co-pilot today in order to ask you any question about what’s happening in your environment, right? So if you’re an MSP, you can open up co-pilot inside of Terra and say, Hey.

How many tickets do I have outstanding today? And can you segment that by customer for me? Or once you go into a ticket, there might be a long back and forth between an end user and a technician. You can hit, generate summary and it’s gonna give you a quick summary of everything that’s being done all over the back and forth between the end user and the technician.

It can also generate a response for you. You can use copilot today in order to write scripts, right? Be it PowerShell or other scripts. You can use copilot today in order to do a smart search. So if you think about, an MSP, which has tens of thousands of endpoints being managed, using an Apple language, I can say, Hey, I wanna see all of the devices that I have, which are on Windows 10 and are not eligible to up, to be upgraded to Windows 11 and are unpatched.

And literally, by [00:27:00] typing that in, you can get a list of those exact devices and then run a script on those specific ones or others, right? So it’s really a powerful tool for the IT technicians. And what we’ve seen now is that it’s. Increasing their throughput, right? So whether it’s the amount of work that they can do in a day, or how many tickets they can field, or how they can get answers about what they’re doing.

But what we’re seeing there is a massive impact, whether it’s on knowledge management, on scripting, whether it’s on kind of general information reporting. The impact that we’re seeing there is pretty massive, right? We’re seeing a, three to five XI would say, output on the product technician level.

That’s what we’re hearing, from our partners.

Erick: You have, as a former MSPI can really appreciate the value and the benefits of copilot. We strive to be more efficient and effective and the technicians really appreciate the feeling of accomplishment when they can actually execute efficiently and effectively and solve issues for their [00:28:00] end user customers.

So I. The big news though, is autopilot. So how is Autopilot unique and different than copilot and how do they work together?

Yoav: So I think I know that people tend to use the word AI a lot, and they use the word revolutionary a lot. But in my opinion, autopilot is truly changing the way that end users are interacting with it.

I think if you look at, the last 40 or 50 years of the way that it has operated, an end user has an issue, they reach out to it, it opens a ticket, or the user opens a ticket that get, puts in a, that gets put in a queue, that ticket gets picked up, somebody deals with that ticket. And that’s just the way that it’s been, right?

And we’ve all operated on how can we make it technicians more efficient? How can we make the teams more efficient? How can you better cues, better automations? But I think what we’ve done at the Terra is we’ve really tried to flip that on its head and what we’re doing today.

We’ve basically created a pro an [00:29:00] application or a product, which is serving the end users first and foremost, right? So if you think about the way that it works, and I wanna make sure that I’m painting the pictures vividly as possible. Imagine, Erick, that you have anm, SP and you have a customer, and that customer has an end user and that end user wants to reset their password because they forgot the password that they set for their outlook.

So what they would do today is they would behave in the exact same way that they have up until now. So if they’re used to communicating over email, they can use email, or if they’re used to going to some sort of service portal, great, whatever they want, any method that they want, they can go ahead and reach out to the IT team.

And what Autopilot is doing is it’s able to, one, autonomously solve between 20 to 40% of the issues typically faced by an IT team. And these are the tier one issues. So these are the things like password resets or software installations, as our partners told us, anything that requires somebody to click [00:30:00] next next.

That’s those are the perfect opportunities. Or somebody that wants to install a new printer, or somebody that wants to set their system clock, or somebody that wants to add a language pack or somebody that wants to, map a network drive. There are hundreds of actions that autopilot can today tape.

So basically the end user reaches out in the same way that they always have up until now that conversation gets intercepted by autopilot, which goes back and forth with the user. So there’s decisions to be made and we all know how users behave. Yeah. You need to be able to understand natural language.

You need to be able to ask your questions in the same way that a technician does when somebody says, Hey, I can’t watch videos. What does that mean? What type of videos? Is it a specific type of format? So what autopilot’s able to say, rich, I understand you’re having a problem watching videos. Is there a specific type of video?

Rich come back, comes back and says, I’m not managing to watch a VI videos. In that situation, what autopilot then needs to do is to look at Richard’s computer because maybe he does have a program which can play in a [00:31:00] DI, and if he doesn’t, then it needs to look at the white listed software as that specific business.

What are the bi, what are the, applications that they’re allowed to install? And if it identifies an application, which can solve the problem, that’s gonna silently install it for that end user. That’s one example. There’s hundreds of examples of autopilot basically solving. A user issue from beginning to end.

