Big Idea

The argument in one line.

The fastest path to a profitable AI product is skipping the polish entirely: ship the ugliest version that works, let real users tell you what to fix, and repeat until you have something people pay for every month.

Who This Is For

Read if. Skip if.

READ IF YOU ARE…

You have a YouTube channel or email list and want to monetize with a product you own, not a course platform you rent.
You have been waiting until you can code before building a digital tool.
You want recurring revenue from your audience without running a coaching calendar.
You have a product idea but have been polishing instead of shipping.

SKIP IF…

You have no existing audience and expect the product to generate discovery traffic on its own.
You need a technical deep-dive on AI infrastructure or APIs -- this covers no-code tools only.

TL;DR

The full version, fast.

Building an AI product does not require code -- it requires a real problem and the willingness to ship something ugly. The four-step loop is: find a specific pain from your existing clients or viewers, build the minimum viable version using a no-code AI builder like Lovable, get feedback by watching whether people return and talking to them directly, then improve and repeat. The best way to sell the tool is to use it on screen while solving the problem your video is about, turning the product into the education itself. Revenue target for a creator starting out: 100 subscribers at $50/month.

Free for members

Chat with this breakdown — free.

Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.

Create a free account →

Chapters

Where the time goes.

00:00 – 01:55

01 · What I built and proof it works

Revenue proof ($130k-$134k/month), overview of StoryOS, and examples of other creators building AI tools.

01:55 – 05:48

02 · Steps 1-3: Find, build ugly, get feedback

The first three steps of the system with real examples: the Discord intro-bot origin, RevTrack feature that got killed, and how to gather feedback with no audience.

05:48 – 10:57

03 · How to build with Lovable (live demo)

Live build of the Channel-to-Buyer Auditor using Lovable. Plan mode first, then YOLO build. App breaks, one-click fix, final score display.

10:57 – 13:07

04 · How to sell: product becomes the education

Sell by using the tool in a video about the problem it solves. Revenue expectations, disclaimer about audience requirement, and CTA to next video.

Atomic Insights

Lines worth screenshotting.

Building an AI product does not mean building the next Facebook -- it means building a calculator or chatbot that solves one specific problem for your audience.
The uglier your first version, the faster you find out if the idea is worth building at all.
User retention is the clearest signal an MVP is working: if people come back daily without being asked, keep building.
Your users will tell you exactly how to improve the product -- you do not need to guess what features to add.
Sometimes feedback tells you to kill the feature entirely, and that saved time is the win.
Selling without pitching means using your tool on screen to solve the exact problem your video is about.
The product becomes the education: viewers want to know how to get access before you even mention a price.
Early adopters who know the product is rough are often the best customers -- they are adventurers, not complainers.
Honesty about imperfection filters out frustrated early churners and attracts people who want to help build something.
100 subscribers at $50/month is $5k/month in recurring revenue -- a sustainable starting point most creators overlook.
Plan mode in Lovable prevents expensive prompt loops that build the wrong thing; always plan before building.
Seeing the first broken version of your MVP teaches you more about what to fix than any planning session ever could.
The creator who ships a useful tool in a video does not need to say buy my thing -- demand reveals itself in the comments.

Takeaway

Ship ugly. Let users build it for you.

WHAT TO LEARN

The biggest obstacle to building a product is the belief that it needs to be good before anyone sees it -- the opposite is true.

Finding a problem worth solving is the only step that requires real judgment; everything after it is execution and iteration.
The ugliest version of an idea that still works is more valuable than a polished version that never ships -- the first generates learning, the second generates debt.
Retention is the clearest early signal: if users return to a rough tool daily without being prompted, the core idea has legs.
Talking directly to users about what frustrates them is faster and cheaper than any market research -- they will name the exact features worth building next.
Feedback that kills a feature is a win, not a failure -- every hour not spent building the wrong thing is an hour available for the right one.
Early adopters who know a product is unfinished tend to be collaborators, not complainers; transparency about imperfection filters for the right first customers.
Selling a tool by using it to solve a real problem on screen generates demand without a pitch -- viewers ask how to get access before the creator mentions a price.
A sustainable product business at small scale -- 100 subscribers at $50/month -- is achievable far earlier than most people expect and compounds over time.

Glossary

Terms worth knowing.

MVP (minimum viable product): The simplest, fastest version of an idea that can be put in front of real users to test whether it is worth building further. In this context, a basic no-code app built in under an hour.
Lovable: A no-code AI app builder where you describe what you want to build in plain English and the platform generates a working web application. Designed for complete beginners building MVPs.
Plan mode: A Lovable setting that keeps the AI in a planning and question-asking state rather than immediately generating code. Prevents wasted credits from building the wrong thing.
Channel-to-Buyer Auditor: A demo tool built live in the video that scores the alignment between a YouTube channel and a sales page URL, helping creators identify why their audience does not convert to buyers.
StoryOS: The creator's own AI-powered YouTube growth platform -- a course plus tools that evolved from a basic Discord intro-writing bot through the four-step feedback loop described in the video.

