$3M/Year From Tiny AI Tools (Here’s How You Can Too)⏐Ep. #228

Unveiling John Rush's $3M AI Tool Empire What is the combined revenue of all of your projects? Speaker 2 It's around 3 million. Speaker 1 That's incredible. Speaker 2 It runs on autopilot. So your bot is doing 1,000,000 ARR. It has been doubling month over month. 110 KA month. Little tools are easy to build in the age of vibe coding. You can build them in one evening. Pretty cool hack. I will show you. It's not a billion dollar idea, but if you're a regular person and you want to build $1,000,000 business, we have so many low hanging fruits. Speaker 1 This is so good. It's just right in front of our nose. So if people say I don't have an audience, doesn't matter, that's not an excuse. What is your go to tech stack for building all of these tools? If you're like me and find yourself bouncing around from idea to idea and testing all kinds of different concepts, this episode is for you. I found this guy on Twitter named John. He has 24 different projects, mostly websites, and they make him $3,000,000 a year of annual recurring revenue. And you know what? Almost all of the things that he's built or talks about can be vibe coded with common AI, AI tools that I talk about on this podcast all the time. So we get super nitty gritty. He gets very specific on how to find good business ideas, how to validate good business ideas, how to launch them and how to grow them. What are his favorite AI tools to use? How does he build agents? How does he build agents for himself and then sell those agents to other people? You don't need employees. You don't need to be a software engineer. You just need to listen to this episode. Please enjoy. Building AI Tools: John's Self-Suffering Approach You have 24 different products plus projects running right now, and you're trying to automate as much of it as possible possible. Speaker 2 Yeah, yeah. When I saw AI happening in 2021, actually pretty early before most of the people, I started automating some of the things already with GPT 2 and GPT 3. And then in 2022 when they launched 3.5, I think I kind of felt that, you know, the, the traction they have on model improvements is so great that I should bet everything on AI becoming really, really strong. So and then I thought like, if that's the case, if that becomes the future, then I should bootstrap and I should run smaller companies and I should create suffering for myself by not having people to delegate to. So that I have to do everything by myself, from accounting to paperwork to design, development, growth, marketing, everything I had to do by myself because in the past I would just hire people for that. And now I did by myself and I learned all those crafts and I sew the paints and I had these ideas for 2080 rule. Like what can I do? Like 20% of the work that gives 80% of the result. And I automated that with AI, starting with operations, then with growth, then with development and with, with everything. So that was kind of my, my story. And now I'm, I'm here with a lot of agents and I used agents to grow my other agents. So it's kind of the loop happening. Speaker 1 So these 24 things, are they some of them agents that work for you and then it's also agent that you're reselling to others as well? Speaker 2 Well, they all start as my internal agent and I run them for a long while so that I'm happy with them myself. And also it's easier to run a project if you're the only customer because you can risk, you can, you know, move fast and you don't have all the users who would complain. That's why I run for a year or two or something, three years internally. And then once I kind of like the whole thing and it's stable and I don't iterate that much, then I bring all my friends because I have a lot of friends who are building startups, and then I launch publicly for everybody to use. So eventually everything I use internally will become public. That's the end goal here. Speaker 1 I love it. You find a problem, you solve it for yourself. You solve it for your friends. If it's still surviving then because surely some don't make it, then you launch it to the general public. Speaker 2 Yeah, yeah, exactly. It sounds easy, but you know, it's the way I create suffering for myself. It's really hard to replicate because nobody really wants to suffer that much. Like I have so many transactions, for example, like a lot of product and transactions. And I still do accounting myself. I hate those days when I have to do that, but I have to do that because then I understand whether the product I'm, I think the H&M building is good enough and, and I can evaluate it. And I, I think when they were somebody else comes into the game trying to build the same thing, like they have too much time. Like I don't have time at all. So I'm, I'm kind of a typical user of the tool, while the other people who would build that thing, they would understand accounting really well and they would kind of, you know, have a lot of time to click the buttons and to use the interface. And I don't have that time. So that's why all my agents are so simple on the surface and still complicated under the hood because my vision is that the end goal for every agent is that I turn it on and I leave it and it just works as great employee that self managed agentic and it asks me whenever it has questions and then goes back to do the work. How John's SEO Bot Generates $1M ARR What's an example of one of your agents that just works perfectly but it's very complicated under the hood? Speaker 2 Yeah, it's like Sao bot for example. I think that's one of the best agents I have. So it runs on autopilot. Like I even forget sometimes that it runs for some of my products, but then I get emails from it where it says now this week I've done this. So I built these articles, I built this how to guides, I built these tools like mini tool. It can even build little products that will drive Sao traffic. Then they can also collect all the news and then build some newsletters and things like that. So it does a lot of the work that Sao department would