How to Build & Sell Managed AI Agents in 2026 (No Code) Transcript, AI Summary & Key Points

00:00 What if I told you you could build, manage, and sell AI agents for small business owners even if you have zero technical experience, zero audience, zero email list, right? All of this is possible. You don't necessarily need those things, and you don't even need to know how to code because my guest today, Phil Goodwin, lives in a small town in Georgia.

00:20 This is not a town that has like a bunch of rich business owners running around. Yet, he still managed to build and sell AI agents as a service for a living to business owners in his town. Now, he has no audience. He has no email list. He's literally a Figma designer by trade. Like, he doesn't even know how to code. And he's going to show us the offer that he's running to sell agents.

00:40 He's going to tell us what he charges. He's even going to show us the exact text message that he sent to land his first client. And towards the end of the episode, he's going to share the one thing that he builds into every agent that got a client back 63 hours in week one. I'm not exaggerating. This is so smart. I've never heard anybody talk about this method before, and Phil shares the whole thing with us.

00:59 So, tune in, and I hope you enjoy the episode. Today, we've got Phil Goodwin on the podcast. We're going to be talking all things managed agents, how we can make money selling managed agents, managing them for clients at pretty attractive price points, and really really strong margins. So, Phil, why don't you give us the the takeaway? Like, what are people going to learn if they stick around all the way through the end of the episode?

01:23 >> Yeah, my goal is people to go away with this really feeling like, I can do this. Um I want people to feel empowered. Um not only with the I'm going to show you guys kind of my stack and how I actually launch these agents and things, but I really want you to go away feeling like, I can do this. Um so, that's kind of you know, that's that's the takeaway I'd love for people to like come out feeling empowered is really what I'd love to see.

01:44 >> And I think that's what people want to hear, right? I mean, most people listening to this are in a position where they're they want to go zero to one, right? We have way fewer people who want to go one to 10, for example. So, I know you've got some some materials prepared. and you're going to share your screen here and we're really going to get into the weeds of like how these agents are built, how you structure the offer, how you found your first clients, which is always everybody's biggest bottleneck.

02:08 So, before we do that, can we make it clear that like one, you're not a tech guy, right? Like you're not a developer, you're not Can you just give us your background so that again, people listening to this are like, oh, like I'm like Phil, too. I can do this, right? >> Right. Yeah, yeah, so my My background is completely non-technical. I'm a designer uh historically.

02:30 So, I have uh I've been in design for a couple decades um designing mobile apps, web apps. You know, I started out kind of I designed everything and then about 2015 went into only product design. So, mobile apps and web apps. So, made a ton of um you know, got a ton of clients from that and you know, met a lot of people in that space. But, I'm not at all technical.

02:51 I've never been a developer. Uh never been like a professional IT person. So, yeah, this is all I'm coming from designer world. Uh spent most of my time in Figma prototyping there. Um AI has been like a total new uh kind of thing for me. I I ran a business uh with a couple partners that we sold last November. And that really kind of transitioned me into, okay, you know, now what do I do?

03:18 And so, that led me in December to really diving into the the AI stream and kind of starting to discover things there. >> Perfect. Well, yeah, again, I The fact that you're non-technical, I think makes this even more accessible to most people watching this. So, let's go ahead and dive in. I know you're sharing your screen right now. We're Obviously, like I said, we're talking about managed agents.

03:37 So, maybe do you want to start by defining that? Like what does it mean to sell managed agents? Like what's the actual offer that we're selling here? >> Yeah, so the whole idea and kind of how I pitch this to to people that I chat with it about, it's like um it's like hiring a a digital worker digital employee for your business and so my kind of pitch is I'll I'll kind of be the manager of that thing and I kind of really act as like a coach with the and a lot of the people the clients that I work with are business

04:06 owners their business owners their entrepreneurs their executives and these also are not really super technical people so they really they really need and want someone to kind of not just not just drop an agent in their lap and say like hey have fun with it Some of them do like to play around with this stuff I've had some clients go wild on playing around with stuff but most of them they want someone to kind of walk alongside them you know give them you know kind of coach them along the way help them build the

