Welcome To Business Unscripted

Dave 0:00

And it has been a minute. Welcome to the Business Unscripted Podcast. We're here to share real life insights, practical strategies, and the honest lessons that we have learned from our own mistakes. Because frankly, we have been in your shoes. So whether you need help with operations, accountability, financial knowledge, sales, or maybe just getting your mindset right, we are here and it's the right spot for no fluff conversations and tools help for you. So grab your favorite cup of Joe. Let's jump into the show. Grab your favorite cup of Joe. Let's jump into the show, ladies and gentlemen. It is another Friday morning, and we're here with the Business Unscripted Podcast. I'm Dave Warden. I'm here with Dwarren Burnhang. We are entrepreneurs just like you. I don't know, maybe you're an aspiring business owner, maybe you've been a business owner for a little while now, and you're in the right spot if you're looking for somebody or a place where you can maybe learn some lessons, learn some of our downfalls, and try not to do them yourself and ways that we're helping others overcome those as well. So welcome. And like I said, let's jump into the show. Duarn, welcome back, sir. I know you've been you've been off for a couple weeks doing some other things. So how are you?

Duarne 1:04

Good, good. Life, life has its moments, right?

Dave 1:08

Flies by quick, doesn't it?

Duarne 1:10

It does. It's the longest thing you'll ever endure. And it's also probably the most endurable thing that you can uh experience as well, in my opinion.

Dave 1:20

Yeah. Uh it's crazy how you know it just feels for me about like you know two weeks ago. Yeah, and now we're now here we are. And it just especially when you're when you're busy and you're doing a lot of things day to day, and then it's like where did where do my where do my 13 hours go?

Duarne 1:36

Oh, I hardly get it, right?

Dave 1:38

Yeah, let me I do you ever feel like crazy? Because I I'll I'll work like 13 hours and I'm at the desk and I'm I'm doing different things, I'll have meetings with clients and you know doing stuff on the softwares and and the applications for them as well. And then I at the end of the day, I'm like, man, what did I what happened today? Like, what what did I actually get done today? You know, because you forget you do a lot and you forget a lot of things. You ever you ever get that feeling?

Duarne 2:03

I do, I do, and I actually like it was yesterday, I think it was. My uh Paige was working in the office with me last night, and she goes, Hun, I'm done. I'm gonna head inside, it's eight o'clock. Yeah, and I'm like, Oh, okay, didn't realize you're so late. And I think I stayed until like 9:15 or something, and then I joined. I was trying to finish something off, and it didn't feel like that. It just you just you're in your zone, you're doing your thing. And it's a little bit different to, I guess, manual labor style work, but you definitely feel it because there's an exhaustion level that comes on. But when you're sitting in an air-conditioned office or a heated room, working away, just focused, building, answering different things, and multitasking constantly, it just feels like you're never getting through the mountain of work that there is in front of you.

Dave 2:59

And well, well, for me, like one of the things that I've started to do is actually at the end of every day, I have it actually say, like, hey, what do we what do we complete today? You know, one of those things that kind of give you a little boost at the end of the day. You may you may feel like you've done a lot, like you had a lot of client meetings,

Where Did The Day Go?

Dave 3:10

you had all this, and so it's like the end of the day sort of wrap-up of, hey, here's what you started, here's what we accomplished, here's the improvements you made, whether it's on an app, a client app, you know, client meetings, here's your way that you help them. It's a great way to kind of keep that motivation going into the next day so that you want to continue to do it. Because it can be overwhelming to like at the end of the day, you're like, man, I worked 13 hours again today. What did I do? Right? Nothing happened. Or the next day, and the next day, it's like you have to build yourself up, you have to give yourself sort of those little kicks in the butt some time to continue to move your process, your goals forward. And I think having those systems in place, like AI, to that can actually look at your notes, look at your code, whatever sessions you did with them, and just have a simple end of the day. Hey, at 5 p.m., take a look at all the sessions we started, what we wrapped up, look at my client meetings today, and just send me a quick report of what I accomplished today, what uh what I helped other people with today. And what's great is that's that boost going into the next day. I think it's huge.

Duarne 4:08

That's really cool, yeah. Because you know, I read somewhere a few years ago that multitasking is actually one of the most exhausting ways of working. And it's just so normal for me, and I'm sure for you too. It it's people who tend to focus on one thing and finish it, then stop, move to the next, focus, finish it, tend to be less exhausted, mentally exhausted at the end of the day than people who jump between tasks.

Dave 4:40

I feel that. Like I I also I I wonder probably at least one or two times a week. Like, I wonder if I just slowed down and instead of having 17 sessions going on the application, I just focused on like two. What would that do? But ADHD Brain Me, right? Definitely is something comes up that's like, oh, I gotta start a new session for that, or I'm gonna do this. And especially when you're trying to start something from scratch, like there's so many things going on and trying to clean things up and move things forward. And so I think where I'm at right now is I think it what the command center that we've been building, I I think it's in a good spot to where I can invite some people in to actually start testing it on that side to get the feedback instead of trying to make it like perfect. Like, hey guys, like there may be some pixels that are off, or there may be some things that maybe not be formatted correctly right now. Just let me know, I'll fix them. I think that is the next step because I could sit at the thing, actually looking at my phone. Oh man, I don't like the way that this looks on mobile, and then you have the freaking iPhone duo that comes out now, and now I gotta reformat my entire application on mobile to be able to work on the duo because I know that I'm gonna get it. So I'm gonna know that I'm gonna be on the phone. So it's like, damn it, there's another thing I gotta think of now. But I think you know, I've also I've like I said, I've always wondered like what would happen if I just like pulled myself back and just kind of tried to do two or three sessions at a time instead of 17.

Duarne 6:13

I do that sometimes. I actually do that sometimes. So I had to do it this week because you know, I got sick on over the weekend because just a very, very long week and a half.

Dave 6:26

Long, drawn out week, yeah, yeah, for sure.

