← Behind the Craft
Behind the Craft
How This Solo AI Founder Bootstrapped 5 Products to 1M+ / Month | Tibo Louis-Lucas
00:43:43
Loading summary
Transcript142 lines
- [00:00]
A
It's very hard to hit a million per month. I never expected to reach that level. I'm convinced right now that just one person can do the job of 20 compared to like 5 years ago. During 4 months, we were shipping 1 new product per week until something stick, 9 product failed and the 10th one just took off. It's so easy to just lie to yourself, to be convinced that you have a successful product because you have, like, very tiny validation moments. But the truth is, if there is no revenue, there's no stickiness in the revenue. It's gonna be very hard to do this when you have a project that is made for one thing and people are actually twisting it to turn into this other thing. That's usually a very strong signal that you have to follow.
- [00:43]
B
All right, welcome everyone. My guest today is Thibault, a founder who's making over a million dollars a month from five separate online businesses. You know, I've been following Thibault's journey for a long time and excited to get him to review all his secrets, to share his step by step playbook, watch him deal with AI live, and hear his take on what separates AI products that stick around versus the ones that just die. So welcome, sir.
- [01:06]
A
Thank you, Peter. It's so awesome to talk to you too. It's been years that we've been talking on X and finally meeting you for real. So nice.
- [01:17]
B
I feel like I have so much to learn from you, so why don't we get right into it? Let me actually share something real quick. So this is your revenue chart, right? And maybe you can start by kind of walking through each of these products. Obviously Red Revit and some of the other ones are doing extremely well. Maybe you can give like a super high level overview of each of your five businesses.
- [01:37]
A
Pretty much everything here is related to AI. I basically started like two years ago. A little bit more than two years ago, I acquired a small product called typeframe. It was about making product videos. And it took some time for me to notice about that. But making product video is something that you rarely do. And so I went through huge churn with these products. And so I eventually pivoted this product from Type Ramp Product video to Revit, which is a tool to make viral shorts on social media. And it was right at the beginning of AI videos. And so it took off multiple times, like at every new model, increasing the capabilities of AI videos. It just got way better. And so it's doing a little bit more than 600,000 per month right now, which is insane with A team of like four people, I guess.
- [02:37]
B
Wow.
- [02:37]
A
I start thinking that outrank, which is like the green line here, can go beyond Revit and just maybe, maybe the product alone can reach like a million per month on its own. Outrank is an SEO product that was initially just about publishing blog posts for you every day. But it, it evolved into much more than this. It's like a full SEO solution for you. It big thanks to this backlink exchange marketplace. It's the most convenient way to grow the domain authority for a new domain and to start ranking on Google. And I don't know if you know about that, but that's the best way to get mentioned by AI right now. If you want ChatGPT to mention about your products, you just need to appear on search results.
- [03:37]
B
Also, SEO still matters. Yeah, totally.
- [03:40]
A
Like, I think SEO is the number one acquisition channel for Revit. So it's, it's totally something that I still spend days working on. Like, tons of people are calling SEO to be totally dead. It's really not the case for me. But what, what, like what you, what you said and what I noticed on this graph is like, Super X is doing very well. That's the yellow curve there. It's doing very well. It just crossed 30k per month. But. But you barely notice the trend because of how the other projects are getting huge. Something that people ask me is why am I starting new project where I can just focus on the bigger ones. My hope is that the little project that have right now can totally get bigger than the biggest I have right now. It happened to me in the past already.
- [04:31]
B
Yeah. I guess you want to have a more diversified portfolio, right? It's like you don't want to put all your money in one stock, you know?
- [04:38]
A
Yeah.
- [04:39]
B
Portfolio, yeah.
- [04:40]
A
AI is kidding. So many products right now.
- [04:43]
B
Yeah. Okay, so why don't we get into it, dude? Let's get into your most successful product, which is Revit. And you share like 14 steps. I just kind of consolidate into five. So I guess the steps are like, find your problem, validate fast, build with AI and ship, do your own support, talk to users, compound distribution and obsess over retention, and say no to almost everything. So that's kind of the high level summary. Maybe you can share Revit's landing page and kind of show us what it's about and then we can talk about how you grew this thing to such a huge scale.
