How to Make $20,000/Month with Claude Code

So, should you as an AI agency owner be looking to transfer over to some kind of retainer base model with your agency in 2026 and beyond due to the Claude Code revolution and AI operating systems, we are seeing AI agencies make leaps and bounds towards much, much cheaper and much faster development for the clients. And particularly small business clients because they're the the most nimble and easy to deliver for. Enterprise, as I've kind of always said, is not necessarily our battle. They're going to be handled by someone else. We're looking at the small to medium-sized business market. And it's a very important conversation to be had right now because with the like we're just seeing the cost of development drop and drop and drop and drop and drop and drop. And as I've said, these new tools always coming out. It's in our favor. We are the winners. And a very, very, very, very, very long time too much complexity for any business owner to go, "Oh, you'll just jump on Claude Code and start doing my my automations myself." This And you're going to point you're going to point out the exception to the rule of like, "Oh, no, but I know my I know there's at least one guy there's We're in a bubble." The 99% of business owners are not hopping on Claude Code and trying to figure it out. So, that brings us to a question of if we have these incredible tools that allow us to not only audit our clients and determine what automations and agents they need, but also deliver it rapidly. And also do the testing and the quality assurance and sort of ongoing management optimization. The sort of full stack of opportunity identification to development to testing to deployment and to everything beyond that in terms of management and then optimization. As I've talked about before here on the channel, we are starting to see the the value of agencies. As I said in 2026, this would happen is that the value of us agencies is going to shift into these more sort of end of the chain services like management and optimization. So, sure now everyone can build this automation and this agent and this because Claude Code's going to make it quite easy, but then you as an agency is going to have to specialize in the management and optimization of it. And also the training aspect as well. I said 2026 is going to be the year of training and I still think that to be true because there's a massive human issue that that needs to be solved. So, with all of this wrapped around, where do retainers fit into that? And this is going back to something that I actually experimented with called Morningside Automation, which was when we had Morningside AI back in 2023. And we were doing custom coded development with the clunkiest Open AI APIs you can imagine. Extremely hard to get anything really basic out of them in the early days. But, there were still there were still people like ready to throw money at it, which was great. But, it was very difficult to get anywhere near the kind of ROI you can get for clients these days in terms of like money that they have to put up and then also in terms of the time for us to put in as a development team. And then of course, the actual efficiency or effectiveness of the end solution. So, instead of doing these large projects, 10, 20, 30, 40,000 dollars, why don't we try to flip it into a Designjoy model? You guys might know Brett from Designjoy. So, he kind of blew up on social media because he was running this one-man million-dollar business. His model was to have a retainer where his clients paid him monthly 3, 4, 5,000 dollars, something around there. They're getting design services and they get to pick what two or three things per month they get designed. So, it's an ongoing thing which is like when you have new design needs, you let us know, they put them in the queue and then he's just doing the design as a designer. We had that same thought of like, hey, could we transfer that model over to automations and agents? And we quickly realized that the tech was just not there. It was way would have to charge so much. We're looking at like 8,000 USD per month per client. If we did that, then how many automations could you give? And then how do you scope a given automation as like large or small? There's a lot of smaller things that were quite difficult to iron out with it. And ultimately, the cost just didn't make sense. It was still taking us too many hours and far too much time with our developers at Morningside to deliver a given automation and the testing and the back and forth that you'd have to charge in 20 or 30,000 dollars a month for you to be in the green. And while we were way too early there, about 3 years too early, I think we are seeing the the rise of these kinds of retainer models like this that allow for that kind of model to actually exist and not only exist, but be very profitable and and valuable for both sides. And businesses get a ton of value out of it. There's a guy called Tyler in my accelerator program and he's been flying out and selling AIOS installs, Claude Code setups for business owners and going in person getting them set up. And then he's coming back home and he's having this Claude Code workspace set up that's integrated into all of the different platforms. And then he's using that as a development engine or development platform to very quickly build automations and agents on top of that AI operating system. So, because he's fully contextualized, because he's tapped into all of the different programs, he's got the right API keys and access, then he is just able to go like charging through these development workflows of just being in the loop saying like using a slash explore command is what we use it in my program. Slash explore, I want to explore how to build this automation for them. And then he's able to walk you through step by step by step by step. And that's the human being in the loop. And if you guys have been playing around with Claude Code, you know how quickly you can go from an idea to a working prototype or even an actual production-ready automation or agent or even just like an app, internal app. When you have these kinds of tools and you also as a founder do not need to have like three developers underneath you. You can literally do this on your own if you've got experience of doing it yourself. The learning curve is