EP41 - Maximizing MSP Helpdesk Efficiency with AI

Eric Anthony:

Hey everybody. Welcome back to another episode of All Things MSP podcast. Now, this week, unfortunately, well, fortunately for Justin, unfortunately for me, he's on vacation and taking a little time off with his family, which we all need to do from time to time. Now, that means that I have the honor of being both the host and the producer this week. So we're trying to record this kind of like it's a live show, but not really. But that way I don't have to spend as much time editing it on the backend. So thank you for your patience. And now let's get on with the show. Today's episode of the All Things MSP podcast is brought to you by Super ops.ai Built for Future Built for growth. All right, so today I have with me no stranger to the show, no stranger to the channel, Tim, coach from Pia. How are you doing, Tim?

Tim Coach:

I'm doing great. I felt like your intro there kind of led into your LinkedIn post today about a work-life balance.

Eric Anthony:

Yeah, well, it's on the top of mind right now, and yeah, work-life balance is something that I've always kind of struggled with as an MSP, especially when I owned my own business. And I think it's something we've been kind of taught some wrong things about it in the past, and so I'm trying to work out how that works for me in terms of my personal productivity and work-life balance. So yeah, check it out. If you haven't seen it,

Tim Coach:

Everybody has to, right? It's an important topic, right? I mean, if we talk about our old MSP days, especially when I was the client, what I found at MSPs is they specifically, their engineers would come in and just crunch all day long. Then it was like if you can manage to shove a sandwich down on your way to the client that you're going to be at till three o'clock in the morning doing whatever forklifts you're doing, then I expect you to be back at eight o'clock tomorrow morning to work the desk and the other stuff. I'm so glad that we've gone so far. I used to say we used to grind 'em into the dirt. I'm so glad we've gone so far away from that kind of a workload and actually trying to provide a work-life balance for everybody and helping them to help them understand what that looks like for them. Because some of us of certain generations just don't get that. It's rewiring ourselves

Eric Anthony:

Completely because that's not how we grew up in our work lives.

Tim Coach:

Very much,

Eric Anthony:

Much doing the grind like you were talking about, and I think that's kind of a great introduction into what PIA does because PIA does affect how we work as technicians in an MSP. So for the people who don't know, go ahead and give a 32nd, what is PA and then we will roll into some more details.

Tim Coach:

No, absolutely, because so much more to expand on going into 2024. I'm super excited. So essentially what happened is one of the largest MSPs privately owned in Australia was growing rapidly, and what they ran into was like we all do help desk really drug them down as far as earning their money back or getting to ROI or getting to that fable positive EBITDA number. So they're like, okay, well in essence, how do we reduce the labor on the help desk? So it turns out they had an extremely smart gentleman working there. They went to him, what do you think we could do? Here he is, I absolutely think I can help with this. And eventually what came out of it was they wrote an enterprise grade AI engine. Now then of course there's, okay, well I got the engine, but what do I do with the engine?

And we tell everybody this all the time. You come by the booth, you see 'em, my videos, I'll tell you. So basically what we do is we use that AI engine to execute various PowerShell scripts to actually resolve the issue. And I think that's probably the biggest key between what you're seeing in the argument between AI versus RPA is at this point, from what I've seen, we're the only product that actually resolves that issue at the time we're doing it, the software's doing it. So the AI is taking the action versus the human taking the action. So we've wrote it all, it's out there for people to use. We'll get on into how people can manipulate our software to go down that RPA route to justify doing their own stuff. We'll talk a little bit about that today if you want to, but that's really what it's about. It's really about making that help desk efficient, consistent, and standardized. One of the things we talk about too is when you're a young MSP, I tell people the profit triangle time, when you're an MSP, that base of your triangle is way out here. It's like, how many things can I do to possibly make money?

What you find is your spread out and your expertise is immensely spread out. But what happens is as you bring the base of that pyramid in, it drives your price up so you become an expert. So what you're really doing is you're really increasing the cost of your services because you've lowered your standardization or not lowered your standardizations. You've scaled down on what you offer and you've specialized. So now what it's doing is it's you an expert because the systems you put in place are handling 85, 90% of the day-to-day, and the rest is managed. So that's when we talk about P, that's what we're really talking about is how do we get something that makes the work super efficient, standardizes it, and gives a better result for the end customer out of the gate. The crazy thing about it's you don't have to be technical to do the software, which is what we love.

