Starting an MSP in 2024: Productivity, Data Management, AI Predictions: Live Show, 17 July 2024

[00:00:46] [SPEAKER_04]: What would I do if I was starting a managed services provider today?

[00:00:51] [SPEAKER_04]: It's the question that dominates my YouTube channel as my most popular content and I can

[00:00:56] [SPEAKER_04]: continue to request update it.

[00:00:58] [SPEAKER_04]: So today we'll dive into that.

[00:01:00] [SPEAKER_04]: Welcome to the Business of Tech Lounge, the live version of the Business of Tech Podcast.

[00:01:05] [SPEAKER_04]: It's Wednesday, July 17, 2024, and I'm Dave Solbowl.

[00:01:10] [SPEAKER_04]: Now I'm going to be taking questions and comments throughout the show so make sure to

[00:01:14] [SPEAKER_04]: put them in chat.

[00:01:16] [SPEAKER_04]: There is a dedicated question section in the show for listener questions that have been submitted,

[00:01:21] [SPEAKER_04]: but particularly with today's topic, I want to hear your comments and thoughts as we go.

[00:01:26] [SPEAKER_04]: I want to thank sales builder, our Patreon sponsor, who support makes this show possible.

[00:01:32] [SPEAKER_04]: Focus on your IT sales workflow with the power of automation and visit them at salesbuilder.com

[00:01:38] [SPEAKER_04]: that's B-U-I-L-D-R.com.

[00:01:42] [SPEAKER_04]: For my anger, I am watching the chat as we dive in.

[00:01:46] [SPEAKER_04]: So our top story today, the thing I want to talk about the most, it's my most viewed content

[00:01:51] [SPEAKER_04]: on the site.

[00:01:53] [SPEAKER_04]: How would I start an MSP today?

[00:01:56] [SPEAKER_04]: Now, I want to give you a little insight into how I prepare my editorials as some

[00:02:00] [SPEAKER_04]: context.

[00:02:01] [SPEAKER_04]: I spend a lot of time riffing, talking and thinking about those ideas and then ideally

[00:02:08] [SPEAKER_04]: to record them and tighten them up.

[00:02:10] [SPEAKER_04]: And normally, I do all this work in private.

[00:02:13] [SPEAKER_04]: I'll do them to a camera, I'll try a couple of versions, and I'll see what happens.

[00:02:17] [SPEAKER_04]: But this one's been percolating long enough that I figured I'd go ahead and do a rougher

[00:02:21] [SPEAKER_04]: version with you live and it could be interactive for those that show up.

[00:02:26] [SPEAKER_04]: And if you get this on the recording, put your comments in and we will discuss them.

[00:02:30] [SPEAKER_04]: So you get the opportunity to do this together.

[00:02:34] [SPEAKER_04]: Now, my biases, I'm a little less enamored of providers who are so focused on their

[00:02:40] [SPEAKER_04]: toolkit and technical stack and selling you their prepackaged, per user, per device,

[00:02:48] [SPEAKER_04]: per month packages.

[00:02:50] [SPEAKER_04]: Because it's been an insight in the way I'm thinking now, but design is being an analyst

[00:02:55] [SPEAKER_04]: and a voice from the show, I'm also a customer.

[00:02:59] [SPEAKER_04]: I run a media company and I use IT to do it.

[00:03:03] [SPEAKER_04]: I think a lot about myself as a small business and the way that I might engage with an

[00:03:09] [SPEAKER_04]: IT services provider in order to get some value.

[00:03:13] [SPEAKER_04]: And I don't necessarily believe anymore that the pure monthly recurring revenue solution

[00:03:19] [SPEAKER_04]: is the win that it might have been.

[00:03:22] [SPEAKER_04]: In fact, as the industry has matured, I've been proven right on this.

[00:03:27] [SPEAKER_04]: One of the things that I thought about a lot recently is my conversation with Peter Koojawa.

[00:03:32] [SPEAKER_04]: Now, Peter and I talked about the way profitability worked in business.

[00:03:37] [SPEAKER_04]: And I want to remind you of a section of that discussion.

[00:03:41] [SPEAKER_00]: If I'm an MSP, it's really important that I not only be doing projects but that I'd

[00:03:46] [SPEAKER_00]: be doing them in a way that's going to maximize my service gross margin.

[00:03:50] [SPEAKER_00]: We did a whole section in last year's annual report which you'll probably remember

[00:03:54] [SPEAKER_00]: that talked about the decline of project service gross margin.

[00:03:58] [SPEAKER_00]: In the impact, it was having an overall service gross margin for MSPs.