And this is everywhere from, technological problems like the one that we just talked about to knowledge problems, right? And we all know that very often it is the catchall for any issue that somebody might have. They’re not, they don’t remember the wifi password or they don’t remember what is the protocol for closing down the offset.

The end of the day, it doesn’t matter, right? An autopilot today is able to be that kind of first line of response, solve problems to end from end to end, and it can also assist technicians. So the second thing that autopilot is doing is if it’s not able to solve the problem, whether it’s for, a technological reason or whether it’s for a permissions reason, right?

Imagine that, somebody comes on and says, Hey, please add me [00:32:00] to the distribution list for the management of the company. You might, even if Autopilot can do that, and it can do that, you wouldn’t want it to do that. So what autopilot is able to do is to go back and forth with a user to follow your protocol and say, what are you guys doing at your business?

What’s the information that you want to collect? And then serve that up to the technician so that they, as my colleague said this morning, it’s like starting on the 50 yard line, right? Instead of starting at zero when the technician shows up. A lot of the work has already been done with that end user.

Erick: I just have a follow on thought Joa, because as a former MSP, like I said, one of the big challenges is the, are those quick, easy, ticket closures, which, autopilot is focused on. But the other component that strikes me as super valuable is the ability to interact with that user when the user is available.

Because as we all know. Having tickets that are open and sitting in the queue, and then by the time the technician gets to it, then the user’s not available [00:33:00] or this or that. So this is an additional value that immediately strikes me as important to a service desk.

Yoav: That, that’s really interesting that you mentioned that because I think that I was struck by the fact that for many of our users, the autopilot assistance was even more valuable than the resolution.

It’s exactly what you are talking about. Meaning, a user and every IT person I never met in my life has heard users complain about slow computers. And I think one of the issues with a slow computer is I might open a ticket and Erick, by the time that you and I connect, it could have been hours or days and the problem’s not reproducing.

And I get frustrated and you are frustrated. So what’s interesting about Autopilot, for example, is if I come in and say, Hey, my computer’s running slowly, I. Immediately we’re gonna leverage the Atera agent on the device, right? And we are going to run a health check immediately, when that user reaches out.

Now, if we can solve the problem with autopilot, great. Maybe it’s just a question of clearing the cache or, restarting the computer. Because even though [00:34:00] the users always say that they restarted it recently, the uptime’s 75 years or whatever, it’s right. But the fact that, I tell you, my computer’s running slowly, it’s running those health checks and if I haven’t been able to solve it, when you pick up that ticket, Erick, you are now going to see, all of the diagnostics, which were done in real time.

And, I’ve just brought it closer to resolution so that when you pick it up, it’s a much, much quicker or less work for you.

Rich: And quick note about that from prior interviews, prior conversations I’ve had with ET Tara and with with your boss, Gil Eckelman the CEO over there, you talk about the value of that, the, the 24 7 immediate access to assistance on your schedule.

My understanding is there are MSPs in the test group for autopilot that have actually been charging more for that service. They, they’re, they’ve raised their rates for the customers who have opted in to the service because they’re getting something that they wouldn’t be getting otherwise.

Yoav: Absolutely. [00:35:00] And we see that, we also see that it’s reducing the cost for that, for a lot of the tier one issues. It also means that they can have more customers per technician. Because if a lot of these things are getting fielded and, typically, we see. Anywhere to be conservative.

I’d say between 40 to 60% of tickets that MSPs are dealing with are relatively simple, rudimentary things. And if we can take a large portion of that off their plate, that just gives the MSPA lot more bandwidth and continuous coverage. And you hear a lot about these question, what, we were just talking to an M MSS P this morning that provides 24 6 support.

’cause they don’t wanna do seven days a week. Because it’s not as cost effective for them. So just being able to extend that. And I think that what’s interesting to understand about autopilot is it’s not a one size fits all genErick model. Even as an MSS p autopilot is gonna behave differently for each and every one of your customers, right?

Because the knowledge base, we work with an MSP who’s in the world of healthcare and they’ve got dozens of customers which are everywhere from nursing homes to other [00:36:00] medical facilities and practitioners. And each one of them sometimes has a little bit of a different knowledge base, right?

And they want to get different answers and we can do that, right? And it’s what’s interesting about autopilot is that it’s. Constantly learning, right? So it’s learning how we’re closing tickets, it’s learning what kind of scripts we’re roaming, it’s learning how we’re resolving issues, and that’s teaching autopilot to behave in the same way in the future.

