I'm shutting down my profitable SaaS. Here's why.

The full story of my journey to €118.67 in lifetime revenue and what I learned along the way.

I'm killing my project, Voiczy.com

Shutting down something you've poured your nights into for a few months sucks. Especially when you have a day job, a family, and you're fighting for every spare hour to build something that might, just maybe, buy you some financial freedom.

But Voiczy is a zombie project. It walks, it groans, it even makes a little money. But it's not alive. And keeping it around is just eating my brain.

So this is the public post-mortem.

It started with a helicopter

I've been obsessed with building things since I was a kid. The moment that broke my 12-year-old brain was watching an old TV show, Airwolf. In one episode, the bad guy plants a "logic bomb" in the helicopter's code. The code waits, and then in the future, it takes over and tries to crash it.

That idea was astonishing to me: you could write instructions that had future consequences!

I didn't have internet, but I had a school library. I devoured books and magazine on programming and computers. I discovered the gods of the time: Jobs, Gates, Allen, Wozniak. I was team Microsoft. Not for Windows, but for the business model: writing code, putting it on a disk, and selling it for millions. It was magic.

I was so deep in the dream that at 15, I photoshopped myself into a photo standing between Bill Gates and Paul Allen.

Could not find the photoshopped photo… but imagine me in between :)

My first "company" was called MK Corporation. I built Windows apps in Visual Basic. My best one was "FaceDesigner"—I meticulously cut out eyes, noses, and hair from family photos in Adobe Photoshop to create a tool that let you build funny faces. I made games. I made computer viruses to mess with the school computers.

I was shipping. Even then.

So what the fuck happened with Voiczy?

Fast forward through a formal computer science & software engineering education and a decade of working for other companies. I got good at building things that made other people money. Millions of users. Complex systems. I learned the game. But it was never my game.

Voiczy was supposed to be different. After dozens of failed personal projects (which I’ll write about them on other posts), this was the one where I'd do it right. I even added a Stripe checkout—something that always scared me off before because of taxes and bureaucracy.

What was Voiczy?

It was a tool for my own son. We're an expat family living in the Netherlands, and I wanted to help him learn Dutch faster. The app used voice repetition and simple picture games to teach vocabulary. It solved my own problem. A classic "user zero" product.

The Stack

I used what I was comfortable with: Next.js, TypeScript, Supabase, Vercel. Simple, fast. The total monthly cost was basically nothing. Like I only have to pay €6 to Microsoft for an email address.

The Launch

I launched on Product Hunt and Hacker News. PH was a ghost town. HN was way better. I got brutal, honest feedback that helped me fix the landing page and kill the bullshit marketing promises I was making. Lesson learned: launch early, even when it's broken. The feedback is worth more than your pride.

The "Good" News

The site started getting traffic. About 15 users a day, mostly from Google. I had auto-generated landing pages for every language combination (learn-dutch-for-chinese-kids, etc.), and the SEO was actually working. People were signing up. A few new free users every single day.

On paper, it looked like it was working.

The Brutal Truth

The free users were a vanity metric. The few, brave people who actually paid... churned.

The product wasn't good enough. It didn't deliver enough value to make them stay. And worse? I didn't have the passion to fix it. Working on it felt like a chore. It was not something innovative enough that I could be proud off. The codebase, which I'd "vibe coded" to move fast, was becoming a spaghetti monster I hated touching.

So what did I actually learn from €118.67 in revenue?

Before we get into the lessons, let's be radically transparent. This isn't some big failure story where I lost millions. This is a small, quiet failure that's much more common.

Here is the all-time Stripe dashboard for Voiczy.

Project Lifespan: Oct 2024 - Jul 2025 (10 months) (RIP)

Total Revenue: €118.67

Total paying customers: 11

Registered users: 128 (all were Google authentication)

Stripe Dashboard (Net volume)

Google Search Console performance report

Every single one of those 11 paying customers eventually churned. That's the real story. Not the traffic, not the signups. The churn. And for some of them, out of fairness, I cancelled their next payment just because I observed they are not using the tool but they forgot to cancel themselves.