[00:04:04] [SPEAKER_00]: And so yeah, definitely you're going to be, you want to be, you want your main business models

[00:04:09] [SPEAKER_00]: manage services.

[00:04:10] [SPEAKER_00]: So you've got to run a very efficient and effective managed service delivery operation.

[00:04:16] [SPEAKER_00]: It's got to be priced or I have etc.

[00:04:18] [SPEAKER_00]: But don't lose sight of the other sources of revenue for your business.

[00:04:22] [SPEAKER_00]: Make sure that you're doing projects really well.

[00:04:27] [SPEAKER_00]: You're maximizing your gross margin there, that your product gross margin is not being overlooked.

[00:04:32] [SPEAKER_00]: That's a significant source of profitability leakage for a lot of MSPs.

[00:04:37] [SPEAKER_00]: It is they are underpricing their products.

[00:04:40] [SPEAKER_00]: So you want to look at the utility of the business.

[00:04:45] [SPEAKER_04]: Totality of the business.

[00:04:47] [SPEAKER_04]: That's doing a lot of heavy lifting there and let's break that out because what Peter's getting at

[00:04:51] [SPEAKER_04]: is when you think about that sophisticated analysis he just did, he's assuming a managed services provider.

[00:04:57] [SPEAKER_04]: Or he's he services company that has a rich collection of revenue that they're working with.

[00:05:02] [SPEAKER_04]: It's not just monthly recurring revenue.

[00:05:05] [SPEAKER_04]: It's also project work.

[00:05:07] [SPEAKER_04]: It's also transactional.

[00:05:09] [SPEAKER_04]: Yes, you may have a good portion of your business be that strong monthly recurring revenue.

[00:05:14] [SPEAKER_04]: But it's not everything.

[00:05:15] [SPEAKER_04]: And oftentimes too often when I'm talking particularly this smaller or less sophisticated MSPs,

[00:05:22] [SPEAKER_04]: they're super focused on that.

[00:05:24] [SPEAKER_04]: And it's clearly a failure point getting to a sizable business that has this mix of revenue is really

[00:05:31] [SPEAKER_04]: important.

[00:05:32] [SPEAKER_04]: And ultimately the more I think about that it's the other two bits that the customer really wants

[00:05:38] [SPEAKER_04]: to think about.

[00:05:40] [SPEAKER_04]: If I think about the core focus of IT services,

[00:05:43] [SPEAKER_04]: it's about providing insight and advice to customers to better their business through technology.

[00:05:51] [SPEAKER_04]: And remember I'm not particularly enamored of cybersecurity when I think about that.

[00:05:56] [SPEAKER_04]: Cybersecurity is built into delivering an effective technology stack,

[00:06:02] [SPEAKER_04]: but it's not a lead thing if I'm a small business owner and I am.

[00:06:07] [SPEAKER_04]: I'm not necessarily going out there and saying the thing today that I want to spend my money on

[00:06:12] [SPEAKER_04]: making my environment a little bit more secure, no I want to spend $1 to get a buck 25

[00:06:18] [SPEAKER_04]: back or a buck 50 or maybe two or four or $10 back on every bit that I spent.

[00:06:24] [SPEAKER_04]: And I want to work with someone who's thinking that way too.

[00:06:29] [SPEAKER_04]: Now I want to use some of the buzz that we're thinking about right now as an example.

[00:06:34] [SPEAKER_04]: Now I recently had Paul on Gionie on the show to talk about some of the feedback that he's

[00:06:40] [SPEAKER_04]: getting from customers.

[00:06:42] [SPEAKER_04]: Let's listen again to what Paul had to say.

[00:06:47] [SPEAKER_02]: There was certainly a lot of interest and I think a lot not a lot of knowledge, right?

[00:06:52] [SPEAKER_02]: And I think there are at least for my experience a lot of misconceptions about what AI is and

[00:06:58] [SPEAKER_02]: what AI is not.

[00:06:59] [SPEAKER_02]: I think product that are branded as AI that are not.

[00:07:03] [SPEAKER_02]: And so certainly we're using AI tools to be more efficient and trying to help our customers make

[00:07:13] [SPEAKER_02]: good decisions about how to leverage AI.

[00:07:17] [SPEAKER_02]: But in my mind, it's still so new and I think there's a lot of potential and a lot of companies

[00:07:24] [SPEAKER_02]: trying to sell things that aren't really AI and aren't necessarily going to help them in the way they

[00:07:31] [SPEAKER_04]: want it to. Now Paul's answer is exactly what I hear over and over and over again when

[00:07:39] [SPEAKER_04]: I talk to solution providers. They're giving me the feedback of yes we've customers that are

[00:07:43] [SPEAKER_04]: interested there's a lot of buzz particularly around AI because that's what we're talking about now

[00:07:48] [SPEAKER_04]: and they're looking to me to help out with that, me being an IT services company.