I’ll give you one example, which we found really surprising. One of our customers had been installing, sorry, one of the MSPs we’re working with, which is our customer had been installing language packs, right? For the end users of one of their customers, right? So they had basically, people wanted to add French, Italian, Hebrew, whatever it is, right to their computers.

And they were constantly finding time to work with the end user, doing a remote session with them, clicking around, finding how to install the language pack on Windows, installing the language pack. And that was that. And what Autopilot identified was that they saw this, it saw the behavior repeating itself again and again.

So it suggested the script [00:37:00] and it said, Hey, here’s the script, which installed the language pack of your choice out of a, a certain set of languages. Then it asks the MSP, do you want to add this script to autopilot? Do you want to add this as something that if the end user comes in and asks for it, we can execute autonomously?

And that’s really powerful because it’s learning to behave in the same way that you are behaving.

Rich: That is super interesting actually. I hadn’t heard about that particular dimension of what it does. I don’t know how familiar you are with this job, but before we get even deeper into autopilot, I’d love to go back and just discuss the backstory a little bit because the roots you’re rolling it out.

It’s going into general availability this week, but the roots of this product go back multiple years to before the official introduction of chat, GPT. J Jock a little bit about what, how did this begin? What did you discover a couple of years ago that got you thinking, wow, we could actually use this open AI stuff to [00:38:00] resolve?

Yoav: So actually, several years ago, Terra had, a first run of the world of AI independently, right? Trying to build model in house, where originally the idea was to find a way to auto tag tickets, right? To classify tickets as they came in, tag them on the standard of the hardware. Is it software?

What kind of issue is being faced with a ticket. And that was a lot of work was done there, but it didn’t really, reach the impact or quality I think that, we wanted. And then Microsoft approached to Terra, we’d been partnered with them for a few years, but they approached to Terra probably three and a half years ago at this point, approached Terra and said, Hey, we’re working this company called Open AI and they’re building these large language models.

Would this be something that you guys would be interested in trying? And, you know the answer was yes. Also, Austria, our CTO started working with the Microsoft team, and this is probably two and a half years ago at this point, right? So this is prior to the public release of charge GPT.

And he saw how much potential was in there and if I remember correctly, but I’m not a thousand percent potential. I think the [00:39:00] first thing that we actually did with chat GPT was the AI based script generator, right? That ability to come in and say, Hey, I want a script which clears the cache, deletes temporary files and resets the computer, right?

And they restart the computer. And basically that I think was the first iteration of what we did with chat GPT. And what’s interesting is as time has gone on, we’re benefiting from the fact that these models are getting better, right? Microsoft Open AI models are getting better all the time. And our domain expertise is becoming deeper and the capabilities that we are giving it, right?

Because I think what’s interesting about, we get asked a lot. How do you differentiate between all these ais and Salesforce is talking about AI and Zendesk is talking about ai, and everybody’s talking about ai. And I think that what we often tell people is the big difference with Terra’s autopilot is that it’s not just responding to tickets, it’s resolving them, right?

And those actions, that ability to really take action on devices, right? To run those health checks, to install the software, to map a [00:40:00] network, drive to clear a tool or whatever it is, those are things which we’re developing. So we are getting a double headwind. We are continuing to build, we’re also, the models are getting better and better, right?

Today we’re looking, if I remember correctly, we’re now up to Chad 4.1, right? And where it’s constantly, the reasoning is getting smarter, the decision making is getting smarter, right? Because there’s a lot of back and forth which happens with end users. And the models need to decide, one. Needs to understand what the end user is actually asking for.

Then it needs to look at all of the selection of tools and things that it has at its disposal. Choose the right tool, the right agent, take action, solve it. So it’s constantly getting smarter, but it’s it’s definitely where we’re benefiting both from the advancements and the general kind of models the chat g PT are building, as well as the specific things that we are doing in house.

Erick: As MSPs are very risk averse, so turning over, components of their service delivery to AI makes them understandably nervous. [00:41:00] Absolutely. So what safeguards or guardrails have you put into place to ensure AI doesn’t do something unexpected or imaginary as the kind Yeah.

That we’ve seen?

Yoav: Yeah, that’s a great question. And, it’s definitely something that we’re talking to our customers a lot about. I think that there’s a few things. One, the way that autopilot is built is that it only has a certain set of actions that it can take, right? So it doesn’t have the option, it doesn’t have access to take actions, which are outside of it, of its realm of possibilities.

That’s the first thing, which is really important to know. The second thing to know is that with autopilot, you cannot take an action for somebody else. So it’s not that I can come in and say, Hey, I’d like to reset Erick’s password. It doesn’t work that way. You can only take actions for yourself inside autopilot.