do. It can also ask me whenever it's not confident about things. It would ask me like I have a list of these things and can you help me to pick what's the best from this list? Like I have 10 items, pick 5 and then I pick five and then it does the rest of the work. So it's able to involve me whenever it's necessary, but then when it learns me and it understands what I want it eventually over time, I'm not involved at all. First you're involved every day and then week and then month and now once every quarter. Speaker 1 Well, it's like training an employee, right? Speaker 2 Yeah, exactly like a good employee. Speaker 1 Yeah. Are you willing to share revenue numbers or anything like that? Speaker 2 Yeah. So your bot is doing 1,000,000 ARR. I think I shared that before and. Speaker 1 Crazy. Speaker 2 It has been growing. It has been doubling month over month before summer and in summer the growth stopped. I think it's kind of been 110 KA month during the summer and now I think it will start growing again in September. Good for you. It was one of the first agent that kind of went into being a real agent, because all the other agents, they say they're agents. Yeah. If you go in there, it's kind of a tool that has a lot of AI in it, and you have to sit inside of that tool and use it. So it's just more convenient UX where with my agent from the start, I went the other way around. So it had no buttons at all. It was just turned on. It was a check box. Put it on and it starts. That was the only baton I had in the first month. And then I was just adding more because some people ask for certain control and I don't want to lose out on agencies and pro users and they want more control. But I'm kind of not really into it. I think I in long term, I believe that people don't want that much control. And the reason they want that much control now is that they have too much time. Like if you're an employee in a company and you have 8 hours a day to do Sao only you have a lot of time and you don't want to have a tool that just has one batch and you turn it on and what do you do then, right? Speaker 1 It doesn't appear like you're adding enough value. Speaker 2 Exactly. Exactly. But in fact, you're adding value because you have an experience and you can help the tool once in a while. And those little moderations will be crucial because it's like the quality of the ideas wins over the quantity in this case. But but people are not ready for that yet. And I took a risk and I went into that early and a lot of people didn't believe that such interface will work, but it work really well and I think it's one of the well known agents. Maximizing Revenue with a Connected AI Tool Portfolio That's amazing. What? What is the combined revenue of all of your projects? Speaker 2 It's around 3 million. Speaker 1 That's incredible. I mean, the 8020 rule is alive and well even, and you can't get rid of the 8020 rule, right? It's everywhere. Speaker 2 Yeah, like 3 products do most of the revenue and then everything else brings the rest. But I think it's, it's possible to grow the rest too. But you know, I'm, I'm not doing these things in a typical way where I want to just maximize the revenue or I want to find the winner and put all my effort into the winner, because that's what a lot of people do who run a lot of tips. That's not my game. My game is to build all these agents that play together. And then since I have them all, I can connect them. And that's also cool, right? Because now suddenly agent, one agent like I have SEO bot and I have another agent that's called Rapify and rapify builds little tools. It's like a vibe coding tool and so your bot can call, repify and ask it to build a tool and send it back. So now I have one agent, you know, asking the other. Like imagine in your office you have SEO marketer and then you have developer and it come one comes to the other and asks for the for help. And I have that across all my agents and that's why I'm not hurrying up in growing them. I'd rather when I have them all running like 100, maybe 1000 customers so that I have enough people to test, to ask questions and to AB test things, but not more because when it's more it's a hustle. It's a lot of support, a lot of legacy to support. So that's my game now. Speaker 1 And you're able to cross sell across all of your projects, right? Like, do you do a lot of that? Speaker 2 Yeah, Yeah, I think that's sometimes that's half of my sales every month. Wow. For new users, For the new users, half of them come from one tool to the other. That's why when they land 1 user in any of my tool, I have some tools that are making like 1K a month, like very small tools. But at the same time, they are so cheap and they have trial and people enter those tools and they might even cancel that tool, but move over to the other tools. And it's very rare when people leave my universe. So if they enter, they, they once you use one tool, the other tool, third tool, it's really hard to leave because it's, it's not locking you, but it's convenient. And if you're busy, it's really convenient. Speaker 1 The one KA month MRR you get from one of your small projects that could be contributing another three KA month MRR in another project. Speaker 2 Exactly because of Cross. Speaker 1 Or or across seven other projects, right? Speaker 2 Yeah, yeah, that. That's why I have so many directories and a lot of people think why do I build directories? Is that that much money there? Like most of my directories are making one KA month, some make 10K but most of make just one Ki. Think like directories are so good at driving traffic from people who would otherwise not land on your website. Speaker 1 You know, with your SEO business, you know, one of the hard parts about SEO, either doing it for yourself or selling an SEO SAS product, is how long it can often take to work. How do you manage churn in that business when people have, you know, they want immediate results and that's just not common in SEO? Have you been able to overcome that, and if so, how? Overcoming Churn in SEO Products and Customer