Resources

Things they pointed at.

05:45toolLovable ↗

04:08productRevTrack ↗

09:39productStoryOS ↗

01:14productJeremy AI (Jeremy Haynes)

Quotables

Lines you could clip.

03:02

“Build the simplest, quickest, ugliest version of your idea called an MVP.”

Punchy permission-giver for non-technical creators scared to ship→ TikTok hook↗ Tweet quote

05:34

“My users literally told me how to make it better rather than me trying to guess.”

Reframes the feedback loop as a superpower→ IG reel cold open↗ Tweet quote

12:22

“Your product becomes the education.”

Four-word distillation of the entire selling strategy→ newsletter pull-quote↗ Tweet quote

12:11

“The irony of this is I'm literally doing this to you right now. Wink.”

Meta-awareness moment -- creator catches himself in the act→ TikTok hook↗ Tweet quote

The Script

Word for word.

Read-along

Don't just watch it. Burn it in.

See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.

metaphorstory

00:00I quit my business to build a new product using only AI. The thing is, I've never coded before. I've never built anything that technical before.

00:08Yet three months later, it was hitting 6 figures every single month. And lots of people have asked, how have you done this?

00:15So in this video, I'm gonna show you the really simple four step system that I used to come up with the idea, to build it, and to sell it so that you can create your own product viewers will beg to buy. And the cool thing about this is pretty much anyone can do it now. So let's just start with a very quick breakdown of what I built for myself using AI and a few other examples of people who've kinda done similar things.

00:37So my product is basically a course plus tools. This, can see on screen right now, is kind of the latest version I'm gonna publish in a few weeks. And the main idea is this, it helps to remove the overwhelm of making YouTube videos, and it walks them through every step of the production process with an AI coach that is trained on my own teaching and my own experience from the last six years of growing YouTube channels for businesses.

00:59It's basically like a little AI me guiding you through the whole thing. There's also a little research tool in there, and I'm working on adding the ability to repurpose your videos into more content, as well as a few other things I'm gonna keep up here for now. Now that's what I made, but here's some other examples.

01:12So this is Mark Simpson. He is an Airbnb expert, and he has built a tool that is making him money right now that's doing a very similar thing.

01:19Kevin here, he's building a fitness app for his viewers that's got content and tools and trainings and his own assistant AI coaching. Jeremy Haynes, he's also built something called Jeremy AI. It's a chatbot his clients are absolutely loving too.

01:32So when I say you can build an AI product, I don't mean you are building the next Facebook or Google or anything that complex. I mean, you're building a very simple digital tool, like some sort of calculator or workflow or chatbot that solves a very specific problem for your viewers using AI.

01:48And then you just charge for it, which thanks to AI has become something anyone can do. So now you know what kind of thing I'm talking about. I'm gonna show you how to come up with your very own idea that your viewers are gonna beg to buy from you.

01:59I'm gonna show you how to build it with a live demo, and then I'm gonna show you how to sell the thing. And all you need is this four step cycle.

02:05You can see it on screen right now. So step one, all you do is you find a problem that you build a product to fix.

02:11So the way I did this was I just looked to what my existing clients were finding hard, and I noticed they were struggling with writing intros. So I thought, right, let's try and fix that problem with AI. Now if you don't have clients, you can survey your viewers, you can hit up Reddit forums, and just ask them what's that one big thing they're all struggling with right now.

02:28And then after you find a problem, you move on to step two of our system, building a very ugly solution. So let me show you what I mean. When I realized the problem, I didn't really know if my idea was gonna be good or bad.

02:40So I didn't wanna spend huge amounts of time making it look pretty or come out with branding or a new logo or website. So instead, I made this disgusting bot.

02:49And, basically, I just put it into my old Discord that I had paying clients in and said, go and use the thing. Look how ugly that is.

02:58Right? All users did was they pop in their intros and look, it would give feedback and kinda rewrite it. So what you do is you build the simplest, quickest, ugliest version of your idea called an MVP, a minimum viable product, and it needs to be like this because if the idea is not good, you don't wanna waste more time building it.

03:15But how do you work out if the idea is good? Well, that's where step three of the system comes in, where you get feedback on it. And there's a couple of ways of getting feedback.

03:24The first simply involves just looking at your product and going, are people using it? And are people coming back to it over and over again? If they are, it's usually a sign it's a good idea.