So there are multiple safeguards in place. We’re working with Microsoft on ethical AI as well, and all of the guardrail that they’ve put in place. So from that perspective, I think that we’re [00:42:00] very much in line and ahead of the market. There. Also, keep in mind that we’re working today with heavily regulated industries, right?

So we’re working with, MSTs and ITDs and the world of healthcare, local government, financial services. So we’re aware of all of this safeguards, all of the all of the accreditation, all of the testing that we need to go through that. That’s one side. But I think, you also touched upon something else, which is there’s a trust issue, right?

And I think that we often remind ourselves that while all of us are probably really excited about AI and have interacted and have tried all the models and all the companies, for a lot of people. Autopilot is the first time that they’re interacting with ai. And we need to be aware of that and cognizant of that.

And it’s, and it was surprising to us in the beginning, I, my assumption was, who wouldn’t want ai? An end user is gonna solve it more quickly. It’s gonna be great. And we went out to meet one of our customers in Florida and we wanted to meet some of the end users. ’cause we constantly talked to the IT folks, right?

IT folks were excited. But when we brought [00:43:00] in one of the end users and we asked her, what do you think? And she goes, I don’t understand why I need this. I much prefer to knock on the door and asked her help. So suddenly we are not really the users, right? We need to understand who are the end users who are gonna be interacting with it, and how do we, how do we introduce autopilot?

Do we need to introduce autopilot ahead of time or not? One of the small changes we made is that you can now name it whatever you want. It doesn’t have to be called autopilot, right? You can call it. Rich or Erick or whatever you want, you can give it a name, you can give it, its own identity.

And we’ve also we think a lot about how do we educate the end users? How do we help our partners, educate their end users, teach them what it can do, what it can’t do? Explain that it’s not replacing the IT team that they’ll always have the option to talk to a human being. But that’s a big part of it and we need to remember that, there’s an adoption curve of technology.

And I would say that, we’re still, in my opinion, in the early adopter phase of, people who are very excited about this. And I think we’re starting to see maybe the early majority come in play into play of [00:44:00] people who can already understand the business case. And the business case is pretty clear with the amount of hours saved and, the downtime prevented and the opportunity costs that people have.

So there’s a lot of potential there. But I think that the market education is key. And I think that it’s exciting because in my opinion, the world of remote RMM and PSA is becoming. Pretty commoditized pretty quickly. And I think that today, the big difference is gonna be an intelligence, who can do this better, smarter and I believe that’s what we’re trying to do.

Rich: I have been tracking autopilot for a good year and a half or so at this point. And a year and a half ago it was still, it was either in or just coming out of the lab where you guys had been testing it against simulated tickets and looking at the success rate and so on.

But it’s been in real world testing both with MSPs and with corporate IT departments for a while now. It may be two different variations of the same question. In terms of those anxieties Erick was talking about, have [00:45:00] you in testing. Seeing any kind of hallucination or misstep by autopilot that you’ve learned from and corrected, and then maybe on the other side, have you seen it do things that you didn’t know it could do?

Has it surprised to the upside in some favorable kind of way? Become a so aware

Yoav: it’s joined the team? We’ve given it a laptop, but no I think I think what we saw with autopilot are the first part of your question, rich. Autopilots are very eager to please. And we hear that a lot from our customers as well.

We sometimes hear from end users. It’s so nice. It’s so funny. It’s so kind, but because it’s eager to please, we found that there was some false positive in the mc king hang, right? Where you would say something like, can you please reset my password to our outlook? And it would say, absolutely. I and then you would ask it, have you reset the past on, you would say yes.

And it hadn’t, right? So we did come across a few of these false positives, which is probably better than the alternative, right? Where it’s taking actions without us knowing about it. But false positives is definitely something we saw in [00:46:00] the beginning, right? We also had an example, I remember six, seven months ago where, we had taught autopilot to be able to adjust people’s signatures, right?

So people wanted to update their phone numbers and their email signatures, and we had given it the ability to do and it kept confirming that thing had been done, but it hadn’t actually done so. So false positives, I’d say is definitely something we saw a lot of. I think that what surprised us was how people interacted with it.

And what I mean by that is. One of our customers, when they started working with Autopilot, the number of tickets that they were getting increased. And that was very odd to us. Why would it increase? We would expect it either to stay the same or for it to decrease, but we actually learned that when people knew that there wasn’t a human being on the other end that was gonna judge them for their questions, they were willing to ask things that maybe they were a little bit ashamed of asking before, had I changed the font on Google Slides or whatever it is, right?

