MSPs in the AI Era: A Roadmap to Success

[00:00:02] I was asked to do a presentation on what solution providers needed to know to be successful with artificial intelligence and what successful ones are doing now. This is that presentation shared for everyone on a bonus episode of the Business of Tech. Success Stories with AI. So my focus today here is I'm going to dive into real-world success stories of AI. Now, obviously, I want to do a little bit of baselining to do that. Why are so many businesses involved?

[00:00:31] Well, from my perspective, AI is moving from innovation into necessity based on its reshaping of industries. If I look at the data, 40% of upper mid-market firms, 40% of core mid-market firms, and 30% of small businesses expect AI-related IT spending to grow by over 25% this year alone, and 86% of global businesses budgeting for AI and R&D. It's becoming part of work life, and especially among knowledge workers.

[00:01:00] This is all about enhancing productivity and competitive edge. But by the way, for you as providers, this signals opportunity to support clients on their journeys. 75% of knowledge workers are now using AI at work, and it's nearly doubling in just six months per Microsoft's Work Trend Index. 74% of small businesses believe that employees using AI tools gain competitive advantage.

[00:01:25] 41% of businesses have restructured their teams specifically to integrate AI capabilities. So leaders are seeing this as a key priority for the near future, and this is indicating a strategic shift in their priorities. For you as providers, this means helping clients maximize that in workflows. 58% of technology leaders predict that AI will be the most important technology by next year, by an IEEE study.

[00:01:53] 86% of IT leaders expect generative AI to play a significant role in their organization soon. 60% of PCs are expected to have AI capabilities by 2027. So I'm going to say, not a trend, central to how businesses are planning to grow and compete. Now, you will not sell AI. I think we need to start with that. You'll see projects and solutions enabled by AI.

[00:02:19] The value of AI is in the practical application, solutions embedded into real-world business needs. I want to baseline here because I want to pull in my first speaker here. Seth Robinson of CompTIA sees AI as transformational when embedded into solutions rather than standalone products, but he draws parallels to other emerging technologies.

[00:02:39] In this discussion around emerging technology, and I think one of the main themes of that discussion is that most of the emerging technology that we're talking about is not a product. It doesn't have a skew. It's not something that comes in a box.

[00:02:54] We've talked a lot at CompTIA about strategic IT and the shift in energy that's been happening from putting a lot of work and investment into the computing platform and starting to put that work and investment into applications. And so applications are getting more and more complex, and you're getting more layers in that application stack or in that tech stack. So AI is one of those layers.

[00:03:20] There's a lot of talk about AI tools or AI products or an AI market, and I think that what we're going to see as companies continue using AI, which is another thing we can talk about, that this hasn't only been happening for the past 18 months, is that the things that they're investing in are these business solutions that start to sound pretty familiar to us, right? ERP, CRM, business productivity, HR, finance, and then all of those things are going to have AI under the covers, driving new features.

[00:03:50] And so there definitely are things that you have to know about AI to be conversant and to be able to use these things in a masterful way. But you're not going to be selling AI to clients. Clients aren't going to be implementing AI outside of implementations that they're already doing. So the potential of AI is not as a standalone product to sell, but it's a powerful component embedded in solutions. So this isn't about marketing AI as a separate service.

[00:04:20] It's about integrating AI capabilities into core offerings that enhance efficiency, improve customer service, drive customer growth, and you do it both internally and with customers. So the approach allows you as providers to leverage AI's benefits while maintaining focus on primary services and expertise. Now, let's get into some specific examples. I'm going to pull two straight from CRN's coverage. The first is the MSP General Informatics, and this is their quote.

[00:04:47] We see a use case in assisting human resources with information for their internal use. Disseminating information to the entire organization sometimes can get really burdensome to the HR department. That's where we started because it made the most sense. And we ended up really being very successful at reducing the number of calls that come into human resources. In fact, we were thinking about hiring an intern just to handle some of that. And once we deployed AI, we didn't have to do that anymore.