Eric Anthony:

Yeah, well, that's what AI is supposed to do, right? AI is supposed to take those boring routine, standardized the robotic process automations and make them more accessible to more people.

Tim Coach:

Absolutely. I mean, so there's some misconceptions about ai, and I wish people would really kind of dig into it instead of just the buzzword, right? Because people hear AI and immediately they all think Skynet. We all go to Schwarzenegger films and oh my God, it's going to take over. The reality of AI is it is still very much controlled by humans,

And we see those with releases. So you can go on YouTube and say, okay, well I want to play the stock market. How can I use ai? Because AI knows everything to play the stock market. I think the last time I checked, it was like 94 million videos on using AI to play the stock market. The problem is 4.0, which is what the majority of 'em are using, I mean, Microsoft, Google, everybody's coming out with their own right? 4.0 was released in March of this year. So we're nine months into old data and we know how much that kills us. The reality of this is those of us that are security conscious are still testing 4.0 to fully engage, and 3.5, which was released in March of last year, is what is currently active in most engines. The point being is humans are controlling the data that's going into those engines. So if the more people understand that the information they're getting is still by controlled by humans, perhaps they come a little bit more comfortable with the idea that there's something doing the work instead of the human, we still allow the human to check it because we have to catch up to the technology, but it's still humans that are controlling what's actually being put out in the quote AI world.

Eric Anthony:

Well, and that's one of the things that really interested me about PIA in the beginning was when you were describing it to me early on, you were talking about how much control the MSP actually has over what the AI does and what the AI doesn't do.

Tim Coach:

Absolutely. I mean, you remember the days, right? Anybody that's in MSP remembers the days, right? You've got that one engineer or two engineers that are super hoarding of the work and the knowledge.

They don't release that well. What we should be doing and what we want to do through standardization is release the ai, taking the knowledge that we have and putting into it and actually allowing it to do the work. But once again, it's scary. It's new. We have to allow the humans to catch up. So the way that we've done that is we just put in several break points. That's completely every MSP can adjust it with their own comfortability level, but we put in points where they can do anything from saying, go on the next action to all the way through letting the actual AI fix the ticket, put it in a completed status, and then going through an engineer review process so the engineer can actually look at what it did. But you have to have those break points because the reality is we still have to catch up to trusting it

Eric Anthony:

And some things because of the complexity of it and things can go wrong during a process. We need those break points to be able to check on the different stages of whatever the resolution is. And so I think that's great that that's what PIA does is allows you to build those in at again, what you said at the individual MSP's comfort level.

Tim Coach:

Yeah, the customization of the product is I spent all of 2023 splashing pia, right? Bringing everybody in. Let's introduce the topic, let's have a conversation. Now, 2024 is going to go much more into the logical and the business and the technical aspects of pia, and that's it. It's like you have to give people that control so they build their comfortability up. But I don't say all the time, once you've done a hundred password resets,

Eric Anthony:

You don't want to

Tim Coach:

Do another one and every time, and they're all perfect, why do another one, right? That's literally people don't think about password reset like it is, and there's a ton of people out there that do this kind of stuff, but a password reset is about 50 clicks once you go through the ticket, once you get into the client management software, once you go into the password manage to get to right click reset password, that is click, click, click, click, click, and every bit of that is toning.

So if I can take that from, let's say you're a super efficient engineer and you can do that in 10 minutes, guess what? I can do it in less than 60 seconds, so I've just saved you nine minutes on a repetitive task that the steps are fairly easy, it just takes time. So I'm helping to remove all that excess out to actually allow those technical people to focus on the stuff they want to focus on, but at the same time elating their clients because it's taken care of before they ever get off the phone. If you remember back to your MSP days, how often in less than five minutes did you fix a ticket with the client still on the phone and they were happy?