- [05:15]
A
The first item that you mentioned, which is, I think way underestimated, is When I acquired Typeframe, it was doing like 1 or 2k per month. What I was initially doing is like, I spent money on this product. Like, I just wanted not to be wrong. And that's. I think that's the typical mistake that most people do, that it's more important for them to not be wrong than being successful. And so what you do in this case is you're trying to force sell. Your primary objective is to sell and not to validate. And I think that's the number one mistake. If you force the selling of your products, you are not listening to people telling you that there is another opportunity that might be bigger there. And this is eventually what I noticed with Revit. Doing product videos was amazing. I really think that you can build a 10k product per month. It's totally something that's worth doing and that there's just a ton of people doing that. But I noticed this other trend and I just pivoted the product completely. I rebranded, bought a new domain, and it has been very successful. Sometimes it's nuts, but the simple facts that I did it multiple times gave me multiple shots at winning, I guess.
- [06:45]
B
This episode is brought to you by Oceanz. I recently hired an executive assistant through Oceanz to help with podcast post production. And it's been one of the best decisions that I've made. He's proactive, picks up new tools fast. What makes Oceanz different is they do all the recruiting and matching for you. They source from Sri Lanka, where the talent pool has strong English fluency and experience at top companies. They reject 99% of applicants. So by the time someone lands on your team, they ready to go. Whether you need help with marketing, operations, finance or executive assistant, give Ocean's a try. Check it out@oceanstalent.com Peter now back to our episode. Can you show us how the product works? It's just making short clips, right? Like B rolls and stuff? Yeah.
- [07:29]
A
Project is getting quite complicated right now in the way that we have like more than 100 tools available to create videos, which I find it insane. One of the most successful use case is music videos. So you basically drop a Spotify link or upload your music. And if you select AI videos, Belvid will create out of the box a full video ready to share on social media, which can even like, you can select a character. Like you can upload yourself there. I think I did it with like I did so many times. You can drop yourself like this and you can have the singer singing the music like this on the video.
- [08:25]
B
Okay, so basically I can just upload my Music and then it'll just make a music video. Like a short clip for me with the characters.
- [08:30]
A
Yeah, that's one of the use case. But like you can have avatars, you can have like a YouTube too short. Like you have a long video, you want just a short video. But something that is starting to work very well is the auto mode where you will basically. I will show it there. You'll basically drop like channels, blogs. I don't like everything that. Where you have new content coming out on a regular basis. And Revel will automatically turn those blog posts into videos. It's super easy to use and it's an engine to create new videos automatically.
- [09:11]
B
Yeah, I can see how this has product market fit because when I talk to video editors, I don't want them to make shorts from my interviews. They charge a couple hundred bucks per short. So yeah, being able to get AI to do this automatically will be a dream.
- [09:25]
A
Yeah. And even, even if the money is not a problem for you, you still have to think about that here. With Revit, you just plug your main channel and Revit will do the work for you automatically.
- [09:36]
B
Got it. Okay, so let's go back in time to when you had the product. Video product. I actually remember that product pretty, pretty well. It's basically just like a bunch of text that animates onto the screen as a video. Right. Like that. That's basically what the old product was. And how did you go from that? Like, I'm sure the first version of Revit did not have all these features. Right. Like what was the primary use case of the first version of this?
- [09:59]
A
The primary use case is the basic thing is that at the time when I did this pivot in 2024, all AI video models were doing just the five second videos. And it was hard for people to just clip those five second videos together to get something consistent. The entire idea of Revit is we will do this automatically for you. Like we will split a long video into multiple scenes and make it coherence with consistent characters with all our engine by just plugging multiple five second videos together.
- [10:41]
B
Got it. Okay, so that was a pretty different value prop from like making music videos and making this stuff now, right? Yeah, more about stitching stuff together completely.