[00:05:16] And while people always say, oh, we're going to lose our jobs, nobody lost theirs. We did see a slowdown in labor growth, end quote. So the insight, AI can reduce repetitive HR tasks and reduce headcount growth without job loss. For you as providers, similar solutions can apply to internal customer support or IT requests. Here's one from a company, Lan Infotech.

[00:05:42] Quote, the low-hanging fruit is definitely co-pilot. But we have had a lot of discussions with customers about how to use AI outside of that. In a co-pilot discussion, I tell customers the biggest benefit that we're finding is in Teams. I say we can re-engineer how they're doing Teams meetings to make things productive and upgrade their post-COVID conference rooms to smart technologies. We always see Microsoft demonstrating it. What did I miss in a meeting? It happens all the time.

[00:06:12] So smart transcription, speaking like that, you're in chat, because we're never all going to be back to the office anymore, end quote. Again, the insight, AI can make remote work more effective with smart meeting technology. Consider how you would use AI to improve virtual collaboration for your clients, like upgrading post-COVID setups to those more seamless remote experiences. And note, AI isn't replacing jobs. It's augmenting tasks and making employees more effective.

[00:06:41] I'm also very much to think about this as an augmentation tactic. Now, I want to pull in another of my guest experts, Colin Britton, who's the COO over at Devicey. And he's talking about the use of AI as a tool that enhances, not replaces human work. And I think the AI promises to go the other way, which is the highest common denominator. So in business operations, you should be able to increase the productivity of your best players

[00:07:09] and increase the productivity of your pool overall. But from a top-down perspective, not a bottom-up. With the AI promise, it starts to become the other way down because the critical thinking becomes much more important when you've got AI there. I'll give you a really quick example. A friend of mine the other day who's a data scientist, super smart guy, he's working on some dashboards.

[00:07:36] He's like, I don't have to write code anymore. ChatGPT writes my code. And he said, you don't have to learn a new language. And it's like, yeah, it was 95% there, minus weak. And it was done. So let's talk about how solution providers are selling AI. And I'm going to give you three approaches. A classic approach, then I'm going to talk about operational transformations, and then I'm going to talk about in-house enhancements. So first, let's do a classic approach.

[00:08:06] And I'm going to tell you, this is fresh stuff. I just talked to a vendor last week at IT Nation, and he brought it up himself. I didn't even ask. He's talking about Microsoft Copilot is boosting his revenue and cutting costs, though higher costs are potentially slowing the adoption. So I want you to hear directly from this vendor. This is Blackpoint Cyber's John Merchinson, who I spoke to last week at IT Nation. A little bit about how you've implemented Copilot in the organization and what it's doing for

[00:08:35] you and where you're finding it's effective. Yeah, I think we're at the early stages of Copilot, to be clear. Fair, but you're the first person to bring it up to me in this context. I'd like to hear a little bit. Yeah, so step one, I think in July was two weeks, two contiguous meetings, first ones in two and a half years. Okay. So the pace on the road that we've been running has been almost too much, which means you can't always be back in every single meeting. But we have a lot of experience. We built this company.

[00:09:05] We know this MSP space. And we've hired so many people, a lot of which need to be trained up and learn the customer base where we're going to. I find Copilot that's great for one for me personally. Summarize my most important emails for the last week and summarize any action items I have on it to make sure I don't drop the ball. The second one, when I'm missing meetings, I find it incredibly important that I can ask

[00:09:32] questions of the meeting because I don't think you have 60 minutes to go back through the whole meeting. Right. So the transcripting, the summarization, the key outcomes, the action items. And this allows me to be a lot more efficient. Now, I think Copilot's kind of interesting because if you rolled it out everywhere and every meeting was recorded, it'd probably be too big, brother. And people need freedom to air their grievances about their boss or whatever was without punishment or something like that. So I don't know.

[00:10:01] We're trying to find the balance on it, but so far, I absolutely love it. Okay. I am addicted to it. So very small, short question. Are you paying for it yet? Yes. Okay, cool. That's important because some people are piloting some people. So that's great. We're absolutely paying for it. Awesome. Cool. So in this survey conducted by Microsoft, businesses with up to 300 employees reported a 12% faster time to market for new products and technologies after adopting Copilot.

