Market Maven digs through Reddit, reviews, forums, and competitor sites, then hands you a report with a verdict: yes, no, or not yet. Real research, your audience's exact words, zero hopeful guessing. Two versions, same coach: a custom GPT and a Claude skill.
Same research coach, two ways to run it. Chat with Market Maven in ChatGPT, or install the skill in Claude and run it right next to your business context. Either way: a short interview, deep research, and a report you can act on. Pick the tool you already live in.
Ask it one of four questions and it goes to work. Every report ends in a committed call, not a hedge.
You only need one. Go with whatever's already open on your screen.
Best for: you already live in ChatGPT.
Best for: your business already lives in Claude.
Market Maven asks 1 to 2 questions at a time, 5 to 7 total, then goes quiet and researches. The wait is the value. When it's back, you get the report and the verdict.
"Here's the truth based on what I found: the problem is real and people are paying to solve it. But they describe it as 'drowning in admin,' not 'needing systems.' Your copy is speaking the wrong language."
Honest verdicts. Even when the answer is "this might not work," you'll know why and what to pivot toward.
You'll have a research-backed report as a PDF. Add it to your Business Brain, upload it to any project that writes your marketing, and steal the Language Patterns section for your hooks and copy. Haven't set up your AI workspace yet? Join AI Skool of Business and I'll walk you through it.
Same coach, but it lives in your Claude from now on. Ask "does my idea solve a real problem?" in any chat and Market Maven takes over.
Works across every chat once it's in.
The Skills feature needs code execution turned on. If you don't see Skills, go to Settings → Capabilities and toggle on "Code execution and file creation" first.
Run it inside a Claude Project that already holds your business context. It skips the questions you've already answered and researches straight from what your business actually is.
Same flow whichever door you picked.
Ask one of the four questions: real problem? how big? what makes me different? or "create my Ideal Client Profile with real market research."
Answer a short interview, 5 to 7 questions, then let it research. It can take several minutes. That's the research actually happening.
Review the draft, ask for tweaks, then get your final PDF with a committed verdict: yes, no, or not yet, and why.
Every report includes your audience's exact language patterns, the failed solutions they've already tried, and the objections you'll need to answer. That section alone writes half your marketing.
No generic "women 25-45 who value quality" filler. Everything in the report traces back to real research, and it will tell you the truth even when the truth is "not yet." That's the point of running validation before you build.
"Coaches" gets generic research. "Female fitness coaches who left gym jobs" gets gold. The narrower your answers, the sharper the report.
The research takes minutes, not seconds. Don't interrupt it or ask it to hurry. Fast answers are guesses, and you can get those anywhere.
If the verdict stings, ask what would change it. That answer is usually your pivot, and it just saved you a month of building.
You don't need both doors. Pick the tool you open every day and go. And if you're validating something new, run the first three modes in order: problem, opportunity, differentiation. Each builds on the last.
Market Maven already lives inside AI Skool of Business, right next to everything else I build for running a business on AI. Join the community and it's all unlocked, no separate freebie hunt every time you need the next tool.