Expectations Well, it's hard, of course, because a lot of people have expectations and then they see somebody on the Internet with a story that they stored yesterday and now they have traffic and they expect the same. But I think in my case, I'm trying to build my products in a way that they also educate people on what to expect. And what is this? Because most people who use my products are not professionals. Like professionals usually go for more complicated products because they want to spend their time by using them. And my products are for busy founders the most my customers are founders who are good at everything a little bit, right? And, and I have this system of emails where I take some little KPI's, little like for example for SEO articles, I send the data every week on what has been done. And maybe there's some views because you don't get the clicks right away, but you get views right away, right or not right away. Within a week, usually you get views and I send the views and then I send the clicks. There's stuff to send to the user to show that things are happening. And that usually works because also it's about the price. If you go for the agency, you pay like 5K or something like that. And, and it's two months after and you don't see the traffic and you, you feel like you've been scammed. And in this case it costs like $20.00 a month. And it's not like you, you're expecting any result for that money. Oh. Speaker 1 Man that's so good. How long are your customers taking to see SEO results with your product? Speaker 2 It's from weeks to month. Like I think Google has been changing a lot in their algorithm. So like 3 years ago you would never see any traffic on your article until like 2 month, maybe something's three month. And now I think the same goes for weeks. So now in the weeks you can see views and some clicks like 2-3 weeks. Speaker 1 What is your go to tech stack for building all of these tools? John's Tech Stack and Google-Based Idea Generation I use a lot of the tools, so my main tool that I use for prototyping is lovable or other like bolt or cursor or pincer. But Lovable is my favorite I think for byte coding. And I use Lovable for prototyping and often I use it to experiment with different ways of building the UI for this product. For example, once I move past the prototype, I use my own AI coding tool called Morsex because that's where I go for serious stuff where it probably becomes more complicated. So it's like 2 tools. So first one for MVPS, the other one is for full product. Speaker 1 What about for automations and other stuff you like make? Speaker 2 I use make, I use Zapier and I use Lindy AI, so I use the whole tree. Speaker 1 All right, we're like getting, I'm like so intrigued by what you're doing. We had talked, John about you, showing us how you find ideas, how you build ideas, and how you grow ideas, right? Speaker 2 Yeah. Speaker 1 Why don't we get into that? Speaker 2 All right, I will just share my screen for that. Yeah. So for the ideas, I have this framework that I can make for myself, and it was pretty popular in the Internet because I think it's very simple. I think the best ideas are the ideas that scratch your own itch. So basically you're doing a job and then you have a problem at the job and you want to solve that and you come up with ideas for that, for solution for the problem. That's what I usually do. 99% of my ideas are born that way and then you have some other options for finding ideas. And usually my second option is to see that people are asking for something and then see whether a lot of people ask for it and then build that. So I will just show how I do both and. For example this is. Speaker 1 So good. Speaker 2 Like if I want to figure out what what people are asking, I do a pretty cool hack. I will show you. So I go to Google and I do this. So website builder that. So you do this and you see the code that doesn't use AI that accept payments that export the HTML. Like that's interesting. Like some people want that. And I actually know a guy who built a website builder that can do that. It became big so, and you can do it farther, like Website Builder 4. And now we see the audience. It can be small business, photographers, artists. That's interesting. Speaker 1 And we're assuming this is ranked by like search volume, right? Oh. Speaker 2 Yeah, exactly. And and if you see it here, that means a lot of people search for it. But you know, we you have the next step here where you can take this into the keyword research and validate, for example, website builder for artists. And now we can go into the keyword research and we can see how much traffic is going there. Yeah. So here we see it's around 10K visitors a month, which is not a lot, but if you look further, you have this one, you have this one. So you have a lot of requests which are also and for the same thing and they're also at each 10K10K10K. So in total it's probably 100K now. So it's pretty good traffic, yeah. And also you can see where exactly people search for and you can sort them based on the monthly searches. And now you see how people frame it because that's also important. And you can see that they that they frame it in basic way. So the website builder for artists. So now you can probably count with the domain idea, which could be in this case, it's probably too long like website builders for artists. But in often cases, you could just find a keyword that turns into the idea like I have this, I've been playing with the other keyword now and it's tux haven't. So I just tested it now before this call and it's a the 100K searches. It's a lot of searches. It's low competition. It means that people don't really place ads for these keywords. Speaker 1 And that's a, that's a very high value keyword, right? Yeah, wealthy exactly are searching that keyword. Speaker 2 Yeah, like people are willing to save money. It means that you can earn money where others are trying to save money, right? And then you can go and test this, all the ideas here. So you first find them on Google and then you see you make