04:32 workflows help them discover different things the agents can do and so that's really what I'm selling is it's like a it's like a combination of kind of coaching consulting but also with this like tool that's empowering their business with like a role that they either have a had a hard time filling in the past or have thought about filling for a while you know normally that's like a central office person or like an executive assistant or like a really specific thing in the business that they haven't really been able to

04:58 hand off to someone else or just a plethora of stuff that they do as an owner that they've wanted off their plate for a while anyway and so this is kind of like hey I can I can give you a digital employee that can kind of help you fill that gap and I'll coach you along the way so that's really what the offering is >> So I love that you specify the coaching aspect right because a lot of people when they think about AI agents they think like plug and play right they think oh I'm going to build this agent I'm going to

05:26 install it in this person's business I'm going to charge them a monthly fee right two grand a month three grand a month five grand a month whatever that ends up being and then they're just going to go off and use the agent and I'm just going to collect a check every month but what you're saying what you're offering is way more in depth you're like well yeah I'm I'm building them that that agent and them that AI employee, but then I'm acting as almost like a manager of that agent.

05:49 Is that fair to say? Like as well as a coach to the business owner on how to use it. So, it's almost like you're taking, you know, folks listening that have listened to this podcast before have heard me talk about like our AI concierge offer, which is done with you consulting. I mean, really your managed agents model is AI concierge, but even a higher ticket, right?

06:11 And even with the deliverable being not just the coaching, but the actual AI employee to go with it. So, I love that. I think that's awesome. >> Yeah. >> Yeah, there's a huge need for that. Like the business owners want, they don't want, a lot of times they don't want you to just build it and hand it off and then walk walk away. Like they need the ongoing guidance, which is why they're willing to pay you a few grand a month.

06:33 >> Yeah, so I, you know, historically how I got into this, I built my own agents, right? Um testing it out with myself. I got one, I got two, I got three. Starting to find out that I don't I really don't need that many, right? And then I was like, I could probably I bet other people would find some value in this. So, I got with a couple friends. I was like, hey, let me give you an an AI agent.

06:50 So, set them up, kind of do it for free, right? For a few people. Kind of start getting some, you know, traction with testimonials, like some case study with them. Um and then started be like, okay, what can I charge for this, right? So, I started doing um just kind of standing up the infrastructure. So, ignore all the coaching, managing the agent. That was not part of my thought process.

07:07 This is back in February. And so, I just just setting up people on VPS on a virtual private server. I would like stand it all up for them, kind of set up the security. Um and then I kind of set them up with the local coding agent that could kind of be their technician for their for their VPS. And then um and then have the agent that's actually, you know, kind of the agent they communicate with in Telegram or whatever their communication platform is they want.

07:31 Um and so, I was kind of doing that. I did like I started like 500, then I did 1,000, then I did 1,500, then I did 2,000. Um but then I start you start finding that like the really the people who are the most you have some of these owner operators who are super motivated and they're like really excited about this tech and other ones they like really need hand holding and so when I hand it off an infrastructure to them it just dies if they need hand holding, right?

07:53 Because I'm not because I wasn't selling my ability like be there ongoing with them. >> Mhm. >> It was just kind of like here's the setup. I'll give you 14 days to kind of ask me questions and kind of be available to you. Um after that like hey, you're on your own. Um so this kind of came out of you know, I kind of see some people that just stop using it cuz they don't really have help to like know how to get this plugged into tools and like how to talk with it.

08:14 Honestly, that's something that I've also learned like this coaching aspect is super um important cuz people also are not used to talking to machines. Um they're used to like asking chat GPT questions now um like treating it like Google. But the whole idea of how do I build workflows with a machine, you know, with a with an intelligence um it's it's weird cuz you'd think it'd be intuitive like I'll just talk to you like I talk to a person.

08:36 But people don't naturally do that. Um they naturally treat it like Google, right? Um so it's um there's like the aspect of like learning how to just talk with it also, which is uh which is kind of a part of the whole coaching and managed agent thing. >> For sure. No, the model makes perfect sense and again, you're you're speaking my language as far as like [clears throat] early on offering it for free, getting those case studies, getting those testimonials, testing out the market, like what are people willing to pay,

09:04 right? Starting at 500 basically raising it $500 every time you got a yes, which is exactly what I preach people do as well. Keep Keep raising your prices, especially if everybody's saying yes. So you're speaking my language. I love it so far. I know you're sharing your screen here. Let's dive into your workflow here and and kind of how you get this up and running for folks.

