I was a teacher. I had no programming experience. I had a problem that annoyed me every single day – and at some point, I decided to solve it myself instead of waiting for someone else to do it. What lies between that decision and the first paying user was neither glamorous nor straightforward. But it was more educational than anything I learned during my formal training.
This is the story of how Tekadio came to be – and what I learned along the way that you won't find in any startup book.
Phase 1: The Idea That Wouldn't Go Away
As a teacher, you spend a shockingly large portion of your time on things that have nothing to do with actual teaching. Creating exams. Grading answers. Recording marks. For every class, every subject, every week. It quickly adds up to dozens of hours per month.
My idea was simple: What if AI could handle that? Generate quiz questions, evaluate answers, formulate feedback. Not perfect – but good enough to take over 80% of the routine work.
I carried the idea around with me for weeks. Looked for reasons against it. Told myself I couldn't build it. At some point, I became immune to the counterarguments – and started.
The Path from Idea to First Prototype
Prototype with Replit and AI
No setup, no server chaos. Built directly in the browser. AI wrote the code, I provided direction and logic. First working draft after two weeks.
First Real Feedback
Three teacher colleagues tried the app. The feedback was sobering and valuable at the same time: the core function worked. Everything else was still rough.
Legal, Branding, Landing Page
Business license, GDPR, privacy policy – the part nobody sees but that's still crucial. At the same time: a landing page that actually explains what the app does.
Launch and First Users
Meta Ads at €0.31 cost-per-click. First registrations. First paying user. No viral moment – but real proof that the product has a market.
What Nobody Tells You About Your First Launch
The launch itself is anticlimactic. You press a button. Nothing happens. You stare at your registration numbers. They're low. You wonder if you wasted a year of your free time.
The first user doesn't arrive with fanfare. They come quietly, maybe through an ad or a recommendation, sign up, look around – and you don't even know why they stayed. But they stayed.
What surprised me most: the product I built wasn't the product users wanted. I had made assumptions. Some were right. Many weren't. The launch wasn't the end of development – it was the beginning of real learning.
The Most Expensive Mistakes I Made
Built too long, showed too late. I could have gotten real feedback three months earlier. Instead, I built features nobody needed because I was afraid of premature feedback.
Technical overengineering. As a non-developer, you tend to say yes to every AI recommendation. Result: an architecture built for a 10-person startup but operated by one person. Simplifying after the fact is painful.
Started marketing too late. I thought: build first, market later. The right approach would have been: build and grow a waitlist in parallel. By launch, nobody knew you. That could have been different.
What I'll Do Differently Next Time
- Build a waitlist before the product is finished
- Get real feedback after 4 weeks – no matter how rough
- Choose the simplest technical solution that works
- Plan for legal and GDPR from the start – not at the end
- Pick one single marketing channel and stay consistent
- Test pricing earlier – free users are no proof of a business model
Was It Worth It?
Yes. Not because of the numbers – those are still growing. But because today I have a product out in the world that I built without a co-founder, without external funding, and without a computer science degree. That changes how you look at problems. No longer: "Someone should solve this." Instead: "I could solve this."
That's the real value of the first launch. Not the revenue. The mindset shift.