In short: Dropshipping, trading, memecoins, my own fashion label – four attempts, four times nothing. What I learned above all: tactics can be copied and lost, systems you build yourself and keep. Since then, I only build systems.
Where I come from
My parents came to Germany from South America. My father didn't speak German when he built an ice cream café here. He made it work anyway – until others destroyed it out of envy. After that, both of them went to work full-time, and that's how it stayed. My father is over 60 today and still working.
I'm not telling this to get sympathy. I'm telling it because it's the only way the rest makes sense. I grew up in a house where you could see that hard work alone isn't enough. My father worked more than most people I know. What he lacked wasn't effort, but something that stays standing when you have a bad day – or when someone else wants to break it.
Four attempts, four times nothing
At 18 I wanted out of that equation fast. So I started, one after another, with everything I had:
- Dropshipping. A store, products, ads.
- Trading. Charts, entries, exits.
- Memecoins. The next big jump that never came.
- My own fashion label. My own name on something you can touch.
Four attempts, four times nothing. At the end there was five-figure debt, and I was sitting with it in my childhood bedroom. I'm writing this as plainly as it was. Debt isn't a badge or a dare you show off. It's a weight, and it doesn't just weigh on you, but on the people you live under one roof with.
I don't recommend anyone follow this path. If this piece has a purpose, it's that someone skips the detour through debt.
The pattern behind the four failures
For a long time I thought I'd just picked the wrong thing four times. Looking back, it was the same mistake four times. Every one of those attempts was a tactic: a trick someone else showed off, that works for a while and then stops working. I collected tactics instead of building something that belongs to me.
A tactic is borrowed. It belongs to whoever invented it, to the market that happens to reward it, or to luck. A system, on the other hand, is something you build yourself, understand and can improve. It runs on a bad day too. And when one part breaks, you swap out that part instead of starting from zero.
That was the real turning point: systems instead of tactics. No longer asking what's working right now, but what I can build that will still be standing tomorrow.
The moment it tipped
The first "clone" came from a simple observation. I watched an AI build a website in a few minutes. And my first thought wasn't "nice tool", it was: this isn't a tool. This is an employee.
A tool has to be operated. An employee you can train, give tasks and place with businesses that need exactly that work. So that's exactly what I did: I offered that clone to businesses. That's where my first clients came from. Not a tactic that's dead next month, but work a business actually needs.
How I build that today and why I call it a "clone" is laid out in more detail in my piece on the idea behind the Clone Looping System. What an AI agent actually takes over in a business, I've written up in AI agents for business (German).
What I took away
1. Output over theory
Back then I consumed a lot and built very little. Today my rule is the reverse: if it doesn't run, it doesn't count. A half-finished system that takes over one real step of work is worth more than ten videos about how you could build one.
2. Don't chase every trend
Dropshipping, trading and memecoins had one thing in common: I was never the one who built the game, I was the one playing along. Today, with every new idea, I first ask whether I'm building something of my own or just playing along again.
3. Stay honest, including about failure
I'm not a guru and I don't want to be one. Anyone who only shows their wins is selling a shortcut that doesn't exist. That's also why I don't make income promises – not here and not in my community. What someone makes of a system depends on them, on their market and on work nobody else can do for them.
4. Know what you're building for
My father is over 60 and still working. I build so that he can stop. That sentence carries me through the days when nothing works, more than any motivation from the internet.
What came out of it
Today I build systems instead of tactics – for myself and for others. With NextCore I build and run AI agents and automations for businesses. Internally, my company runs on its own AI operating system, which I write about in ESIS: my AI operating system. And with Jarvis I'm building my own AI clone.
In sales I hold a view that surprises a lot of people: in the conversation itself, I think emotion matters more than systems. Why that doesn't contradict everything above is in Why emotion beats systems. How AI helps around the conversation, I describe in AI in sales (German).
If you want to start with AI services yourself, read Starting an AI agency (German) first. It also explains why I don't name income figures.
Who I am today
Contrarian, founder, systems builder, son. Not a guru. I'm 22, I work in the German-speaking region (DACH), and every day I build things with Claude that would have needed a whole team a few years ago. The four failed attempts are part of that. I don't hide them, and I don't make them bigger than they were. They were the price of understanding the difference between a tactic and a system.
More about me is on About Mario De Luca.
Mario De Luca Founder and systems builder from the German-speaking region (DACH). With NextCore I build AI agents and automations for businesses, and I founded the Clone Looping System (CLS). Here I write about what I've built, tested or messed up myself. More about me · All journal posts