- [10:51]
A
But that's still, by the way, the core engine. The core thing behind it is the plugging together many, many very small sequence and making sure that the characters, the scenes, the environment is consistent between all the scenes.
- [11:07]
B
Okay, so you had this other product and I guess the other product for product videos was more Targeting founders and product people.
- [11:16]
A
Right.
- [11:17]
B
And it was making like maybe one or two K a month.
- [11:19]
A
Yeah.
- [11:20]
B
So how did you go about like trying to validate this new stitching videos together opportunity? Like you just tweeted a lot of stuff out and just like try to find customers or how did you, how did you do it?
- [11:30]
A
I think that the number one, the number one answer to that is like maintaining a constant communication channel with your users. It's super hard to do. And like, let me give you an example with, with Tweed Hunter, my first successful product, what I did was the support button inside the app was redirecting users to my Twitter DMs. And so it was a little bit overwhelming. Like I had all those DMs coming about people telling me that there was a bug there or there was an issue here, they would request new features. But I was the guy building the product and I was the guy hearing all days about the problems, the pains and the little friction that the product had. And that was pretty much the same thing with TypeFrame when I got it. Just maintaining the communication with the users is the one thing you need to truly understand the core problem behind the product.
- [12:31]
B
Yeah. So you kind of got a fire hose of feedback from the users and then eventually realized that the product video thing was not the primary use case. People actually just wanted to, because like you said, all the AI models are generating 5 second stuff. So people just wanted a longer videos without having to system together themselves.
- [12:48]
A
Yeah. And when you have a product that is made for one thing and people are actually twisting it to turn into this other thing, that's usually a very strong signal that you have to follow.
- [13:00]
B
Got it. And the other tip that you shared is like you want to charge for the product on day one. Right. Like you don't want to give it out for free.
- [13:06]
A
Yeah.
- [13:06]
B
Just to kind of validate.
- [13:07]
A
Yeah, yeah. That's because like I very big mistakes that I did. Like I think like 10 years ago, like when I got out of college, the first thing that I did was creating my first startup. And I think at that time it was much more important to me to appear important at like family dinner. And so saying things like I'm managing a team of 10 people, I just raised 200k, it would make me appear as a successful person. But it has no strong foundation about a successful product. And so if you reverse things and you only rely on revenue, you ensure that you have validation. And what the number one thing that I wanted to avoid is like spending two years on a product and at the end having nothing.
- [14:06]
B
Yeah, dude, I think for a lot of people, like, you know, there's a lot of things on Twitter about like I raised X million dollars and like my ARR is Y or something. And a lot of it just kind of bullshit, man. Like, like you raise so much money, but do you have proper market fit? Like you don't have product market fit, you know?
- [14:21]
A
Yeah, it's, it's, it's so easy to, to just lie to yourself, like to, to, to lie about, to be convinced that you have a successful product because you have like very tiny validation moments from, from time to time. But the truth is if there is no revenue, there's no stickiness in the revenue, it's going to be very hard to build a successful business.
- [14:43]
B
So basically when you start this thing, you just want to spin up a landing page and then have like a paywall and then have like a product that's focused on a core use case, which is like stitching videos together. Like the initial version, right? Yeah. Do you want to show us like the landing page?
- [14:58]
A
I can show you. I think that something that might make sense is something like this. So I built this AI Reddit humanizer. It's incredibly simple. Like it's a one page feature. You just plug contents and analyze it. But it's enough to notice if people use it. It's enough to start ranking on SEO, it's enough to start like distribution based product developments. Which means. I see, is it enough to start a little thing that can grow? Is there usage behind it? Is it enough to start a conversation with user about what did they like and what should they improve or what would they rather use? I built a ton like this, honestly, like tons of small experiments. This is a way to rewrite and validate your tweets. Yeah, I just built a ton.
- [16:08]
B
Okay. Okay. So basically. Okay, got it. And the landing page, like can you just pop a landing page? Cause dude, you built so many landing pages that like I feel like you have a lot to share about landing pages too. So like the landing page is like. Yeah, like maybe walk us through this landing page and why it works at the high level.
