I have been building and promoting AI based SaaS for a while now. I have gone through a lot of trial and error. Some of it went well. Some of it did not. If you look at my previous articles, you can see the logs I kept while working. This time I want to share my honest thinking on whether I have actually made money building SaaS through vibe coding, and whether there is real potential in it.
Here is my conclusion up front. It is hard for a solo developer, and the system is designed to favor corporations. I know you want to create great results with vibe coding and turn that into profit, and I want the same thing for myself. But I think it helps more to give you the realistic picture. This does not mean it is impossible to make money. My conclusion is simply that it is never easy.
The Illusion of Automatic Profit
If you watch YouTube these days, you will see many videos claiming someone made thousands of dollars a month by building a single app with AI. I fell for a few of these videos myself before I started building, and that is exactly why I wanted to write this column. Even if they exaggerate, they keep promoting it that way because it still makes them money. Most YouTube advertisements claim that you do not need to know how to code. They advertise that the app builds itself if you just explain the features you want to the AI. They talk as if you can just click a few buttons, upload it to the app store, and watch the profits roll in automatically.
When the Real Work Begins
I spent some time working as a web developer before this. Before trying it myself, I thought I could build it relatively easily since I already had that development background. But once I actually started building, I realized that creating an app does not end with writing the code. The real work begins only after the code is written.
 |
| The simple custom link builder I created using AI |
The service I built was a lightweight SaaS. It created a customized link page in seconds, solving the restriction where Instagram and TikTok only let you put one link on your profile.
Using AI definitely sped up the process of building the basic screens and features. In the past, I would have had to Google and write the code line by line. Now, if I explain the task to the AI, it quickly generates a somewhat working result. Looking at just that part, it feels like an era has arrived where anyone can build an app. But I realized that an AI generating code and you operating a completed service are two completely different problems.
The Maintenance Nightmare
The code written by AI does not always come out clean. In my experience, there were times when it made conditions unnecessarily complex, or left behind code that was not used anywhere. There were times when I asked it to fix one specific feature, and another feature that had been working fine suddenly broke. When I asked the AI again to fix the error, it sometimes just slapped a new conditional statement on top instead of properly organizing the code.
As these modifications repeat, the initially simple code becomes increasingly complex. It might work right now. But it can become very difficult to find where a problem occurred when you try to add or modify features later. Simply put, development is possible, but maintenance becomes a nightmare. Even a simple app was this challenging. I felt it would be truly exhausting to commit to a large project all by myself.
The Tail Wagging the Dog
 |
| The reality of monthly server costs draining money with zero revenue |
To do vibe coding, you need to pay subscription fees for services like Claude or Gemini. You also need a server to make the app accessible from the outside. Database costs will go up as the number of users increases. I am currently using a Render server, and it costs about 14 dollars a month in maintenance. My revenue is still zero, but money is already slowly draining away just to keep the service running. It is literally the tail wagging the dog.
Many independent creators fall into this same trap of building without considering the hidden costs of maintenance and operations. I discussed this specific trap in detail in my piece on What Every Solo Founder Gets Wrong About the AI Era.
The Hidden Walls of Launching
Just because you somehow finish an app does not mean you can immediately upload it to the Play Store. I worked very briefly as a web developer, but I did not know much about the process of an individual launching an app from scratch to the market. When developing inside a company, the service is often already on the market. Developers focus only on their assigned features. They rarely get the chance to directly experience the launch policies of an individual developer account.
 |
| The hidden Google Play testing requirement for new solo developers |
But I looked up the Google Play policies. A new personal developer account created after November 13, 2023, must run a closed test with at least 12 testers for two consecutive weeks to apply for official release permission. This condition does not exist for corporations. It only applies to new individual accounts.
To put it simply, a solo developer has to not only build the app but also personally find 12 testers and maintain that test for two weeks. You have to reapply for production access even after the test is over. When you consider account verification, test preparation, the 14 day test, and the release application, an individual developer is looking at a month just for the launch schedule after finishing the app. The deployment period will stretch even further if testers drop out along the way.
A company might be able to test through internal employees or QA testing staff. But finding 12 testers is much harder than you might think for an individual developing alone. You could ask friends or family. But it is a burden to keep asking people who have no real interest in the app to participate for two straight weeks.
The Final Missing Pieces
There is an even more difficult problem than this. Even if you build an app and launch it on the market, your revenue is zero if people do not use it. There are already countless apps on the Play Store that connect profile links. From a user's perspective, there is no reason to install my unfamiliar app when they can choose apps with high download numbers and stacked reviews.
 |
| Planning execution and marketing are the real skills needed for success |
This is where coding well stops solving the problem. The people who really make good money with vibe coding are not just good at development. They are people equipped with the planning ability to find what people want, the execution power to finish a product to the end, and the marketing skills to promote their created service. The YouTube ads about people making thousands of dollars with vibe coding are true, but you have to realize these conditions come attached. Saying anyone can easily make big money while skipping this entire process is basically telling a story that leaves out the hardest part.
If you are wondering why your finished product is sitting empty despite all your hard work you should read my breakdown on Why Your Vibe Coding App Has No Users. It covers exactly what happens when you rely entirely on generated code and ignore the actual market.
Building True Leverage
I am not telling you to avoid vibe coding. I actively use AI myself, and it allowed me to build services much faster on my own. But the greater the expectation, the greater the disappointment. If you expect crowds to swarm in and profits to roll in automatically the moment you upload a single app, you will quickly give up when faced with a reality where even launching is hard and users do not show up.
Using AI is not a competitive advantage on its own. Anyone can use similar AI tools right now. The most important thing is accumulating your own know how. You need to learn how to instruct it to get the results you want, how to solve problems when they arise, and how to deliver on the needs people actually have. It might look like an insignificant experience at first. But as that know how continues to pile up, I believe it will become your own unique weapon and a serious asset that others cannot easily copy.
I plan to keep adding features and improving the app I am currently building. I will honestly record the process here to see if I can actually launch it, if I can attract users, and if I can generate revenue.
The greatest asset you will ever build is not the code generated by a machine, but the unique resilience and deep know how you carve out for yourself along the way.
For deeper reads on tech, business, and the shifting economy, follow The Techtonic for weekly breakdowns.
Latest Insight: [Why Vibe Coding Is Built To Fail]
Trending Insight: [Stop Learning and Just Start Copying in the AI Age]
Comment