So, here are the painful, expensive lessons that data bought me.

Business & Product Lessons

1. Your users are not stupid. Your UI is.

   For weeks, I had zero paid signups. I thought my product was just a bad idea. I opened PostHog and watched session recordings. I saw at least 5 different people finish the demo, get dumped on the dashboard, and then just... get lost. They would rage-click around the page, looking for how to subscribe. I made the demo flow interactive with a clear CTA at the end, and my paid conversion went from literally 0 to 1-2 new subscribers per month.

2. A single user on live chat can save your ass.

   I added a Crisp live chat button. One day, a user messaged me: "Hey, I can't sign in." I checked. Supabase had paused my inactive project, and my entire backend was down. That one user saved me from days of downtime. Talk to your users. They are your best, free QA team. Beside that, they bring valuable ideas to the table, that could be their actual need.

3. Pricing can't fix a value problem.

   I was a coward about pricing. I started at €9/mo and got scared when nobody bought. I dropped it to €3/mo and a few people started subscribing. But they still churned. A lower price can get a user in the door, but it can't fix a product that isn't valuable enough for them to stay.

4. Ship faster. Your biggest risk isn't failure, it's wasting your time.
   I waited months to launch, trying to polish features. I should have launched on Hacker News in week three. The most valuable, painful, and useful feedback I got came within hours of that post. Every day you delay your launch is another day you're willingly working on something that your future customers might not give a shit about. Launching isn't the end of development; it's the start of finding out if you're wasting your life.

Growth & Technical Lessons

1. I did weeks of SEO work for zero paying customers.

   This one hurts. I spent weeks setting up a system to auto-generate pages in 20 languages. It got me 10-20 organic visitors a day. The number of paying customers I can attribute to that effort? Zero. All that work, all that traffic, resulted in €0. Traffic is a vanity metric. Revenue is sanity.

2. Stop "playing business" with analytics.

   I had Google Analytics, PostHog, and Heap. Why? It felt like what a "real" startup would do. It was a waste of time. I was drowning in data but blind to the one thing that mattered: the funnel. I couldn't easily answer "how many users finished the demo, clicked subscribe, but didn't pay?" I should have had one tool and three critical events: demo_completed, subscribe_clicked, payment_succeeded. That's it.
   I still recommend PostHog for its session recording capability, it brings a lot of clarity to see how your actual customers are using the product.

One might ask: what would you have done differently?

1. Launch in 2 weeks, not 5-6 months
   I wasted months polishing a product that the market was only kinda "meh" about. I should have shipped a minimal, ugly version the moment it was just usable. The goal isn't a perfect product; the goal is to get your first real user feedback, and first credit card, as fast as humanly possible.

2. Have a 60-day "kill switch" policy
   I let Voiczy drag on for almost a year out of emotional attachment. My new rule: if a project doesn't show a clear signal of product-market fit (like paying users who don't churn, or people telling their friends about it, or becoming trendy) within ~2 months of launch, it's possibly a zombie. And you have to shoot zombies in the head. No mercy.

So in short, launch fast, fail fast, go on with the next project, and continue the cycle until you find a right market that at least give you $150 per month that could be the better indication of success!

So, What's Next?

I'm not sad about killing Voiczy. I'm excited. Because I’m going to open up the space that Voiczy took in my mind, and devote it to a new project that I’m planning to build.

For a total cost of about €60 and a few months of my nights, I bought myself a masterclass in building, launching, and failing. The €118 in revenue was just a bonus. The real return is the knowledge.

This post is the final step of this project. And it's the first step of the next one.

I'm done building in the dark. I'm doing this in public now, under the name @PolderDev. I'm taking these lessons and starting the journey over.

The goal is to get to €1k MRR. I'll share everything: the ideas, the bad code, the good code, the first dollar, the first bug report. All of it. My next idea has some AI in it… so let’s stay in touch if you’re interested!

This is Day 1. Thanks for reading the post-mortem. Now the real building begins.

Follow the journey on X: @PolderDev

---