So we saw that and we saw, for example, that a lot of people were update, were uploading screenshots. So people were getting stuck in Excel, for example. This is [00:47:00] an example that’s really vivid in my mind. Somebody had an Excel error, they screenshotted the Excel error, they put it into autopilot, asked it what they should do, and Autopilot gave it, gave them the right answer, which was really cool.

And I, I was just really surprised by that. So we’ve seen people use it in ways that we didn’t necessarily expect to begin with. We’ve also seen people use the knowledge base around questions, which aren’t necessarily to do with it, right? People are putting in the, the office protocol for closing the office or when are the federal holidays or whatever it might be.

Because, once you start getting a taste for this kind of ability to interact with something which is giving you the right answer at the right time in an easy way, anywhere you are it’s just self perpetuated.

Erick: So you’ve talked a little bit about the types of tickets that.

Autopilot can handle. You’ll have two questions where, so the first question is, do the end users know that it is not a real person behind it? They know that they’re [00:48:00] interacting with an ai. And number two you mentioned the kind of the tier one type of tickets. So where do you see, the future?

Will it expand to close more and more tickets, maybe level two tickets? So first question do the end users know that it’s an ai? And number two, how far can you extend autopilot?

Yoav: So in terms of, do the users know from architect? Absolutely. If anybody asks auto product, are you a real person? It will tell them that they, it’s not a real person.

Do our partners, the MSPs that we’re working with, are they telling their customers I don’t know. I don’t wanna speak for them. I think you can name it. So for example, I. We’ve seen people call it Billy for some reason, or we’ve seen people give it, like first names of human beings.

So I don’t know. Our, some of our customers may be doing that. We’re definitely not endorsing or otherwise, but I think we, we are very clear. We use our, obviously autopilot for ourselves as well internally and obviously it’s very clear to everyone that, through traction with ai in terms of what’s the future and what else can autopilot do.

[00:49:00] I, I would say that probably is the sky’s the limit. I also would say that the definition of what a tier one issue is changes right. From business to business. I don’t think that there’s one agreed upon definition for what constitutes a tier one issue. The way that I look at it, there are, in the vast majority of businesses that we’re working with. Tier one is the simple recurring, you follow a checklist kind of thing. And then the tier two are things that require more judgment. So I’ll tell you this, I think that the next step for autopilot is more and more actions on the local device, and then the next step is interacting with other as systems.

And we’re doing that already today. And I’ll give you an example. We work with a bunch of email quarantining tools. There’s all of these if something’s a suspicious email, it gets quarantined in a service and then the user gets notified. So what we’re doing there is we’re basically craving integration where an end user can come in and, rich, you can come in and say, Hey, I’m waiting for an email from Erick, which hasn’t arrived.

Then Autopilot can go and check whether it’s an Ion scales or it’s one of the other systems which are doing email [00:50:00] quarantining and stuff like that. They can check, it can find whether that email exists, and if it does, it can release it, it can release it back into the inbox of the user.

Or think about, different identity providers, right? Integrating with an Okta or integrating with any of these kind of systems. So I’d say the next step broader pilot is building more and more integrations into the ecosystem, which isn’t just cloud operations and taking action on the devices that themselves, that actually being able to interact with the broader ecosystem of technologies, which are being used by, the MSP or any business.

Rich: In any major decision. An MSP, makes their, they’re factoring in risk and reward. And we’ve spoken a little bit about the risks and all of the measures you folks have in place to minimize, if not eliminate those risks. We’ve touched on some of the rewards, but as you think about the MSPs involved in testing autopilot to this point in time, what kind of impact has it made on, the the pace of hiring and payroll [00:51:00] growth?

Like the ability to even raise prices. I was talking about what kind of business impact has autopilot been having for MSPs.

Yoav: So we shared in a few places. I think the first thing that we see is that it’s able to solve. In the, let’s say, 20 to 40% of the time that was spent on tickets prior.

And the reason that I differentiate here is there are some tickets that are very quick to, you solve ’em very quickly. There are some tickets, which take more time, but today we’re seeing it able to eliminate after 40% of the time that was spent on tickets prior. So I’d say that’s the that’s the very basis of kind of what autopilot is trying to do.

And then the way that it gets translated by our partners, one, we’ve just seen an increase in ratio, right? First and foremost, we’re seeing that our partners can have more customers, they can have more endpoints with the same team that they have today, right? So they don’t need to grow the team, they don’t need to hire additional people.