[00:10:28] Their study also showed that security analysts were 22% faster with Copilot for security and 7% reported their work being more accurate with the tool. In a recent report by Forrester, early adopters of Microsoft 365 Copilot among small and medium sized businesses are reporting those significant benefits and they justify the additional $30 per user per month. The study surveyed over 250 small businesses worldwide and found that organizations using

[00:10:55] Copilot are seeing a 6% increase in revenue and a 20% reduction in employee attrition. Additionally, more than half of respondents expect up to a 20% cut in operating costs. And furthermore, 75% reported improved employee satisfaction and over 80% anticipate better customer retention. I want to have a counterpoint. From the information, the 365 Copilot has received a mixed response from customers due to performance issues and high costs, leading some businesses to pause its adoption.

[00:11:25] Some companies like BlackRock and EY are expanding its use and overall adoption remains around between 0.1% and 1% of Microsoft 365 users paying for the feature. So I think there's room to figure this out. We've got examples, we have uptick, we have data that it's happening, and we see slow adoption. So I think, I like that, it's a space that we can dive into. So next, I want to take a quick look at how we might be able to change operations for customers.

[00:11:53] And I want you to tell you, this one example is coming right from the field. I spoke with Uzhar Ahmed, who's actively deploying automation solutions to his customers. He utilizes AI for automation in his business process automation firm, Cottonwood Automation, by implementing AI agents to handle customer interactions. The AI is trained to understand the business processes and can engage with customers through phone calls and emails. By leveraging the AI, his team ensures that customer inquiries are properly addressed,

[00:12:22] even during after hours and when calls go unanswered. The AI can handle customer complaints, provide information, and even identify potential business leads. The automation improves the customer experience and streamlines operations by ensuring no calls are missed and allowing the team to focus on some high-value tasks. Let's hear from him. So I was really against AI at the start for using it for home service. I'm like, you don't want to outsource the customer service aspect of your business.

[00:12:49] But I was working with a few developers and for my other company, Instabike, which is like my testing grounds for everything I do, I have a new AI to start to call. And Dave, it is so good. It is so good that I'm just like, okay, six months from now, I don't think I'm going to need anyone answering calls anymore. Really? People can't even tell they're talking to a robot. If you don't know, you don't know. Right. And it's got the ums and the ahs and inflection points and little jokes here and there.

[00:13:19] And it also can be like our logics too. So like, for example, we had a customer cancel a job within 24 hours and we have a rule with a job that's canceled in 24 hours. And the customer out logic my AI to be like, hey, you did AI away with it. But instead of being upset, I was like, that's impressive because it worked outside the script. But so tell me a little bit more there because you're literally putting AI agents then talking with customers. So we were missing too many calls.

[00:13:48] So if it's four rings and no one picks up, it goes to the AI. After hours and voice mail go to the AI. And the AI is fully trained in our business. So it knows like everything that we do and everything that we don't do. It asks the right questions. And then if the customer is a good fit and they're ready to be booked, it'll pass that information to our operations team through Slack. We'll give you a summary of the call and everything you need to know, the next steps. It's incredible. I love it so much.

[00:14:18] He's really leaning in. So he's focused on the customer experience and he's freed up a bunch of resources for other tasks. I also want to highlight Asana, who's introduced an AI teammate designed to enhance workplace productivity by offering advice and plans based on the team dynamics and historical project data. The AI model aims to assign tasks to individuals with the most suitable skill sets, such as pairing designers with relevant projects.

[00:14:46] And those early adopters, including a marketing organization, have reported that the AI is effectively generating customized marketing content and standardizing their workflows. Now, that's two ways. And the third way I'm seeing is around specific ways within your organization. Now, I want to start with, Isaya has introduced AI automation functionality called Cooper Bots for its entire suite of RMM, PSA, backup, and other solutions.