Duarne 6:28

Yeah, we had a client over, so there was you know, a lot of work with that client, you know, so he felt value while he was over as with the time he had here, and you know, just making sure I was there and with him as well and present. But as part of that as well, like I I I took a day off, I took two days off this week, uh just sick in bed, and I felt really guilty for it. And I thought I could lay in bed and I could do something work-related. You know what? I'm just gonna keep on top of my email from my phone. That's enough. And I'm gonna watch the new season of The Gentleman. So bugger it. That's what I'm doing.

Dave 7:10

There you go.

Duarne 7:11

And you know, Joey Richie's series. It's it's fun, it's a little bit outrageous.

Dave 7:14

I mean, I think a lot of people have that feeling, right? Where you know, it's just like I'm sitting like last night, right? I was watching the Cleveland Browns, and I'm just like, I have to do something. Like, I feel like I'm I'm missing out on awake time when when I have a lot of goals, right? That are or a lot of things that are on the agenda that I need to accomplish, not just for myself, but also right with my clients. And so I feel like I might be letting them down or letting myself down and getting behind. And plus the last two weeks, like we talked pre-show, the last two weeks, I have felt there's just been some things going on that I can't control that I've had to help out with to get other people caught up and do things for them. Then I feel like I'm behind them trying to catch up. So I think for me, like a shift it really comes down to like for your situation, what works. And you kind of have to make it like you watching, right? It's gonna be a lot different in your situation than what Dorne or I might do in our situation. So for me, it's like I my decision is I've been working till seven, eight, sometimes nine o'clock at night. And then when I I'm missing, I'm missing right my baby, you know, because she's in bed. And you know, by eight or nine, she goes to bed. She's only nine months old. So for me, I'm like, well, I'd rather maybe let's start waking up, you know, set my alarm at four, four thirty. Let's get up, you know, by five, get into the office earlier so I don't feel as guilty leaving at maybe six, six thirty, and I can catch some time with her on the way home. So that's been my goal the last week and a half, and I think it's been really well with some of these other things that have been happening. But that's what I'm gonna try doing because it's I I miss that time. Like I want to, I want that time back. And because you know, we you're your two kids are grown too, you know, starting to grow it up. Well, that's once you miss it, they're you know, it's gone, so you don't get it back.

Duarne 8:56

So absolutely, you know, I mean I've gone into the room many, many times, and my two-year-old's already in bed by the time I leave the office, right? Okay, there's wait, what is it?

Dave 9:08

What would you say that? Is it board borgor? Borgor or Bourgore? I don't know, whatever. Hello, sir. Welcome, Borger or lady.

Duarne 9:14

Hi there.

Dave 9:19

We will answer them for you. But yeah, go ahead, buddy. Sorry.

Duarne 9:23

Yeah, no, and and you know, and I sit there and I go, I want to just check in and see see how she's doing, but she's asleep, right? So I have to wait until the next morning to catch up with her, and I try and catch up with her before I start working, you know, or being well. My other daughter, on the other hand, seems to stay like staying up until midnight, one o'clock in the morning, two o'clock sometimes. So I definitely get plenty of time to spend with her, the little one-year-old, because sleep is not something she's really that interested in at this point. So I think there's that balance you have to have, but like you talk about not getting everything done. I've got what do we got here? Yeah, I can't even here we go. No, I can't reach it. There's hang on, there we go. I gotta pile these mini PCs, right? Yeah, pile of these things sitting on my desk. One, two, three, four, five, six, seven, eight. Now, these are all meant to be turned into AI agent teams.

Dave 10:37

Right.

Duarne 10:37

And so far, I built out four teams for the process, and I just have to build the machines now. Now, all of these are dependent on my room next door here being ready to start putting them in and posting them next door. Right. And I'm taking actions towards getting that done, but I feel like my feet are dragging because it's not ready yet. And the reality is life and work have to be completed in between. And I think it's gonna be I've got to fit in this work, but at what cost? What do I have to sacrifice to be able to get that extra work that I want to achieve in place?

Dave 11:27

And I I think it's a real it's sometimes like a reality reminder, like not like a reminder, I guess it is a reminder for ourselves that we have to everything outside of our world is moving so quick, right? Like AI changes so quickly, there's a lot of new announcements, new new things. And so when we're in our offices, we're like, man, we gotta move at lightning speed just to keep up. And I think we kind of have to remind ourselves that like we're not those teams, right? They have billions of dollars that they're playing with, but they have uh way more things in than any AI team could ever help you do. And so you have to compare you can't compare yourself to that. Like you have to really compare yourself to like the other small business owners or other, even maybe even medium sized, depending on what you're trying to grow to, and say, am I staying ahead of that game? Not am I staying ahead of the AI game, because that AI game is always gonna be lightnings ahead of where you're at and where you're gonna grow.

Duarne 12:25

And and I think the biggest piece is realistic with what your goal is, which a line in the sand is like for me, correct? It's about making sure that my team have enough work to keep them busy and working on projects moving forward, that

Daily Wrap-Ups And Task Switching

Duarne 12:40

my clients' projects are all up to date and on target, that I have our next generation of products and services planned out and in action to be built. There's just you know, there's always something, and this is why I I've never really had that feeling of like, oh, I'm bored when I go to work or when I start working. For me, it's always like, all right, I could take a day off from the office, but I know there's still gonna be work there for me to do when I get to the office. I lost a week that I lost two days sick this week, plus I first day back, I spent the whole day dealing with an unexpected client-related concern, which I had to get onto because it was a very important client. And right at the end of the day, when I'm feeling very drained, I get a message from another client saying, You've been shortlisted for a project, you're one of two. Right. And like that sort of brings you right back into the mix and go, Oh, you know what? This is why I do it. Yeah, this is exciting. I'm happy it gone.

Dave 13:44

And I think you just have to find like your why, as a lot of people say, right? You gotta find your purpose. And I think if you're strictly in it for the money, you're you're literally it you're gonna get overwhelmed and you're gonna get exhausted, and you're gonna get worn out. I think ultimately you have to figure out like what is it that is your driving force, and it can't just be the money. Like, what is it?