09:25 >> Yeah, so I'll kind of start with like the high level of like hey, what I build is like I I sell managed AI employees to small businesses and it always starts with one, right? The idea is that the offer kind of starts with like the starter kind of cost to setting everything up, and then there's like a monthly ongoing per agent. That's kind of how I do it.

09:40 Now, I know you've had Nick on the podcast. I know he's kind of like a hey, 5K a month, unlimited agents, you know, unlimited tokens, all that kind of stuff. I've taken a little bit of a different approach, and I'm like, hey, there's going to be a setup, and then there's like an agent fee, like there's a cost per agent. Cuz most people don't need, you know, uh a ton, right?

09:57 Um so, I'm basically starting with like I sell them one agent, um and typically that falls into like more of an administrative role cuz those are a lot of the like these tasks kind of suck for owners. They don't want to do these things. But, the high level is like there's a client request or a scheduled job, right? Then it goes to what whatever the agent is.

10:14 It utilizes company, you know, knowledge and the tools we have connected. Sometimes there's some kind of approval gate that we put on there to require approval, like an email send or something like that. And then the work's done. Um and so, I, you know, inside the agent, each agent kind of gets its own job description, you know, obviously memory, sometimes do an email address.

10:34 Um not all my clients actually want their agent to have email addresses. Um and then obviously connected apps, dedicated VM computer or VPS, uh and then kind of monitoring ongoing. That's also part of what I'm offering is I'm going to watch this thing and monitor it. If it goes down, hopefully you'll never know, right? Um if it goes down, you do know, I'll fix it immediately, right?

10:52 But, everything's got like things on it that I'm that I'm watching. I should get a ping if something happens. >> And quick point on that, right? Because that was, you know, back when Open Claw came out, for example, that was everybody's biggest gripe with things like Open Claw is like it goes down so often. You're spending like you can't be nearly as productive using the agent cuz you're spending 50% of your time troubleshooting it.

11:12 So, a lot of times, yeah, and and when people like I still get this pushback all the time, like, oh, who's going to pay five grand a month for you to manage an AI agent for them? The person that doesn't want to spend five hours a week inside Claude Code or Codex troubleshooting the agent, right? Like a big part of the value that you're offering is, hey, if it goes that when it goes down, cuz it will go down eventually, right?

11:36 They always do, I'm going to be the one fixing it and realistically, it's going to get fixed before hopefully before you even notice." So, that that's a huge benefit and that reason in in and of itself, people are willing to pay a few grand a month for. >> Especially if they've tried it themselves, right? They they know that it's bad. You're you're like new people to this, they have no idea that's ever been a problem.

11:55 So, but if you've had people that have like tried to do the AI agent thing themselves and got frustrated cuz it broke all the time and they just gave up. >> Yeah. >> It's an It's a good selling point for them, for sure. >> For sure, yeah. >> Um and we don't have to go through all these use cases, but I just wanted to show these are actually these are real use cases from current clients that I have that I just kind of had, you know, put together a little list here.

12:15 So, around like operations and admin stuff, um content knowledge work, executive assistant type work, technical and recurring work. These are all like my my clients have been in um my current Manage Agent clients are in like construction, um and and like coaching. It's like business coaching, um construction and then some of my past clients have been in more like property management, um and uh gosh, miscellaneous engineering, um so kind of all over the board, real estate.

12:45 Um so, these are these are like some of the, you know, kind of real-world sort of applications that we've found that people are using, you know, using the agent for. >> And let's let's maybe touch on the top one to three from each of those categories there because that's a big I wouldn't even say like objection, it's more of one of those things that when you're like when you're talking to a prospect, especially somebody that hasn't really used AI that much yet, a lot of times the issue [clears throat] they face is they

13:14 just don't know what they don't know. They're like, "Well, yeah, I think I need an AI agent, but I don't even know what I would have it do." So, let's talk through some of the use cases here to kind of get people's minds, you know, spinning and and so they can have some ammunition in their back pocket when they go to pitch a client on this. Like, "Hey, here are some things that it can do."