[00:15:16] These bots use AI to automate a bunch of different IT management tasks. And likewise, last week at IT Nation, enhanced their Asio platform doing exactly the same thing with AI bots that allow you to do that. Now, Transputech, which is an MSP based in London, created a tool called NeoAgent that uses AI to automate the resolution of help desk tickets. The tool is delivered via Microsoft Teams in a bot,

[00:15:44] and it's helped reduce their ticket resolution time. Now, I want to give you a big one too. So, I want you to hear about this one, which is Friday AI. And so, you know, rather than us look externally for support engineers, we actually look internally and innovated, and we built something that we codenamed as Friday AI. And it was actually just called Friday. And yeah, long story short, you know,

[00:16:11] it allows us to scale ultimately to several thousand locations or cost a huge footprint, which is the national US. And we were ultimately managing that footprint with a matter of three guys. And that was because Friday AI was doing most of the heavy lifting for us and enabled us to do that. So practically, like, I mean, how does it work? What does Friday AI do? Well, ultimately, we have an appliance that we plug in on-premise that takes all of the management aspects and learns your environment.

[00:16:40] And without getting too technical, we can certainly go deeper. But it sets a baseline of normalcy. What are all my devices? How are they communicating in my environment? You know, which maybe destinations are they speaking to out in the internet and in the wild? Once it understands the environment, it basically goes into supporting it. And if it sees any anomalies in your environment, like a device failing or a POS terminal, not able to process transactions where it used to,

[00:17:08] Friday jumps in and it either, one, creates a support ticket, ultimately identifies what's the problem, when it happened, ultimately the root cause of the problem, and, you know, creates a support ticket, sends it in. But it can also do something that we've been actually really excited about from the beginning. It takes the actionable next steps to resolve the problem itself. And it does that by either maybe making a configuration on the fly. Maybe it does it by power cycling a device,

[00:17:36] maybe shutting off an upstream switch port, things of that kind of nature. Yeah. Okay. I got to go one level deeper on that because I'm somewhat technical, a little rusty here, but I know my audience is screaming going, okay, tell me more. So I'm guessing, and then I want you to fill in the details. So it's an appliance, it's watching the network, it's doing network packet sniffing, it's doing some interpretation to understand what's going on. It's got to be given information too.

[00:18:05] So that's as far as I've guessed, how right am I? And now start filling the rest for me on like how it actually works technically. Yeah, absolutely. Keep in mind, this was made for a tool that was allowing us to scale. So if, you know, any configuration that needed to be done on premise was major overhead for us. So the, the, one of the objectives was, how do we make this not just another tool that has to be managed or at least managed in a very light way. And so to answer your question on how it's working is you plug in the appliance.

[00:18:34] If it's a DHCP enabled environment where it can get an IP address on the network, then great. It starts drawing some simple logic. Hey, here's my local subnet. Let's do a quick scan, right? And not only listen to like pinging and layer three information, but also listen to like what we hear in the broadcast domain, ARC information, DHCP requests. Ultimately learn the Mac addresses and the devices that are in my environment and then go from there.

[00:19:02] We have that idea of ultimately discovering your environment. That's really our first phase. But then we start to look at the data and say things like, okay, if we can communicate to other networking devices in your environment, we can actually speak to it. And we can say which devices are plugged into which ports. What's the destinations that some of the devices need to hit to be functional or have been hitting over the last X amount of time, right? It draws conclusions and learns the environment from that respect. That makes sense?

[00:19:32] It does. Now, so how much of that is the unit doing itself with like the large language model, maybe being on the device? How much is in the cloud? Walk me through a little bit about how that portion of it works. Yeah, absolutely. So everything that I described is not on any large language model, right? This was, we built this pre, like we've been in production with this since 2017. I shouldn't say production, but in development since 2017.

[00:19:58] And, you know, there was no chat GPT or any sort of LLM concepts, right? And so for us is how do we take data, crunch that through what we call our machine learning algorithms and draw conclusions and set baselines on that, right? And so what we're doing is we're ultimately taking very simple data. We're not even looking at payloads or, you know, what data is being transmitted. We're just looking at like, for example, network overhead, you know, source and destination information.

[00:20:28] What ports, network ports are a device communicating at in the environment? You know, how much bandwidth, round trip time, HTTP get header information, things of that nature. And so once we learn devices are in the environment, Friday can ultimately create traffic and simulate that by bouncing it off a device, depending on what ports it's listening to. And based on those variables that I described, we can ultimately draw conclusions.

