From Training To Use Cases, MSP CEOs Sound Off On AI

‘If you're an MSP and you're not in here talking about AI, I feel sorry for you because there's somebody else knocking their door that is talking about AI, and your customers are wondering why you didn't come out and do that,' says Corey L. Kirkendoll, president and CEO of 5K Technical Services.

Dipping Toes In The AI/GenAI Waters

Thanks to ChatGPT and especially to Microsoft Copilot, managed service providers have in general been exploring how AI and GenAI will impact both their businesses and that of their clients.

However, a panel of top MSP executives told their peers during this week’s XChange NexGen 2024 conference in Houston, that care is needed so that neither side succumbs to the pressure to jump all-in with the new technology.

The panel, moderated by CRN Senior Associate Editor CJ Fairfield, included Corey L. Kirkendoll, president and CEO of 5K Technical Services; Joe Ussia, CEO of Infinite IT Solutions; and Brent Yax, CEO, Awecomm Technologies.

The panelists agreed that education is the key to developing a winning AI strategy given the potential pitfalls of rushing into the technology.

“From the top down within the organization, you have to make them aware of the risks,” Ussia said. “You just can’t turn it on, because that's very dangerous. We start in the pre-sale cycle that’s baked into our process, and we spend time educating them.”

However, Kirkendoll said, don’t wait too long.

“If you're an MSP and you're not in here talking about AI, I feel sorry for you because there's somebody else knocking their door that is talking about AI, and your customers are wondering why you didn't come out and do that,” he said.

The panel brought a keen understanding of what MSPs need to know when investing in AI, said Omar Simmons, managing principal at P C Ambulance 1, a Philadelphia-based MSP.

“They talked about implementing it into your own environment first so you get more of an understanding before you try and send it out to your clients,” Simmons told CRN. “Some of my clients are more savvy than others. So the educational part of AI is so important to me, more than all the other parts because without clients actually knowing about it, they’re just gonna be out there vulnerable.”

An important point the panel made is that AI can save businesses time, but humans still have to check the results, Simmons said

“When GenAI first came out, my first thought was, ‘Oh, this is so dangerous,’” he said. “AI can impersonate anything or anybody, and it's not just voices but pictures and even whole videos. So, yeah, I'll have to do more of it myself before I feel comfortable with it. But I'm anxious to do it now after hearing the presentation.”

There’s a lot MSPs have to do to get ready for AI and GenAI. Here are some more questions and answers from the panel.

What are you doing to train and educate customers on AI? Are you monetizing that?

Yax: We package everything in with our monthly service fee. We don't really have a line item necessarily for it. But training is a huge focus going into 2024. We really need to build out a better training organization to go out to our clients and help them better utilize technology. Because of that, we have a ton of programs. We've built a lot of tool sets. We've built some training applications. And AI is no different, really.

I don't think people truly understand some of the implications of going down the path of AI. You have to train on it from a data management [view]. If you do AI, you have to identify what data is there. You have to categorize all of it, classify it, secure it, test it. Clients don't know that, so they jump in and hop into some generative ChatGPT or something like that and start dumping data into it. Or they'll turn on Microsoft Copilot and let it scour all the data they have. You have to really teach them and get them ready to be able to actually roll those out. We have training programs in place. We also partner with an employer organization in our town and we do all their AI training for all their members, just because it's such an important initiative.

Ussia: It really is about the initial conversation. From the top down within the organization, you have to make them aware of the risks. You just can’t turn it on, because that's very dangerous. We start in the pre-sale cycle that's baked into our process, and we spend time educating them. We do have managed clients that are ready. We've gone through every single one of our customers as a part of our menu of services, and we've made sure that things are locked down within Active Directory, et cetera, that everything is secure. … We have a full readiness assessment process. The education is ongoing.

Kirkendoll: We train on data protection and things that go with it, and we also work on the business side from a culture perspective because when people start hearing the word AI, they start thinking, ‘Is my job is going to get taken away?’ The first thing we do is educate our customers on setting a strategy of how to communicate AI internally, of their plans as an organization, of how they’ll roll it out. We want to kill that noise right off the bat: ‘Yes, we’re rolling out AI, and things will change. Some things will happen. And this is what we're setting up to do.’ We’re changing the conversation. We still have a technical conversation, but we actually help them on the business side.

One of the things we do is come in and say, ‘We want to solve a specific business problem.’ We interview different people inside the different organizations. For example, one of the biggest problems for sales people is taking a call, documenting it, clicking the CRM, and moving forward. We show them that we can take their calls, their Teams or Zoom or whatever, grab their notes, summarize it, and dump it into the CRM so they can continue to sell, they love it. That's how they see that value. We can give them back 40 percent of their day. That's a huge value they can turn that straight into making more money and selling more.

Yax: There’s an ROI to that. It's either going to be some tangible dollar amount created by top-line opportunities, or eliminating some costs internally. First, you have to identify what the ROI is. Sometimes when you're saving time, people don't recognize that as ROI because they still have the sunk cost. You have to look at it as, are you scaling currently? Is the business actively growing? Because saving that time so you're not hiring as fast as you would have, that's where you're going to recapture those dollars.

Kirkendoll: We want to train our techs. They don't like to do things like taking notes. So we actually have an AI trained on what 5K does and how we do it. When the tech inputs the problem, what they saw, and the resolution, the AI generates a response about what we're going to do so they can literally cut and paste it, put it back into the ticket, and save that time. This also allows us to create documentation from that problem. It actually generates the SOP (standard operating procedure).

How does AI impact a company’s culture?

Ussia: There's two ways of looking at that question, internally and for our customers. Internally, being propeller heads, we love the tech. We've had AI built in some of our tools for a while now. We like to be early adopters. We were on our own account. We learned it, and we moved forward. Our staff learned this is going to actually enable them and not take jobs away.

