On Building

For the past seven years, I've been building things — apps, tools, side projects that somehow turned into real products. But as a self-taught, learn-as-you-go kind of maker, I've made plenty of mistakes along the way. I've also picked up a few things worth sharing.

If you're someone who wants to create something meaningful — something that outlasts you — my another later essay might resonate.

If your only goal is to "make money," this probably isn't for you. There is a entire different strategey for this.

Everything here comes from real experience.

Mono Card

This one taught me the most painful lessons: thinking about the problem.

When I started, I had zero clarity on positioning. I just wanted to build a fun little toy that grabbed link content and turned it into shareable cards. It went semi-viral early on, but purely because people thought it was a novelty — not because it solved a real problem.

This confusion bled into naming too. I originally called it "DrawLink," and it took three months to realize the name didn't match what the product was becoming. I rebranded to Mono Card.

FlowFerry

FlowFerry is a fairly straightforward product in terms of form factor. My mistakes were almost entirely about growth.

I got lazy with payments and promotions. I relied too heavily on invite codes for everything — running a campaign basically meant handing out free codes. As a result, I barely ran any real promotions through in-app purchases and missed a lot of growth opportunities.

Lack of users means lack of feedbacks. It's not totally bad, but I would appreaciate it if I managed to find more users in early days.

ClipMemo

My biggest mistake here was underestimating the product's potential.

For over two years after launch, users kept reaching out to ask if I was still working on it. During the early days, I received heartfelt messages from people thanking me for building it.

But because ClipMemo didn't scratch my own itch — I built it, but I didn't really use it — I never found the motivation to keep pushing updates.

Notes

• Not everyone uses Chrome. Pay special attention to Safari users. Safari has its own quirks in CSS rendering and security policies that can silently break features.
• A rough start is the best start. The hard part is never beginning — it's continuing.
• Keep going for a few years and you'll notice something: most of the people who started alongside you have already quit.
• Don't fear the quiet days — the ones where nothing's happening, no inspiration strikes, and you feel unproductive. Those idle stretches are exactly where the next breakthrough is quietly forming. Live your life. Pay attention to the world.
• The goal of a product isn't to make users fall in love with the product. It's to make them fall in love with themselves. Word of mouth — the ultimate growth engine — is born when users feel good about who they are while using what you built.
• Technology isn't a moat; delivery is. I've seen plenty of what I'd call "clumsy developers" — people who can build a "product" easily enough but have no idea how to get it from their IDE onto a user's device.
• Taste is your real moat. A solid delivery pipeline can put you ahead of most people. But in that final top 1%, what separates you is taste.
• In UI design, good design is invisible — it feels intuitive. Bad design, on the other hand, is immediately obvious.
• If you're building on the side, don't let anxiety consume you. Let things grow organically. Accept that you can't do everything well at the same time. Sometimes imperfection has a beauty of its own.
• Keep a craftsman's mindset: you won't always be winning or getting what you want. On the other days, build your character.
• Adding features is easy. Removing them is hard. Practice restraint. You'll always have something new you could add, but knowing what not to add — and what to take away — is the most valuable skill.
• People actually read their email. Even the stuff that lands in spam.
• If your product is content-driven, email subscriptions are essential.
• Instead of "let AI rebuild the whole product," I prefer "rebuild it with better UX and sharper taste, and use AI to accelerate the process."
• If your product is content-oriented, play the long game. Build evergreen, not viral. Time will prove you right.
• Remember to use AI to check the typos. I saw 20+ significant typos in Detail when someone reminded me there were.

You might be wondering: so what did I actually get right?

Honestly, just one thing: I started building, and I never stopped.