- [16:25]
A
So the entire purpose of the landing page is to drive people from a simple visit to a sign up. That's basically it. So social proof, like right now we are running into a lot of trust issues, especially with AI. So social proof might be the most important components of your landing page. So that's why I always try to have those kind of things, like testimonials from users, number of users, basically numbers that validates the products.
- [17:02]
B
Got it.
- [17:04]
A
What we did on outrank and because the product is quite successful, is we built a few success stories from real users who had huge business impact from the usage of outrank. And we are very happy to showcase them on the landing page.
- [17:23]
B
I see. And then talk about the product.
- [17:26]
A
Yeah. And by the way, most people just jump on the product, I think a little bit too soon. Talking about the product, the problem itself and how it's hard to overcome. It is. I think it's always forgotten.
- [17:41]
B
Yeah, got it. Okay. And then for this product, you let people try out for free and then eventually you kind of pay one of them. Right. To get more use.
- [17:50]
A
Something that might not be obvious for many, many people is. And including myself, like until a few weeks, I didn't know this was possible. But I just want to show you something like. So this animation that I showed earlier, it was one shots on cursor by Gemini Pro 3.1. And it's insane to me that those beautiful animation, it's just a prompt that you can put on cursor and it's gonna do the work and you can just ship it. And so we're going from a world where you had this huge never ending backlog of IDs to right now my backlog is totally empty. I have no idea what to build. And the one thing that get into my backlog is just give it to AI. It's done in 10 minutes. It's insane.
- [18:45]
B
Well, you still have to do the work to promote it, right? AI can build it pretty quickly, but you got to promote it and try to get users. Yeah, totally.
- [18:52]
A
Yeah, that's the hardest part. And earning trust of people is incredibly hard today.
- [18:58]
B
And just to understand, like you have all these ideas. You wrote that you should just pick a problem that you personally have. Is that kind of how you come up with all these ideas or do you like do some research?
- [19:07]
A
Yeah, totally. So the one thing that I think people should do more is look at the product that you are already paying for and look for a way to pick one, replace it with your own products and try to build something that better fit your use case. Because your use case might be a little bit more specific to the more broad use case that the tool you are paying for is trying to fix. So by going a little bit more specific into your own situation, it's very likely that many other people have the same weirdest situation as you.
- [19:52]
B
Yeah, and there's like a lot of SaaS products out there that are like, you know, just, they're like really expensive. Like you don't have to pay that much. You can just pick a more niche use case and just try to. Yeah, completely disrupt them. Yeah.
- [20:04]
A
And I don't think you should. Like it's, it's very hard to hit a million per month. I never expected to reach that level. It's completely fine to get like 10k per month. You have like a huge, it's huge success already and you can totally live with that. And so by just trying to copy an existing product and to niche down to specific type of people, you can totally get this milestone of 10k per month and live, be free. Like be completely free. No boss and enjoy life.
- [20:44]
B
Well, the only boss you have is a lot of people dming you with their feedback. That's the only boss you have. True, true.
- [20:50]
A
Yeah.
- [20:50]
B
Before we get into the AI part, I have some more questions. So you have a pretty big audience on social now. You can just tweet something out. Hey, I built this free tool. Or like, hey, I built this landing page. Check it out. Right? But you mentioned that your primary channel is still SEO for a lot of these products. So like how does someone think about SEO? Like, you know, beyond the basic stuff, like you know, try to, try to put a keyword in a title or something.
- [21:13]
A
Like how do you think about having an audience is. Is giving you shots, but it doesn't make an access unsuccessful product successful. Like I'm not convinced of that. Like I, I have so many failed project right now that I'm convinced that you cannot force a product to work or, or it will work for like two weeks and then that's, that's it, that's the end.
- [21:39]
B
Got it.