Eric Anthony:

Yep. Password reset is a perfect example, right? Because that is such a prolific problem. You get so many help desk tickets per day to do that, that there are purpose-built systems out there just to solve that problem. Absolutely. Now, what you guys are doing from what it sounds like is you're taking that model and applying it to other repeatable processes so that you can actually take this even farther and for the other tickets that are also coming in consistently be able to apply the same type of logic

Tim Coach:

And people are like, okay, well how do you do that? Well, it's easy. I have a nine figure MSP that's full of ticket data. Okay, let's take the last hundred password reset tickets and figure out what the most efficient way is that we did that, and then let's script it out and then have the AI run the action versus the human. So now we're becoming faster and freeing up those assets, the people that are more important to the job, but still elating the client that the work's getting done faster, but the technician, the person that lives on the phone all day long has a better quality of life.

Eric Anthony:

Now, you guys have some statistics on your website that I wanted to go ahead and pull up. 50% reduction in service ticket workload. That's huge.

Tim Coach:

It's massive. And where we're going to get more into the logic of this year is to better explain these numbers. So 50%. So our sales process is, I love it personally and I love it from an MSP mindset. I love it from a buyer's mindset. So what we do is we go to the MSP, we do a discovery, right? It's 30 minutes. All we're doing is seeing if we want to hold hands, right? Hey, I met you. You look all right. We had a great conversation. Let's see if we actually fit if we're going to be friends. So what I say by doing that is we go in, have a discovery, make sure that they're a fit with us and we're a fit with them because we don't want to lead somebody down the wrong path. The second part of that as we move forward is we actually take a dump of their tickets.

So we go out, we get their tickets with their permission of course, and then we run 'em through our software and we go back to them and say, Hey, out of this amount of work that you do, here's the amount that PIA does and it looks like it's going to save you about this much time. So we do that before there's ever any kind of sales paper in front of them. Then the next part, and this is a critical part, especially now, once they sign the paperwork, we do the work, we onboard them. So we've all been in those areas where the vendor says, you know what? They're spending $5,000 a month on our product. There is no way they're not going to implement our Well, you and I know there's no advantage. They'll pay that $5,000 a month for two years and never implement the product.

Eric Anthony:

Unfortunately, yes,

Tim Coach:

But they're not getting the benefits of the tool. That's the reason why. And there's a lot of things that we pull from the MSP life because we spurred out of an Ms P and a lot of us worked at MSPs and it's like, if I don't have the time to do it, I'm not going to buy it. So we just took that off the table, we'll implement it for you, and then day one out of the gate, what you'll see is somewhere between 20 and 30% of that help desk. And then over the course of the year as we're learning more, we'll interact more with that client. Depending on what the client wants to feed us, we've seen well in the 50 if not above the 50% range of those repetitive tasks that nobody wants to do. So that's how we get that 50%,

Eric Anthony:

And that is important. I mean, a lot of places, you call it time to revenue, in this case it's probably time to efficiency in terms of implementing. And you're right, so many MSPs don't implement the tools that they buy because of bad onboarding. And so I think it's great that you guys are more engaged in making sure that onboarding happens properly.

Tim Coach:

But now on the side that we're on now as the vendor, that makes sense to us. It's like as the vendor, I need to have the responsibility of onboarding these clients because I can guarantee that our clients see ROI, they see the tool working functionally. The day we sign off on onboarding is done, how many vendors, software, packages, whatever it is, is out there that guarantees that. And the reason why I can guarantee it is it's because it's my people doing the onboarding work. It's my people stress testing the MSP software and it's my people making sure that all the little bits and bobs have got to be adjusted this much and that much for every client is done by the end of onboarding.

Eric Anthony:

And you and I could probably have a whole conversation about the effects of customer experience and how that onboarding helps get that customer experience off to a really good start, but we just don't have time for that today, unfortunately. And then these other statistics lean into what you've already talked about, the eight times faster ticket resolution, the 30% uplift in customer service satisfaction, which is just a result of that eight x faster ticket resolution. And then this is the one I'd like a little bit more because this is something that's being talked about a lot is because technicians and especially security professionals and things like that are such in demand right now, what is this 40% increase in employee retention about?