Duarne 14:06

It's emotionally breaking if it's just the money. Because you're always feeling you always have a pressure on yourself to try to grow and get more money, like absolutely, and you know, that one client that's gonna leave, and you have to find the balance. How much time do I spend keeping them happy? How much time do I spend with them? How much, you know, do I neglect my other clients to make sure that this client's got the right amount of service being offered to them? Yeah, there's that there's all these balances that have to be put in place. So, you know, and at some point, I mean, like we've had that realization, like my team, I'm really happy with my director of ops. He's doing a great job. He's got the team you know rowing in the right direction. We've got a steady plan to bring AI into the into our team, whether it's through agents and coding assistants. And it's happening, it's all happening, and it's not happening at ridiculous speeds, it's not happening straight away, but it's all happening, and I'm seeing quality results. And we're now at the point where I'm going, okay, unlike a lot of other businesses, we're ready to bring on another two developers into our team, like physical real people developers. Where a lot of other companies at the moment are going, we're not hiring a developer, we're gonna just use AI. Well, we're like, well, no, we've got a lot of stuff that we want to do. We're gonna bring another developer into the mix to use that AI and build more and make more, which I think is great. And it's funny because from a technical aspect level, from some of our other clients who work in the MSP space, there's a product that got released recently in the last I don't know, a couple of months, and it's a level one tech support agent, which basically does a lot of triage, follow up, auto, instant replying for level one related tickets. And it's not the cheapest, but it's still probably equivalent of running two or three people because of one, right? And I was looking at going, that's really cool. I think it's gonna have a bit of pickup with a certain type of customer, and then there's gonna be the other type of customer go, I'd rather have people deal with that stuff. There's a personal aspect to it I want, and I want to have people deal with that. So I think we're always gonna have that split where one side's gonna say AI combine you know, hybrid or fully, and the other side's goes, I'm just happy just doing with people doing that, with maybe with a little bit of AI if they have to, or automation, but I want people involved in that.

unknown 17:22

Right.

Dave 17:23

I think I think we're always I don't think we're I mean, I hope we never get 100% AI because then it'll be like what the Terminator if we if we're ever get there. I don't think it'll ever be there. I I tell my clients, any of myself, you have you're always gonna have to have some sort of human integration involvement to make sure it's going in the right direction. Even if it's something that is working autonomously, you want to have some sort of audit trail. You want to have some sort of process that a human is looking at the output or looking at the audit or looking at you know, findings, whatever it is, to analyze that. That is experience. It's not just like, oh, hey, Jane, uh, by the way, can you just run that and just tell me, you know, tell me if it if it's working? Like, yeah, somebody without knowledge of what it should be doing is gonna say, yeah, it's working. It's it's got it's done 17 runs. Like, is it active? Is it is it uh saying what it should be saying? Like, is it actually generating results? And I think that's that's the part uh that I think people overlook is that just because you you put out like let's say like an outreach engine or an outreach agent uh to have conversations, uh well, how many of those conversations are actually generating right are actually uh coming to fruition? You know, because if not, then maybe you need to change your messaging, maybe you need to you know maybe change your approach, change your ICP. I don't know, but you have to do something, every situation's different. And if you just set it and forget it, you're never you know, you know, that's what I think a lot of people are into right now. Well, I tried that. I'm like, okay, how many times did you like review it? What do you mean? Like, once you set it up, how many times did you go back and like look at it and test it and you know make sure that it's doing what you wanted it to, or and have it update? Oh, I never did that. Well, then it's never gonna work. That's like trying to, you know, take the first version of anything you do and expect that it's gonna be perfect. It's never gonna happen.

Duarne 19:10

It's like writing a book, right? I mean, most people tend to write a book and then rewrite chapters, rewrite paragraphs, remove chapters, change different bits and pieces before it even goes to a publisher. And that's normal. And people just assume that you can do, you know, and it and I think a lot of this to do with the hype from a lot of content creators out there is one-shot code vibe coding, one-shot prompting, you know, just trying to get it right the first time and without having to do any feedback. And this is something I find interesting because it like Claude, for example, asks a lot of questions when you try and make something like it forces you to answer questions, right? Codex is doing it more now, too, in the latest versions. I actually don't mind it because a lot of the time it makes you think about something you may not have thought about before. And I like I have always described AI as an over-eager intern. Like, give it the instructions, it'll run away and do it, but it's like cutting a sandwich in half. We've probably all seen that, you know, video where the guy's like, hey, cut a sandwich in half, and then it instead of cutting it straight in the middle, it goes and cuts it through the middle of the sandwich, or it tries to cut it, you know diagonally, or you know, in the middle.

Dave 20:32

Yeah, you know, like there's so many different ways.

Duarne 20:34

What are you doing? That's not how I'd cut a sandwich. We have assumptions, and I think it's the same with people, right? Like we're talking about earlier, like one of my techniques when I have a staff member not doing something that or you know, they're not performing as well as I thought they may have performed on a task that I assigned them. I find myself asking a lot of questions and probing and getting on a call and watching them do something and realizing, oh shit, you're doing it that way. Cool. Did you think about doing it this way? Let me show you how I do it. And then you see that light bulb moment for them, and then suddenly you realize, okay, you've got a good way of doing this. I want to learn from you. Right. And

Get Feedback Before Chasing Perfection

Duarne 21:18

you start to see a difference in the way they do output. Now, AI is exactly the same. I do the exactly the same thing with AI. I do it with clients. I have clients who are doing something one way, and I say, Have you thought about doing it this way? Oh, I didn't know you could do it that way. Thanks for sharing that.

Dave 21:36

Well, I think that's a part 20 minutes a day off my I know a lot of people miss because I miss it sometimes, right? And and I and I realize, oh man, I'm so dumb. But part of it is that first one or two times, giving it the exact step by step by step. Instead of just saying, hey, I want you to go out and do this, like figure it out for me. Stopping in that first one or two times, literally saying, Here's what I need you to do. I need you to go to this website, pull this report, you know, drop it in here. And then when it's done, say, Oh, and all right, now take that whole process you just did and turn it into a repeatable skill or a system routine, whatever you want to call it. And that I think is the part that sometimes even I miss because I'm just like we talked about earlier, right? F 17 sessions going up, I'm going boom, boom, boom, boom, boom, between them all, and I'm just like, you know, move it. And so even when like looking at the app, I'll give things, I'll give it a screenshot. So some pains of my my life right now. Right. So you're building the app. The mobile version to me is important because I think a lot of people are going to be on their their mobile phone while they're while they're on a command center. So I take a I'll take a screenshot and I'll say, Hey, like there's this issue. Like go fix it. And I highlight the issue. I circle it and I set it back. And then it it says, Oh yeah, we fixed it. Well, on the preview, it looks fine. But then when I go to put it in a prod, you know, production, it's still there. I'm like, it's still there. Like, what are you doing? And I feel like I have to like literally tell it sometimes step by step. And then it's like, oh yeah, I didn't do that part. Sorry.