a list. So I usually play with it for an hour. I make a list in the Excel for like 50 or 60 ideas and then I bring them back into the keyword research tool. And then I see their traffic. And in the end, you have a picture what you should work on, because if you don't do that, then you're kind of in the dark and you have no idea. You're guessing. Yeah, you're guessing. And the problem with guessing is that it's like in stock market world, whenever you guess, usually you're doing something wrong because that's how our mind works. The intuition never helps in this case. Speaker 1 I would take this a step further, John, like So what if you're only listening to this? What we're looking at is just the auto complete on.google.com website builder 4:00 And from the top to bottom it goes small business, free photographers, portfolio artists, nonprofits. So you can go to the keyword tool. In the artist example, there were 92 keywords that had over 100,000 searches per month. Cool. I would do that for like the top ten small business artists, whatever. And then I would go to perplexity and I would have a prompt that says all I want to know, I want a grid. I want to know how many website builders are there that specialize in being 4 artists. Like not a Squarespace, not a Wix. It needs to be a niche website builder that is 4 artists. I need to know how many there are. And then on the second column I need to see like what's their estimated monthly traffic using similar web or something else, right? Run a deep research prompt on all ten of those. Website builder 4, XY or Z. You get a table, you put it in Google Sheets, and then you take the search volume and how competitive it is and put it in column C&D. And then you can clearly see not only where is the search volume, but how much demand is there for each. And then you can easily sort and say, Oh yeah, this one has the most demand and the least supply. I'm going to build here. Speaker 2 Yeah. And the thing here is that it's very often search IDs are not really heavily targeted by anyone because it's a small niche. And if you're AVC backed company, it's, it's too little for you. Like it's not a billion dollar idea, right? But if you're a regular person and you want to build $1,000,000 business, that's of the market, right? And that's why like we're now living in a world where where we have so many low hanging fruits like that, because the VC backed world, surprisingly is all going after the same ideas because they have to measure the market and, and that's super important. And all, all these other ideas are just unserved. And there are a lot of them. And I see now a huge movement where, you know, solopreneurs and indie makers and, and a lot of the people are just, you know, targeting this. Speaker 1 This is so good. It's like it's just right in front of our nose. It's just google.com, but it's it's like a prompt. It's like a Google prompt, right? And looking at it exactly. Speaker 2 Yeah. How to Validate Product Ideas Before Building Them And so I do that, and my next step is usually to validate again. So basically when I built something, I'm trying to convince myself that I shouldn't build it. You should behave that way because if you try to. Speaker 1 Like a negative bias just to be exactly, yeah. Speaker 2 If you're in love with your idea, then you will do the opposite. Whenever you see the data, you will interpret the data in the wrong way. You will think that well, it's not that bad. It's not that bad, yeah. But you should try to kill your idea because there are so many ideas. You rather kill, you know, 20 of them and 21st works out and then you don't waste time building it, right? Because building it takes a lot of time. So then my next step is usually I go to my Uniggar platform and you can go into any other vibe coding tool. But I recommend to use like websites builders for websites rather than vibe coding tools because the website builders are made for landing pages and they have good SEO optimization and things like that. So go for Framer, Unicorn web flow, Don't go for lovable or BALT or others for this because they are made for apps, right? For where you have the user login and things like that, because people have mixed those things. And I think this is really important. Like you have landing pages and you have your web app and for landing pages, use landing page builder, right? So I usually go here and then I kind of pick a wait list. So because I haven't built a product yet, so I just pick a wait simple wait list, for example, this one and then I go for create and then I say website builder 4 or this Yeah and it will just build a little landing page with a wait list. Now, yeah, it has to be super simple. Like your goal here is to just see whether anyone is willing to put their name or like here it puts phone number. I don't need that. So I just remove it. Yeah, name and e-mail is enough for me. So that's it. So now I have my landing page show it to potential customers and C normally don't want to join wait lists because it's silly, right? But still, if you show it to 20 people, at least one person would join because some people are cursed and they want to get to know about this once you launch. So it's it still works like you're gonna lose a lot of the people who otherwise would sign up if you had the product. But right now, our goal is to understand whether people are willing to put any effort into consuming your product. And this is good enough. So I share this on social media, on Reddit, on replies. So I search on social media for people rather than talking. Speaker 1 About this question. Speaker 2 Yeah, yeah. Or sometimes not even asking about the question, like sometimes you can be a bit more spummy. If you can't find the actual questions, you can find the right people. So in this case these are musicians and you can just search for anything around musicians and software. Doesn't have to be the website builder can be something else right? You can search for that and then you can reply the year and it's not exactly on point but it's not that far. It works