The biggest change wasn’t the idea, it was what I skipped. I was learning Next.js and didn’t know how to properly do auth, sessions, and accounts, so I left auth out entirely.
The first version of Crush Analyzer had:
- no accounts
- no saved state
- I had to do the backend work manully
Users just pasted real chat conversations and got feedback. I wanted to see if people actually cared before automating anything. Only after repeat usage did I wire it up to APIs.
What I learned:
- low friction beats “proper” architecture early
- people will tolerate rough edges if the value is clear
- doing things manually taught me more than building features
Next project I will do the same - focus 100% on the value bring to the user even if it is not real time and automating later when I seee traction.
Kamil_KKA•1h ago
The first version of Crush Analyzer had: - no accounts - no saved state - I had to do the backend work manully
Users just pasted real chat conversations and got feedback. I wanted to see if people actually cared before automating anything. Only after repeat usage did I wire it up to APIs.
What I learned: - low friction beats “proper” architecture early - people will tolerate rough edges if the value is clear - doing things manually taught me more than building features
Next project I will do the same - focus 100% on the value bring to the user even if it is not real time and automating later when I seee traction.
Happy to answer questions.