[00:07:54] [SPEAKER_04]: Now that's the problem we really want to solve. How can we apply the new ideas, new innovations and

[00:08:04] [SPEAKER_04]: more and more and more and more and more they drive to better business outcomes?

[00:08:08] [SPEAKER_04]: Now what I've said before the last time I put together my what I would do if I was building an MSP

[00:08:13] [SPEAKER_04]: video or if I was starting an IT services company today. The medically I'm not necessarily going

[00:08:19] [SPEAKER_04]: to fall out. I'm building on what I presented before. I'm very interested in that productivity

[00:08:24] [SPEAKER_04]: leader, right? I think that the difference here is making a difference around delivering value.

[00:08:30] [SPEAKER_04]: How can you improve their business? How can you work in their production layer? The stuff

[00:08:36] [SPEAKER_04]: that makes the money, their intellectual property, how can you work there and how can you help them

[00:08:41] [SPEAKER_04]: improve it? Can they get better business outcomes? Can they make more money? Can they save money

[00:08:45] [SPEAKER_04]: on delivery? Can they expand into new areas? Can they deliver new technology ideas? Can they make

[00:08:50] [SPEAKER_04]: their people more efficient? Can they get more work with the same number of people? Can they make

[00:08:56] [SPEAKER_04]: their people a little bit more competitive? I think there's an area of doing all of that and it's

[00:09:01] [SPEAKER_04]: working in the productivity layer. I think that portion of the stuff that I've talked about prior

[00:09:06] [SPEAKER_04]: is still very relevant but there's something new here as well. And I want to break this out by

[00:09:13] [SPEAKER_04]: highlighting from a recent interview with Yasser Ali. Now here I were talking about AI governance

[00:09:19] [SPEAKER_04]: in the context of working with business. Again, let's pull from that and listen and talk more.

[00:09:25] [SPEAKER_01]: The AI governance starts becoming a very relevant topic around what information

[00:09:32] [SPEAKER_01]: classification do you have? That is going into third-party tools at one point. On the other side,

[00:09:39] [SPEAKER_01]: how the prompt is being used by the end user to get that information and ask a question,

[00:09:45] [SPEAKER_01]: are we at risk of disclosing more than we should have? And what we have seen with

[00:09:50] [SPEAKER_01]: Jajipiti prompt engineering prompt attacks, that's not necessarily a safe bed that once you have

[00:09:57] [SPEAKER_01]: a prompt, it's safe. There is a lot of hidden risks available and it'll take some time for those

[00:10:04] [SPEAKER_01]: to be resolved. But even without those, understanding what information, what people have access to

[00:10:11] [SPEAKER_01]: from a business context or risk context or sensitivity context is the problem where

[00:10:17] [SPEAKER_01]: data security and DLP is going to be forefront of any organizations looking to adopt the AI?

[00:10:24] [SPEAKER_04]: Now that's the insight there. The idea that if we're going to talk about using these new technologies

[00:10:29] [SPEAKER_04]: like an AI, that there's a lot of governance and there's a lot of potential risks there. And the

[00:10:34] [SPEAKER_04]: new twist for me, the new development over what I'd said previously about starting an MSP

[00:10:40] [SPEAKER_04]: is this idea of data management. Two often providers talk about data backup and sophisticated ones

[00:10:48] [SPEAKER_04]: talk about compliance. But the real frontier, it's data management. The idea here is is that you really

[00:10:57] [SPEAKER_04]: do have to spend some time working with customers and putting the data together in a way that

[00:11:02] [SPEAKER_04]: makes sense. Make sure they're only collecting what's relevant, make sure they understand what

[00:11:07] [SPEAKER_04]: their collect and why not collecting too much data and making sure that it's all management

[00:11:14] [SPEAKER_04]: maintained in a proper and diligent way. Now this is something that's really necessary. It helps

[00:11:19] [SPEAKER_04]: with compliance. It also reduces your cyber risk. It's really good for liability and for privacy

[00:11:27] [SPEAKER_04]: laws. By putting time in on data management, you're helping organizations with a lot of relevant

[00:11:35] [SPEAKER_04]: needs right now. And it's a precursor for artificial intelligence anyway. You don't just need your

[00:11:42] [SPEAKER_04]: data to be organized. You need to make sure you have the right data not too much of it. And make sure