13:32 >> Yeah, so my most recent managed agent client, he's a kind of a one-man, you know, runs a lot of his business, family business. It's a directional drilling company. So, he has a lot of central office type needs. A lot of that stuff's around like setting up new jobs, like uh getting uh like all of his like field kind of contractors he has like reporting hours and they do logs.

13:52 Um and so he's kind of like constantly managing, you know, and going between QuickBooks and like field logs and job, you know, PDFs, contracts, folders. So, I've actually been super impressed. So, every time these models like come out with a new there's like a new kind of jump forward, I get more impressed with the agent's able to actually handle. I am One thing that I do that's um may maybe it's unique, I don't know if it is or not.

14:16 When I set up a client with an agent, I actually set up a like a group chat in Telegram. So, it's not just I don't just kind of leave the client to talk with the agent by themselves and I can't really see what they're doing. Um they do have that, but I also set up kind of this group chat where it's like myself, the client and the agent. So, I can they can see me interact with the agent and they can kind of learn how I talk with it.

14:34 But, I also get to see things they request and I can tell them like, "Hey, is that a you know, like this client would like ask his agent like, "Hey, search my inbox for an email from, you know, John Smith and if you you find it, there should be an invoice on there for job 3370. I want you to download it, put it at a create a new job folder for that job, put this PDF in that job folder and then take the information out of that PDF and put it in the Excel sheet.

14:59 Also, save that to the Dropbox folder and then send me the link." And I'll be like, "Dude, I don't think the agent can do that." Like in like one like one fell swoop, but sure enough six 10 minutes later the agent brings it all back and it's like all checked off. I'm like, "Dude, I'm I'm like dumbfounded that the agent was able to pull that off." Like, you just gave it like a insane number of stuff to do in a single command and it does it.

15:19 Um and so that's that's a recent thing I've been seeing, but like multiple things on this list, it will like knock off in a single like prompt from from the client. >> And another thing I love about this model, especially when your client is someone who hasn't really used AI before, especially if they haven't used a genetic AI, is like take the example that you just gave, that five, six-step workflow for that drilling client.

15:45 Once he Once he sees that happen one time, like if he gave that prompt, 6 minutes later agent comes back, yep, it's done. That guy's mind is blown. Like, he's probably not going to churn because he's like, "Oh my god, that's something that I've been doing myself for every job for the last 20 years, and it takes me 20 minutes per job, and half the time I don't do it because I I either hate it or I forget.

16:10 Like, that like, you know what I mean? Like, that's just such a lightbulb moment for most people that if if you can set up those conditions for it to work properly, they're just they're they're not going to go anywhere. Like, they're they're bought in at that point. >> Yeah, and once he once a client experiences that, they start their mind starts going to, "Well, what else can I do that I didn't think >> most common next step.

16:30 It's like, "What else can it do? What else can it do?" Yeah. >> And so, like one of these things on here, organize thousands of accounting, vendor, and HR files. Like, he had his agent go back and organize his entire Dropbox, and was like, "Dude, I've been putting this off for years. Like, going back through all my job file, like organizing all this.

16:44 I just had the agent do it." And so, his first week, I mean, I've never seen someone save so much time like in week one of having an agent. I think he did about 63 hours of like saved time. Um, I I tack on what a value tracker. So, this is like kind of a pro tip. >> That's smart. Yeah. >> if you have the agent kind of like have a ledger, um, like the the agent that you launch, like actually have a value ledger, so it logs every single task it does, and it assigns an amount of like human time and dollars to those tasks,

17:16 and I have it at the end of every week, I have it give a report to the owner of the agent. It'll say like, "Hey, this week we did, you know, this amount of tasks. It amounted to, you know, about this amount of time, which equated to this amount of dollars." And then it'll also say like, "Hey, we did these tasks we don't have dollar equivalents for time, you know, tell us you know, tell me how much time would this normally take you."

17:36 Right? So, every week they get like re reconfirmed like, "Oh, okay, this week I saved $1,700." Right? Last week I saved $6,300. Right? >> Dude, that is so smart. I'm I'm stealing that. Like for for anybody in the audience listening to this, that I mean, that's genius. I've never heard of anybody doing that. Now that you say it, it sounds so obvious.