Tim Coach:

What it narrows down to is it's allowing the people that are technical like our technicians, we bring you in as a 0.5 or as an intern and hopefully we're going to grow you to a God level engineer or a tier three or whatever you want to call it. But the problem is, is they get stuck in the mud because there's too much work. People all the time say, I want to be proactive. Well, you can't be proactive if you can't see the light. So what the software actually does is it allows non-technical people to run the software and still fix the ticket. I tell people all the time, I don't care if it's your most knowledgeable tech, excuse me, or it's your grandma. They're literally going to get the exact same result out of the ticket every time because we've standardized it. The AI engine is the one that's executing the scripts, and you don't need that.

And that free your technical people up on one side to do the things that they got into tech for. I want to learn the crazy stuff. Security's huge, but it's like maybe they want to be a security expert. Guess what? I've now freed you up on the other side of that. What we see is we see people entering the field that are not technical people, but enjoy their job because they're no longer intimidated by the tech. So that makes sense. Perfect Examples of that is one of the testimonies on the website. We got this young lady a little over a year ago. She was a dental receptionist. Nothing wrong with that. It's a good job. That's what she did. Now though, she went to work for an MSP on average with no technical knowledge. She's doing over 42 tickets a day.

Eric Anthony:

That's crazy.

Tim Coach:

And she's doing it running the software. So what does that do? That means that someone that doesn't necessarily know the bits and Bobs doesn't necessarily know the OSI model. Where's the problem? At most of the time, it's charity keyboard. We all know that. But the point being is they don't know that, but what they're able to do is they're able to advance their career into something that's more sustainable and they're able to perform a job that makes other people happy because you're fixing an issue for 'em. So they become more satisfied in their career because they're helping people without having to have a whole bunch of knowledge. And then the people that have that crazy OSI level model knowledge is now able to do the things they want to do in their career instead of answering the phone all day long.

Eric Anthony:

That's awesome. Yeah, it sounds like because what can happen quite frequently is you have so much tier one work that your tier one texts can never get out of tier one work. And now you're talking about being able to bring people who are less educated than your tier one starting tech intern level people and bringing them in and allowing them to do tier one and getting enough tier one stuff done so they can actually progress and learn how to do tier two.

Tim Coach:

And it's bringing in people who love to be quality service people. I want to be customer facing. I love the interaction because guess what, even though we're not as bad as the old joke of how do you tell an introvert, they're looking an extrovert introvert from an extrovert, they're looking at your shoelaces, not theirs. That used to be our field, the best way I can put it. We are not there anymore. Our people have advanced, but they still like to be in the weeds. This allows them to be in the weeds and allows the people to be customer facing that want to be customer facing.

Eric Anthony:

That's awesome. And I think that's very true about what's happening with customer service, with customer experience in this industry. We are having to level up that part of the game. You can no longer just be a good technician. You have to have good people skills as well.

Tim Coach:

And that's the thing is think about the ability to take somebody that's a tier one but a really good people person and you're training them into that next level. But now all of a sudden I have a nontechnical dispatcher, which I will always argue for, but now you have a nontechnical dispatcher going through all of the mundane pieces of the day because that's what, like what you've just done is you've taken that one that has the technical knowledge and you've freed them up to actually be in that relationship and actually be a resource for the client. We've said for years, we deserve a seat at the table when they're doing their yearly budgets, we should be there. They shouldn't make any advancements in their business. If they want to grow their business 20%, they have to have us there. Guess what? We haven't had time to be there.

It's products like this that free our people up to be there to have those conversations because they are trusted, because they're the ones that's been fixing those issues every day. I always tell people the MSP wants the client to be faithful to them, and I hate to break the news, but they're not. They're faithful to the people that save their butt day after day day. And when that person goes from saving their butt with password resets and new user logins to, Hey, you know what you going to add, you're looking at 10% growth. Your server capabilities where their cloud or on-prem are not going to handle that. So we're going to have to look at a couple options too. So you can pre-plan for the growth, and I'm not coming to you saying, Hey, you're going to have to drop another 50, 60 grand a year to allocate for that growth that nobody told you about. So we're actually advancing the technical people to build those relationships with our clients, and we're making our technical people happy so they stay with us.

