Transcription
Hello, welcome back. Today I will be talking about the best type of business in the entire world. If you've had anything over 95 IQ points and have been alive at all in the past 20 years, you will know consciously or subconsciously in your head that pretty much every single giant business in the last few decades has been this exact type of business. Facebook, Amazon, Apple, Google, Netflix, OpenAI, Anthropic, you name it. I mean, this type of business is a software business.
There's a reason why software businesses are just [ __ ] Goliath hyperscalers that that that just end up becoming huge huge gigantic businesses. And I I I really encourage every single one of you if you're trying to make your first dollar online to really stay away from trying the physical route, you know, the commerce and products route and really trying to give software a real swing. Of course, this is uh biased coming from me. I mean, I'm the guy who has built software products and made a bunch of money off them. But, I mean, there's just a reason it's so easy. You never hear of a one person clothing store. You never heard about a one person candy store that made millions of dollars. But, there are countless stories on the internet, mine included, of essentially one person or a very, very small team getting to millions of dollars with software businesses. And it's just so much easier at every single level.
In this video, in my attempt to try and convince you uh to start your own software business, uh I'll be talking about a few things. I'll be talking about the cost comparison between software businesses and physical businesses. Next, I'll be talking about the ease of starting. And third, I will be talking about how how you just don't have that much time left to start something like this.
So, first let's talk about the costs. I mean, I'll give you an example. I'll give you the example of myself and my first product interview coder. And I'll give you the example of little Timmy. Little Timmy wants to just like every other ABB in California or or or every other white guy in a garage with his friends wants to start a clothing brand. So little Timmy goes off and starts a clothing brand. Timmy and his friends, you know, like they're trying to sell hoodies. They they they contact a few manufacturers from China. His brother knows a guy and they go back and forth. Shipping times, international shipping times take a while. But hey, if you want the cheap clothing cost, you got to buy in bulk from China. So no other way around it. 2 weeks uh 2 weeks in, they get their first shipment. Cost maybe a couple hundred. is they check out the hoodie, they don't like this hoodie. So, another two weeks comes by, another $700 down the drain. They get a couple more blanks and oh, they don't like this hoodie. One more time, two weeks round trip, and now they're down a couple thousand at this point. They've also had to hire a graphic designer to find a print. But, you know, a couple thousand, maybe like a month or two, and now they have the hoodie, the blank that they want. They have the prints that they want to put on the hoodie. They have a little Shopify website up. So, okay. So, great. So, now Tim Timmy's ready to sell. He's down a couple thousand, and it only took a few months.
On the contrary, my story with interview coder, when I built interview coder, it cost me $13 for the domain. And I think uh everything else was free. Hosting is free on Burcell. My database is free on Superbase. And all the code I have to write is just code that I write by hand. Um in fact, I I I use AI to code. So So maybe there's like a $20 Open AAI subscription there. And after that, it was literally free. It cost me $33 and I coded it in maybe like about a weekend or two. So, it cost me $2 weeks and $33 versus $2,500 and 3 months to start selling our product.
But let's move beyond that, you know, like like Timmy and his friend have a generational run. Their video goes viral marketing their hoodie and they now sell 1,000 hoodies. Let's go. 1,000 hoodies, we just made $70,000 of profit. Except that's not actually $70,000. You know, before that 1,000 hoodies, before those 1,000 hoodies were sold, they had to have the inventory. Otherwise, their shipping times would be 3 weeks. And nobody's waiting 3 weeks for a hoodie from a new brand. So they contact China and they bought the,000 hoodies beforehand. Maybe 45 bucks a hoodie for a really quality hoodie. That's $45,000 down the drain contingent on them having sold the hoodies. So they're down like a lot of money just buying the blanks in the first place. They have to ship everything. Uh they have to package it up. Shipping costs. Uh like who who's going to do it? I mean like like it's it's certainly like like not just going to be done by a robot for them. We don't have the robots for that yet. So they they wrap it up, they ship it, they have to print out the labels, they have to do all this [ __ ] And you know, when someone doesn't like it, they got to process the returns cuz who doesn't have free returns nowadays? So about 10% of that cost maybe goes to free returns. And whatever marketing cost they put into marketing the thing, they got to pay for that. So Timmy and his friend are trying to sell 1,000 hoodies. They made $70,000. But in exchange, they had to put up something like $50, $60,000 upfront. So all that all that 1,000 hoodies, $75,000 in profit to net what? [ __ ] like $10,000. It's ridiculous.