[00:20:56] This device, it's a actual video surveillance camera. It's an iPad. It's a POS terminal. It's a Windows server. It's a, you know, a list of about 50 different things that we can categorize. TJ really makes you think about the question. Have you considered what parts of your operations might benefit most from AI augmentation? Now, I want to start by saying, do I think most of you will build your own technologies? I do not. I want to be very clear on that.

[00:21:23] Building a services organization really differs significantly from building a product organization, particularly in terms of business model, focus, and the financial data. And to give you some data from the recent service leadership report, since 2009, the average managed services provider has driven an 11.7% compound annual growth rate in total revenue, 17.2% CAGR in recurring revenue, and 25.8% CAGR in approximate company valuation.

[00:21:52] The growth trajectory demonstrates the popularity of service-centric businesses. Services organizations, like MSPs, focusing on delivering intangible value through expertise, ongoing support, and that requires a different operational approach compared to product-centric businesses. They typically have higher sales, general, and SG&A costs, around 30% of revenue to account for the additional resources and complexity in service delivery.

[00:22:17] Now, in contrast, product organizations concentrate on developing, manufacturing, and selling those tangible goods and software. They can operate with lower SG&A costs, potentially as low as 13% of revenue for companies with 90% product. The financial dynamics also differ. Product companies may have lower ongoing costs once the product is developed, and services organizations have more predictable recurring revenue streams, but are also needing to manage utilization rates and service delivery costs.

[00:22:46] So really, just know they're different, and it's very hard to do both. Now, I also want to highlight another area that I think is really key in all of this, and I think it's one of the active AI opportunities right now, and it's data governance. Effective AI needs reliable, secure data. So let's hear from a data expert, Avi Perez of Pyramid Analytics.

[00:23:09] In particular, the area I wanted to get your insight a little into is that I like to think a lot about how we make sure clients and clients implement this responsibly and effectively. And it feels like there's a distinct gap between those product developers that are making the LLM empowered and supercharged products and the end customer.

[00:23:31] Do you think about the role that those service providers are going to play in making sure that that is done responsibly and correctly, and specific services that you think are important there? So there's no doubt, and I think you start at the beginning of your question. Is there a gap between – I don't want to call it hype because we've all seen the functionality. We've all played with it so hard, but it really does what it looks to do.

[00:24:02] But making that work, certainly in a business context, there's definitely a gap there. It's still a bit of a chasm to grow up. But interestingly enough, at least in our opinion, from the firmness perspective, it's not necessarily an issue of the LLMs themselves that have to do. It's all the software that will incorporate the LLMs. That's the chasm that you could close. It looks a lot harder, and it is a lot harder because there's that gap.

[00:24:31] Then continuing that sort of thought process, who's going to close the gap? Well, part of it is the software vendor. Again, it's a pyramid story. But the entire success of Pyramid's gap-closing strategy is the implementation of the data strategy, which brings us all the way back to the service provider. Exactly. So to put it all together, it'll look something like this. The LLMs are absolutely phenomenal in terms of what they can do.

[00:25:00] I mean, their interpretive capability is mind-boggling. It's just amazing. It's a game-changer for what was there, let's call it, two years ago or three years ago. Then you need the vendor and the vendor software to use them properly to make good use of them. This is a case of Pyramid and in our industry of all the other vendors. But none of it matters. None of it will really go anywhere without somebody feeding Pyramid and ultimately the LLMs of good quality structure, good quality data,

[00:25:29] and a deployment that is meaningful and reasonable. And from that perspective, it all sits on top of, to your point, service providers to make that happen. It's otherwise garbage in, garbage out. And that's ultimately the story. And that's ironically the same problem that we had last year and five years ago and a decade ago. All the cool tools on the market are useless without the right people knowing which buttons to push to turn them on.

[00:25:58] It's seldom is random people in a company. It requires the degree of expertise. It could be very sophisticated users in a company. And by and large, it's normally service providers and consultants to come in to make that happen. That's pretty much the story. It's been like that for years. And now we have the next era. It still requires that kind of sort of lifting to make it work. So in order for this to all be realized, there's a step before.