03:33So you can see my intro bot got used thousands of times, and it was getting used daily, which told me, alright. I probably need put more time into this. And then the second way you get feedback is you simply talk to your users.

03:44So back then, I had a Discord chat, which was great because it meant my clients would use my bot and say, Ed, this thing sucks at hooks. Oh, the intros are too long. I wish it did something else.

03:54So the result of all of this was I could see that the product and idea had legs and my users literally told me how to make it better rather than me trying to, like, guess. But what if you don't have any clients or users to get feedback on? Well, here's another option for you.

04:09I also have a software called RevTrack, and I had this idea for a new feature. It was like this little sales page bot that when people would come to your site, our tool would help them work out if they should buy it or not, like a little AI salesperson.

04:24But before I went and spent ages knocking up a polished version, I created a really quick basic one. And then I went to my community wall and I posted this, basically asking my viewers, could you just quickly try it out and just let me know what you think? Twenty four hours later, you can see I had a ton of comments from people basically saying what they thought of it, what didn't work, what they liked, what they didn't like.

04:44So now I looked at this and went, does it have legs? Should I keep building it? What do I need to improve?

04:48And as a result, we actually decided to scrap the feature, which saved me a ton of time. Now if you don't have a community wall, you can just email your list, and you just say, I've made this little app. Send them a video explaining how it works, and basically say, tell me what you like about it.

05:02Tell me what you don't. Try to break the fringe. Ask what's frustrating and ask, would you use it again?

05:06And then if you don't have email, you can still do this. Just go to Reddit or a school forum, just anywhere your viewers hang out and share it with them and ask for the same kind of feedback. Now now once you've got your feedback, you move on to step four of our system, and you just action that feedback and improve the product.

05:19And that's it. You then just get people to test it again and get some feedback, make improvements, and just keep repeating this cycle over and over again until it gets better.

05:28And this is exactly what took my scrappy little Discord bot and turned it into a platform like this that's helping hundreds of businesses now. I have no idea where this thing will be in twelve months. But as long as I repeat the system, I'm sure it'll work out fine.

05:41So before I show you how to sell this in a way that is perfect for your YouTube videos to promote, how on earth do you build something like this? Well, what I'm gonna do right now is I'm gonna build a minimum viable app to show you how easy this is to start.

05:55We're gonna use a platform called Lovable. So Lovable is basically a platform where you type in, this is what I wanna build, and the thing makes it for you. Now there's other platforms.

06:04I'm not gonna argue about which one's the best. I just like Lovable because it's it's a very easy one to create an MVP with fast for a complete beginner. So the first thing we need to do is use our system and start with a problem that we can build a little app to solve.

06:17So for this example, I I know that a lot of my clients make videos that are not in line with their offer. What that means is the channel is solving one problem, the offer is solving another.

06:26So when people go from the channel to the offer, they don't buy because they don't care about what the offer's selling. It's a very common mistake, very easy to make. I'd love to solve it, so this feels like a good place to start.

06:36So that's the problem. Now what we do is move on to building it. And to do that, I am just gonna tell Lovable the problem that I wanna solve.

06:42I'm gonna tell them my rough idea. I'm gonna say, can you just help me work out a plan to build a tool that could solve this just like I was talking to a human who knows about this stuff? Now before you hit build on this thing, just change this button here to plan mode.

06:56You don't wanna be having planning conversations in build mode with Lovable because it'll just build and it'll cost you more and slow you down. So Lovable will then probably come back with a few questions for you. All you do is just answer them.

07:08Have a little chat with it and just throw ideas around, and then eventually, it's gonna spit out something like this that's a bit of a plan. So we can see the plan it came up with. It's saying channel to buyer conversion order turn.

07:18It suggested that we've built it in different phases, not in one big prompt at a time. And it's also suggested how we can maybe help solve this problem.

07:26This is very smart. Now at this stage, have two options. You can read this document.

07:29You can edit a ton. You can keep chatting back and forth, or you can do what I do when I'm building an MVP and just go YOLO, hit the build button and see what comes out.

07:39And I actually think for a lot of people, for an MVP, the YOLO option, this is the easiest way to start. It's not the best, but for me, what I found as a complete beginner was once I saw the first version, I suddenly realized, I know how to improve this.

07:52I know what to fix. Without that, I have no idea what the heck was going on. So this here's what Lovable made for me with a little ten minute chat.

07:59Didn't really go into, like, styling or bother with anything other than the basics. There's no logo on there. So now what we do is the fun part.

08:07We try and break it. Now I know that sounds a bit funny, but what you wanna do is pretend you're a user to see if anything doesn't work or if it makes sense.

08:16So let's pretend we're a user for our app here. The first thing is it wants me to put in my channel URL, so I'm gonna do that. Uh, and I'm gonna hit this button, and it's gonna go and fetch the details from my channel.

