Making an app is easy

There is a trend shaking the internet, and most probably the whole side hustle movement. “Create an app with AI in 17 minutes and make $50,000 every month.” It sounds appealing, right? Easy money. I have created a few apps and I can share a few opinions on the reality.

Create an app with AI…

I must admit that this part is true. I do have a technical background but I’m not a developer. Tried that way multiple times but coding is just not my talent. With the help of AI I was able to start developing (vibe-coding) applications that work.

Forget the free tier of ChatGPT. In fact, forget conversational AI at all. If you want to create reasonable code you have to invest in coding agents. I’m personally using Claude Code but you can grab any available coding assistant. The advantage is that they can interact with your system – create and update files, call commands, use MCPs to connect with other tools. It feels almost magical.

A coding agent will enable you to write code. It’s always better if you know how to code and use the agent just to assist you, but even if you have very little coding experience it will help you create the app. Start with something simple. You will not create another Instagram on the first run. But it really works.

…in 17 minutes

Well, yes and no. Coding takes time. Even vibe-coding. You can create a scaffold for your app in a very short time. To be reasonable and not to sell sensations – you can create an MVP (Minimal Viable Product) within one day. Which is already really good. But it will not be a very good app. It will be buggy. The design will suck. There will be no distribution channel. And Google and Apple will have very strong opinions on placing the app on their marketplaces.

Having a reasonable product will take you one to two weeks of intensive work. You do not need to set up a complicated structure of AI agents where each is specialized on a different task. Good results can be achieved with an instance of Claude Code and Claude AI. Make Claude AI write the prompts for Claude Code and cut some corners with that. But be prepared to invest a lot of time into the development process.

The trap of AI slop

In fact, your vibe-coded app will be yet another AI slop. Yes, we should get used to this term even in the context of software. The app will probably work fine but it will look terrible. You can use tools like Stitch to help you design the user interface, but design is one thing where AI struggles a lot. It’s very similar with pictures. Midjourney can create beautiful pictures, but there is always something a bit off with them, right?

You have to switch your role from vibe-coder to designer. And that’s a completely different league. So, to make your app stand out from the crowd of gazillions of other vibe-coded apps, you have to dive into design and mostly… typography. Because typography is 80% of the design. Making your app look and feel good will take another week or two. Trust me. All the YouTube videos and courses take time to process.

Marketplaces

There’s a good chance you will try to push your app to Google Play and/or the Apple App Store. Without a store presence you are cutting your chances of reaching billions of users. You can maybe do it for Android phones, where it’s quite easy to install a rogue .apk downloaded from your website, but it’s almost impossible to reach Apple devices without an App Store listing.

Google and Apple have very strict rules for accepting your app into their store distributions, and handling it is yet another discipline. Signing the app, building the right bundle for the right devices. If you want to be present on both platforms, you need to have two building environments (PC and Mac) and you have to fulfill all the legal and bureaucratic requirements. You also need to build a testing pipeline – basically a group of people who will be willing to use your app for a certain period of time and give you feedback which you will then implement into the app. And if you think you can work around this step… you really can’t.

Oh, and did I mention that you need to pay for the Google and Apple developer programs to get access to the stores?

The app is done, now what?

But let’s assume you went all the way to public store presence. You bought a coding agent, developed the app, fine-tuned the design, climbed the bureaucracy mountain, and persuaded your family and friends to test your app. The sweet day has come and you published your app. Congratulations. It’s there. In both stores. Sitting. Waiting. And nobody is coming to download.

That’s the reality for 99.9% of current app developers, because marketing is much more important than the app itself. You need to tell people your app is there. You need to convince them to click. You need to convince them to download (yes, these are two different things). And that, my friend, is yet another mountain to climb. In fact, it’s a much bigger issue than the development, design, and store presence itself.

You need to start with ASO – optimizing your listing for the stores themselves. Your keywords have to be right, your screenshots have to bang. Then you have to think about how to show your app to as many people as possible. Relevant people. That means thinking through and implementing a marketing campaign. Facebook, Instagram, TikTok, Reddit (lol), landing pages… you name it. The more content you create the better.

You still want to make the app?

That’s great. If you don’t get intimidated by the steep hill in front of you, and you start building your app, you will have the time of your life. You will learn and stack so many skills that it will blow your mind. You will find a few rabbit holes along the way you can nerd into – be it the development process, AI agents, typography, or social media marketing. You can spend months honing each skill, and it’s absolutely worth it. You will also meet a few soul mates in the process. I certainly did. If you are brave, go for it.

Thanks for reading. Feel free to send me an email if you have any questions, ideas, or just want to be in touch. Looking forward.