Duarne 23:03

Like this is why I say there's always gonna be some sort of skill set to do this in future.

Dave 23:10

Yes. And there's always gonna be human involved, I think.

Duarne 23:13

Because oh, and and you know, it's that makes me think of something funny, right? So I've got this client of mine, and I'm watching them work with AI, and they're sitting there and they're like, it's not doing it yet, it's not done yet. It's like it's an agent, just leave it, let it go and do its own.

Dave 23:33

Yeah, you don't have to watch it, right?

Duarne 23:36

Like, and this is what we I was helping them with Grockbot, and they were getting really frustrated. Well, anyways, they got to see me working on my seven screens here, building something for them of a similar sort of environment, and they would see me just set something and then go work on something else, and then come back to it like 40 minutes later or something. Okay, cool, there we go. Make another change, jump back and do something else. And they're looking at me going, This is why you're so calm when you're working with AI. I said, What do you mean? Well, you're not sitting there waiting for it to finish. I said, Oh, hell no. I said, That's like watching someone paint your house or something.

SPEAKER_02 24:19

Yeah, watching a paint dry done in five minutes.

Duarne 24:22

I said, I don't want to do that. I said, No, no, no, I just now now there are there are some things like so if you want something done quickly, I think a lot of people don't even realize they can do this.

Dave 24:33

You have to make sure, like, make sure you're on the right effort level as well. Like, oh yeah. Um, you know, if you if you work on a default that's like high effort level, that means it's it's rechecking itself, it's it's it's bouncing different variations. If you want something just quickly responded to, make sure you change it to like low effort. Doesn't mean you're gonna get, I mean you're not gonna get double checks, you're not gonna get some corrections, but you're gonna get a pretty good result, especially if you want something like quickly. So make sure you're also checking the effort level on your session. If you are gonna sit there and watch it instead of watching paint dry, you know, because I think a lot of people they put it on high sometimes and they're like, why is it taking like 20 minutes to run this? And it's because, well, it's it's rechecking itself, it's double checking, it's doing some variations, and it's gonna analyze which is the best output and then give it to you.

Duarne 25:20

For me, I work a lot in medium. That's what I do, or like medium. If I'm coding, maybe on some stuff I'll do high or very high because I want it to recheck its own work a little bit more often. But I mean, depending on how you build these things out, you can create teams where you can have one guy just smashing code out really, really quickly as an agent, then another one coming along, double checking, QAing it who has a higher setting, and then sending that work back to the original guy to go and bang out again, and that's probably more accurate to an actual person doing coding, right? They're not putting as much QA and thought into it initially, they're just banging it out to get it out there and make it function and work. Then the QA tells them what's wrong and then they go and fix it. But I mean, it goes with life too, right? I mean, how many times? And it was something I saw when I was watching someone using one of my clients using AI, and they're like getting really frustrated because it was making it asking for stuff, it was asking questions, and it was asking for direction, and it was coming up giving the wrong answer. And I'm like, ask it this, and I'd ask it a question to verify its detail and uh on how it came to the conclusion it did, and basically show your working, tell me how you did it. And whilst it was showing it's working, I was like, There's your problem, it's it's can assumed that tell it not to assume that, and then suddenly when they didn't assume that, oh my gosh, every that's what that's the result I wanted. So they just have to learn how to communicate better. You can't just assume it's gonna know exactly what you want.

Dave 27:01

That's like like you like you already talked about, right? It's it's it's it's a it's an individual intern. How would you communicate to a new employee the same way? Like if the employees are giving the output, okay, how do I instruct the employee to give me what I want? And that's essentially how you have to look at your AI, your team, your session, whatever you're whatever you're building.

Duarne 27:19

So yeah, I mean it's like if you had a special custom order at the at a restaurant and you need your meal cooked a very certain way or a particular way, and you just said, hey, give me a s, you know, give me a steak and veg. It's like, how would you like your steak cooked? Great. What type of cut of steak would you like? What sauce would you like? Would you like the sauces be poured over or would you like it separate? What sort of vegetables would you like? Would you like those? Would you like your potatoes fries, roasted or grilled? I mean, we we accept we we expect in life certain levels of questions for different scenarios, and I think people just overlook it sometimes to go, it's AI, it should be quick. And sometimes it's like, well, no, you want a better result, and then you just answer a couple of questions. And we're answering questions in our head all the time, even if it's not out loud. It's just when we have to do it out loud, it feels like work, it feels like effort. And that's my I think that's the biggest thing I've seen over the last month is watching people working with agents. They just assume these AI agents are going to do magic work for them without the instructions, without the skills, without actually giving them feedback. And I've gotten to the point now where this client who was a bit was interesting to watch them working with AI, they're now going through and saying, Hey, no, what I meant was this. Right. And I'm listening to them on calls, and then I'm like, good, that's the right feedback. And they'll come to me and go, hey, by the way, they it got I got this weird output earlier. This was happening, so I told it to do this, and it seems to be working now, and I told it to save that as a skill. Good. That's the what you should be doing. So as people start to learn with that, they start to realize it's just like training a person, it's just like training a person to do the job. The only difference is they probably have a better memory in an AI than a person's gonna have. They won't deviate as often if you set them up with the right skill set and the guardrails in place to do it. But it's it's all it comes down to how much effort are you willing to put into the training of your team, whether it's AI, whether it's an a person, how much time are you willing to put into it to get it right? And we've witnessed this with you know, staff, virtual assistants who have been assigned to clients who don't put time into training and then sit there and complain that they're not getting the results they want. And that's like, well, what training did you do in relation to that? Well, I just figured they would know how to do it. Did you know how to do it the first time you tried?

