AI Killed The App Economy

0:00Right now is the easiest time in history [music] to build an app, but the hardest time to make money from one. In just two years, new app releases have nearly doubled. And in the subscription app universe, the world of habit trackers, calorie apps, budgeting apps, AI wrappers, and tiny tools people charge $10 a month for, new launches have exploded more than seven times over. But while all these new apps have been flooding in, app downloads have

0:27been falling for five years straight. [music] And the data shows that people aren't trying new apps, but somehow they're spending more money on apps than ever before. So, the question is, where is all that money going? Well, apparently, not to the new apps. Nearly 70 cents of every subscription dollar is still going to apps that existed before 2020. And the apps that launched in the last year, the biggest wave in a decade,

0:52are all fighting for over just 3 cents. So, we have more people building than we've seen in a decade, fewer people downloading than we've seen in years, and a whole new generation convinced they can build the next big thing with Claude, Cursor, [music] or Replit. But the money is still going to the apps that already own the audience. Because AI didn't kill people's willingness to pay for apps, it killed the scarcity of being able to build [music] one. And when anyone can build a simple app, the app economy

1:24completely changes. One, two. [music] The everyday person went absolutely crazy for the first iPhone release. I'm buying a hundred thousand dollars worth of iPhones. Now, the competition, not so much. That is Steve [music] Ballmer, the CEO of Microsoft at the time, laughing at the iPhone on television. $500, fully subsidized with a plan? I said, that is the most expensive phone

1:53in the world. Which [music] is insane to think about in retrospect. And the wild part about that statement is that he wasn't that wrong at first. SO WHAT IS THE $64,000 QUESTION? In 2008, the iPhone 3G turned into something much bigger than anyone could ever anticipate. The original iPhone took 15 months to sell 6 million units,

2:16but the 3G did 7 million in just three. That's because that summer Apple launched something brand new alongside the 3G, [music] and that is the App Store. When it opened in the summer of 2008, there were only about 500 apps. And within the first weekend, people downloaded over 10 million copies of them. And this was 2008. I was 13, and barely anyone I knew actually had an iPhone yet. But overnight, that phone became the deciding factor whether

2:46[music] you were cool or not. Underage drinking became legal with iBeer. Spending a dollar on a virtual Zippo lighter was just the thing to do. You could wave around a lightsaber. You could stare at a fish swimming in a digital koi pond for some reason, [music] and it was super sick. But it wasn't just these stupid novelty apps. You could hold your phone to a song playing on the radio, and it would tell you what that song was, which in 2008 felt absolutely crazy. [music] You could spin the entire planet with your finger. You could play what looked like [music] a real console game on your

3:20phone. And suddenly, everyone just had these supercomputers in [music] their hands. And every week, someone was making it do something completely new. And in the midst of all this hype, a few saw this as a once-in-a-lifetime opportunity. [music] A developer named Steve Demeter built a simple $5 puzzle game called Trism. In two months, it made him a quarter of a million dollars in profit, [music] and he quit his job writing ATM software for a bank. Another developer, Ethan Nicholas, built a tank [music] game called iShoot. iShoot hit number one and made him $600,000

3:55in a single month. I Fart. $27,000 in a single day. And how could we forget I Am Rich, which was a $1,000 app that's [music] pure function was to literally just flex. And surprisingly, eight people bought it before Apple took [music] it down. And this was the dream. You didn't need Silicon Valley, you didn't need a publisher, you didn't need shelf space at Walmart. You could build one quirky app, put it on the App Store, [music] and maybe it would change your life. And for the next few years, that dream kept coming true. Instagram, 13 employees,

4:35sold to Facebook, you know the rest. Temple Run, 100 million downloads in its first year, built by a husband and wife team. And we all know the story about Flappy Bird, making almost $50,000 per day from [music] ads at its peak. And these apps, these stories, are a huge part of why I became a mobile engineer. I thought I just needed one great idea,

4:58one app, and my life could change. By the time I was old enough to actually build apps, the era that I fell in love with was already over. And you can basically put a date on it. By late 2012, just 25 developers, 24 of them game companies like Zynga and EA, were taking half of all the app revenue in the United States. 25 companies, half the money. By 2014, the largest developer survey in the [music] world found out that half of all iOS developers were living below what they called the [music] app poverty line,

5:35earning less than $500 a month per app. In that same year, this billion-dollar research firm published maybe one of the most brutal sentences ever written about the App Store? Less than one in 10,000 apps will ever be considered a financial success. [music] And as quickly as it came, the land grab was over and the winners had already been decided." But still, I think it's interesting to reflect on why that first era even worked at all. And you probably already know the answer. Building was hard. In 2009, making an app was this very mystical endeavor. Because not only did you have to know how to code in 2009,