Eric Anthony:

And that makes a lot of sense in terms of, you're right, we haven't had the time to do those things relationship wise, which we know are important. We've been telling MSPs for years that it's important, but you're right, there just hasn't been the time. I know that you guys have had a lot of announcements lately, notably your integration with Halo PSA and with Lion Guard, so congratulations on those. Thank you. Now, this primarily works from the PSA, correct? Because it integrates with the ticketing.

Tim Coach:

Correct. It only works through the PSA, so we can't sell it without the PSA already being involved.

Eric Anthony:

Okay. So which PSAs do you guys integrate with at this point?

Tim Coach:

So right now it's ConnectWise, Autotask, obviously we now Halo and then ServiceNow. Now, if you know anything about ServiceNow, that's a giant piece of software and usually pretty massive. So those are the ones that we're integrating with now. We're constantly integrating with more. That's the benefit of the show is when I fill out who I'm talking to, I fill out what their PSA is. So my engineers know in the background as they're leading that roadmap, Hey, who's next in the PSA world that we need to connect with? It just so happened this last one was Halo was winning the race. So Halo got put in

Eric Anthony:

Halo's winning quite a few races recently. So

Tim Coach:

They are

Eric Anthony:

Kudos to them. We actually had them on the podcast probably about two or three months ago. In fact, it was right before you guys announced.

Tim Coach:

Yeah, and that's the thing is there's so many good tools out there. I was a ConnectWise guy for years, so I love things about ConnectWise. I was an Autotask guy for years and I love things about Autotask, but we have outliers in our industry that don't want to be bleeding edge, but they want to be leading edge. Well, as we looked at the PSAs halo's, the one that's really got the hammer down quite figuratively and quite intentionally. So it just made sense that they got the call up next, and obviously we'll do more as we're coming along, but it was just made sense because that's who the people wanted. So if you're a good vendor, once again, that comes from an MSP background. You'll listen to your clients and you make adjustments based on their needs, not what you think they need.

Eric Anthony:

So what is next for pia?

Tim Coach:

So obviously 2024 is going to be a Bain, but we're switching it up this year. I'm switching up the conversations. The conversations need to become more business directed and logic directed. So we know everybody loves their choices. So we've released a couple of things here lately to do that. One of those being our smart forms. So how do we take more time off of the tech? Well, instead of the tech sitting there filling out the form for whatever the cause is, we've now created smart forms. We white label that. You can send that out to the client. You can actually monetize that in a way that we could talk about, but then the client can sit there and do, unlock an account. I need a new user setup, whatever the scenario is, put that stuff into the form and then when it comes into the system, our AI engine will read that form and automatically integrate it into the ticket and the work being done without any interaction from the human.

So once again, we're freeing the human up. The next thing I really love about this thing is we've implemented triage so we know how time consuming triage can be set down there and work with it. So what we've done is we actually have a way that when your customer facing tickets are coming in, we'll run all of them automatically through our model and it will automatically triage it. And then when the person goes in to fix the ticket, the resolution form has already popped up and all they got to do is hit go. So now they're not entering into information. We actually have a way using leveraging smart forms and triage together that it will completely automate the entire ticket all the way through. And then currently, once again, because we've got to have that human factor, we've got to make the humans comfortable, which we should.

You always got to have checks and balances. We all actually put that in a completed status. And then one of your engineers can go do and do an engineer review every day to say, okay, well what did it do? How did it do it? Every hour, whatever you choose. But we're still very much putting it in front of the human to know what's going on because you've got to be able to have checks and balances in the system. So now it's not sitting there with an engineer saying, go following the red bouncy ball. Now here's the form, here's the triage, here's the fix, the work's done. All I'm doing is reviewing and completing it.