And the second thing, and this is something which is a little bit less tangible, but I would say no less [00:52:00] important, is the fact that you can take the team that had been focused today. On taking care of a lot of these tier one issues and redirect their focus into more strategic initiatives, for your customers, right?

And those can be anywhere from hardware issues to software migrations, cloud migrations, whatever it might be. And that’s where we are seeing a lot of value, right? We’re basically saying, Hey, we can take a lot of these rudimentary things, which you’ve been spending time and money and attention on, right?

We know what it’s like for MSTs to hire and to retain talent. We think about the benefit here, not just for the owners of the MSTs, right? We also look at the benefit for the people who are working at these MSPs. Nobody got into it to rest, to reset people’s passwords or to, install software.

So if we can give a tech, a tier one technician, if we can give them superpowers through copilot, and we can solve issues for the end users through autopilot, and they can now, uplevel their skills. So they’re working on more interesting things. I. We’ve seen that there’s [00:53:00] actually a benefit in terms of the employee happiness in terms of the team, which are, working at these MSPs because they feel that what they’re doing is getting more and more meaning.

It’s more and more interesting. It’s less repetitive, less mundane, less boring. So we’re seeing that. And then there’s the pricing side, as you mentioned, right? So with a few of our partners, we know that they have, charged their customers directly, right? So we see it in two places, either charging their customers directly saying, Hey, I’m gonna give you 24 7, 365 days support, and it’s gonna cost you x amount of dollars per end user.

That’s one option, right? And the other option is to basically say, I’m just gonna reduce my cost basis, right? I’m, as an Ms P, I’m gonna reduce my cost basis. And the best ones are doing both, I would say, right? They’re both able to charge their customers a higher premium for their services because of all of the things that you’ve mentioned and, at the same time, make their own business a lot more efficient by having more endpoint, more customers with a smaller team.

Erick: Yeah, this is the promise of agentic ai with [00:54:00] autopilot and then copilot for the end users and then copilot for the internal IT teams. You’re collecting a tremendous amount of data. So my question to you is, how do you see leveraging all of this data in a way that allows the person responsible for managing service delivery or the service department or the MSP business owner to be more strategic?

In, and I’ll give you a couple of examples. So one sales opportunities, like we have a user that is opening a tremendous amount of tickets because their, machine is just a piece of junk, or there’s some. Trend that isn’t really readily visible, that’s happening over time in the environment, that now the data can surface and say, Hey, there could be an issue over here from that perspective.

Second scenario would be, how well is my technical team performing? Like, why does [00:55:00] it take Erick twice as long to close these tickets than Rich? For instance, does Erick need a little bit more training? Does he have access to the tools and things like that? And then third, when do I need to hire more technical help to address the volume of tickets as I grow?

Three specific examples. I’d love your feedback.

Yoav: I think you know a few things that kind of came to mind as you were speaking. One, today. Across the terra, we’re seeing over 6 million endpoints. Not all of them are on autopilot quite yet, we’re seeing over 6 million endpoint. So we have a tremendous amount of data, but what’s important to mention is why we do not share data between any of our customers, right?

So there’s only so much that we can learn between customer to customer. ’cause every customer’s data is segregated. This is a big issue and an important issue, right? Training the models and nobody wants, their secret source to be used by their competitor or to be used by another business.

And we’re very careful in terms of how data and insights are housed. I think that a lot of what you are talking about is [00:56:00] also while we’ve insights and being preemptive and really understanding what the data is telling us. And I think, we’re on the way there, but I can’t tell you that today if you switch on autopilot and it’s immediately gonna tell you, Hey, you should hire another two te tier two technicians.

What we are seeing, what we’re able to show with autopilot is we can show trends. We can show, for example, are there specific types of tickets which AI is solving for you more quickly, we can show you how your response times are changing. We can show you which customers are opening, more or less tickets on different types of topics.

I think that the other thing is using, this is more on the copilot side rather than the autopilot side, but exactly what you mentioned, we’ve seen a lot of our MSP partners basically setting up reports which get populated automatically using our AI to tell them when machines, for example, are about to be out of warranty, right?

So if we can pull over API warranty data and we can find which machines are about to be out of warranty and we can make that easily accessible and we can show them, for example, which machines are also not Windows 11 ready. That’s surfacing, those insights for our partners that they can go out to their [00:57:00] customers and talk to ’em about it.

But I think we still have a way to go, in terms of being able to a lot of the things you’re talking about, a lot of the things that we are talking about internally as well about, how we can we be even more insightful, even more preemptive in terms of giving our, our customers information and insights rather than information about where they should be focusing and where there are additional opportunities.