On the contrary, my generational run with Interview Coder, when I first sold my first thousand subscriptions for $60 a month, you know how much I kept of that? I kept about $59,500. In fact, there were almost no costs at all. The only costs were the API costs. API cost being like the different like like the different AI endpoints that I'm hitting or the different and also like like I had to upgrade my database at one point. So, it was about $20 a month to upgrade my database and it was about uh $300 in OpenAI credits. So, that's it. That's literally it. And I did actually make $60,000 from selling 8,000 subscriptions. And I kept $59,000 of it. And you know what else I didn't have to do? I didn't have to go and I didn't have to pack up the subscriptions cuz you know, all you got to do is hit the download button, hit the pay button, and it's software. So I I don't have to do anything. Getting my thousandth customer was just as easy as getting my first customer. It costed nothing. I didn't have to open a physical location. I didn't need a place to store a,000 subscriptions. I didn't need any of the things that traditionally Timmy and his friend would have needed.
And even if they are to hypers scale and this this is like the the where the math really gets brutal. Let's say Tim Timmy and his friend are really really hypers scaling, you know, like 5x per year. So let let's let's just imagine, you know, it's like year two. Timmy and his friend make $20 million selling hoodies this year. That's a [ __ ] ton of money. $20 million. Like yo yo, that that's an inconceivable amount. And next year they say, "All right, let's go sell $100 million of inventory this year." Someone has to again pay upfront for when you are expecting a $100 million year. You have to pay upfront for the expectation of selling $100 million worth of hoodies. So now you got to have $100 million worth of hoodies bought. And you cannot fail because if you fail, you bought hoodies in anticipation for selling $100 million worth of them. And if you didn't, then all that cost of goods and services goes to [ __ ] zero.
And you know what I did when I was selling my millionth dollar on Interview Cutter? I was doing literally [ __ ] nothing. I I I had already had the marketing going. Like the marketing at this point is [ __ ] word of mouth. I'm sitting there doing literally nothing for this company. Maybe I'll change a landing page every once in a while, make a Reddit post once in a while. I do nothing. It scales faster and faster because as more people use it, more people are talking about it. I don't have to dump a bunch of money into ads. It just scales naturally on its own. That's the thing is when you do software, things can scale naturally on their own when you are anticipating a 5x year. You don't need to pay upfront for 5x cost. It's all almost essentially free. And still to this day, my margins are something like [ __ ] upwards of 90% on interview coder. Every dollar I make is essentially a dollar a keep. And you're just never ever ever going to see a physical product like that except for luxury goods. Uh with the sole exception of luxury goods. If you're not pricing your products at like [ __ ] $4,000 for a handbag, then you're never ever ever going to see the same marginal returns as software.
So for that reason alone, if you are not sitting on maybe $100,000, the very very strong marketing strategy, and you're 100% sure that you can get a bunch of sales on this clothing brand, then you probably shouldn't start a clothing brand because you'd just be much better off, you know, if you wanted to make money if that that was the goal. not not not to generate a lot of revenue and [ __ ] donate it all to meta ads. If if you didn't want to do that, you want to keep the money to yourself, you're much better off starting a high margin software company than you are starting a clothing brand. Cuz the clothing brand ultimately, even if you make a lot of money on paper, you're just not going to keep all that money. But software, you really, really do get to keep all that money. You don't need to hire a gigantic team. You don't need to hire like a million sales salespeople and closers and operations and [ __ ] facility factory workers. You just don't need any of that. So for that reason software is uh just just just preferable.
The next reason next big big reason why software is just better than physical products is the reason of software is pretty easy to make right now. And I say this right now as in right now when I say right now I mean January of 2026 this first year. Claude Opus 4.5 got released a few months ago. And for those of you that don't know this is Anthropic's latest coding model and it is [ __ ] giga good. For context, when I built interview coder, my my first coding product, I was using claude opus 4.0. Last year, like almost exactly one year ago is when I built interview coder and I did not write a single line of code in that product. Every single line of code there was written by AI for a model that I am telling you right now is about five times worse than anthropics latest opus 4.5. And to this day, interview coder is still the market leader. It's still printing me a lot of money and there's been functionally no work extra work that I've done to it. So, you have this product that is the market leader that was built entirely with AI over the course of two weekends that has made me millions of dollars. And to this day, nobody's ever built a better product that just surpassed it. It's still the market leader. It's still the one making the most money. There's so many little micro niches like this where you can just build something. Nobody will compete cuz nobody's willing to nobody's willing or interested to compete in that particular space. And you can just eat it all up because coding is so easy right now.