- [21:40]
A
So you can like, you can hack your way to 5k or 10k per month, but if you want to grow beyond that, you need a more sustainable way to grow. And that's why SEO is my favorite acquisition channel because it's very recurring by default. It's a constant flow of new users that are really interested in what you are doing. The way I think about SEO, especially right now is like two main components. The number one component is domain domain rating, domain authority. It's very hard to rank for anything with a very new domain with a domain rating below 20, basically. So that's why outrank is so valuable. It just, it can take any domain and by giving you some backlinks without spending too much money, you can get this, this domain rating quite high.
- [22:49]
B
Okay.
- [22:49]
A
Second components is that AI snippet on Google is going to make hard to have your blog post to rank. But what AI snippet cannot replace is tools. And by building tools which answer specific queries, you can rank and leave like a small preview of what your product is. Okay, let me give you ads. So here are my tools for revit. I have a specific page for every single one of my tools. Like if you want to, you have an audio file and you want to turn it into video, I have a tool for that. You just drop your audio file and we'll create a video.
- [23:39]
B
Got it.
- [23:39]
A
But that means that every single one of those tools, it's an opportunity to rank on search engines. Yeah.
- [23:48]
B
Like every use case, right?
- [23:50]
A
Yeah. And for every single one of them, it's very hard for Google to give an AI snippet that would correctly answer to the request of a user. AF types like audio to video.
- [24:03]
B
I see.
- [24:03]
A
It's very AI resilient, you know? Yeah, yeah.
- [24:06]
B
Because this is not just content, it's actually like something you have to use. Yeah.
- [24:10]
A
If you, if you look at like a. Let me show you, like photo AI from Peter Lovers. Yeah. So I, I hate his landing, but he's doing pretty much the same. If you look at like, oh, there's
- [24:24]
B
so much content man here. Wow.
- [24:27]
A
All of that. They are like very specific tools or very specific landing pages that aim at ranking at AI Easter photo.
- [24:36]
B
I see.
- [24:37]
A
So this guy is very famous for having a huge audience, but still he's spending a lot of time on trying to get SEO traffic with those specific pages.
- [24:49]
B
Do you see what I mean? Okay, so you just kind of do keyword research by what are the popular terms that people search for and then try to rank. Try to make this dedicated landing pages for the keynote.
- [24:58]
A
One way to do this is let's go to Google and Let's search for AI Video Tools 4 and just look at everything here. Like every single thing that comes up here is a good keyword to try to rank for.
- [25:14]
B
Got it.
- [25:15]
A
If you are onto SEO, you can do the same SEO tool 4. All of those are like good landing pages to build.
- [25:23]
B
How important is it for the domain to be like the same keyword? Because he has photo AI, but your revit is not like video AI or something. How important is the domain? Yeah,
- [25:36]
A
It is important. I think it matters. I have competitors which have a very relevant domain name regarding AI videos, so it's easier for them to rank. But so I gave up on ranking for AI video. I totally gave up on that because it's Way too competitive. I'm going after the like nicher keywords and I'm doing a pretty good job at that.
- [26:03]
B
Got it, got it. And like with, with AI, it's like very easy to spin up these individual use case landing pages. Right? Like you can probably just tell AI to look at all your code and like list all the landing pages. You just spin up and just go do it.
- [26:16]
A
Yeah, yeah, totally. Which by the way will make it harder and harder for to differentiate. So if you want to rank right now, given all the competition that you're going to have, you're going to need amazing contents, you're going to need awesome performance instantly. Backlinks. Yeah, you're going to need a lot of things.
- [26:42]
B
And just real quick, what tool do you use to monitor where you're ranking for different keywords? Just like the Google tools or you have some special.
- [26:51]
A
No, I'm using Ahref. It's pretty classic.
- [26:54]
B
Okay.
- [26:54]
A
It's, it's expensive, but I think it's worth it.
- [26:56]
B
Got it. Okay. All right. All right, dude, well, why don't we get into the building part? So now. Yeah, I would love to understand how you as one person run five businesses. Like maybe you can share the slack first, like how you talk to your team or.