And that’s definitely, I don’t wanna say too much yet, but that’s definitely something that, you know we see as a key differentiator. Generally, as I mentioned before, I think that we believe that the people gonna win, the company that’s gonna win the space is gonna be the most intelligence, and that’s where we’re putting our focus.

Rich: Okay. Very interesting. Tantalizing in there and I’m sure we will be speaking with a terror down the road about where that is going and how it materializes until they just very interesting technology appearing at a very interesting moment in the evolution of AI within managed services.

Yeah. For folks in the audience here who wanna get in touch with you, wanna learn more about, at, wanna learn more about autopilot, where should they [00:58:00] go?

Yoav: One, my email is ya, YOA [email protected]. Feel free to reach out and at.com. If you go on our website from this morning, we have a brand new, beautiful website out there with a really great video all about autopilot, which I highly recommend everyone watch.

I every single IT person that we’ve shown this to it resonated with them. So I would encourage everyone to take a peek and yeah, reach out. We would love to show you what we’re doing and we would love to see if we can help as many MSTs as possible.

Rich: Alright, fantastic. You have Suze General Manager for the US at Terra.

Thank you so much for joining us on the show here. Folks, Erick and I are gonna take a quick break here. When we come back on the other side, we’re going to share some thoughts about autopilot and this very interesting conversation here. We’ll have a little fun wrap up the show. Stick around. We will be right back

and welcome back to part three of this episode of the MSP Chat [00:59:00] podcast. Such a super interesting technology. We’ve talked about AI and managed services multiple times on the show, Erick and what they’re doing at ConnectWise and Kaseya right now. What companies like PIA are doing a Terra is further out there than anyone else in the industry than I’m aware of down the road towards and now officially having.

A, at least 40 ish percent of the time, fully autonomous AI technology out there. Meaning it sees that a service request has come in, it triages, it addresses, it resolves. It closes the ticket, it documents. Start to finish. No human technician participation. I’m not aware of anyone else out there who is doing that or is even talking in a serious way about rolling techno, fully autonomous, truly agentic tech technology like that out.

Anytime soon. I at the beginning of March, I was a guest on [01:00:00] a webinar that a Tara did about autopilot uncompensated. A Tara is not a channel mastered client, but I’m just interested in autopilot and I agreed to be a guest on this this webinar that they were doing. And one of the things that came up was the speed with which AG agentic AI was going to become mainstream within the world of managed services.

And my take on that, basically Erick was. The answer to that question has much less to do with the technology than it does with the MSPs. Already based on what we’re hearing and seeing from from a terra, the technology’s kind of there as far as we can tell The question is the comfort level that MSPs have and trusting tickets to to ai, start to finish.

No, that’s gonna take some time, and I think that’s gonna be what determines how far a gentech AI goes in managed services. How quickly? I do believe [01:01:00] that at some point in time, it is an inevitability that you’re going to see most level one, level two work being done by agent AI and managed services.

What’s gonna d determine how that happens, how quickly it happens? Is more about that that comfort issue the MSPs growing more confident in their ability to entrust AI with the wellbeing of their clients and therefore their business.

Erick: Yeah, it’s very interesting, rich. So a couple of thoughts from the conversation.

Obviously they had very early mover advantage. To be in the room with Microsoft, at the very, before chat GPT and working with Open the folks at Open AI and for the, for that period of time really gave them an advantage. And I think that, from a, from an MSP’s perspective I think that what yo have said about, everybody has a different definition of level one, level two.

Yes, it’s true, right? Because, we know [01:02:00] MSPs that are the, third, level three escalation from their internal IT doing co-managed IT stuff, but. Even if you have a le a third party escalation, you still have a level one, level two, level three within that if you’re operating, based on best practices, right?

So it’s the ability for the AI to learn and adapt. And I agree that the pace with which the adoption will happen will be really based upon the MSPs feeling comfortable that they have put in place, and not just a terror, but anybody, the safeguards that allow them to control what AI can and can’t do, what data it can and can’t access and things like that.

And then ultimately be the, I think the inevitability factor that you just mentioned, I agree with and here’s what I’m thinking now, rich, is will end customers leave their MSPs because their co competitor has a [01:03:00] better gentech AI experience than theirs. Think about that.

Rich: Yeah, that is an interesting question too.

And and you, there are all sorts of different directions you can imagine. A gentech AI going in the context of managed services and it, it’s delivered via chat now with autopilot. Are there like video agents that people are interacting with down, and does the end user prefer that experience or not?