[00:26:29] You have to be ready with the data. So this year and 2025 are going to be the year of data readiness, helping organizations structure their data for the AI of tomorrow. I think the big opportunity around AI is the idea of working with organizations in that consulting mindset of doing data prep, data inventory, and doing the part that makes sure we're all ready for the different security models. One of the big things about all the LLMs is you can't necessarily let them loose on all your data because they won't have any concept of security.

[00:26:58] And they'll give it up to anybody who asks. So you have to start thinking about that. And you have to start prepping. So for 2025, your big opportunity is getting businesses ready. Remember, before we can harness the power of AI, we need to ensure the data readiness by organizing and prepping all the assets. Imagine for a moment deploying a customer-facing AI bot without security planometers in place. All that sensitive data could be mishandled or exposed.

[00:27:23] And I tend to think that it's going to be using the kind of intellectual frameworks like a NIST framework or a risk assessment framework to put that all together. And what we're going to see in 2025 and 2026 will be those more advanced programs. So I put data readiness as a key component of AI governance. It ensures that data used for AI is not only available but also secure and compliant with regulations.

[00:27:48] And like shadow IT, AI governance requires a lot of vigilance. It ensures AI usage aligns with organizational goals and it complies with all those legal and ethical standards. So let's pull in another expert. This is Colin Graves. You brought it up first. People in process felt like it was significantly more important in this than technology. How much of this is data governance?

[00:28:14] How much of this is about putting together that policy and having ways of managing it? What does that look like from a data governance and how much is it data governance? Data governance is a huge piece, but it's often... Let's face it, Dave, right? Maybe the folks who are listening to this get really excited about data governance because they're IT professionals. Customers don't get super excited about data governance and left to their own devices. Oftentimes, that chapter of the book gets skipped, right?

[00:28:43] We want to sort of... Let's skip ahead to chapter seven. Oh, who's this character? I've never even heard of them. And why did he just die? There must have been something I missed. The data governance piece, right? If you talk about account structures, if you talk about the flow of data within those systems, and if you talk about the access controls, really it's about solidifying the data quality within an organization. And when most people hear the term data quality, they think,

[00:29:12] okay, does this field in this database say the right thing? That's a correct assumption about what data quality or data governance is. But really, it's about how do I create a mechanism that causes this data to be updated and trustworthy over time? Because when you're implementing data capabilities, you really have one shot, which is why we like to start with that crawling as opposed to running.

[00:29:41] Because customers have their way of doing things, sometimes outdated, sometimes very inefficient. But the worst thing that can happen is you implement new systems and the governance and the controls and the quality aren't up to snuff. And Dave takes a look and he goes, yeah, I don't trust it. I'm going to go back to using this 17 gigabyte Excel spreadsheet I have stored locally on my computer. And you're not going to convince me otherwise.

[00:30:10] The mission has failed at that point. So data governance is a core tenant of creating that trust within these systems. And that's what we talk about a ton at North Labs is if you don't have trust, or if you lose trust within your data program, you might as well scrap it. It's over. Okay. Because it's going to be really hard for people to come back. Interesting.

[00:30:35] Now, the other dynamic that I would think is really important here is we talk about data quality and getting to a level of... We use phrases like funliness and organization. That feels like an unattainable goal, right? The idea of having perfect data is probably unattainable. And by the way, even if we manage to get there for one fleeting moment, the moment users do anything, there's probably...

[00:31:04] It's no longer perfect. Like, how much should we think about this in terms of perfection? And how do we think about where we're actually trying to get to from a data cleanliness? And I'm putting that in air quotes for listeners. Seth? It's a great question and something I get asked about a lot. The thought by a lot of people is, well, I can't have a data program until my data cleanliness is achieved.

[00:31:33] That is an impossible goal, like you talked about. If you think about an average manufacturing organization, for example, we work a lot in asset-intensive industry. So manufacturing, industrial, supply chain, oil and gas, etc. The average number of systems in an average company over the last eight years has nearly doubled from right around 60 or 65 to more than 130 systems.

