Published May 31st, 2025
You ever see some post like "How I made 10K per month with new AI SaaS" and think, "wow, I should make one". Well, I don't blame you, I'd say this is the ultimate developer dream. To build something that solely yours and it generates enough money to let you quit your job. To be on some beach, eating lobster while your AI SaaS makes you passive income.
But...This isn't as easy as it's made out to be. I want to go over where the actual barriers are and some things you can do to get over them.
Also note, everything here is just my opinion and based of my personal experience.
If you've been on X (formerly Twitter) in the recent years, you'll notice that a lot of developers have moved to creating their own SaaS project. For some, they've succeeded but I'd say for the large majority of devs, most project never even reach 100 users.
Plus, we've seen solutions like V0, Lovable, Bolt, Chef or even Google's FireStudio come on to the market — essentially letting you create apps really really quickly using AI. And so, of course, this reduces the barrier to entry when creating a project especially for monetary reasons.
So, a bunch of posts like "Making 10K a month with AI" or "I built a 10K SaaS in 2 weeks" have started popping up. And of course it's possible but it's extremely unrealistic.
Sure, these new AI platforms automate the coding, but not the architecture, security, marketing etc. So anyway, let's actually talk about the lie. We're going to use my personal project for this.
Recently, I thought I should create my own SaaS. I saw all those posts about how people started making 10K after like 2 weeks of development. So, I took something I really enjoyed, which was journaling and decided to solve my problems with it.
Now I used Daybook as my main journal, but as dev, I found the UI super frustrating, I hated the lack of features, I hated how much was behind a paywall. But this is great because now I can fix their problems and give my journal app a reason to exist.
Check out https://jadebook.app/ — that's my SaaS I made
So, we have an idea, what's the first thing we do? We research about how many people are also unhappy with the current available options. And we take a not of all the problems they are facing so we can solve them.
Then we start testing, what tools are available and how we would architecture the whole thing. Then we build it. Pretty easy.
But here comes the first problem: You only see the ones that succeeded.
If everyone could start making 10K per month simply by using these AI tools, the market would quickly fill the void effectively rendering the method useless. What that means is that the chances of success are like they've always been: really low.
But survivorship bias acts against, we see all of the winners but none of the users which causes us to have ideas and scope that are largely unrealistic.
Note, this doesn't mean you'll fail. Instead, all it means is that you're going to face a lot of barriers that you initially did not account for. Plan to spend at least 5x the effort you initially assumed.
Vibe-coding is the practice of using AI agents to generate software based on natural language prompts, allowing users to focus on the desired outcome and overall "feel" of an application rather than writing explicit code.
Now, this has been a huge trend and is result of all these new AI coding platforms. But a lot goes into a SaaS like encryption, storage, performance, security, rate limiting, legal stuff, etc. But current AI solutions simply do not handle these things.
Take Jadebook as an example, I implemented semantic search which is a pretty big features. We have to handle encryption, embeddings, searching with multiple filters, and more. There are so many edge case, places where security concerns could show up and more. What happens if the AI decides to fetch journal entries without the user id filter?
What about the duplication of code? Or inconsistent code patterns? Not to mention the lack of code awareness and understanding. All of the problems that come with vibe-coding are going to act as barriers.
Luckily, a quick way to mitigate some of these drawbacks is simply having knowledge about programming and infrastructure plus putting in the effort to learn what the AI is doing.
For a business, you basically need 2 essential things: product and marketing.
We've already discussed vibe-coding, let's talk about the general pitfalls impacting most developers.
Our job, generally requires us to be risk-averse, meaning we try and avoid bugs, write code in a very specific, readable manner and follow document and guidelines to make sure our code is good.
The problem here is that we're building for people and not ourselves. This requires putting in the time to add creative elements to engage users, add onboarding, and think outside the box.
But that's not what we've been trained to do. Over years, we develop a sub-conscious barrier that keeps us within guidelines but business is opposite of that especially when you're starting off.
So, you're going to be battling your instincts the whole time. Which brings us to the biggest problem of all.
Have you ever wondered why UI designers and developers are 2 different roles? It's because marketing and design require you to have absolutely no guidelines and think outside the box all the time.
Most devs can make the product but if no one knows about it then you'll never get any users. And what does marketing require? Making content, posting on social media, branding, understanding a bit of psychology, constant iterations and the best of all, absolutely no clue what will work and what won't.
It takes ages to get good at marketing, and you'll post a bunch of stuff that won't work or it'll get no view, maybe it'll even look horrible. It's going to be so painful to having to sit down and create something with no boundaries or structure.
This is a tough spot to be in. On one hand, you're a genius but your inability to convert that genius into actually money is frustrating as hell. "Marketing" will most likely be the make or break for your SaaS and it's the most difficult to do especially for developers.
The goal of this article wasn't to discourage you from making a SaaS. In fact, even if you don't get any users, if it's something that's genuinely useful then you could use it for yourself or put it on your portfolio.
But if you're planning on making money using SaaS, you need to understand that there's a reason so many fail. Understanding what barriers you'll face can help you prepare in advance.
For example, if you know you've never marketed before, start making content when you start, that way by the time you actually get to launching it, you'll be more comfortable posting things online and might already have an audience.
The solution is simply to work hard in a smart way over a long period of time.
If my journey resonates with you, or you're curious to see the fruit of these lessons, I invite you to explore https://jadebook.app/
Also, thanks for reading :)