Kervan Yolda Düzülür: Building an App the Turkish Way

There's a Turkish proverb my family used constantly growing up: Kervan yolda düzülür. It roughly translates to "the caravan forms on the road." You don't wait until everything is perfect. You start walking. The rest figures itself out along the way.

I built an iOS app with zero Swift experience using exactly this philosophy. Got almost 1000 users. Made exactly zero dollars. And I'm completely fine with that.

This is that story.

What Does "Kervan Yolda Düzülür" Actually Mean?

The proverb comes from the old caravan trade routes. When merchants set out on long journeys, the caravan didn't leave in perfect formation. People joined at different points, loads got rearranged, roles shifted. The caravan organized itself while moving.

The closest English equivalent might be "you figure it out along the way" or "learn by doing." But those feel weaker to me. The Turkish version carries something extra: an acceptance that disorder at the start is not just okay, it's expected. The road itself is the organizer.

This is how I was raised. This is how I build things. Not with a grand plan, but with a first step and the stubborn belief that the next step will reveal itself.

It Started With a Word Puzzle

A while back, I built Puzzle Koala, a Next.js word puzzle site at puzzlekoala.com. It started as a side project, the kind of thing you build on a Sunday afternoon because the idea won't leave you alone.

Puzzle Koala taught me something unexpected. I loved building things people use to learn. Not productivity tools, not dashboards, not another SaaS. Learning tools. There's something deeply satisfying about watching someone get better at something because of a thing you built.

That's when the idea hit: what if I built something to help people learn German?

I'm Turkish-German, living in Augsburg. I speak both languages. I know firsthand how hard German is to learn, the articles alone (der, die, das) are enough to make grown adults cry. And most language learning apps are either expensive subscriptions or gamified to the point of being useless.

So I decided to build Deutsch Wunder.

I Had Zero iOS Experience. I Started Anyway.

Here's the thing. I'm a web developer. JavaScript, React, Next.js, that's my world. I had never opened Xcode. Never written a line of Swift. Never dealt with App Store provisioning profiles or whatever fresh hell Apple's developer portal has in store for newcomers.

But I'd been here before. Years ago, I didn't know React either. I learned it by building things. The first version of Deutsch Wunder was born from the same instinct: start walking, the caravan will form.

Why not React Native or Flutter? Honestly? Because I wanted to learn something completely new. Cross-platform frameworks would have been the "smart" choice, the safe choice. But kervan yolda düzülür isn't about safe choices. It's about trusting the process of figuring things out.

So I opened Xcode, stared at SwiftUI for about ten minutes, and felt like a complete beginner again.

It was terrifying. It was also exactly right.

Claude as a Programming Tutor (Not Just a Code Generator)

Here's where it gets interesting. I had Claude on my side.

Not as a code generator. That's important. There's a massive difference between "write this for me" and "teach me why this works." I've written about how AI can build and break software in equal measure. With Deutsch Wunder, I used Claude differently.

I'd write something in Swift, and it would be wrong. Not crash-wrong, but structurally wrong. The kind of wrong where it works today and becomes a nightmare in three months. Claude would explain why it was wrong, suggest the iOS-native approach, and walk me through the reasoning.

"You're thinking in React patterns. In SwiftUI, state flows differently. Here's why."

That kind of feedback is worth more than any generated code. It's the difference between copying someone's homework and understanding the subject.

Did Claude write some code for me? Of course. But the valuable part was the teaching. I went from zero SwiftUI knowledge to shipping a real app with 350+ quizzes, 26 learning journeys, 4 games, and support for levels A1 through B2. Not because AI did it for me, but because AI helped me learn fast enough to do it myself.

1000 Users and Zero Dollars. That's Fine.

Deutsch Wunder launched. People downloaded it. The numbers grew. Almost 1000 users. People were playing the games, working through quizzes, and some even messaged me saying the quizzes genuinely helped them while learning German.

Revenue? Zero lifetime purchases. I asked Claude to take a look at the database. The answer came back fast: "78% of the app's content is free. No wonder nobody buys it."

Fair point, Claude. But here's what an AI doesn't understand: that was never the problem I was trying to solve.

The Goal Was Never to Exit

I'm not Linus Torvalds. I'm not running a foundation or changing the world with open source. But with what I know, I try to give as much as I can.

Deutsch Wunder isn't the first time I've built something and given it away. I've been running webdesign-freelancer.com for over 6 years. It's a directory for webdesign freelancers in Germany, helping them connect with potential clients, giving them a visibility boost because I've been ranking in those keywords for a long time. Over 120 freelancers signed up. I've never charged a single one of them.

I open-sourced three WordPress plugins on GitHub: Easy Order Management, Bricks Recruitee, and FluentForms Date Restrict. Not because I expected anything back. Because someone else might need them, and sharing costs me nothing.

Deutsch Wunder fits the same pattern. I built it because I wanted to learn iOS development, because I know how painful learning German can be, and because most alternatives lock everything behind a subscription. The app is a one-time purchase. No recurring charges. You buy it once, you keep it forever. And yes, 78% of the content is free. On purpose.

Could I lock more content behind the paywall and "optimize conversion"? Sure. But that's not the caravan I'm forming.

What Deutsch Wunder Looks Like Today

So what did all that caravan-forming produce?

An app with 350+ quizzes covering grammar, vocabulary, reading, and listening. 26 structured learning journeys that take you from complete beginner (A1) to upper intermediate (B2). Four games that make the painful parts of German actually fun: Article Blitz for mastering der/die/das, Scramble Rush for sentence structure, Speech Champion for pronunciation, and Plural Parade for the nightmare that is German plurals.

It supports six interface languages. English, German, Turkish, French, Spanish, and Russian. Because if you're learning German from Turkey or Spain, you shouldn't need to navigate an English UI first.

Is it perfect? No. I'm still adding content, still fixing rough edges, still forming the caravan. But almost 1000 people are using it to learn German right now, and some of them tell me it's actually helping. That's the metric that matters to me.

The Caravan Is Still Moving

I run Wunderlandmedia, a one-person web development and SEO agency. That's the business. Deutsch Wunder, Puzzle Koala, webdesign-freelancer.com, the open-source plugins, those aren't the business. They're what happens when you like building things and believe some of them should just be out there.

Kervan yolda düzülür. The caravan is still forming. I learned Swift along the way. I learned that 78% free content makes for a terrible business model and a pretty good language learning app. I learned that people will message you out of nowhere to say your quizzes helped them pass a test.

You don't need a perfect plan. You don't need a CS degree in iOS development. You don't need to monetize everything you build. Sometimes you just need a first step and the stubbornness to keep walking.

The caravan will sort itself out. It always does.