17:58 But like, that is freaking genius. I love that. Continue. >> Yeah. Yeah. So, that was that was like week one for that guy. It was like a $6,300 week. Um so, I have another client who does for the content knowledge work stuff. This is another thing I haven't used AI agents for, but he has and it's short-form video editing. Um So, he's used his agent a lot.

18:16 He'll put his like iPhone videos in a in a Google Drive. And then he'll point his agent he uses a I can't remember what the short-form video software is called. But he has his agent interfacing with a specific software and he'll tell the agent, "Go grab this video, use, you know, such and such software to like cut it up for me and put the captions in."

18:36 And he said right now it gets about like 80% of the way there, which still saves him a few hours. Like he's trying to put out eight to 10 of these short-form videos a week. So, for him to be able to drop it to his agent and and it takes longer. Video editing takes a lot longer right now for the models, at least for these agent models. And I'm having all of them run on 5.6 soul.

18:54 So, GPT 5.6 soul. He says his is like a 45-minute 30 to 45-minute thing, but the agent comes back and it's like 80% done for him. And so, he's like just kind of tightening up the edges of it. So, it's still saving him like a ton of time. Um That's a unique use case. I haven't heard anyone else using. I mean, I'm sure a lot of people use them for video editing, but it's not something I really do.

19:16 So, I know video editing is pretty time-consuming. >> Yeah, that's one that again, like it's going to be very difficult when it comes to editing for AI to get you 100% of the way there. Like very rarely is it going to be able to, but if it can get you 80% of the way there, I mean, legitimately saving you multiple hours per week. Like I know how time-intensive and and just tedious it is to um edit short form.

19:40 So, that's that's a great use case. >> Yeah. But, I mean, the rest of these we don't I mean, we don't have to go through all of them, but, you know, a lot of stuff that people know that AI agents are good at, research, you know, prospecting. Um these are all like really good things. I found that it's really good to encourage clients to not just have the agent um kind of do like if you're doing a research project and you're trying to or you're trying to or you're trying to prospect for a specific um kind of niche or

20:06 specific service, have the agent actually drop it into like an Excel file or something. Don't just have the agent like store it in its own markdown. Um I think I that was something that it took me a while to like it for it to click in for me. I'm like, oh, I need like a source of truth, you know, where this thing can like write to that's I can then go in and edit that's like, you know, something more like a a table.

20:23 Um and I can kind of go through and check it off. And the agent also can refer to it, look at it. Um but that's like a it was like a little thing that occurred to me like a few weeks ago. I'm like, why haven't I not thought about doing that? I started like tracking my own payments and things in an Excel sheet and have my agent double-check it when a payment comes in, log it for me.

20:40 It's an easy place for me and the agent to like use together. Um so, that's like a kind of a a small thing, but having the agent actually generate human-readable things for yourself, especially if you're prospecting. >> For sure. Well, and on the idea of prospecting, right? So, let's talk about client acquisition, how you found some of those first clients because just Just it clear to the people in the audience, right?

21:06 Phil, you don't have some like huge social media audience, right? You don't have some big uh email list. You don't You know, you don't know I mean, you It's not like you have hundreds of people that you could call on at any given point in time to try to sell them these services. So, let's talk about how you found those initial clients and how you're going about acquiring new ones.

21:26 Cuz again, that's most people's bottleneck. So, do you want to shed some light there on like how did you find that very first not just the free client, like the case study client, but the very first paying client? >> Yeah, typically what I've what I've found so far So, obviously the free ones, those kind of were like friends, right? So, if I'm I'm getting the friends in there to to test, does this work?

21:48 Is it giving them value? Let me get Let me get some like quotes from you, right? About how this is helping you. And then a lot of times, those friends would actually refer it to somebody who had actually could charge for it, right? So, um So, then the free ones become like the referral engine for the paid ones. Uh and I've I've noticed like I've Dude, I think I've done 20 about 21 setups um so far.

22:15 Uh a lot of those kind of at the beginning were like close tight-knit network friends. And then from there, it was like they told a person and it's kind of expounded out like I'm, you know, I'm in a small town in Georgia. Um this is not It's not like a super like wealthy town. But the These people who are here, they have friends in Atlanta and they have friends that are in Dallas, right?