As a matter of fact, probably one of the best coders, programmers of our generation, the best manual programmers of our generation is this guy named Linus Torvalta. He built uh single-handedly built Linux and um Git. Linus is the operating system which pretty much the entire world runs on. Any any computer you see that is not an Apple computer is going to be based on the operating system that he wrote. He also wrote Git. It's essentially like a like a digital journal of all the code that's ever been written. All the code that's ever been written will it be journalized uh via Git. Um, and he wrote both these softwares, which pretty much every software engineer in the world uses. So, so he he is like the [ __ ] goat of our time. Him very recently has been on record using Google's anti-gravity, which is a coding IDE, a coding editor, an AI coding editor by Google to write his code, and he's saying, "This is better than me. This is doing a better job than me." So, the [ __ ] goat of our time, the LeBron James of programming is now on record saying like, "Hey, this machine scores better baskets than me." So, if you don't think that software engineering is over right now, then you you'd be wrong. Like, it it it is over.
And right now, like like what that actually means for you, the person watching, like presumably you're not some [ __ ] senior software engineer at OpenAI. You're not one of the best in the world. A big reason why most people are hesitant to jump into software products. In fact, I say the single biggest reason is cuz they think it's too hard to start a software company. It's too hard to build a software product. I don't know how to build Uber. I don't know how to build Airbnb. It's very, very, very easy right now. Claude Opus 4.5 has never, it's never been easier. Just Google quad code, figure out how it works for two minutes. And literally within seconds, you can have your own version of I'm not even trolling Uber. You could literally code Uber in a couple minutes overnight. You could code your own thing overnight with one sentence. You could code your own docu sign and start selling that. There are so many industries in this world that are just ripe for disruption. So many software companies to start that nobody is starting.
I'll tell you another crazy fact of the the of the times that that that really just goes to show how how how few people are actually doing this thing. In the last 3 months, there have been 24,000 apps published to be approved and distributed on the on the app store. 24,000. And of those, the number of apps that have generated over $100 in revenue is 700. That's three less less than 3% less than 3% of all the apps, all the developers in the world who do all this work trying to get their approved to the app store like like all all the back and forth. Oh, fix this, fix this, fix this. Okay, finally, finally, can I get approved? Yes, finally. And when they finally get approved, 97% of them don't even put in the work to get a hundred people to pay $1 for their product. That is crazy. Most of these people are just quitting when the [ __ ] journey starts. And 700, this number 700, like really get into your head how absurdly low that is. For all the cracked people, for all the 8 billion people in the world and all the cracked engineers you think are out there, there's only 700 who in the last fiscal quarter, the last three months, have built something that has made them over $100. That's such an abysmally low amount. Your competition is 700 dudes. These 700 dudes made a $100 building and selling apps. And I if that doesn't just show you like how giant of a market that there is, that there can be cases like me, someone who has using worst models, the 5x worst models, built products that have gone to millions of dollars in sales, and now the models are 5x better, and there's still only 700 people in the last 3 months who have cracked into the the the who have tapped into the fact that, oh, I can just build something and sell it. your competition has really never been easier. It's never ever ever been easier to make a software product. And it's also been just very easy to market the thing. Like it's just an absurdly no number. Anybody can realize like, yeah, if I really tried, if if gun to my head, my life was on the line. Yeah, I bet I could figure out how to get a 100 people to pay $1 for something. So ultimately, right now, the models are much better than they were when I built my product with AI that made me a bunch of money. you too can use the same models like the much better models to build your own thing that can make you a million dollars and the competition is just low.