- [27:12]
A
Yeah, so I have all the channels related to Feather. I have all the channels related to Outrank, everything related to Revit.
- [27:22]
B
But how many people do you work with? Because there's like a lot of channels. Dude, I think you have more channels than I have in my company. Slack. Yeah. How many people do you work with?
- [27:32]
A
I don't know. It's hard to answer to be honest. I have five projects with revenue right now and it's hard for me to imagine going much further than that. So of course I have some secret things going on that I cannot reveal. Like maybe new project coming, but I don't see a way right now where I could go from five projects to like 20. I'm in this process, by the way. Like I'm trying to see like, how can I just put myself out of the a question to like scale a little bit more. Should, should this go through like more hiring?
- [28:17]
B
I don't know, do you have like full time employees or you have like contractors building this stuff or. Because you gotta do a lot of marketing, right? You gotta do a lot of.
- [28:24]
A
Yeah, yeah, I have contractors. Not a lot, to be honest. I'm, I'm trying to keep the team very lean. Very, very, very lean. Like I'm, I'm convinced right now that just one person can do the job of like 20 compared to like 5 years ago. And so it's much easier for me to have less people to talk to, ask them to use the latest AI tools. And so that way we are very fast to adapt to something new. And if we need to pivot, it's. It's way easier.
- [28:58]
B
Got it.
- [28:59]
A
Like, I, in the past, like I used to have a 20% team for one product. And if you want to shift direction, the insane energy that you need to spend to convince one person one by one that the new direction is the correct one is just insane. Instantly you go from the nice boss to the guy who has no vision and has no idea about what he's doing with his company. I don't want to go through that again.
- [29:33]
B
Got it? Yeah. Imagine a couple of thousand people, company dude. That's why it's so hard for these big companies to turn around. Yeah, totally. Okay. In terms of actual development, you just use cursor or codecs. It's pretty straightforward.
- [29:47]
A
Honestly. I have a very basic setup, so I'm using cursor. I'm constantly experimenting with the best models. Right now my go to models are GPT 5.4 for like all the, like the complex stuff. Everything that does not comes down to UI. And for UI stuff I would switch to Gemini 3.1 Pro. I think it's. It's much better when it comes to UI stuff. So that's. That's it. And it's crazy because like two weeks ago it was a completely different set of models, like with or proofs being my favorite. So it's jenning so fast insane.
- [30:34]
B
And in terms of how you build new features, you just get to plan a little bit, make a spec and then just go, just build. Is that how it works? Yeah.
- [30:43]
A
Yeah. Basically I ask the AI to come up with the spec. I usually have an id. I will talk with the AI directly on cursor. It will come up with specs. One time I was like 50%. There's a 50% chance that I will just discard the specs. I don't like the feature like the way it was specs. I just don't like it. So I will not even try to modify it. I find it very complex to make the AI change routes. And so I would just start over. I would just go, go back to my id, twist it a little bit and run it again. And I will do that until it comes up with specs that I'm comfortable with. It seems legit. I would want to use this feature and then I will run it. And sometimes I'm very surprised by the outputs. Something that I did very recently, it was quite new to me is it came up with a spec, but it was hard for me to understand, like, what it would be. So I asked the AI to come up with a prompt that would generate an image preview of what the feature would look like. And so it generated these prompts. I took that prompt and put it on Gemini Nano Banana 2 and it generated this awesome image of a dashboard, showing me exactly what the fitting would look like. And that made me realize that, okay, this part, I really like it. This part, I don't like it. And that was just the way to visualize it. Visualization with AI, it's very hard. And that was a nice way that I found.
- [32:40]
B
Yeah, same. Yeah, it's way easier to look at the designs and figure out what's about than like, some words.
- [32:45]
A
Yeah.
- [32:45]
B
So you should try this tool called pencil.dev. it generates designs directly in cursor. Check it out. Yeah, pencil.dev. yeah, I interviewed the founder too, but it's like a visual way to generate designs.
- [33:01]
A
Yeah.