22:36 They have friends that are in North Carolina in some bigger cities. And so, I start getting referrals for people that are all over the country. So, then I'm start Now I'm doing AI agent, you know, setups for people all over the country. Like And it just started with like me working on doing some free stuff for friends. Um So, I honestly, it's all come through network.

22:54 And then I've started to post online, like you said, I probably have like 1,500 followers on X. Like I'm not like a I'm not a big influencer or like I don't I don't really work on trying to build a following count follower count, but I am starting to try and post like helpful things that I think would help other people who are in my position of like just start doing it.

23:10 You just start If you just start doing it, like I've got a friend here locally who's trying to get into doing similar things. And it's really easy to get in your head and try to like think you've got to have this whole plan in place where I've got to I've got to map it out exactly how I'm going to do it, who I'm going to be offering it to you, what exactly my offering is, and like yeah, I got to button up all these things before I can offer it.

23:30 And you don't. Like you can just start doing stuff and uh and you'll like start learning and you'll start finding the stuff that I actually like this part. I actually don't like this part. So, I know that you're a big like the AI concierge thing. Like I'll do an audit, I'll I'll recommend some tools for you. For me, I'm like, dude, I actually don't want I like I don't want to do the like bunch of different tools, right?

23:50 It's not really my thing. I don't want to set up someone up with Claude CoWork and like develop plugins or or develop uh skills and stuff in Claude CoWork. I'm like I I really like the idea of like setting them up with an agent. So, I've you know, I've done some of the like Claude CoWork stuff, right? In the Claude stuff. But I've found through like playing with that stuff, I'm like, okay, I actually that's not really my thing.

24:09 I like this thing. Um so, I think it's it's a start doing it, right? Um and eventually someone will ask you, what do you charge to do XYZ? And you're like, "Hm? Charge? >> [laughter] >> So, it will happen. >> A couple good takeaways from that. So, one, I mean, I totally agree that there like there's hundreds of different AI services or offers you could sell.

24:31 But your point, which I agree with completely, is like, well, just because a particular offer might be popular or you can charge a lot of money for it, doesn't mean it's going to be one that you personally like to fulfill. So, the the advice I'm hearing from you is find the offer that is both lucrative that you can charge for and one that you enjoy fulfilling, right?

24:50 Is that fair to say? >> Mhm. >> And then your second point is you let the the free work kind of be your pipeline for the paid work, right? So, you like how did again just tactically for people in the audience, let's say somebody's listening to this and is like, "Okay, well, yeah, I know a couple of business owners here in my town or, you know, maybe from my gym or from my church or parents of my kids' friends, for example, I could probably reach out to them and offer free work, but I I feel like a lot of people even

25:22 get stuck there. They're like, "What do I even say?" So, like what what is that text message that you would send to that contact that you know, even offering free work? >> Yeah, the honestly the free work sometimes is not different than like the hey, I'm just trying to like test some paid work. But it would be like, "Hey, I'm starting to experiment uh with AI agents.

25:43 Um and a lot of times I've had like conversations with some of these people that I'd be texting, right? Like at a gathering or something. We've talked about AI, right? And I might text them like, "Hey, I'm starting to test out uh a managed agent service where I like I install an agent for you. I kind of can kind of coach you through building workflows.

25:57 And I'll be there like once a month we'll have a call, we'll talk through what's working, what's not. Would you be interested in like being a guinea pig in this with me? Um I'll charge you less. Like I'm, you know, but I I just want like the experience and the reps. Would you be willing to do that?" Um and most of the time they're like, "Sure." Right?

26:15 They'd rather like kind of be the guinea pig for you than pay like high dollar for somebody who's already got it figured out. So, it's uh it ends up being like a win-win. Um but that's that's normally what those conversations look like. Or it's like literally a conversation at a kid birthday party or something and I'm talking with the dad, right? And I'm doing the same thing.

26:32 >> that I love that perspective so much because you're like you're putting it out there like, "Hey, you are a guinea pig." Like you're not trying to be something you're not. You're not trying to be like, "Oh, I'm the AI expert and I've got it dialed in and this is what we're going to do." You're basically like I love that cuz it's so transparent. You're like, "Look, I'm I'm getting into this space.

26:49 I'm starting to build some cool stuff. I'm starting to take on some clients, right? So, I mean, for the people listening, you don't even have to like, if you know enough to be dangerous, like if you built this for yourself, kind of like how you started, you could charge from the start, right? But, you're not going to charge probably five grand a month.