[00:11:47] [SPEAKER_04]: you understand what's in there so you can determine who gets access to what. So now where I think

[00:11:55] [SPEAKER_04]: starting an MSP and 2024 really comes down to is it's not just about focusing on the productivity

[00:12:02] [SPEAKER_04]: layer, but it's also understanding the data needs of the organization. And every organization has data

[00:12:09] [SPEAKER_04]: needs. I think these two work together in concert to make you really effective and deliver some

[00:12:17] [SPEAKER_04]: business value. If you think about automation and helping businesses with their business process automation,

[00:12:23] [SPEAKER_04]: you not only have to have a good how old of the productivity layer, but you also have to make

[00:12:28] [SPEAKER_04]: sure that data is there to empower them. You've got to make sure that you're presenting the right

[00:12:32] [SPEAKER_04]: data for analysis to be done. If we're going to have any conversations about what artificial

[00:12:37] [SPEAKER_04]: intelligence might do for a business, we have to make sure that we can extract good data from the

[00:12:43] [SPEAKER_04]: organization anyway. Now, it doesn't necessarily have to be amazingly organized, but we do have

[00:12:50] [SPEAKER_04]: to make sure that we understand what's there. And oftentimes businesses don't even know what data

[00:12:55] [SPEAKER_04]: they have. So for me, data management is the new area. It's the area that I'm really focused on

[00:13:01] [SPEAKER_04]: thinking about as the area we want to focus on. So if I were starting an MSP today, it would be focusing

[00:13:08] [SPEAKER_04]: not just on the productivity layer, but also on data management and data governance, particularly

[00:13:15] [SPEAKER_04]: as a precursor to helping them with artificial intelligence. And I would be leveraging a lot of

[00:13:22] [SPEAKER_04]: the frameworks and organizational structures that are out there that are being published already.

[00:13:27] [SPEAKER_04]: A lot of the larger organizations have already put together these ideas about where you should be thinking

[00:13:32] [SPEAKER_04]: about the ethics and frameworks are of artificial intelligence. Even removing artificial

[00:13:37] [SPEAKER_04]: intelligence itself, the frameworks are incredibly useful. And you can build an entire practice

[00:13:43] [SPEAKER_04]: about working with customers to make sure that they are prepared for this. And you get value

[00:13:49] [SPEAKER_04]: even if you don't layer in advance stuff. And we know watching the market that there may be

[00:13:55] [SPEAKER_04]: a lot of fake AI stuff that isn't necessarily where we're looking for. We can get real value now

[00:14:01] [SPEAKER_04]: out of organizing understanding and being ready with the data that we have. And I think this is the

[00:14:06] [SPEAKER_04]: area that if I was starting a new MSP today, that I would focus my time on. Now I did this

[00:14:14] [SPEAKER_04]: sort of live and rift a little bit and I would certainly love to hear from listeners. If you've

[00:14:19] [SPEAKER_04]: got a comment, put it in chat if you're watching live or put it in the comments below if you're

[00:14:25] [SPEAKER_04]: watching the recording. It's an area that we certainly can talk about. But if you've got a question

[00:14:31] [SPEAKER_04]: on something else, I'm also here to take for it. If you're short for one of my shows and bring

[00:14:36] [SPEAKER_04]: your question live, you'll get a live response. Really do enjoy taking the questions and answering them.

[00:14:44] [SPEAKER_04]: Of course, if you want to submit it ahead of time, you certainly can. And Q&A is a chance for you

[00:14:50] [SPEAKER_04]: to get involved with the show. Now I will be keeping an eye out for any last minute questions.

[00:14:56] [SPEAKER_04]: I haven't seen any yet. So instead, I'm going to give you a preview. Can you predict employee turnover

[00:15:04] [SPEAKER_04]: by using AI? I spoke with Tyler Hockman and now he's the CEO of four enterprise. And he

[00:15:11] [SPEAKER_04]: discussed their AI workforce analytics platform that predicts employee turnover before employees

[00:15:18] [SPEAKER_04]: themselves are aware of their intentions to leave. Now the technology uses a combination of external

[00:15:25] [SPEAKER_04]: internal data sources to make those predictions. Let me give you a preview of that interview.

[00:15:31] [SPEAKER_04]: We're going to start with it within overview of this because you've got a technology

[00:15:36] [SPEAKER_03]: that's predicting employee turnover. We take into account a variety of external and internal

[00:15:44] [SPEAKER_03]: data sources in order to create an amalgamation of data that can then be used to predict both

[00:15:51] [SPEAKER_03]: on an aggregate basis and an individual basis whether or not people stay or leave in a business.

