The Korean Market Opportunity Is a Payment Window
Most writing about Korea as a market is about size. Population, smartphone numbers, how much people spend online. Fine for a deck, not much help on a Tuesday afternoon.
The useful version is narrower. A Korean customer finds your app, likes it, taps checkout, sees a bare credit card form, and closes the tab. That's the gap. Not demand, not ideas, not even language. The payment window.
What Korean checkout looks like
If you've only sold to US customers, your mental model is one card field and maybe Apple Pay. In Korea the payment window shows KakaoPay, NaverPay, Toss Pay and card, and people pick whichever one already holds their saved identity. An app offering only international cards reads as foreign, and foreign reads as risky when money is involved.
This is also where most AI app builders quietly stop. They generate a card checkout and call payments done.
Every DontCode project ships with all four in its payment window, one-time or subscription, priced in KRW by default or USD if you need it. There's nothing to enable per provider. We hold the merchant keys and verify each payment server side. The only thing that can actually stop payments going through is your payout account, and that's a form, not an integration.
Login is the same story
Kakao login matters in Korea the way Google login matters elsewhere. It sits next to Google, GitHub and Apple in sign-in settings: paste the client ID and secret from the Kakao console, press Save, and the button is live once the app rebuilds. No callback handler, no env vars, no wondering where the secret goes.
Once accounts work, retention is where Korean apps tend to get lazy. Set up the subscription expiration nudge on day one: a scheduled job at 9 AM KST that finds subscriptions ending in 7, 3 and 1 days and sends a different message for each. It catches churn before the card declines. Boring, and it pays for itself.
The part nobody can do for you
Sphinx showed up on Hacker News this week with 239 points and 81 comments on its thread. It's a Git hook that pauses your push and asks you a question about what your agent's code actually does when things go wrong. Answer wrong, it explains why and gives you another scenario. Its tagline: you can outsource your thinking, but not your understanding.
That's right, and it points at something about platforms. Payment plumbing isn't yours to understand, because it isn't yours. Neither is session handling or SSL. What's left is the part a tool genuinely cannot take off you: whether a Korean buyer trusts your checkout, whether your Korean copy sounds like a person or a translation, whether the price in won feels normal for what you're selling.
That's the real Korean market opportunity. Not some market nobody noticed. A market where the setup used to eat your first three months, and now doesn't.
If you want to see what's handled before you commit to anything, the feature list is public. Or start something and skip ahead to the part that's actually yours.