Concept Coach

Stress-test your business idea before you invest time or money

Describe your business idea and Concept Coach will challenge it, surface weak assumptions, point out genuine strengths, and help you figure out what to test before you commit more time or money.

See what Concept Coach gives you

Real output from an actual run on the idea below. Part of the result is shown; nothing is reworded.

The idea

An app that connects people who want to learn a skill (like cooking, guitar, or coding) with local experts who can teach them in person. Think Airbnb for skill-sharing. Teachers set their own rates. We take 15%. Starting with cooking and fitness in my city. Evidence so far: Spoke to about 15 people who would want to teach; 4 said they would pay for lessons. No sign-ups yet. Two similar apps in other cities shut down last year. Founder: I'm a product manager at a tech company, decent at building products. No experience running a marketplace. Have $30k in savings to invest.

Real run, 2026-10-08 — one of the tool's own built-in examples.

Verdict: Clear concept, core demand unproven

The marketplace structure is clearly defined and the starting scope is narrow, but the evidence gathered so far does not establish that enough learners will pay, or that enough bookings will occur to sustain a 15% fee model.

Biggest risks

Learner demand is unestablished — Before building further, identify potential learners through community channels, local groups, or personal networks and ask them for a small deposit or pre-payment toward a first session. Whether they convert from interest to commitment is a materially stronger signal than whether they say yes in conversation.
Marketplace cold-start dependency — Manually broker sessions end-to-end without any app — connect a willing teacher with a willing learner, handle scheduling yourself, and collect the fee. If you cannot produce successful completed sessions this way, building the platform is premature. If you can, you have a template and early evidence to build from.

Strengths

Supply-side early interest
Narrow starting scope reduces coordination burden
Product management experience is relevant to early iteration
Capital available for testing before a larger commitment

Questions to answer

Do you know why the two similar apps in other cities shut down, and does any part of what ended them apply to your market or model?
Of the four people who said they would pay for lessons, how many would commit to a real session with a real teacher at a real price right now?

In a Nutshell

When: Before you quit, incorporate, or spend real money.
You give: Your idea in detail — what it is, who it's for, how it makes money — and how far along you are.
You get: The specific ways it fails: market size, willingness to pay, cold start, moat. Ranked, with what would have to be true for it to work.
The edge: It assumes the idea dies and works out why. Feedback from friends tells you what's good; this tells you what kills it.

Good to Know

The more specific your description, the better the autopsy — 'Airbnb for skills' is less useful than a detailed description of how the marketplace works and how you acquire both sides
Add your founder context — a critical risk for one person may not apply to another with the right background or network
Viability score is honest: most ideas are 3-6. A 4 means 'fixable with serious work' — that's actually useful information
Kill questions are the most valuable output: these are the hypotheses that MUST be true for the idea to work. If you can't answer them, the idea might be dead
Next steps prioritize validation over building — the goal is to find out if the idea works as cheaply as possible
Run the autopsy again after you've done validation work — the score should improve as you answer kill questions

Questions people ask

How do I validate a business idea before building it?

Test the riskiest assumption with real behavior, not opinions: ask for a deposit, sell the service by hand before building software, and find out why similar businesses failed.

What is a concierge test?

Delivering the service manually — matching people, scheduling, collecting payment yourself — before building a product. If you can't make it work by hand, an app won't fix it.

Does Concept Coach say whether my idea is good?

It says what your evidence supports, what it doesn't yet, and the cheapest tests to find out — not whether to go ahead.