6:18you had to know how to code in Objective-C. That's literally the worst language ever. Coding took months, it took money, it took a team of years and skill. And most people with a good idea had no way to act on it. The challenge was building. And the moment that millions of developers flooded in and corporations figured out mobile, that scarcity was gone. And the money consolidated around whoever owned the

6:46charts, the ad budgets, and the users. But even then, in the naivety of my early 20s, I still thought I had a golden ticket and I too could still build an app that could possibly change [music] my life. I spent most of my early career actually building out apps, mostly for my employer, for their

7:05clients, and a couple for myself. [music] A music streaming app, a healthcare auditing app, an app that monitored if other apps were infringing on your privacy before this was such a worrying thing. Mark, give me the zuck. And this was all throughout 2016 and early 2017. [music] And I personally watched every single one of those app ideas die. Not because those ideas were necessarily bad, I think some of them were genuinely good products. We spent a lot of time building them. They died because nobody ever found out they existed. [music] The clients that I built for could afford to pay my company tens of thousands of dollars to build an

7:45app, but what they couldn't do was solve the part that came after that, which was the marketing, the distribution, and getting a single human being on Earth [music] to care. And over that year, watching all of these apps fail made me come to a really harsh realization. The chances of me building an app that made it big without an audience, without me doing sales, without me doing marketing were effectively zero. And to be honest, I didn't want to do anything other than code. So, I gave up on what felt like hitting a lottery ticket, and I just decided I would build apps for businesses that already had traction and

8:21[music] collect a paycheck, because that's what I enjoyed doing. I enjoyed coding. And I've been doing this for almost a decade while friends, family, [music] and every guy that I've ever met at a bar pitches me their app idea. Because I was the scarce resource, I could build and most people couldn't. And then,

8:39almost overnight, it [music] flipped. A couple years ago, I canceled a $5 a month subscription to a workout tracking app. Not because it was bad, but because on a random whim, I decided to build it on my own. And this was late 2023, early 2024, and I had AI helping me in these early Copilot and [music] ChatGPT days. And even then, using these early AI tools, it still took me about a month and a half to replace this $5 app. Now, fast forward to just last week on a random Friday evening, I decided to give that old app a facelift. I added AI macro tracking, a feature that other

9:23apps charge 10 bucks a month for. I finalized my GPS running feature that I started and never finished. And I re-hauled the entire UI and architecture and had AI tie it all together. The front end, the back end, the database layer, bada bing, bada boom, new app. 20,000 lines of code refactored with just a handful of prompts using Claude's [music] flagship model, Fable. All while I clicked some rocks in Runescape. The point is, in early 2024, replacing one $5 app cost me 150 hours. But in mid-2026, building two features that could also be considered standalone apps cost me an evening. And as a mobile engineer, I could have always built

10:06this. The point is, I just never would have. I never would have justified the weekends that it would cost me. And that experience is happening to millions of people all at once. Except most of those people aren't engineers. Matter of fact, 63% of people of Vibe Coding right now have no development background at all, which is why app submissions being at

10:27its highest in a decade isn't a mystery. And that barrier that protected me, that protected every developer of my generation, didn't get lowered, it completely [music] evaporated. Part of me is worried that this is a direct threat to the career that I've spent the last decade building, but a bigger part of me is excited [music] because at the very least, I never have to pretend to be interested in someone's app idea again. But if everyone can build apps, [music] then what happens to the business of

10:57selling them? I think it's important to acknowledge that whenever there is a gold rush, the safest money is usually in selling the shovels. And in this new software gold rush, the clearest winners so far aren't necessarily the people building apps, but rather the billion-dollar platforms selling us [music] the tools to build them. Meanwhile, inside of the App Store, the money is still going to where it was already going even before AI, which leaves the new generation of

11:27builders in a very strange position. They're paying a monthly subscription to build their app so they can compete against companies that already have the customers, the reviews, the recognition, and ultimately the trust. And that is the part that AI hasn't solved. AI can build the app, but it can't make anybody care about it. So, as software becomes abundant, attention and trust becomes the scarce part, the hard and ambiguous problem to solve. Just think about exactly what it takes to get someone to subscribe to your app. They have to find it, and then they have to be comfortable handing over their data, and eventually their credit card to software made by a