27:05 You're going to charge, I mean, your first client you said was 500 a month, right? >> The first ones were just like infrastructure set up and then, "Hey, I'm see you, right?" And that would be like a yeah, one-time 500 bucks. I'll set it up for you. Enjoy it. >> first the first actual like managed agents clients where you're meeting with them, like the once a month, you're actually managing it, how much did you charge for that actual first managed client?

27:31 >> So, the first one um that I did, so I I only got into the managed thing about a month ago. So, this is a pretty new thing for me. Um the first client that I got, um I didn't even charge him for a setup. I was just like, "Dude, can I just give you an agent and I want to like work through stuff with you?" And I did like 250. Um like 250 bucks. I'll It basically covers like any infrastructure cost and like maybe like an hour of time, right?

27:55 Uh a month. And um cuz a lot of these systems, once they're in place, they can run themselves. Um and so, he was he was kind of like number one. Number two was about the same. Um and I could They were two different industries, so I kind of wanted some experience in each industry. Um and then my most recent one, uh which is kind of like the first one where I started getting into, "Okay, I want to do a setup."

28:16 It's like a $1,500 setup. And then I upped it to like 500 bucks a month to like have a one once a month call. Um you know, we'll have our Telegram group. So, my goal is to probably get to a couple thousand dollars on a setup and then, you know, maybe a thousand dollars a month per agent sort of situation. Um right now I'm trying to build that value, like the value prop that like I have I can back up.

28:38 So, I haven't found anybody locally who'd be willing to pay like five grand a month for like me giving them unlimited agents. Um but, for me it's like I'm trying to figure out the sweet spot around where I can take on enough clients where it supports what I need and also um is I'm providing the value that I feel like I'm like they're getting that value from me, right?

28:59 Um So, if that that's like the thing I'm trying to figure out right now is really sort out in the market that I'm in which is like a small town. Um you know, obviously you can you can get clients in bigger cities and live in a small town. But the my immediate network where all this stuff is kind of starting this isn't there's not like big money, right?

29:19 So, charging charging somebody who's running like a drilling company as a family business 500 bucks a month. It's like he has to think about that, right? It's like it's kind of like not money I planned on but he's like but if you can replace a central office person for six grand a year I'll 100% I'll I'll take a couple of those and I'll tell all my friends, right?

29:35 So, um so that's kind of that's that's sort of the starting place where I'm coming from on the manage side. >> And and I love that. And again, the point that I'm trying to make here to the audience is that it's okay to make it clear to that those first clients that you are that that they're guinea pig, right? They're like, "Hey, I'm charging you less than I would normally charge because I'm still trying to work out the kinks of this stuff myself and I'm still learning this stuff myself.

30:00 But one thing I can promise is, you know, I'm going to provide you a high level of service and I'm not going to charge you nearly as much as I will be charging two months from now, for example." So, just the whole the whole premise of like again, client if clients know that they're a guinea pig, some of the like a lot of them are okay with that, right?

30:18 Because they're getting a deal. So, you don't need to be an expert. You don't need to start by charging five grand a month. You can charge 250 a month or 500 a month to because you're you're getting paid to learn at that point, right? >> 100%. That's exactly >> And would you say that I mean when it like if somebody watching this wanted to take that step one and wanted to like, "Okay, I've I built an agent for myself.

30:39 Like I've spun up a Hermes or an open claw. I feel confident that I can at least do the same for a business owner and provide some level of value. Do you think that they even need to do that for free for one or two clients, or can they go straight to that pitch of "Hey, um you're a guinea pig. I'm going to charge you a lot less than I would normally charge."

31:00 Like can they go straight to that paid work, you think? >> 100% Yeah. Yeah, 100%. I I mean, ideally you're not doing free work, you're losing money. Like you're You don't want to be paying for the infrastructure. >> You just want to cover your costs for that. >> Yeah, at least cover your costs is the goal. So, even when in the people when I said like people I did stuff for free, it was just like my time that I was giving for free.