pretty well if you do like 20 or 30 replies like that and then you do DM's to the same people and then you maybe find their emails and send called emails. So eventually if you do 100 actions towards a cold out region and cold replies, you will see either 0 signups for your wait list and if it's zero, it's a good sign. It means that it's not that important of a problem for people so they don't care that much. That's good. Speaker 1 Yeah, if you get 0, if you get 100, it's good because you're learning and you don't want to waste time on a project that nobody wants. Speaker 2 Yeah, yeah, exactly. But if there's one or ten or five, then you have an answer. Like if it's ten, then it means people really want it. And sometimes you actually hit the right spot and a lot of people say no because that doesn't exist. And people all wanted it. And then this, they sign up. I had that. I was gonna ask. Speaker 1 Did you have any examples of that happening to you? Speaker 2 Yeah, so I had a case with when I build dev hunt. So Dev Hunt isn't similar to Product Hand but for Dev tolls. And I just shared a tweet where I said that I want to build this new launchpad only for developers because Product Hunt doesn't serve developers as it serves the marketers. And if 100 people sign up for this, I will build it. And I had 100 people signing up and people were actually sharing the link to their friends telling them, hey, come help. Speaker 1 Me, they want for this. They wanted to help you because they wanted to use it. Speaker 2 They wanted to reach the 100 marks. They build it, right? So they wanted to help themselves, actually, not me. Yeah. And that's right. The best thing here, right? So, and I do that every time. So basically all my products, even before I build them for myself, I still kind of test them on public using these methods because I know I'm not building something that I will never be able to launch to the public and works really well. So I do it through social media and I'm kind of found a way to do it through social media even before I had followers. Now I have a lot of followers, but in the past I had almost none and I still found a way because the interesting stuff about social media is that right now they don't really respect the follower count count as I did in the past. So if your content is interesting, they will show it to a few people and if they find it interesting, they will show it to more people and eventually it can blow up even without followers. So you should try that. You should always try that and you should test. You should test it for messages on your kind of launch tweet. But if that doesn't work, you can also do ads. You can send paid traffic to your tool, and paid traffic works too sometimes. So in the past I was using Google and Meta for that. But as with everything else, I end up building my own tool for that. That kind of does it for me. Speaker 1 Has there ever been a time when all of your indicators were shouting yes, yes, launch this, and then you did this research stage or you actually launched it and it just did not hit? And if so, why do you think that happened? Speaker 2 No, that never happened really. I had the opposite. I had the opposite where everything's sad that the ID is bad, like there were no sign ups and all my friends were saying that this is a bad idea. But when I launched, it worked out. So I had the opposite. But if I get the validation through this matters that you've shown, I go for the last step. Actually there's one, one more step that's just to make sure 100% is going to work. I ask for pre bookings. So I give people 70% or 90% discount if they buy now. And surprisingly, I had cases where a lot of people sign up for a toll, a lot of people, but nobody wanted to pay. Because you know why it happens? It happens because sometimes you hit certain audience, for example, students, they would be signing up for the thing and you would think, wow, a lot of people, thousands of them. But then you ask them for a dollar and they don't want to pay because they're used to finding things for free, right? And you never know because you have no idea. You have some emails in your list and you have no idea who are those people, right? That's why if you ask for a payment and you don't get at least 10% of your wait list to execute it, it means that it won't be easier after, right? And I had cases. It's like that. Speaker 1 If I put up a landing page for a $20.00 a month tool that people are signing up for and they show demand for, and then you launch it and less than 10% of people actually pay for the thing, it's not a good sign. Speaker 2 It's not a good sign. It's a very bad sign because it it means that you're missing the whole thing, like you're not selling to the right audience. And that's a big problem because it's not something you can just change because if you go and after another audience, it's a new game. Yeah. You have to like, maybe all this traffic you saw on keyword research was from the wrong audience, right? Yeah. And you don't know that because it's just numbers. And that's kind of the problem. Like you get excited about something because there's an interest, but it's wrong people and you can't like now you have to kill the idea because most likely be the right people. Don't search for it. Speaker 1 What if you're just priced too high? Like it is the right audience? 