[00:32:03] 130 systems. And we think that is a plausible goal to go ahead and make all of those systems clean. Never going to happen. The way that I suggest doing things is, even if you do have 130 systems, God bless you, There are probably two or three that are the nucleus of your organization.

[00:32:26] Your ERP, your CRM, your manufacturing execution system, your invoicing and accounting and financial systems, whatever that might be. If you start with those, even in an imperfect state, the nice thing is while you're taking that crawl approach, it can almost serve as a litmus test for those internal systems.

[00:32:50] Before it's rolled out to the entire organization, you can get to cooking a little bit, look at the data that gets output by these data capabilities and go, That doesn't smell quite right. Seems like there's a little bit of drift. Now we have a pointed area to attack within our core systems from a data management and data cleanliness perspective. And we can go and that becomes a much easier conversation.

[00:33:19] Okay, this subset of this system isn't quite right. Let's put some time and attention into that. Get it up to snuff. And then we'll put alerts or guardrails in place in the future. So if there is drift, we're informed about it and can do something about it.

[00:33:39] But that's really the core push that we make within customer accounts is like, if you think that you're going to have 130 pristine systems or 15 pristine systems, you're out of your skull. But if you need clean areas of certain systems, now that becomes a lot smaller bite of the elephant. Gotcha. Okay, that's a good way of framing it.

[00:34:05] So AI is going to impact how we approach productivity, customer engagement and operational efficiency. And I want to, again, position rather than marketing AI as a standalone thing. The successful implementations have a theme of embedding AI into existing services. And so they're driving gains without diverting any focus. I've pulled some real world examples for you from companies like General Informatics and Land InfoTech, who show that AI is reducing repetitive tasks and enhancing collaboration.

[00:34:33] While those Microsoft co-pilot ones are focused on speed to market, security and analysts, and employee satisfaction. The integration emphasizes the value in specific goal-oriented applications rather than any of the big broad-based promises. Now, alongside with all of this, data governance is going to be an essential portion of this, ensuring data security and alignment with goals.

[00:34:56] Data readiness, prepping, organizing, and securing data is going to be the foundation of anybody's effective AI strategy. Experts like Avi Perez are pointing to that critical period for businesses to focus on data prep and those compliance frameworks, such as like ones from NIST, to manage risks associated with AI. The focus of data lays the groundwork for advanced AI applications in the coming years.

[00:35:20] So as orgs seek not only to harness AI's benefit, but also manage the impact responsibly with secure and compliant parameters. AI's value lies in targeted, personal use and not broad adoption. Data readiness is key. Now, I would be remiss if I didn't end by telling you how much AI went into this presentation.

[00:35:41] All my data points, all retrieved by the AI in Notion, where I keep all of my stories, which are all summarized and coded for me in AI. My conclusion here? All in interaction with ChatGPT. The presentation was then provided to ChatGPT to edit and offered areas to rewrite. I asked it to reorganize drafts, where to tighten, where I might have missed.

[00:36:07] AI was my real-time partner just even in creating this conversation with you. So your takeaway, start with clear, focused use cases, invest in data readiness, and build AI into existing workflows rather than as an add-on. And as you prepare for AI's impact, think about where your organization can lead and not just adapt. It's going to be a journey. And those who invest in the right foundations today, I believe, are going to be the success stories of tomorrow.

[00:36:37] The Business of Tech is written and produced by me, Dave Sobel, under ethics guidelines posted at businessof.tech. If you've enjoyed the show, make sure you've subscribed or followed on your favorite platform. It's free and helps directly. Give us a review, too. If you want to support the show, visit patreon.com slash MSPRadio and you'll get access to content early. Or buy our Why Do We Care merch at businessof.tech.

[00:37:06] Have a question you want answered? We take listener questions, send them in, ideally as a voice memo or video to question at MSPRadio.com. I answer listener questions live on our Wednesday live show on YouTube and LinkedIn. If you've got a comment or a thought on a story, put it in the comments if you're on YouTube or reach out on LinkedIn if you're listening to the podcast. And if you want to advertise on the show, visit MSPRadio.com slash engage.

[00:37:33] Once again, thanks for listening and I will talk to you again on our next episode. Part of the MSP Radio Network.