31:18 They still paid for all of their stuff, right? They still paid for their infrastructure, they still paid for their AI brain. Like they're still paying for all that. Um so, I would definitely try and make sure you're at least covering costs and some of your time, um even if it's at lower margin. Yeah, you can still do it uh at at a low rate and skip the free phase the free phase.

31:38 >> Yeah, and I and again, I think that's a great That's a great point. You don't have especially if you've done this for yourself, like you don't have to charge for it or you don't have to do it for free, I guess, to begin in the beginning necessarily. >> For sure. >> Well, that's that's awesome, Bill. I mean, this has been great. I think this is super tactical.

31:52 I think if anything, this has kind of showed people who are considering getting into this business that again, you don't have to be technical, you don't have to be some AI genius. Like you just need to know a a little bit, enough to be dangerous. And I mean, you're non-technical and you're seeing a lot of success with this. So, I guess one of my final questions would be So, if somebody listening to this did want to get started with this model, let's say they're like, "I'm ready to go."

32:17 What would be your your best advice on like step one? What what should What should their step one be if they want to get started with this? >> I think just pick some kind of infrastructure that you want to use. So, I I'd like Orgo. I know that you guys have had you've had Nick on. I think you probably used Orgo. >> That's what we use, yeah. We use Orgo as well.

32:36 >> Yeah, I mean, it's the easiest like way to get set up. Orgo has like pre-built templates and things. So, that's like the easiest thing is like go start, you know, plug into something like Orgo. I like to like connect my agent into Orgo. So, I actually do very little in Orgo. The my agent handles like duplicating templates, creating workspaces, you know, setting everything up for me.

33:00 Um and I mean, that it's as it's as like simple as that to get going. Get your first client. Orgo I think will give you like a computer for like 29 bucks a month. There's probably like There's probably some promo codes floating around. Um also, and uh it gives you like your first computer pretty cheap. You can get get somebody on boarded pretty easily.

33:21 Um And then And once you have that first rep of like, oh, okay, I can do this. This is This isn't that hard. Um And And you do, if you know a little bit, you know a lot more than the person who has never done this. So, there is value you're bringing to the table, even if you feel like, I don't really know that much. You know You know more than you think you do.

33:39 >> For sure. And you really only need to be one step ahead of your target client, right? To be able to sell this effectively. And I'll I will also vouch for Orgo as well. I mean, that's all all of our internal agents are hosted on Orgo as well. I've done interviews with the co- one of the co-founders of Orgo, Nick. So, what we'll do is we'll link that video with Nick in the description and in the show notes as well.

33:59 That's a really good uh follow-up to watch after the episode you're watching here. And for anybody that wants to try out Orgo because like Phil said, it's it's basically a platform that gives your agent a computer. And they [clears throat] built They're building Orgo specifically as the platform to deploy managed agents for clients. So, if you're serious about this business model and you plan to deploy agents, Orgo's really the place to do it.

34:23 We've got a code. We'll put this in the description, too, but we've got a link. And if you you got to use both the link and the code Corey, but you'll get I think it's 20% off or your first 3 days for free and then 20% off your first 3 months of Orgo if you use both the link and the code Corey. So, we'll put that in the description, too. But, I mean Phil, awesome episode, man.

34:48 I learned a lot. I think the biggest takeaway, like you said, was to build in that basically that like weekly ROI tracker. And ideally one that pings your client every week is like, "Hey, you know, great week. This week you used your agent for this amount of time, you saved this amount of money or you saved this amount of time or you made this amount of money in terms of time um based on how much you use your agent."

35:09 Like that That's genius. That right there is going to quantify to them why they should keep paying you every month. So, really good really good takeaways there. Anything you want to point the audience towards, anywhere you want to tell them to follow you or connect with you or hire you? >> Yeah, I'm uh on X I'm FHL Creative. Um that's kind of my design background kind of bleeding through my my username there, but this this service is called Sidecar that I started.

35:34 So, the website's runsidecar.com. Um and the managed agent service is like Sidecar SMB. And uh we're kind of working on an enterprise level, too. So, the enterprise would be more of like the $50 million plus business versus the small business. But, yeah, man. So, those are my those are my areas. >> Awesome. Well, we will put those in the description as well and in the show notes.

35:56 So, everybody, thanks so much for watching. We'll be back in a few more days with another episode and Phil, thanks so much for the time. >> Yeah, man.