We have employees outside of North America, so there are language barriers and whatnot. We found tools like Copilot and ChatGPT help our communication and writing. Everyone knows that this is what it's supposed to look like now, and they use the tools to keep things consistent and clear. A little trick that I tell my staff is to put everything at a grade-9 reading level, and AI it does a great job of that. So it's easy for everyone to read in plain language.

With our customers, it depends on education. Similar to what you guys said, we take on each department within a company. We will do training, including education on the benefits, explain to them that this is not going to replace their good jobs. It does reduce future hires. We've already seen that. It makes people better at their job. The majority of people want work, to do a good job, and if you give them something that's going to make their lives easier and make them more efficient and professional, they typically embrace it. Some of the older generations will struggle with that because they don't like change, they don't want to do new things, they don't care, or they don't want to put the effort into it. So we're trying to figure that one out.

Yax: When you look at rolling out AI, the challenge is people sometimes identify with their job on kind of the value that they bring to the organization, and AI can displace it in some cases because it's able to do some of these workloads faster and more effectively with higher quality. So you have to look to have open conversations with your teams and your clients, let them know that this is going to happen. When somebody feels their identity is threatened, they have fear, and fear affects culture. People start having a negative view of the organization that spreads through their teammates. So you have to have upfront conversations. For example, we have an on-site team that runs equipment, sets it up, gets clients going. We have a new employee checklist. When a client of ours hires somebody, they have a ton of things they have to do to deploy a machine. That's a huge portion of what these people do. So we have to prepare them, tell them the next steps, how they're evolving within the organization, what new things they're going to be responsible for, what's their value going to be at that point, so they are not as connected to the building of those machines.

Ussia: We were own customer first. We have a little bit of luxury with the certification side of our business, our GRC (governance, risk, and compliance) practice. We have multiple ISO certifications like SOC 2, etc. So we are very conscious of how things are being implemented. We documented thoroughly, and we share that with our customers, ‘This is how we did it for ourselves.’ So we made a case study of ourselves. It's still really early in the game, so to have a lot of customer testimonials at this stage is tough, for sure, but I will say that the ones that we have deployed have been happy to talk with new prospects or other customers. It's getting easier as we go forward, but I'm hoping 2025 becomes the year where we'll have an abundance of customer testimonials to help us.

Kirkendoll: It’s hard to sell and be excited about something if you don't have any experience. And so doing things first with us has been great. But we’re very targeted in the nonprofit area, and so we spend a lot of time there. One of the challenges clients have is grant writing. How do I write grants? It's very time consuming. It takes a lot of research and a lot of data crunching, which is perfect for AI. So our goal for those guys is to leverage AI and LLMs.

Are you seeing customers ask about AI for specific tasks?

Yax: We're seeing some specific, and we're seeing a lot of general, like, what is this? What should we prepare for? How do we know if this is going to impact our business? We do the training. We do readiness that leads to a proof of concept of some sort. An easy one is configuring Copilot or something that gets sales and marketing onto something generative. Those are pretty easy ones to get them to understand. And then we're digging a little bit deeper and going into specific things, and they're getting the idea to come to us now. ‘Now we want to be able to do this. Now we want to be able to do that. How can we automate this to help me better leverage whatever it is?’ They’re coming up with those specifics after they see it and start doing something.

Ussia: We support a lot of legal firms. The irony is, law firms, believe it or not, are the easiest win when it comes to this. It doesn't make any sense to me. I don't know why, but our law firms that we support, a lot of them have been very anxious to get on to things like Copilot. With us being a Microsoft shop, that's what we push a lot of, obviously. But they've seen options. They've seen our tools. We have tools within our PSA connected into other AI platforms. RPA (robotic process automation) is another big one. Within our customer portal, clients are not seeing how we've enabled RPA for them for employee on-boarding off-boarding. Their HR departments using it. We built it into the RPA in the background, so it goes out, provisions, licensed, commissions, etc. So there's a lot of this and that within our RPA. But more importantly, when they look at the outside, they're seeing it as an extension. For example, they've already been interacting with our customer portal. We have an AI chat model that's doing well now. It's been deployed for seven or eight months. Best thing ever. We thought our help desk employees were going to be worried about their jobs. But because we explained to them, this is going to be more efficient and showed them how it works, they embraced it. Now it’s the best thing ever. Even our phone system. We've enabled some AI to do metrics and to analyze conversations, and if there are issues, it actually will go into our PSA and flag it, put the recording in for us, and assign a ticket to myself or our business department so we can take a look at what's going on and why. It gives us a summary of what went wrong.

Kirkendoll: This is the time where you really have to be out there first to talk to [clients] about what AI is, what it's all about, and how it's doing. I’ll tell you, if you're an MSP and you're not in here talking about AI, I feel sorry for you because there's somebody else knocking their door that is talking about AI, and you customers are wondering why you didn't come out and do that. And one of the things I would tell you is everybody's getting emails now from all of their vendors saying, ‘Hey, we're updating our privacy policy.’ What they're telling you is that you're giving them permission, if you don't respond back, to throw your data into our AI models and do things and use your data to do certain things. But the customers are not reading that. They're looking for you to tell them what's going on. How is this happening and what's happening. It's time for you to come in and be that strategist and say, ‘Here's what it is.’

Last week, I was at the BoB conference, and I think the CEO of HPE said it best: Everybody, no matter what you do, you have to have a major in whatever it is that you do, and a minor in AI. And I was like, cool, because it's exactly right. Everybody needs to understand it. And he said everybody has to learn to coexist with it. You have to teach your customers that it's going to change the game, but you're going to have to learn to coexist with it and figure out how it integrates into your mission. And they’re looking for us to help them through that process.