12:06complete stranger. And those pre-2020 apps aren't winning because their code is better. In a lot of cases, it's actually a spaghetti mess. They're winning because they've had years to become familiar, to pile up the reviews, and ultimately build that trust. [music] And with these new tools, the software and the simple apps on the app stores

12:26can be copied within an afternoon. [music] However, the credibility can't, which is why having a good idea is mostly meaningless nowadays. You've probably heard that distribution and trust, essentially, is the new moat. So, if trust is so important, then how do you get it? Well, traditionally, you buy it with advertising, targeted ads, sales teams, [music] sponsorships, anything that puts your product in front of your target audience over and over again until your product stops feeling like a stranger. There's this old marketing rule of thumb called the rule of seven, and it's this idea that someone needs to encounter your brand or your product around seven times before they are ready

13:09to buy it. Point being, familiarity gets built through repetition, and big companies can simply afford to buy that repetition over and over again. Most independent developers, however, cannot. So, a lot of them stopped trying to, which brings us to the Indie Hacker. If you've spent even 10 minutes on Tech Twitter, then you have definitely seen this crowd. The way of the Indie Hacker is to build small, launch fast, and share the entire process [music] online as a way to build trust, transparency, and to gain attention. This movement isn't new, by the way. It started in the mid-2010s [music] and blew up in 2023 alongside the AI

13:51boom. And nobody has shaped the Indie Hacker playbook more than the folk hero himself, Peter Levels. If you don't know the story, in 2014, Peter challenged himself to launch 12 startups in 12 [music] months. Now, most went nowhere, but one of them became Nomad List, which is a business that would eventually generate him millions of dollars, I believe. And that story became the Indie Hacker dream. But there's part of Peter's story that tends to get compressed, and that's the decade he spent building the distribution and a trust machine. Peter has posted more than 150,000 times on X. And in his own words, he spent a third of his life building that

14:31distribution, and he paid for it with an absurd amount of time. And he's not the only example. Another example is the developer behind Brain Rot, an iPhone app that helps stop [music] you from doomscrolling. Before launching it, the developer, Yoni, posted a video about his entrepreneurial journey every single day for more than 400 days while [music] still working his full-time job. So, the question is, does building in public actually [music] work? And based on these examples, clearly it can. But this idea of build something and post about it, I think makes the strategy sound deceptively simple. In practice, it means hundreds and hundreds of ignored posts, uncertainty of whether you're

15:14actually reaching your target customer, and then the discipline to keep experimenting when nothing [music] seems to work over a long enough time frame. And in Yoni's case, recording, [music] editing, and uploading a video every single day for over a year straight, which is very hard for 99% of [music] people. Even for me as a content creator, I find it hard to post one

15:39video [music] every single month. However, that being said, I think it's important to remember that building in public is only one way to find customers, and I don't think it always attracts the right ones. [music] And as some Indie Hackers have noted, often the people following the journey are the founders also trying to build an audience of their own. They may be interested in the process [music] without ever actually needing the product that you're building. I think

16:05building in public can be great. However, I think it is just one tool to solve the problem that people can't ignore anymore, [music] which is marketing and building that distribution machine. And with marketing, distribution, building trust, this very intrinsically [music] human thing you're trying to tap into, there is no universal playbook. Which is why when you look at any Indie Hacker community or subreddit of someone posting about how they hit [music] 10K MRR, the number one question people ask is how did you get [music] users? But whatever your strategy is to get users, you still have to solve the major problem of getting enough people to

16:45trust you to download your software and hand over their credit card. Now, what makes the original App Store story so unusual is that for a brief moment, developers barely had to think about solving that trust problem themselves. Between 2008 and 2012, building software was scarce, [music] but distribution was abundant because there just weren't that many apps. And then all of a sudden, there were millions of people actively looking for reasons to download

17:14something on their brand new iPhones. And the App Store didn't completely eliminate marketing, but for a few years it came remarkably close. And that combination has basically never existed before or honestly ever since. But for a lot of individuals, AI recreated that excitement of that first era. And now millions of people can finally [music] turn their ideas into working software without going to school and getting a

17:40four-year computer science degree. [music] But what it didn't do is recreate the empty App Store or the millions of customers waiting to see what the next [music] developer would invent. It only multiplied the number of apps competing [music] for attention. Now, marketing has always been hard. And AI didn't suddenly create this distribution problem. It simply made building so much easier that distribution is now impossible [music] to ignore. And building software and selling software are becoming two very different challenges. So, for anyone trying to turn an app into a business, sales and marketing are no longer the work that

18:21happens after the product is finished. Thanks for watching.