08:26And look. That worked. It's pulled in some data.

08:29Happy days. And then I'm gonna put my sales page URL in, and I'm gonna hit calculate. And it broke.

08:36Okay? So what we do is we just click fix, and then we wait, and it says fixed.

08:41So let's just try again. So I'm gonna put in my channel and the sales page, and, yep, this time it's worked. And this is cool.

08:49It's giving me a high score to say that my channel, my sales page, stand a good chance of converting people. But it also says, these are the problems my channel solves, and these are the problems my sales page is saying it solves.

09:03Here's what's working. Here's what you might need to fix to get more sales from your views. Oh, and then, uh, it's given me some content ideas, which suck.

09:09So I wouldn't use them. I'd probably delete that feature from here, but honestly, right now, you kinda wanna not worry about the little things that are not perfect because I think this is good enough as an MVP to move on and get feedback from.

09:21So actually, if you wanna play around with this, I'll leave a link below, Roast it. Break the thing. Comment what you think about it in the comments below this video, and you guys can see in live how this works.

09:31Now, at this stage, think it's pretty natural for the creator to panic and worry about sharing something that is so unpolished. Because I imagine they'll be thinking, what if it sucks? What if people think my main product will be useless and what if they complain?

09:43But here's how I get around this. So when I launched StoryOS, the app I built, it is a little bit iffy. It's still a little bit clunky until the new version comes out.

09:52So in my marketing, I just said, look, this thing is going to break. It's not the best yet. If you're easily frustrated, don't sign up.

10:01Just wait. If you like an adventure and and you wanna help me produce something that will help other people, come on in and buy it. And I've only had one real complaint, and it was actually my fault because I missed an email.

10:14Now what's cool about this is when you come in with that mindset, the people who come in, what I'm finding are, they're little adventurers, and they're quite excited to report bugs and give me ideas. So they're just adding value and paying for it at the same time. So the trick is to not hide and pretend this is supposed to be amazing.

10:31Just be honest, and you'll find that early adopters actually get a kick out of this too. Anyway, when you're happy with your demo, you just hit this publish button up the top, and it's gonna give you a link that's a bit ugly. You can link it to your own fancy domain layer, but there's no point in the testing phase because it takes more time.

10:47So let's now say our feedback's coming and we fixed a few bug. Now what I would do is I would say to Lovable, build me a login and a payment system. That's it.

10:57So let me show you how to sell the app that you just built to make some money. And I think I probably need to start with a disclaimer. So I've managed to make a lot of money from this because I've built a following for years.

11:07Don't expect to launch an offer and make hundreds of thousand dollars every month. That's very unlikely to happen. But the exciting thing about this is you can sell your tool for maybe $50 a month.

11:17And if you get a 100 people signing up over the year, that's 5 k in your bank account every month. And I think that's a great place to start. It's also a more sustainable way to grow a business too.

11:25So here's how I'd sell this thing in your videos in a way that I think will get you a ton of sales and actually make your content more interesting too. I've been doing this for years. It's actually helped make me millions without ever really needing to say buy my thing in a video.

11:38So what you do is this. Because you followed my system, your product solves a problem your viewers care about. So the next time you make a video about that problem, all you do is you use your tool in the video showing people how to solve the problem.

11:52That's it. So when I was developing my AI tools, I made a video using a tool I built for clients in this way, and I said, look, you can use AI to solve this problem and just showed it off. And then everyone who applied to my program after that said, I want that tool.

12:06How can I get it? Now, of course, the irony of this is I'm literally doing this to you right now. Wink.

12:11So as an example, if you have a health tracker, what you do is just show people how to track their health using it. If you had an AI chatbot that help people bake, you'd show them how it would give you better baking instructions. Your product becomes the education.

12:24It is so effective for sales and hype. Of course, you can email and you can post social media and you can build marketing strategy around this, but I don't actually think there's a better way to sell a tool you've built than just using it in your video. That does bring us to a problem because in order to make big money from this, you really need to build up your audience, which is hard.

12:41But there is an easier way to dominate your niche now that's already helping businesses make insane income. So watch this video next. I'm gonna show you exactly how to biddle biddle, how to build a channel that could one day make you a million dollars.

12:53By the way, if you're interested in building these kind of things or products and want more on it, can you let me know in the comments? I don't really wanna be the app building guy, but I do like sharing what I'm working on. If it's inspiring for you and you want more of it, just let me know below.

13:05Maybe I will.

The Hook

The bait, then the rug-pull.

He quit his own business, had never written a line of code, and three months later his AI product was clearing $100,000 every month. The opening frame is not a claim -- it is a Stripe dashboard.

Frameworks