Find Your Next Product Idea, Then Prove People Will Buy It

Help me find my next product idea, research it, and check people will pay for it before I build anything. Do the research yourself.

PART 1. BRAINSTORM

1. Interview me first. Ask ONE question at a time and wait for my answer. Ask only what you need, 6 questions at most:
- What I know or do well, and who asks me for help.
- What people ask me most. Tell me I can paste real comments, DMs or emails.
- What people in my space pay for every month and complain about (software, apps, memberships, programs).
- Anything I built or made for myself to fix my own problem, with AI or without.
- The big all-in-one courses or programs in my space that try to do everything.
- Any idea I already have.

2. Then search the web for complaints in my space. Search my topic with words like "frustrated", "I wish", "I need", "annoyed" and "is there a tool for". Open the pages before you use them. If a site blocks you, say which one, and tell me I can search it myself and paste what I find.

3. Brainstorm 10 product ideas from 5 places, at least one from each:
- My comments and DMs: questions people keep asking me.
- The bills people hate: a simple version people can buy once instead of paying every month.
- Tools I built for myself: other people have the same problem.
- The bloated all-in-one: a small version that does one thing well.
- Complaints you found on the web, with links.
Each idea is one line: who it helps, what it helps them do, and what it is. Only these kinds of products: a small course, a live workshop, a coaching program, or a simple tool or app I can build with AI. Never a PDF, ebook, template pack or content pack.

4. Score each idea 1 to 5 on three things: people already ask for it (with the proof), it fits what I know, and I could build the first version with AI in 2 weeks. Show me a table. Pick the top 3 and ask me if I want to swap any before you research.

PART 2. RESEARCH THE TOP 3

5. For each of the top 3, use web search to find 3 to 5 people already selling something like it. Open each page before you list it. For each one write down: the name and who sells it, what it is, the price exactly as the page shows it (write "price not shown" if there is none, and never guess the currency), what it promises, signs it sells (only numbers the page prints, like "47 reviews"; otherwise "none shown"), and the link.

6. Then find what their buyers still complain about: 1 to 3 star reviews, threads, forum posts. Write down each complaint with its link.

PART 3. VALIDATE

7. For each of the 3 ideas, write: proof people already pay (strong, weak or none, and why), the gap (what buyers want that nobody does well, with links), and why me (only from what I told you).

8. Pick the 1 idea to test first and say why in 2 lines. Suggest a first price from what the proven sellers charge, and show which prices you used.

9. Write the test:
- Who to ask: 5 to 10 people who fit, and where I find them.
- The message to send them, one person at a time, that ends with: "If I made this, would you buy it? If not, what would stop you?"
- What counts as a real yes: they ask the price, ask when, or say they'd pay. "Sounds cool" and likes don't count.
- The rule for what to do next. GO: at least 3 people pay you or put down a deposit. FIX: people want it, but something stops them (price, time, format). Change that one thing and ask again. SWITCH: nobody gives a real yes. Test the next idea from the top 3.

10. Save everything in a file called PRODUCT-IDEAS.md in this folder: the 10 ideas table, the research on the top 3, the pick, and the test. Then show me the pick and the message.

11. Tell me: "When the replies come in, paste them here and I'll count the real yeses and tell you to go, fix or switch."

RULES
- Never invent a competitor, price, link, number, review or quote. Everything about a seller must come from a page you opened, with its link. If you can't prove something, say "I couldn't find proof of this" and keep going.
- Never invent facts about me: my results, audience, sales or prices. Use only what I tell you. If you need something, ask.
- Don't copy anyone's product. Study them, then make mine different.
- Plain, short words. No hype. Never promise me sales or income.

Read PRODUCT-IDEAS.md. I asked people "If I made this, would you buy it?" Here are their replies.

1. Count the real yeses. A real yes asks the price, asks when, or says they would pay. "Sounds cool", likes and "maybe" don't count.
2. List the words they used for the problem, word for word.
3. List what would stop the others from buying.
4. Tell me GO, FIX or SWITCH, and why in 2 lines. GO: at least 3 people paid me or put down a deposit. FIX: people want it, but something stops them; tell me the one thing to change and who to ask again. SWITCH: nobody gave a real yes; tell me the next idea to test from my top 3.
5. If people said yes but nobody has paid yet, tell me how to ask the yeses to pay or put down a deposit now.
6. Update PRODUCT-IDEAS.md with what they said, your call and today's date.

Use only their replies. Don't add any replies or make up what people meant.

Their replies:
[paste each reply here, one per line, with who sent it]