- [33:02]
B
Nice.
- [33:02]
A
Yeah, I like it. I would.
- [33:04]
B
Okay, so I guess like just running your companies, like, the building part is not that hard anymore, right? Because you have all these tools to ship stuff for you.
- [33:13]
A
Yeah, well, it used to be, but yeah, it's definitely not anymore. The hardest part is to come up with new ideas. Right now. I think we are really coming from a world where ideas were not very valuable to like, right now it's getting very, very valuable to have good ideas.
- [33:33]
B
Yeah. Just have a couple questions since you have five products. Like, even just running my little new newsletter, I get these people sending me emails like, hey, I want a refund, and hey, I want this and that. So I'm just wondering how much of that time that takes up doing support or did you find a way to automate that too?
- [33:56]
A
Yeah, so I have two answers for that. So first of all, I have tons of shortcuts. I will type refound on my keyboard and that will automatically generate a full text saying that I have refunded you and the money is going to be back on your account in like five days, and that will forward the request to the correct person on the team. And so the refund will happen. So, like, automation is something that you can. You can like, you can do so many things if you. If you spend time on that. And my basic rule is, like, if I start doing the Same thing twice or more. I will find a way to automate it.
- [34:42]
B
Like you're just typing in slack and use some sort of scale to do it or something.
- [34:46]
A
Yeah, yeah. Some kind of that, like. Yeah. And the second piece of information is that I just hire people like for customer support on revit and outrank the product got so big that we just hired people to handle that. And so it doesn't take me that much time right now. I really often get DMs about this product, but I actually enjoy that because just it keeps me inside the loop a little bit. And so I keep hearing people talking about that.
- [35:20]
B
Yeah, you want to have the constant stream of feedback just to recap the playbook.
- [35:26]
A
Right.
- [35:26]
B
You have an idea, you want to solve a problem for yourself and then you spin up a landing page and like maybe like one specific use case, like a product for one specific use case. And then like how do you decide like how much to charge or something? Like do some research or you just ask people how much are willing to pay for this?
- [35:43]
A
I do it. It's fun because I do it the other way. Like I, it's. It's. It's more like I want to find an ID where I know that I will be able to charge people from 50 to 100 per month.
- [35:58]
B
Okay.
- [35:59]
A
So I changed the ID, I change the AI usage. I change like I adapt the products to this amount because. And then so you like, you may ask why this amount? But I feel that's the perfect amount from 50 to 100 per month because it's not too much. So people do not ask me for codes before signing up subscribing. But it's high enough so you have high quality customers and low churn.
- [36:33]
B
Yeah, yeah. If you charge like 20 per month. A lot of people just like. It's people who are just like churning all the time. It's like pretty annoying. Ask for refunds and churning.
- [36:42]
A
Yeah, I see like so many people charging like nine per month. And it's good but like it's, it puts you into the position of a cheap product, you know. And I want to be the premium product. I want to be the product that you enjoy paying a little bit more for because it's going to be high quality. I find it much more enjoyable compared to being the cheap product. Having tons of DMS about the cheeky product that you have.
- [37:12]
B
Yeah, it doesn't sound very fun. Doesn't sound very fun. I mean they only pay you $10 a month. Is it really worth respond to each other yeah, but, but dude, like how do you know if a product has like what is your bar for good monthly retention or I guess it depends on the use case, right? Like it depends on the use case. But like is it like you know 50% come back every month or like keep paying or how do you know if you have traction versus you just give up?
- [37:39]
A
You know, I would consider everything better than like 20 churn, like monthly churn. If your churn is better than 20%, I would consider that there is something to do if you're like, if your churn is 40%. And so basically, basically a subscriber is going to stay like just two months. It means that your product is really not sticky and it's going to be very hard for you to get successful product because you don't really realize that at first. But there is a metric called max monthly recurring revenue, like max mrr which basically computes your churn and your acquisition and tells you that you're going to have the S curve until a certain point and you will not be able to go higher than this monthly revenue because of that churn. And so if you want to improve your revenue with that very high level churn you will have, you will need crazy high amount of acquisition which is definitely not sustainable. So I find it like way better to pivot iterates until you get something which is like below the 20% massive level and then start thinking about acquisition. And honestly I say 20% because we are in the AI world and so people are going from one product to another. But like 5 years ago people would only consider product with below 5% churn.