gradual enough with Claude and just sort of helping you to figure it out that there is this incredible opportunity for a very, very lean agency, perhaps even just one person, where you are saying, "Hey, look, this is going to be two and a half grand, three grand, four grand." And Tyler's selling these ongoing automations retainers and and packages for two and a half grand per month, 2,500 USD. And I still think that's quite low, but he is working his prices up over time. But, for that kind of offer in a business owner's mind, they're looking at like 2,500 dollars. Now I get some AI guy in my corner and I can ask questions to all the time and he's going to build me whatever I want. And there's a new thing or two every month. There's like one or two new automations every month. And what Tyler's doing is when he gets there in person, he is building out the first part of it. So, not only is he setting up the AI operating system, he's building out the first use case that they've decided on prior to the actual flyout. Say, what is the one thing that if I can solve it, it will make this worthwhile, make this package, make this time you've spent. So, not only is he setting it up, but he's proving and validating the value of it right in front of them, setting up the first system in front of them. And then he goes home and he's like, "Great. Now on a on a retainer basis, what are we doing next?" Okay, put it in the queue. You can add your requests. And then when you really think this through, it doesn't take a genius to realize where this goes. But, you will literally have a fully contextualized AI development platform where they will come and make requests. It will get kicked into an automated workflow. It will research the whole code base in the business. It might go and ping messages to certain people to get more information. That would be a bit more advanced. But, it will be able to plan the whole build. It'll be able to request test or dummy data. Or that's part of the input perhaps. Like you've got a whole queue and it's like as they put requests in, you'll go, "Hey, would be good if we could get this this once they move into like we've got our prerequisites received." Then you can kick it into literally an automated development pipeline to build it out, to test it rigorously. And then you can say, "Hey, look, yes, we think we can do this. That'll be 5,000 dollars." They click a button, they pay, and then literally in a few hours they have it up and running or the next day you're walking them through how to do it. It's just shrinking the timelines down so much. The time to value is getting so short and the difficulty and technical ability needed to deliver these kinds of outcomes for clients is just massively dropping due to Claude Code as well. And so, going back to the offer standpoint, a business owner looking at 2,500 dollars or 3,000 or 4,000 or for even 5,000 or 10,000 depending on how much proof you have to back it up and how big how sort of how many automations you're packing into the package. They look at that price and they're like, "I'm already paying a marketing agency like three or four grand a month. Yeah, I mean, that that seems like a great deal. A AI guy in my corner, automations every month for like two and a half or three, four, five grand. That's that's great. That's like the best band-aid. I don't have to fork out tens of thousands of dollars up front. It's just like bang, there we go. We're going to get the first thing guaranteed. They're going to come out and set it up or they're going to solve the first problem for me and that's guaranteed. Like there's just so much offer engineering that can go on here. And like I was talking to Tyler about, there is two directions you can go from there. And that is I want to build a retainer base agency with that's that's very wide and I have lots of clients and they're all paying me a small retainer. Or you could think, "Hmm, more clients, more headache." Why don't I go deep? And you can get some incredible case studies if you just keep chopping through and you start to stack onto those retainers. 2,500 might be the start, but for everything you build for them, you are adding on management and optimization fees on ongoing basis. And if you start to niche down to a specific niche or industry or type of business, this stuff is going to compound ridiculously fast with the learnings from one carrier over to the other. Hey, Claude Code, can you export this so I can give it to my other client? Bang, hey, here's this from another client, please install it. It is just going to be like the craziest flywheel of growth and learning for you as an agency owner when you can get into this kind of retainer base model and you're just stacking on those. Hey, for for the system that we just built for you, it's going to be another 300 dollars or 500 dollars per month to maintain it. We'll also chuck on an extra 500 bucks for optimization where we'll be improving it over time. And so, these for these main systems, you'll be able to set up a delivery package or like, hey, this is we've sort of already figured out how to optimize this given system. But, yeah, getting in the weeds, but I think you guys will get what I mean here. And it's extremely exciting. You can have four clients paying you 5K per month. And you just constantly tacking onto those retainers. And you're only worrying about four people. You're literally having to bang out one, two, maybe three Claude Code related automations, which we know how easy those things are to do now, per client and you're making 20K per month. I think this is the most exciting time to be in the AI agency space, to be building with AI. And so, I thought I'd make this one just to paint a bit of a picture of where you guys can go with your agency now that we have these things. But, the key piece is automated auditing and automated development. If you can get systems and workflows built into your own cloud code to deliver these kinds of things uh rapidly, then you're going to be in a very, very good spot as an agency. So, if you guys want to see the full video breaking down how to uh sell AIOS um based off a lot more information on Tyler's story, I'll put a link up here if you guys want to check that out. But, aside from that, that's all for the video, guys. Thank you so much for watching, and I'll see you in the next one.