7% of people buy it and it would be higher if you were priced at $9 a month instead of 19. How do you know when that's the case? Or do you? Speaker 2 Yeah, you offer them a discount. Speaker 1 Let's say even with a discount like it's normally at 49 and you're offering it for 19 or 9, right? Like let's say that is the case. Speaker 2 They will tell you so you offered them a discount and they tell well you know that price would work if that was one time. But I won't pay every month because this tool doesn't look like a tool I should pay every month. For example, I had cases like that. For example, I have listing bot where it lists you on directories and 1st I want to launch it as monthly subscription. But people said that why should I pay every month. You just kind of do the work once for me and that's it. You can talk to people and usually if they walk through your process where they join your wait list and you pitch them the discount and pre booking, they would usually, not all of them, but some of them will tell you back that it's too expensive or it's wrong pricing. It should be once or the opposite. So it should be once. It should be every month because I had the opposite too. I had this tool called count visits. It's for analytics and I went for a lifetime though there and people said that I don't want to pay $100 at once because maybe I don't like the tool after all, so I'd rather start with $9. But they told me so and I changed things. Like if you do this, like imagine if I put just my payment link like everybody else does. Then if people don't want it, they just leave. You have no idea why they don't want it. But if you get them into your wait list and then after a while you send them the payment, then you're sending it through the e-mail. So you have the trap on the e-mail with them and they will answer. Speaker 1 There's OK so much I want to ask. All right, so you, you can validate with paid ads. So if people say I don't have an audience, doesn't matter, that's not an excuse. A, the algorithm rewards good content whether you have 0 followers or a million B you can use paid ads to skip the line. What other tools do you use to validate demand? And has there ever been a time where the opposites happened and you never launched something because your tools just told you it wasn't there and you think that you you avoided a failure by doing so? Speaker 2 I haven't had cases where I completely gave up on that idea, because usually my first task for the idea is whether I want it myself or not, and I never go on if I don't want it. And if I want it, then I still want it even if others don't want it, right? But now I have to figure out why I want it and others don't and it takes time for me sometimes. Like I could put the project on hold for for a year just because I want to slip on it. And eventually I ask people and ask people and ask people and my intuition tells me. Also I do one thing that's not common. Leveraging SEO for Product Validation and Automated Content Like people usually go for SAR traffic pretty late. So because it's a long playing game and you shouldn't start with it because they will play back on entry month, right? And you should start with something else at first. But in my case, since I have so many products and they're all not pressing me on time, I have all the time in the world. I usually start SEO work on the day one. In that case, I can just post the product and wait for a few months and then come back and see whether there is traffic from SEO. And that's interesting. So SEO traffic is a new source of validation because you know, one thing that you have to also validate with the product is not just if people want it, but also if you're able to drive traffic there. Like you may build something that everybody wants, but then it doesn't mean that you will find you will win the attention of the people that so that they know about the product and, and a lot of people don't understand that that that's even more important because I rarely seen people who didn't manage to build a product that people wanted. Speaker 1 It's funny, I I've made the same mistake of spending months and years saying, yeah, I just, I don't do SEO because it takes so long to pay off one year later, yeah, I just don't really do SEO because it takes so long to pay off 1. Speaker 2 Year later. Speaker 1 Dude, if you would have just done it any of those times, it would be paying off by now. When it comes to converting a percentage of your wait list, how much does time play a role? Because if you collect a wait list on August 1st and then you don't launch the thing until April first, you're going to convert a lot lower percentage, right? Because it's less relevant, less top of mind. How quickly do you like to launch things after you start getting emails from a wait list? Speaker 2 It's not important at all. So I felt that's important when I was running my first products and I felt like the time was pressing and people are waiting and all those things. But basically in the start world, if you can do something once, you can repeat it, right? Like, if you're able to drive 100 people into your wait list this week, you can repeat that next year, too, unless you're building something that's about Christmas and it only works on Christmas, Right. But if there's no stuff like that, you can repeat that again. And that's why, like, you shouldn't really think about your first wait listers as real users. Like, I hope it doesn't sound lame, but you're kind of using those. Yeah. They're validated. Validate. Yeah, exactly. And they don't put that much effort into helping you to validate, too, Right. And once you validate it, you can, you know, come back in a year and maybe all of these hundred people are not interested anymore because they were founders and now they went back for full time jobs, for example. But it doesn't matter. You just do the same thing again. And then you have hot yeah weight lifters and then you go on with them. That's why like I, I would rather be worried about me doing things on the right time. But then like, for example, like you said about SAR, like if you don't do it early, then after a year you think you should have done a year ago. And so it's compounding and you should do it as early as possible. But for some things, you can do it at any time and then redo it and it's all fine. Yeah. So here's how it work. So basically I give it my URL so it does the analysis and then it learns the audience, the keywords and here it found all the keywords that will drive traffic and they have low competition and it made a plan for the next articles. So now if I click proceed and I pay it will just start working and every day it will pass on article and I come back in two months. Speaker 1 Is this web-based or is it an app? Speaker 2 