So those are probably the biggest things. So now we've got full automation. It's not just a partial automation with what we call engineer assist, which is falling the red bouncy ball next server 2000 where we had to hit next 20 times because none of us ever read anything that they put in there. So those are probably the biggest ones we've released. And then the thing that's always been there, but I haven't talked about it a lot, so it's probably on the guy that talks, it's probably on his head and his shoulders that people don't know about it. We actually already have it and have always had it built into our system. That system from that RPA standpoint, you're interacting with it. You want to write your own stuff. We've always had that integrated this year. I'm going to be bringing that more forward. I want people to see, yes, we're using AI and it's meant to be automated.

However, if you look on our website, we have 60 packages, and depending on the package, it's either on-prem, hybrid or in the cloud. So basically three per, but there's people out there that still have things that they need to do. So we're allowing them to either A, take our package and manipulate it if it makes the best sense for them. We have over 300 activities that they can adjust right now. It won't overwrite the actual package. It'll add a package. We don't ever want to mess up the source code, right? It's like the old firewall days. You don't want to firewall, you save the config, you make your changes, you save the kg config, you put it in the run mode and you kind of hold back. It's like, is it going to work or do I revert? So kind of that same scenario, but we've always had it in there where people can make their own packages. I'm super excited to bring that forward. And there's, like I said, there's 300 activities already. They can manipulate ours, they can create their own. And the best thing that I love about this thing is it is no coding. You literally go into, it's the right side of the form. Here's the things that I want to do. Here's the trigger parameters, here's the activities that I need it to do Based off that,

Here's the fix. And once you hit the inner key, all the code automatically writes for you. We've already written all that,

Eric Anthony:

Right?

Tim Coach:

So when we say no code, it's literally you're putting in parameters that any of us would put in, but we have to get past. But the different version of RMM manipulations and programming, right? That's what the RMM, the RMM did a great job of us being able to go in there and do scripts and do stuff like that. We got to move past that. We don't need the RMM. It's all done within the software, but allowing people to create their own packages super easy and they don't have to know code. All they have to know is what the parameters are looking for.

Eric Anthony:

Nice. So if somebody's interested in finding out more about po, where do they go?

Tim Coach:

The easiest place to go is to go to our website, which is PIA ai. Don't be afraid. We're an Australian based company. AI is real. And right there, it just says, sign up for a demo, but this is what I want to press. If you're interested, let's have a talk. We're going to have a 30 minute conversation to figure out if we fit, because we don't want to waste your time because you don't have enough of 'em to this. So that's what you're seeing by the statistics. That's what you're seeing by the software. That's what we see by the lack of people to hire. We need to stop wasting time. The vendors need to start being more proactive for the MSPs. So that's how we approach it. How can we make sure that at every step of the way, you're a fit and you're taken care of and you just get to use the product?

Eric Anthony:

Alright, thanks Tim. That was awesome. I'm glad to see that PIA continues to evolve and is really at the forefront, not the bleeding edge necessarily, but the leading edge like you described it, of this incorporating AI and automation into the pieces that we haven't seen those things added to yet. Thanks very much for joining me today and hopefully I'll see you out on the road this year.

Tim Coach:

Eric, as always, it's a pleasure to see you. It's a pleasure to be on the show. I appreciate every minute you give me me. I know your time is just as valuable as some of those technicians.

Eric Anthony:

Well, I am kind of burning the candle at both ends, but I'm having fun doing it.

Tim Coach:

You've got to have fun, right? Let's go back to the first of the show work life balance. You've got to have fun. You got to have fun.

Eric Anthony:

Yep. All right, well, I'm going to close this out and I'll come back and talk to you in just a second. Alright. Hey everybody, thanks again for watching another All Things MSP podcast. I promise Justin will be back for the next one, leaving me free to do more production stuff on the back end. But until next time, have a great one. And as Justin is very fond of saying, we'll try and do better next time. See you. And don't forget our sponsor for this episode, super Ops ai, the AI powered PS A RMM platform designed for fast-growing MSPs built for future, built for growth. From your host, Justin Escar and myself, thank you for listening to the All Things MSP podcast. Join the All Things MSP Facebook group or follow us on LinkedIn, Instagram, and YouTube. The All Things MSP podcast is a biz POW LLC production. And even though we drink a lot of it, this podcast is still not sponsored by Liquid Death.