All sorts of interesting questions. I will just say if we are both correct that this is an inevitability, that first mover advantage you were talking about is really important because a terra is multiple years in. On developing fully agentic, fully autonomous technology. There are, a lot of other companies that are doing semi-autonomous kind of technology out there.

And the question is just, how quickly are they gonna be able to ramp up their research and their capabilities and so on? If there is some sort of groundswell that starts to appear the companies like Kaseya and et cetera that are telling [01:04:00] me that we’re not planning to do that anytime soon are saying that based on what their partners are telling them, I don’t want that.

But as that comfort factor kind of grows and all of a sudden MSPs increasingly do want that because they’re being outcompeted by the people who already have it. There are gonna be a lot of vendors out there that are gonna have to maybe if they’re not already developing this in r and d, gonna have to, make up a lot of ground in a hurry.

I.

Erick: Yeah. We always talk about, how long do we think it’s going to take before this is now the common theme. We’re using Ag agentic AI as an MSP, and it is part of our deliverable. And, the interesting insight that you brought about MSP charging more for that experience.

’cause you are giving them a higher SLA response time through this, I used to think, yeah, 18 months or so before we really see, damn, I don’t know that, that time window is shrinking for me. Now, let’s see what happens in 2026.

Rich: If there’s anything I feel dead certain about with respect to AI and [01:05:00] managed services is that everything you think is going to happen is going to happen faster than you think it is going to.

And autopilot of the classic example of that basically where yeah, I would not have predicted six months ago that we would be where we are now with AgTech AI and managed services. Folks, that leaves us with time for just one last thing on this episode of the show. And regular listeners and viewers will recall are referring to a technology called Lego GPT on the last show as the killer app.

We finally have the killer app for ai. Out there, and it’s helping people design and build interesting, sophisticated Lego structures. I think we have finally found the killer app for Google Maps. Erick and it comes to us in the in the form of what’s called Internet Road Trip, which I’m reading about this in TechCrunch.

The author of this piece is referring to Internet Road Trip as a massive multiplayer [01:06:00] online road trip game, essentially, where a bunch of people have gotten together in a shared Google Map session and every 10 seconds, whoever happens to be who, whoever happens to get there first decides, do we go straight?

Do we make a left? Do we make a right? And they’re making their way down the road wherever it is. This Google Maps presence happens to be interesting unanticipated use of Google Maps. I’ve gotta stay, it doesn’t sound like the world’s most engaging online game, and yet it does sound like a slightly more engaging and interesting use of Google Maps than just show me how to get to the grocery store.

Erick: I’m waiting rich for the prize. Somebody’s gonna take this and create a competition, like the Amazing Race or whatever, and, I can see it. I can see, oh, this is where we get people engaged and get excited. There’s a reward. There is something that incentivizes these folks to, become sims in real life and maybe win [01:07:00] some stuff.

What do you think?

Rich: You’re exactly right. That’s what turns this from just a way to kill a little time between tasks into something that could be a popular game. Yeah. You know that we have put that idea out there, folks that is free to you to use if if you choose to. And just drop us a line if you do so we can watch how it comes together.

Speaking of coming together, it is time for us to bring this episode of the NSP Chat podcast together for you. And we thank you so much for joining us. We’re gonna be back in a week’s time with another episode of the show. Until then, I will just remind you that this is both a video and audio podcast, which means that if you are listening to the show, but you’d like to check us out on video, what you wanna do is go to YouTube, look up MSP chat.

You’ll find us there. If you are watching us on YouTube, but you are into audio podcasts, go to Apple, Google, Spotify, wherever it is, you get your audio podcast. You’re probably gonna find us there too. And wherever it is you find us, please subscribe, rate, review. It’s gonna help other people find and enjoy the show [01:08:00] just like you do.

This show is produced by the great Russ Jones. It is edited by the also great Riley Simpson. They’re part of the team with us here at Channel Mastered. They can help you create a podcast if you’d like to do that. And podcasts are. A tiny, little, itty bitty piece of what we do for our clients at Channel Mastered.

If you wanna understand everything we do please visit www.channel mastered.com channel. Mastered has a sister organization called MSP Mastered, that is Erick working one-to-one with MSPs to help them grow and optimize their business. And you can learn more about that organization at www.mspmastered.com. So once again, we thank you for joining us here.

We’ll see you in a week. Until then, please remember, as we always say, you just can’t spell channel. Without [01:09:00] MSP.