Yeah, yeah, it's web. It just looks like it's web. Yeah. I think SEO was probably my cheat code in the beginning. Like now a lot of people use it. But in the past, only rich startups used it. And that was the only poor founder who is doing SEO because I had the tool that I built myself. But I think it's like for people who want to test a lot of ideas like 3456, this is one of the best way to test it. You just invest a little time into SEO for all those ideas and you'll leave them hanging for a few months and you come back and you see and you'll clearly see the difference. Maybe you won't have each traffic, but you will see that some URLs get 0 traffic. It means that there's no way you can compete. Like with some keywords now, they're like rich people paying a lot of money and you can never win that. And with other keywords, surprisingly, you win easily and you get a lot of traffic and you can grow on that. And it's sometimes easier to just test whether by building the articles than actually doing this, you know, long keyword research stuff. And because the keyword research, as I told earlier, you never know who is clicking that, which is also dangerous. Like you might see some numbers, but maybe those are wrong people. But if you drive traffic into your blog, you can have little banners within your articles with the wait list. So that's kind of another way to test your wait list. You just put the banner and it says, by the way, if you want to try this website builder for artists, leave your e-mail and we will not notify you. And that's another way because you will see someday data out of it. Speaker 1 Now with this is does it integrate with like web flow and other web builders so you can publish these articles without having to touch it? Speaker 2 Yeah, yeah. It just syncs with framing, Web Flow, WordPress, everything so you basically never have to come back if you don't want it. It will try to nag you and bring you back to to help. But if you ignore it, it will just like like an employee who will ask you questions and if you don't answer it, they will just go and and do as they think they should. Do it right. Oh man. Speaker 1 That is so cool. OK, so we've talked about validating. Well, we've talked about ideating. Validating. What about growing? Five Essential Strategies for Growing Your AI Products Yeah. Speaker 2 So the growth part in my world comes from a social media. Social media is the best way to grow for small products right now. The best part about social media is that it's not just the people you bring directly from social media, but also. The people who learn about you through social media and start talking about you start including you into their blog posts, their newsletters, their directories. So basically if you become popular on social media, you get so much attention from everywhere and all the other people become your marketers for free, like influencers start talking about your tool and you don't have to pay them. So I think like whenever you start working growth, you should try social media and see whether you have talent for it. If you have talent, maybe that's the only thing you should do because the app side of social media is bigger than everything else. I think you will get SEO traffic by doing social media because your product will become popular. People will put it on their lists and directories and you'll get backlinks and that's why you'll get the traffic back. So kind of universal. And the other thing is SEO as I just talked. So you should make sure that you have enough backlinks. You should grow your domain rating by having backlinks and you should generate articles or write them yourself or use any of the tools available for that. And a lot of people think that that's complicated. They should outsource that to agencies and stuff like that. I think if you have money, you should because that will they will do a good job, but it doesn't take that much time. I think spending one hour a day in one month you can have decent output from your SEO efforts. Yeah. And then #3 is partners, affiliate partners. So you should sign up for something like TALT or First Promoter where you're able to create this portal so that your partners could sign up there, get unique links and the code and they could promote your product and earn money. And you should be really, really generous on the Commission. Like a lot of people have this like 10% or 15%. You should be crazy generous, like go 50%. Like if you don't have costs on your product, go for 50%. If you have underlying costs, maybe go for 25 third. And then you'll see people actually, you know, promoting your your product. I have a lot of people promoting it and it's one of the best investments I made. So that's the number 3. The partners who promote your product for for the Commission #4 is we should place your product in front of the people wherever you want, wherever you can. It has to be everywhere, on directories, launchpads, forums, reddits and social media. You should find all possible conversations on Quorum and other places where people talk about it and place it there. So that's really important. Two reasons. So number one reason, people may just stumble upon your product there and become your customer. But #2 which is even more important. LLMS crawl those websites and they crawl those comments and those articles and those listicles. And then they may actually add that into their training data. And then they will reply with your product when people ask. And that's really big. Like I have some of my products that suddenly just blew up because they were in the training set for the latest LLMS. And now I have like almost all traffic coming from ChatGPT and, and all the others because I think they all use the same training set. Like it was surprisingly, you know, simultaneously when I got the traffic from ChatGPT and Claude and Grog on the same week. So I think they're buying the training that data from somebody. Speaker 1 Interesting. Speaker 2 From the same person, from the same company, Yeah. So. And then #5 you can build little tools that are