And the the final point here it's it's very closely tied to the second point. The final point is you don't have much time. The flip side of this 700 very low competition not many people are building and selling at the same time. Like the flip side of that is that 700 is very low and that number is going to grow. In fact, this number is growing. If you actually look at the number of published apps on the app store, that number is growing tremendously. It's it's it's never been easier. when you sit back and think about how many people could do it cuz like literally every single one of your friends, every single one of your friends who you talk about making money with, like all of them could in theory build an app, publish it to the app store, and make $100. They could put their own $100 in, you know, like [ __ ] like $100, that's that's a testing budget. Like I just want to make sure that the payment work works. I'll put $100 for that. And the amount of people that are able to do that, like in the future, they will because you you it's very likely that you actually haven't heard about Cloud Opus 4.5. this thing that I'm telling you about Claw Code. It's the first time you've ever heard about this. If that's the case for you, which it likely is, and then it is likely the case for all your friends. But this technology, it's coming, man. Once everyone gets access to it, once everyone realizes that, oh [ __ ] the method that Roy employed, the method that I employed a year ago, it's actually still not patched, anybody can go out and do it, then they're going to patch the method, like everyone is going to tap in. And when that number of people who have made money on the app store over $100 becomes from SE goes from 700 to 70,000 that's when you need to start worrying cuz now 70,000 people that that's some competition, you know, like you got to there's 69,999 people you got to beat there. But right now there's not. There's 699 people. There's 700 people roughly that you have to beat out and that's a very very feasible thing. And by the way, if you beat out 700, you become the most profitable app of the last fiscal quarter of all the app store, the number one app in the last 3 months. If you beat out just 700 people, that's absurd to think about. How many apps come out every year that you use every like like how many apps do you use? Like there's at least over 20. Like every year there's some [ __ ] Giga Goliath billion dollar app that's been made. Candy Crush came out one year like like Tinder Hinge like like like Twitch Twitch, Discord. just so many so many apps that are printing hundreds of millions of billions of dollars and to be number one of that fiscal quarter essentially guarantees you a spot is like you know like extremely rich you've made it and all you got to do is beat 700 people right now right now is the moment you will not have this time next year honestly I'm surprised that it's even gotten this far cuz really I thought last year when I when I published interview quarter I really thought the method was going to get patched up instantly but it didn't get patched up you really just don't have that much time left.
To sum it up man software businesses are just at every single level more like the the the numbers just make better sense. You end up with more money in your pocket. It is easier than ever to build software. Uh and the reason why it is an actual method and like it is an actual tactic like like exploit right now is because it's not patched yet, but it will get patched. And you know there's so many people who just live life not hearing about opportunities and not jumping on the opportunities. You know, hearing that GameStop is running up like ah whatever, bro. Like like people have already priced it in like like it's it's it's not going to run up anymore. They heard about Trumpcoin but was like, "Ah, whatever. This memecoin thing is all a scam." They they just like forget the opportunity. There's so many people who let opportunities just pass by them. Do not let that person be you. Right now, the opportunity is not in cryptocurrency. Like like the meme coin, the [ __ ] lottery ticket, the the [ __ ] invention of fire, the railroad of our like the gold rush of our time is in building AI apps in Silicon Valley. Even in Silicon Valley, there are very few people who are actually building AI apps. Everyone, every smart [ __ ] is trying to sell some variant of a shovel. There's actually few people who were on the ground in the trenches building AI apps. There's only 700 of them in the past three months who are even half serious about this thing. So if you genuinely want to make something of yourself, you genuinely think that I am the guy who will, you know, like conquer the world one day, you have any sort of ego or ambition, right now is the moment. January 2026, the gold rush is in AI apps. Go build an AI app and sell it. It's never been easier. The method will not be here. In 2027, you'll look back on this as another, oh, just like [ __ ] meme coins. Like, oh, I wish I wish I wish I did that. You know, I was going to actually buy some Trumpcoin. You know, I was actually going to buy some Dogecoin. You know, I heard about GameStop when it was at like [ __ ] 14 cents. Don't let that be another opportunity like this. Cuz these opportunities when they come, you got to [ __ ] aggressively try aggressively try and grab them. If nothing else, go to Claude Code and just try and figure that [ __ ] out, bro. Cuz if you can, you will be giga rich. And don't try to start a clothing company cuz, you know, that [ __ ] is a pain in the ass. you don't end up with much money. And it might be cool and fun to think about, but it's just at the end of the day, you know, like like what what what is all this about? You know, like I I start a business cuz I want to make some money. You want to make some money, build some software. It's easy and it will get patched.
All right, that's it for this episode. Thanks for thanks for stopping in. Uh I will see you guys next week. Talk a little bit more about startups and software.