Dave 30:36

Well, no, but I think I think the problem with that piece is twofold. One, they haven't even thought through any sort of specific repeatable process that they want their VA to

Capacity, Delegation And Time Guilt

Dave 30:52

do before they bring somebody on. And two, they've waited until they're already overworked and have no time in order to try to bring somebody on to help, which is why I always suggest you need to understand a good 70-80% of your capacity, and that is when you should be bringing somebody in and be able to pass that stuff off because then instead of being at 100% and trying to get back down to 80%, you can get at 80% and free up and maybe drop down to 50% to do more to actually grow your business. And I also think it comes down to a lot of people don't understand what those metrics are when they need to bring somebody new on. Right? You know, I'm I'll I'm gonna have a meeting with a new client soon, and they're they're they they they're blown up. They went from like zero in 2024 and they're gonna be like four million dollars this year. And so part of their stuff is we don't know when we need to bring out a new, you know, sort of next level staff member. Honey, I love you. So thank you. And part of that is is setting up those metrics to realize not even when you're 80%, but like when your staff gets to be at an 80% capacity level, that's when you should probably bring somebody else on, not when the whole staff that's there is 100%. What's that 80, 85? Now that's that next trigger. What does that look like? And then what's the process?

Duarne 32:19

That's really solid advice. And it's interesting because like one of my management team has probably got to the point where they're probably hit that 80%, and we were asking them to take on a little extra responsibility short term due to a situation that occurred with another staff member. And as part of that, I said, I said, well, how about this? Let me train, let me help you train up an AI agent team. And what I want you to do is just list off all the tasks that would be better if you didn't have to worry about doing them on a regular basis or a very repetitive for you, right? And let's see what a team would look like doing those tasks if possible. And we ended up building out a foundational team concept for him, and he's going back to refine and improve that. But basically, we're talking about a team that can help with about 30 to 40 percent of his daily stuff, which is incredibly helpful for him because that means that he goes from being 80% capacity down to something like a 60% capacity, you know, at a bare minimum, and he can focus on getting you know, spending the time training that team to take on more responsibility, and then bring we can bring on some real individual staff again to help pick up some of the other slack as well.

Dave 33:41

And I think the the biggest like you hear us talking about like teams and stuff like that. I don't I don't think you have to like get started with team. Like if you're listening, it's like, man, agent teams, I don't want agent teams, I haven't even gotten anything done. I think the biggest piece that we've talked about a lot is what are those one or two things that are going to save you an hour or two a week? Those add up, you know, an hour or two weeks, hundred out a hundred hours a year, which is you know, over two and a half weeks. What could you do with two and a half weeks? Some people want that on vacation, but it frees up a lot of your time to then start focusing your attention on some of the other stuff that you should be doing, and then you can grow, compound, build on top of it. But the biggest misconception I think a lot of people have when it comes to building something like you're talking about, or that you know we've been helping people with as well, is the time it takes up front, especially if you don't have those processes documented already. Like if you don't have like the biggest the the best thing you could do for an AI agent is to literally have a documented process in a Word document or a markdown file or whatever. Because then it literally can follow that every single time. But if it's in your head or if it's in your employees' head, you can't get that across to an agent every time. You know, so you got to get those things written out, you got to get them written down. That's the biggest thing you can do. And so there is a lot of upfront time. And then a lot of times I've heard stories of people like, yeah, I just took so much time up trying to get it started. I just I just stopped doing it. It's like, oh, so you just that's like that's like basically saying I put some money in the stock market, I didn't see 20% return, so I stopped investing and took all my money out. And then two weeks later, there's a big stock price you know jump, and you would have made hundreds of thousands of dollars. You know, that's essentially what you're doing every single time you stop training your agent because you're you're just gonna continue to fall behind.

Duarne 35:35

Absolutely. And it's interesting because one of the things you find is people try and chew off too much right away and start with something really small. You know, one of the smallest things you can do, you don't need an agent to do it. Link your email into ChatGPT, Claude, whatever, Gemini, and ask it to run a brief every day of things that need a decision from you, and ask it for a morning email brief which summarizes anything that you know in your email that's critical and needs attention, and then list it down, you know, and have it list off any of the thing that needs your attention. And at some point, what you'll start to see is you wait for that email of the morning, you check it, you respond back to the AI telling it what to do with the rest of it. Prepare a draft for this, this, this, and this, say this, set for this email off to this person, ask them to handle that for me. This one's already been handled via a message chat that I had earlier today or a phone call. And you find like you're just knocking things on the head really quickly. And the big thing is you're not having to open your email to a ton of emails in there, and then take it one. Then once you're happy with that, you can get at one next and go, hey, you know what? Anything that's coming in from this email address, I want you to auto-archive that for me. Or, you know, can you unsubscribe from you know that mail list for me because I don't really find any value from it? Or can you check these three mail lists that I'm on? And you know, and if it's talking about X, Y, or Z, summarize what it's talking about, and then tell me in the email. And if I find it interesting enough, I might go and open the mess, open up that actual email and have a look at it myself. But just little things like that, like that's something they can you can do right away just using acron jobs and scheduled tasks built into your platform, your regular AI platforms without an agent, and that can save you, you know, 10, 20, 30 minutes.

Dave 37:44

Yeah. A day, sometimes, yeah.

Duarne 37:47

Exactly.

