What I might have gotten right
独立产品上线，也许我做对的事- For my first indie product I picked a niche market I know: a piano practice tool. I know the setting and the user profile is clear, but the group is small and leaves little room for error.
- Tools serve the output. I used Lovable for the first version because it was fast, and switched to Claude Code in phase two when I needed to; I started with Supabase’s built-in email and added Resend when that wasn’t enough.
- For a functional product, build the smallest feature set and get into the real market early, without spending much time on UI, the name or the logo; then do some “wild user research” in subreddits to get to know real users.
1. Choosing where to start
I first thought of several directions. After thinking it over myself and talking with my “virtual strategist,” I picked something I know well: a practice tool for a niche market, piano learners.
The upside: I know the setting well, and the user profile is very clear
The downside: get it wrong and it collapses easily, because a small group leaves little room for error
2. Reasonable goals
I understood early that the first project was largely about learning, so I adjusted my expectations.
3. Tools serve the output
For the first version I chose Lovable, because it was fast. Only when phase two launched did I switch to Claude Code, because by then I needed to.
The first version used Supabase’s built-in email service. In phase two I found it wasn’t enough, so I connected Resend.
4. MVP -> MKT
This concept from more than ten years ago still feels very useful now: a product should get into the real market as early as possible. So my first version was roughly the smallest set of the features I had in mind. That’s how I avoid living inside my own imagination.
5. How to weight resources
I didn’t spend much time polishing the UI, thinking up a product name or designing a logo. For this product, users won’t pay because it looks a bit nicer, so early on none of that matters. For a functional product, finishing early and connecting with users early matters more.
*If a product is selling its looks, that’s a completely different story
6. Finding Reddit as a place for wild user research
Maybe this only works for target users who are very concrete, but finding the right subreddit can really help you understand them.
Written in Chinese and translated into English with the help of AI. Read the original