index a82395d..7b705e9 100644
@@ -1,7 +1,9 @@
---
status: raw
tags:
-- search
+- ai
+- research
+- product
title: idea tester
type: idea
updated: 2026-04-11
@@ -10,4 +12,8 @@ visibility: public
# idea tester
-automated idea validation and market research.
\ No newline at end of file
+an automated system for validating ideas and running market research — takes an idea as input and outputs a structured assessment: who has this problem, how bad is it, what exists already, what would differentiation look like, how big is the market, what would the MVP be. the goal is to compress the "should I build this" research phase from days to minutes.
+
+the pipeline would involve web research (competitors, similar products, market size estimates), synthesis of what problem the idea actually solves and for whom, and some model of what makes an idea worth pursuing (difficulty, differentiation, market size, personal fit). ideally it would be opinionated — not just returning information but giving a clear recommendation with reasoning. one challenge: market research that LLMs can do often surfaces known information; the genuinely useful signal is often in qualitative interviews and obscure niche forums, which are harder to automate.
+
+this is a meta-tool for the ideas wiki itself — every raw idea here could run through idea-tester. connects to [[b2b-competitive-analysis|B2B competitive analysis]] (competitive research is a subset), [[stress-testing-suite|stress testing suite]] (validation later in the cycle), and [[project-suggestor|project suggestor]] (which recommends ideas based on who you are). the [[overnight-app-grinder|overnight app grinder]] could use idea-tester as its first stage — validate before building. also connects to [[idea-extraction-system|core idea extraction]] since you need to know what the idea actually is before you can test it.
\ No newline at end of file