Dave 37:49

I mean, so and and the biggest, so I had a question, I was at a networking event Tuesday, and a gentleman asked, he's like, how do I what do I do? How do I just get started? Like and you know, I tell people the two list thing, so I'm not gonna repeat it again. You can watch one of the previous episodes about the two lists that I that I suggest you make. But he's like, I don't know what what it could do for me. I'm like, well, tell me what is what's a lot of stuff you're doing manually right now. And he's like, Well, I gotta send emails, I gotta do follow-ups, I gotta, you know, grab this report and have it run an analysis, and then I get the output, and then I have to drop it in this other system. And I'm like, Well, you just highlighted a bunch of stuff you could potentially do. You know, he's like, What do you mean? I'm like, What your follow-up? How do you follow up? Emails. I said, Well, you could have some sort of AI that is connected to your email that analyzes and knows your contacts, knows who your clients are already, knows who prospects are. And once they, you know, send an email, it can analyze what they're asking. You know, are they asking for a renewal quote? Are they asking for an update? Are they asking you for you know insight into a change that they're possibly thinking of making? And it can even start doing that research for you. So that way when you come to your email box, it says, hey, here's this email from John. John was asking about his insurance, you know, kind of coverages, and here's here's his current coverage, and here's the change he was potentially looking to make. Here's some suggestions, right? And here's here's a response, review it or send it. Right. That that number one, that's one right there that you could potentially do. The the follow-ups on renewals, like uh instead of sitting here and waiting for it to pop up in your system, you could already have it set up where it's like, hey, six weeks out, send an email to somebody or do the analysis ahead of time and start getting ahead of the game. Now you're you're gonna be seen as like, oh my god, that guy never misses a beat because he's uh on renewals, he's given me the information, he's given me updates, he's already looked at other solutions for me to consider. And I have another four weeks to think about it, great. Instead of a lot of times, what happens is it's maybe two or three weeks before renewal, and you're like under pressure to figure out do I want to change, do I want to switch or stay the same? And most people just stay the same because they don't want to switch, even if it goes up. And so he's like, Oh, I never thought about it like that. I'm like, Yeah, but there's a lot of even if you're not doing like the PHI, because I know like insurance or investing, things like that, there's a lot of stuff that there's regulations around, but there's still some stuff that you can do that your day-to-day operations are that don't go necessarily into a client's account or pull information or you're not putting it out there. Or you can invest, like if it's a if you're a smaller company, invest in

Choosing Models And Managing AI Costs

Dave 40:32

a local model, like just invest in a company that can serve everybody on your local machine so you know the information's not going anywhere. And then you could use that to help you run your operations day-to-day instead of trying to use like a claw or you know anything like that, because they're getting better and better and better every single week.

Duarne 40:49

They certainly are, and there's devices out there that you can plug in and use as companion devices to use their localized AI capabilities now. It's getting much, much more powerful, much more interesting, and I think it's really changing. And if you look at like what's happened over OpenAI this last week, got an email saying, hey, your $200 a month plan is halving the amount of usage that we're offering you. Yep. We now have a $500 plan if you'd like to step up to, and it offers this, this, this, this, and this. Oh, we're introducing this new function called dots, but all I'm reading is well, that's disappointing. I'm now only getting 10 times instead of 20 times my usage for the $200 a month.

Dave 41:39

And the five times the $500 is what the $200 used to be at the 25 times.

Duarne 41:45

Exactly.

Dave 41:45

Which is freaking so they raised the price $300 and gave you what you were getting.

Duarne 41:51

I look at that exactly, and I look at that and I get a little bit frustrated, and then I sit back and I go, hold on, what am I actually achieving during the day with this setup that I've got? It's still probably worth it, but if I could do some of this stuff local, I'd much rather do that. I'd rather do local because it's a case of if I build my entire business model around delivering services with myself and my team with multiple costs, multiple licensing costs involved in my solution. If they pull the rug out from underneath me and then double the pricing, then that I've got to be aware that that could happen. And if I'm running local models that can do the majority of what I need and frontier models when I have to for the other portions, I can control my costs a bit better. I get better security, but I can control my costs a bit better as well. And this is something which you know I mentioned to you I'd seen coming for about eight or nine months now, and it was a concern for me. And sure enough, I'm starting to see it happen more and more as you know, Claude pulled support for agents very early on when agents came out, unless you're using a localized Clawed agent in Code Work, for example. But uh external agents are not supported unless you're using API. Grockbot coming out with their agent first solution, great, but you've got to spend the money for it. And you know, it's not their basic plan, doesn't really give you that much. But all the I mean, in the end, I mean when you when you think about it, it it really does come down to for what you get when you start breaking it down. Like you can't get an employee that can do as much as one of these agents can do for a $200 spend.

Dave 43:41

It's just or even I mean when you think about it, like a $500 spend. Like, yes, it sucks that what you're currently paying $200 for is gonna be $500. But when you and sorry, it's uh you get five times more because I think now it's 20 times for the 200, 500 is gonna be 25 times. So you're actually gonna get a little bit more, you know, 20% more than, or sorry, 25% more for $300 more. But even for that, like the output that you could potentially get, if you're using it for productivity, you're using it for things that are saving you time, you're never gonna be able to spend or or get that out of an employee. Like that's like a couple days three or maybe a week out of an employee, and you're getting an entire month that is non-stop, really. And at 25 times, open AI, they're still probably gonna give you some resets. So it's like you get 25 times, but then you also get to like reset it two or three times a month, which you know is actually crazy, right?

Duarne 44:43

You get even more than that, really. Exactly. So they super value, I think they've super got super high value in that instance, and it's all about the perspectives you that we have when we look at those. It's a little different than going to the gas station right now, where what you used to put into your truck, it's about you know 60 to 80 percent more now to get the same amount of gas in your truck, but you can't go any further with it, at least with AI. We can still deliver a lot of results, and we're getting better models that can do more and we they can do it quicker. So less tokens last longer, do more, create more at better higher quality results, and a little bit more autonomy. So I think there's lots of good trade-offs with it, but we do have to be aware that you know local models have some more advantages now, and it's probably worth looking into the investment into the hardware to run some local models, even if you still plan on using these frontier models. Because when you start putting a $500 license on for yourself, it's okay. But what if you've got two or three staff and you want them to be just as productive as you are? Are you getting paid the $200 or the $500 a month to help them get to that level as well? I mean, I've I've done it with my team and I've noticed big results, which has been incredible. Originally, I was like, oh, do I want to spend that money? Are they going to use it? And sure enough, I've been looking at the results and the outputs, and I'm actually really quite impressed at how they're utilizing it.

Dave 47:00

I mean, you hit it, right? It literally comes down to how are you using it? Are you are you paying for it and then you're utilizing it to you know get you better pictures and better images? Okay, that's definitely not the best way to use AI. But some people are doing that right now. It really comes down to your situation, your own processes, and how are you going to utilize it to be more effective. So it'll be interesting to see what happens long term if some of the other companies start following suit, like anthropic claude, things like that. I wonder it'll be interesting how like open router, you know, they start changing their tokens, or if it is it just I wonder if like the API usage is gonna change and go up, or is it just the subscription charges that are going up? It'll be interesting to see what happens there.