[00:16:00] [SPEAKER_04]: OK, tell me what kind of data are you using to do the predictions? We can predict externally

[00:16:05] [SPEAKER_03]: but if you couple that with internal data and this is data that we're very cautious about privacy

[00:16:12] [SPEAKER_03]: and so this is data that every employee knows is being collected, everything from your utilization

[00:16:17] [SPEAKER_03]: into your schedule adherence. You know the basis of what forms your job performance, we can use that

[00:16:23] [SPEAKER_03]: coupled with our external data to then determine you're likely to have a sane or leave in that

[00:16:27] [SPEAKER_04]: your business. Now that's a preview my Patreon supporters already have this video. If you're interested

[00:16:34] [SPEAKER_04]: and you can listen now by being a supporter on Patreon, patreon.com slash MSB radio, it will drop

[00:16:41] [SPEAKER_04]: for everybody on the weekend on YouTube and the podcast feed. So if you're interested I encourage

[00:16:46] [SPEAKER_04]: you to go listen. Now we have gotten a comment, I want to go ahead and talk about that a little bit.

[00:16:53] [SPEAKER_04]: This listener feels you can predict and get an OK estimate for the numbers by using AI.

[00:16:58] [SPEAKER_04]: I would agree with you. In fact one of the things that I think is most interesting about this space

[00:17:03] [SPEAKER_04]: of what we're able to do with some of the new incoming AI tools is do better data analysis.

[00:17:09] [SPEAKER_04]: If you're regular listener of the show, you know I've been reporting on a number of breakthroughs

[00:17:14] [SPEAKER_04]: particularly recently about research into disease detection that is exactly this idea

[00:17:19] [SPEAKER_04]: that you can do predictions and get much better ideas of outcomes by applying AI. I am 100%

[00:17:27] [SPEAKER_04]: with you on this idea. But a pre-cursor to doing it well is exactly what I was talking about earlier

[00:17:33] [SPEAKER_04]: and it's a run data preparation. And I think the vast majority of those audiences are not

[00:17:41] [SPEAKER_04]: in fact for those that have not been able to come out today at 5 pm on today's show of the

[00:17:46] [SPEAKER_04]: business of tech. I actually talk about how only 17% of enterprises have an enterprise wide plan

[00:17:53] [SPEAKER_04]: for AI. That's not necessarily a bad thing for those of us that are in the advice business

[00:17:59] [SPEAKER_04]: because it means there's a lot of customer looking and just like you're right saying you can predict

[00:18:04] [SPEAKER_04]: and get good estimates for numbers with AI in order to do it you've got to get them ready from a

[00:18:10] [SPEAKER_04]: perspective. So I'm 100% with you and it's why my thinking is going the way that it is. I'd love

[00:18:16] [SPEAKER_04]: to hear from you if you've got additional thoughts and comments do go ahead and put them in chat

[00:18:21] [SPEAKER_04]: if you're watching the recording put them in and I'll respond to them. Now, I do want to thank

[00:18:27] [SPEAKER_04]: sales builder who's our Patreon sponsor who makes this show possible. Focus on your IT sales

[00:18:33] [SPEAKER_04]: workflow with the power of automation and visit them at salesbuilder.com that's BUILDR.com.

[00:18:41] [SPEAKER_04]: And vendors, you too can get your name mentioned on the live show. It's a simple monthly

[00:18:46] [SPEAKER_04]: subscription visit patreon.com slash MSB radio for more. And listeners, the number one thing

[00:18:53] [SPEAKER_04]: you can do to support the show is share it with a colleague like Sharon follow in your favorite platforms

[00:18:59] [SPEAKER_04]: but make sure to send it to somebody who you think gets a little bit of value at it. And if you

[00:19:04] [SPEAKER_04]: want to support us directly, you can visit our Patreon with our give what you want model. You set

[00:19:09] [SPEAKER_04]: with the content as worth and you get early access including that interview that I just talked about.

[00:19:15] [SPEAKER_04]: If you've got a question and you're listening to the recording send it in at question at MSB radio.com.

[00:19:22] [SPEAKER_04]: Really appreciate the discussion. I hope this has been valuable for you and I'll continue to

[00:19:26] [SPEAKER_04]: iterate on this idea. This one with a little bit of comments is going to end up being a full editorial.

[00:19:32] [SPEAKER_04]: I look forward to the next conversation and thanks for joining me on the Business of Tech Lounge.

[00:19:38] [SPEAKER_04]: I will see you next time.