- [39:21]
B
Got it. But okay, I guess last question about this. If you keep pivoting your product, what about your existing paying customers? Like you know, like for example if I pay for the product demo, video product and then you pivoted to revit, does that impact your reputation? Or you just tell them, you just say hey, I changed my product. If you want a refund, you can have a refund.
- [39:42]
A
Yeah, I try to treat them nicely but I don't consider them in the decision. I really try to avoid considering them as a blocker to take the decision about pivoting. But then yeah, of course I just lock them out of the platform and just do a full refund without question ask. I try to anticipate the refund.
- [40:08]
B
Do you also sometimes just build a landing page without a product and try to validate demand or you always have some sort of product behind it, even like an mvp?
- [40:17]
A
Yeah, I did that A lot. So before I started Tweet Hunter, like the product that got acquired, we had this crazy phase where during four months we were shipping one new product per week until something stick. And so we were just building the landing page, tweeting about it. See the traction that we had. If no traction, we would just go to the next product. If traction would just build the product, try to onboard the people who sign up on the landing page and repeats and like, got it. Nine product failed that way. Like, got just no attention. And the 10th one just took off.
- [40:57]
B
Got it. That's great, dude. Dude, I love this conversation. You've inspired me a lot. I'm gonna be building all kinds of shitty products on the weekend now. See if it. I mean, the thing is, I have a big audience, so even as a shitty product, maybe people will just try it.
- [41:12]
A
You know what? Like, it's insane because. So I used to think that the quality was above everything else. And so by. By building more and more, I really realized that building. Focusing on quantity instead of quality leads in the end to better quality. Like the. The highest quality products on a series of like many, many products is going to be higher quality than the products where you will be only focusing on one product and try to achieve higher quality.
- [41:46]
B
Wow.
- [41:46]
A
That was actually the result of an experiment that I think Harvard did. Like, this guy splits a classroom into two. One was just doing one essay and he told them focus on quality. And the other group do like 10 essay each. Focus on quantity. And the best essays were coming from the quantity growth. It was insane.
- [42:11]
B
Is that because you just get more reps or like you actually are. You can validate your ideas faster?
- [42:15]
A
I think it's because you never know. Like, you never know what's actually interesting. And like, I had no idea. Like, I would not have bets on my successful products in advance. They were surprised to me too.
- [42:31]
B
Even after doing all this, you get surprised by what is.
- [42:35]
A
Yeah, I'm still very surprised. Like, I constantly. It's. You need. You need to. I think you need to change your mindset and see every project release as an experiment. And it's totally okay if it. If it's a success, but it can totally fail. And that's okay. You just go to the next experiments and continue.
- [42:56]
B
Yeah, same thing for like YouTube and content you like. Some. Some things I think are going to be super successful. Like don't take off and some random video I make. Take off. It's.
- [43:03]
A
It's.
- [43:04]
B
It's like very random.
- [43:05]
A
Yeah, completely.
- [43:06]
B
All right, man. Well, this is super helpful. I think this is awesome. Like, I think it's going to inspire a lot of people watching this. Where can people follow you? You have a newsletter, right?
- [43:14]
A
Yeah. I think that the best place to find me is Tmaker IO it's like your personal website? Yeah, that's my personal website. That's where has my newsletter. I have my project here and ways to find me on X and LinkedIn. That's my base.
- [43:31]
B
Incredible, man. Incredible. You're free. You're making a million dollars a month and living in France. It sounds like a pretty good life. So hopefully other people can join you. Yeah.
- [43:40]
A
Thank you very much for having me, Peter. That was amazing.
- [43:43]
B
Yeah, thanks.