Duarne 47:52

Yeah. And I think we're also getting to that point now where there's there's a lot more emphasis on the governance around these models. We're not seeing them jump out as quickly as they have over the last 18 months. Right. Actually, we get they're taking a little bit more time again to come back out, and we're not seeing such great big jumps.

Dave 48:14

Unless the competition, unless competition releases one and then they got to push something out. So it's like 6.1 or whatever for Seoul jumped out like five days after they reached released six. And so it was pretty crazy. But that was the only one recently that I saw like jump out really quickly.

Duarne 48:28

Exactly. And I think it we and I think some of that comes down to also fixing some bugs because they might have dropped it too early, sort of thing, as well, on something. Could be true, yeah. But I mean, at the end of the day, we're we're talking about it's really hard to keep up with all of the different models and the versions and stuff. What was working three months ago is probably still perfectly fine for most people today, and these newer models just make it a little bit better or a little bit more affordable on the token level now as well. But I think you're right. One thing I think people need to be aware of is regardless of what models you're using, just educate yourself a little bit about the light, you know, the low, medium, high, very high professional use cases so you can get the results that match the output because you can save some tokens just by rolling it back. It might automatically say, Hey, do you want to use medium version or high version? And you may be only needing low. I've made the mistake where I've been working on five different tabs at a time. One of them was something I needed on pro. I went back to another one, I started doing simple stuff like writing emails, and it's on pro. I'm like, oh geez, that's a well.

Dave 49:41

Yeah, because they don't change, they just go to by the last setting.

Duarne 49:44

So exactly. And it's like, I don't do that. Quick, roll it back, roll it back. Right. And that gets, you know, so there's lots of different tips and tricks, but one thing I think's interesting is just beware. A lot of YouTube influencers out there are gonna sit there and tell you, this model's the latest, bestest model, it's the best model they've ever released. And then the next model comes along, they necessarily have the same thing.

Dave 50:08

You know, just they're always gonna be better than the than than the previous version now. Like they are, typically. And and that was that was one of the other questions the gentleman had at the networking event. He's like, Well, which one do I know how to use? I'm like, in reality, you're probably gonna use a combination, like, because they're all they're all good for different things, depending on what you're trying to do. But don't try to go for like the newest shiny, it's like start with one and get confident, comfortable with that one, and and you'll be good. Like a lot of people don't necessarily need to have like two or three or four different versions. They can get one, run it, get it going. And even now, like you said, like I mean, I've never I haven't played around with grockbot just because I don't have the time for it because I'm doing everything else. And that was the biggest thing for me was you know, I have to start just picking my lane, and my lane is doing stuff you know, myself, not grockbot. Because grock bot is also like for people, and and and same thing with like Muse or you know, like open I just came out with dots. Those are for people that are like trying to understand agents without understanding agents that don't want to understand the back end. And and they're for simple, simple use cases, they're not being you can't really tag team them into a platform or to multiple services for you, like it's got to be done on their app, in their system, their way. So that's a bit and bad.

Duarne 51:26

Yeah, and it's a bit awkward to work with them if you're used to doing regular agent stuff outside the ecosystem. They do some nice things, but they also they're I I think they're they're good for people who are tipping their toe in and wanting to try agents. It gets them started, and at some point they're gonna be looking for the door, going, well, what's next? How do I get to that next level? And that's when they come out and start looking for things like Hermes

Start With Everyday Business Problems

Duarne 51:52

or you know, systems that are integrated into multi-agent platforms, etc. But I mean, like you've got your platform you're building there, the command center, which is fantastic for most small business owners because it solves all the initial problems they've got. Gets them started with AI without feeling overwhelmed. And I think that's the sort of direction a lot of small businesses probably are better off heading in. Because small businesses are just like us, they are they are so busy with the day-to-day, they don't have time to be dealing with everything else, but you know, learning new technologies.

Dave 52:31

Like I mean, a lot a lot of the pains I hear from networking, prospecting clients is how do I manage my email better? How do I make sure I'm following up with people? How do I, you know, maybe propose, you know, proposals in a more timely fashion, get pricing back? How do I, you know, do more social media? How do I do more posting? How do I do more outreach? And those are the biggest pains that everybody has getting started.

Duarne 52:59

Yeah.

Dave 53:00

So you're not alone. If that's you, and you're like, man, I just feel like I'm alone, you're not alone, you're not. So that's what we're aiming to solve for a lot of people, I think.

Duarne 53:09

So I think so. I think that's right. And you know, if you're watching these videos on social media, YouTube, and all those other platforms where you've got a business owner sitting there talking about how they're running their team, and you know, that it's been GrockBot's been such an amazing, you know, revelation for them, and it's just made their ability to run their team so much easier. You've got to remember the platform's only been out for a month and a half. I mean, video production quality on those videos are pretty high. They're probably all pre-produced before the actual event, and they didn't really even use the platform. So, you know, I'm not saying it's propaganda, I'm just saying it's marketing.

Dave 53:50

Yeah, yeah, all that stuff is marketing. Yeah, they didn't plan that right then and there. Like they've been in planning for that event for six months. They might start planning for the event the minute the last one ended.

Duarne 54:01

So don't get don't get caught up in the hype and thinking all that sort of stuff. Go and talk to people like you know, Dave, myself, other people have actually implemented solutions that have learned from the the failures of implementing these solutions as well. And you know, and and just see. And what was it 12 months ago when we were talking about AI platforms? Back then we were talking about pick one, learn it, become really good at it. You don't need to go and use three or four of them, even though there's tons of them out there at the time, and all of them have different use cases. Yes, we will use it for different use cases, but even us, we start to peel it back at some point and go, you know what, this is probably gonna do 80% of what I need it to do. I'm gonna focus most of my time on this platform, right?