free to drive traffic to those tools. And then you channel the same people into your paid tool. Like if you're running website builder for artists, maybe you should build a little tool which is like a cover generator for your album. That's kind of probably good keyword, right? Because they. Speaker 1 Just a leave magnet. Speaker 2 Yeah, exactly. And they would go there, they would generate their album, they would enter their e-mail to receive the PNG in a good quality. And in the e-mail in the bottom you would say, by the way, here's also my other product. Check it out if you want. And the chance that people will click is really high because they were satisfied with your first product. Because that's the difference. Like when you're cold presenting, when people see your banner for a first time, they don't trust you because they see things all the time. But now they trust you because they like your product. So kind of these 5 are my my favorite ways to grow at the start. And then you can add a lot more. But I think these five bring like 80% of the growth. How Directories and Influencers Thrive in the AI Era What do you think about directories in the age of AI? Is it still a great opportunity there? Speaker 2 Yeah, yeah, I think so. Because like directories that are made by people who have influence will be really big. It's just like if you go to Instagram and an influencer has a list of anything, car influencer has a list of car shops or tires or things like that, you're gonna trust that and you will be willing to follow those links and buy that and share to your friends, right? And I think the same going to happen with AI. Like AI is now in a place where it has so much information, it doesn't know how to pick the right one. Like Google in the old world tried to use the number of incoming links as a way to run the websites, right? But it's so easy to game that. And now AI systems don't know how to rank the websites. They're just using the same system Google is using and that's easy to game. And over time I think AI will like we'll already see that they start using Reddit now. And when they use Reddit they also take the articles with certain amount of app boots, right? So they want to filter out trash. And I think the directories are going to help humans to filter out trash. And like in every category, there will be so many items that how do you choose the right one? And directories can help with that. And special directories made by people who have social media influence. I think that's kind of mix here. Speaker 1 And so there's an opportunity to partner with people that have influence, right? Because most people listening or watching are not influencers, right? That we can all slide into their DMS and, and plant that idea in them because they're certainly not thinking along these lines. Like, hey, you like plant, like they say, invent the disease and sell the cure, right? Hey, I see that you have this list of all these things. Like, do you realize that directories are going to become more relevant than ever in the age of AI? Just so happens I can help you with that. Let's talk about splitting up this business somehow. Speaker 2 Yeah, I mean, that's exactly my model, but I do it the other way around. Partnering app with people who own attention to build a product around that attention is great. And you can do not only directories, but you can also do little products, little tools. Little tools are easy to build in the age of vibe coding. You can build them in one evening. So now how do we compete? We can all build that same little cover generator for an artist. But if that power generator is built by $0.50 or M&M or there is some connection there, that's different thing. It stands out from all the others, right? And if we make a wild gas, that AI will make it easy to do everything. If that's the gas or most of the things, then they'll be difficult to compete because everything is the same as good as the other things, right? And now you can compete through reputation and the reputation can be either built by yourself, which is also not a bad idea because like we're not influencers, all right? When you say influencer, you think about somebody huge with millions of followers. But if you pick something really small, like like something we're only, you know, 10,000 people are interested in this in the whole world, probably there's nobody talking about that. And if you just start talking about that, that will be enough to go viral within that little group, right? That's one way. But the other way is, of course, to find the influencer. I think both are possible. I wouldn't say that if you're not influencer, you shouldn't even try. I think you should just redefine what's influencer and it can be something very small. Speaker 1 That's so good, John. Thank you. This was awesome. Where can people find you and your products and whatever you make? Speaker 2 I'm on Twitter and LinkedIn and sub stack. I love to be helpful with my products. I love to be helpful with my content that I share on X. It's John Rush XI love to be helpful with advice. So I think I'm probably the only person with so many followers who on support every single TM I ever received and every single reply if there was a question, of course, because there's a lot of spam. And I think I will stop doing that by the end of this year because it's overwhelming. But I will still do that for the next three months. So if you will need help, I'm open. Find John Rush and Support the Podcast My DMS are open. OK, thank you. Thank you. Speaker 1 Hey guys, if you're still listening to this, it's probably because you haven't had a chance to take your Airpods out, You're still mowing the lawn, you're still driving, what have you. If you're still here with me, I would really, really love and appreciate a five star review on Spotify, Apple, or wherever you get your podcast. It would mean a lot. If you want to go the extra mile, share this episode with a friend that might have an interest in starting a business. It would mean a ton. Hope you have the best day of your life today.