Research the market before yourself

Founders research their idea (is it cool?) when they should research the market (who pays, for what, and why now?). The RESEARCH stage is where the idea stops being yours and starts being a product for other people.

10 minutes of research that saves 6 months of coding

Three questions, answered with evidence, before any code:

1. Who else sells this? If the honest answer is "nobody," that's not a moat — it's usually a sign there's no demand. Name the three closest alternatives, including spreadsheets and doing-it-manually.
2. What do they charge? Pricing is the fastest read on whether a market has money in it. If every competitor is free, ask what makes you believe this category converts.
3. How do customers find products like this? A great product with no distribution path is a hobby. Name the channel — communities, search, marketplaces — before you need it.

Your competitor's 1-star reviews are your feature list

Reviews, forum threads, and support forums are free customer research. The complaints people leave about the incumbent are the wedge you enter through — in their words, not yours. Collect ten verbatim complaints and you have your positioning, your onboarding copy, and your first three features, all sourced from the market instead of your head.

Size a niche you can actually win

"The market is billions" is a research failure. You need the honest, unsexy number: how many people have this specific painful problem, can you reach them, and would ~$29/month sting less than the problem? A winnable sliver beats an unwinnable ocean every time.

Where the OS enforces this

• The playbook makes the research write itself down — market proof, competitor notes, and the wedge, as fields you have to fill before the stage advances.
• The AI assistant pressure-tests your answers — the second opinion you can't get from the mirror.
• Research recorded as evidence counts toward the gate. Research in your head counts toward nothing.

Next: Evidence before code.