Dave 54:50

You know, I mean, and back then we didn't have like if you think about it, we didn't have cowork, we didn't have Hermes, we didn't have any of that, even a year ago, like it was such a different environment than it is today, and it's gonna continue to evolve, continue to improve, continue to grow. So, like you know, you have to kind of figure out what's the best option for yourself and just get started. I can't I can't stress that enough. I think so many people just like drag their feet because they just I don't know what to do. It's like just do something, like just start with something. Think of something you do on a regular basis and see if it can help you just go. I think we're here where we are right now because Dworn and I we started with ChatGPT when Chat GPT came out. So, like we have always been in it since it started. So you don't need to be Dwan and I right now. Like you need to be me and Dwarne three years ago when ChatGPT came out and said, Hey, let me test this, let me see what's happening, right? What can I get out of it? Like, that's what you're looking to do.

Duarne 55:53

We didn't even pay for the accounts back then. We're just using the free version. Yeah, like I don't know if there's enough value for me to be paying for and we we we thought twenty dollars a month was a lot.

Dave 56:03

We're like, wait, why would you pay twenty dollars a month for this?

Duarne 56:06

I know, right?

Dave 56:07

Shit, I paid I I paid 20 times out of month now admin.

Duarne 56:11

Oh it's crazy, right? It's so funny because like today I had a cabler come to my house and he was doing a quote for me and he saw a few things and he was talking about how my command sent a room next door that I'm building out with computing for AI localization computing. And he's like, Wow, this is incredible! This is a lot of hardware. I said, Yeah, I said it is. He goes, I think I want to learn some stuff from you. And I'm like, What sort of stuff are you thinking? He goes, I don't know. I just think I need to learn some stuff from you. Yeah, and I said, I said, Well, look, you know, how have you thought about just like having something that could do what I mentioned earlier with the email? And he was like, Oh my god, that'd be amazing. That'd be so helpful. And then I said, What if you had something could check your messages for you on your social media platforms and do the same sort of thing? Oh, that would be so helpful. Okay. What about if you had something that could help you with your quotes, where you could record a message telling it what you need to put in the quote and have it go and prepare the quote in a draft format on Zoho, which is his platform he's using, and you can go and review that and make notes to it and adjust it. He goes, Oh, that would save me like half an hour a quote at least, maybe more. I said, There you go. I said, You've got a use case for a little team of agents if you were able to put the time in. I said, but it's not a I said, this is just from one conversation with you. I'm sure there's a pile of other stuff you could get, you know, you could get some more benefit from. And I said, Imagine if you were to implement this across other key players of your business. What's because he's got a team, I said, what sort of results you could get, and what sort of resources could be freed up within your business to go out and look for more business and help new clients and get quotes out quicker. And he's his mind was blowing his oh my gosh, let's have coffee soon. Okay, let's do that. But it's that's the sort of conversations that I'm having with people, and their eyes are just becoming awake, and they're really seeing what's capable of being done. And we're not talking about rocket science, we're talking about like just really small things from our side of things to set up. It's not even complicated stuff from our side. This is just small things which have huge impacts for the average business owner, right? Which is incredible. That's the part that I love about it. I mean, you can have big impacts with with starting with small things, and once you get your but just keep your mind in once you actually dip your toe in, it becomes very, very addictive to look for what's that next thing that I can work on. What's that next thing? My one-year-old's in the background watching it flapping. I love it.

Dave 59:08

No, I mean, I think it that's it. It's it's what's what's small something to start that's gonna have even a medium impact, right? A medium large impact on your time. That's where you want to start with. So all right, wait, we went, we're about an hour, you know, which is our typical time frame. So Dwarne, what's something besides what you just mentioned, what's something that you want people to potentially take away from from this episode today?

Duarne 59:37

Well, I think uh first of that, it's good to be back. It's been a couple of weeks. So I appreciate the being able to have these sort of conversations again. And for me, I think it's never too late to get started, whether it's with AI, whether it's with creating processes in your business, whether it's you know, just going out and starting a conversation with a new client. So just be prepared that it's gonna take a little bit of time. Anything that's worthwhile typically takes time, it's not gonna happen overnight, and just be willing to, you know, just chip a little bit off at a time and solve one little problem at a time, and you celebrate those wins. You're gonna like I look forward to my three emails a day that I've set up for my morning and afternoon email summarization and action list and my decision brief. I love that. Those three emails, I look for those, and that's what I basically focus my day around.

Start Small And Build Confidence

Duarne 1:00:32

So I think once you start with this, you're gonna find yourself really appreciating the tools that you can you know automate and help you run your day. And if they don't work, that's when you start stressing out. You're so used to them at that point. Like, where are they? I don't need that. Right, right, right. So, yeah, get in and try. Start try start solving one little problem.

Dave 1:00:56

I know I I mean for me that I guess a piggyback on that a little bit, right, is you know, making sure that you're reminding yourself that what you're doing is you're making progress. It may not feel like it in the moment, may not feel like you there has been much progress, but just remember like every single day that you're doing something, you're making progress. You're either learning something new, you're you're putting, you know, a couple next steps into your goal. And I think that's that's the piece that you have to keep reminding yourself on. Like goals aren't made by giant leaps, goals are reached from small, all the small actions that lead to what looks like a giant leap from everybody else. That's because I didn't see all the all the small steps that you put in day in and day out. And you have to remind yourself those because you're the one, you're the only one in it every day. So you have to remind yourself to continually put in those small steps, otherwise, you're gonna feel overwhelmed because you're not making a giant leap. So that's what I'm gonna leave y'all with today. Again, we do this every single Friday morning around 8:30, depending on how often Dorne and I talk pre-show, maybe a couple minutes late. But with that, we hope you guys have a wonderful and amazing rest of your weekend. If you have any questions anytime, feel free to drop them down below and we will answer them whether it's live on the show or we'll respond to them afterwards. But we hope you guys have a wonderful and amazing week. Keep pushing forward, keep reaching your goals, do all that fun algorithmy stuff for us if you could. You know, like, subscribe, share, especially if you found something that you got out of today's episode. But until then, Darren, thanks for joining me, buddy. And uh, we'll talk to you everybody talk to everybody later. See ya.

Duarne 1:02:29

Thanks, guys. Take care.