Before you build

Find the riskiest assumption in your startup idea before you build it

Crux turns your idea into a ranked assumption map, then gives you practical experiments to test demand before you waste weeks building the wrong thing.

Built by We Scale Startups, using the same assumption-testing process we use with early-stage AI and B2B SaaS founders.

What you get, from one real idea

Paste an idea in. This comes out.

Input idea

“Route-level flight reviews by flight number — when several flights cost the same, travellers can't tell which one is actually better.”

Typed in 2 minutes. No pitch deck.

Experiment

Launch pages for specific routes ("BA vs easyJet, London → Berlin"), drive a small amount of search and community traffic, and measure clicks and email captures. Crux deploys the test page and drafts the ad + outreach copy.

Pass signal

50 page visits, 10 email signups, 3 qualitative replies within 7 days.

Decision

Build, narrow, pivot, or drop — based on evidence, not enthusiasm.

Get my assumption map

Ideas do not die from bad code. They die from untested guesses.

Most founders don't need more features. They need to know what must be true for the idea to work — and which assumption to test first.

Building becomes procrastination

Writing code feels productive, but it delays the uncomfortable question: does anyone actually want this?

Generic AI gives generic validation

A chatbot can list risks, but it won't reliably turn your idea into a ranked testing plan with pass bars.

Founders test the wrong thing

Most early tests validate opinions, not behaviour. The real risk is usually demand, urgency, buyer, or channel.

From idea to experiments in minutes

No pitch deck, research doc, or product spec needed. Just the idea, in plain English.

1

Share your idea

What it is, who it's for, what you already believe or fear. Messy is fine — Crux handles messy.

⏱ 2 minutes
2

Get your assumption map

Crux breaks the idea into hidden assumptions and ranks them by impact, uncertainty, and testability. #1 is your crux.

⏱ ~60 seconds
3

Launch the right experiments

Practical tests with pass bars — landing pages deployed with one click, plus outreach copy, survey questions, and ad angles.

🚀 Live the same day

What you get for every idea

Optimised for one outcome: the fastest possible evidence on whether to build, pivot, or drop.

The crux, named

The single assumption that makes or breaks the idea — with the reasoning shown. Validate it and everything downstream gets cheaper; invalidate it and you've saved six months.

Assumptions, ranked

Importance × uncertainty across desirability, viability and feasibility. Test from the top.

Experiments, designed

Falsifiable hypothesis, one metric, a numeric pass bar — smoke pages, ad tests, outbound tests, concierge runs.

Pages & copy, generated

Test pages live with one click, plus outbound messages, survey questions, and ad copy — conversion tracking built in.

Human review in the pilot

Every pilot assumption map is reviewed by We Scale Startups before you spend money on traffic — AI leverage with operator judgement on whether the experiment will actually settle the question.

Why not just ChatGPT?

ChatGPT can give you a list of risks. Crux gives you a decision system — one workflow that extracts the hidden assumptions, ranks the riskiest, designs practical experiments, and helps you decide whether to build, test, pivot, or drop.

Generic AI chat
Crux
Gives broad advice
Gives a ranked assumption map
Lists generic risks
Identifies the riskiest assumption
Suggests vague validation
Designs specific experiments with pass bars
No decision framework
Build, test, pivot, or drop guidance
Easy to ignore
Deploys testable next steps — live pages

Where it sits

Just build it

3–6 months

  • Feels productive, tests nothing
  • Sunk cost compounds
  • Market verdict arrives last

Accelerator / course

£500–5k + months

  • Frameworks, not execution
  • Cohort pace, not your pace
  • Experiments still on you

Crux

Free map · paid pilot

  • Ranked assumption map in minutes
  • Experiments with pass bars
  • Smoke pages deployed for you

Generic AI chat

£20/mo

  • Tells you it's a great idea
  • No metrics, no pass bar
  • Nothing goes live

Test 5 ideas before you build them

For founders, creators, and operators sitting on too many ideas. Submit up to 5 ideas and for each one you get:

  • The riskiest assumption, named
  • A ranked assumption map
  • Practical validation experiments with pass bars
  • Landing page, survey, or outreach copy where useful
  • Human review from We Scale Startups
  • A clear build, test, pivot, or drop recommendation

Or start self-serve: get your assumption map free →

Questions idea people ask

Won't ChatGPT do this for free?

Ask ChatGPT about your idea and you'll get encouragement. Crux gives you falsifiable assumptions with scores, a ranked order to test them in, experiments with numeric pass bars — and then actually deploys the smoke-test pages. The difference is evidence, not opinions.

What does "automatically implemented" mean?

Smoke-test experiments go live as real hosted landing pages — headline, benefit bullets, signup capture, view and conversion tracking — the moment you click Launch. Ad and outbound experiments come with copy ready to paste; you keep control of budgets and sending.

Do I need traffic for the experiments to work?

Yes — a smoke page with no visitors proves nothing. Each experiment includes the ad copy and outbound messages to drive traffic. £50–100 of ads or 30 targeted messages is usually enough to get a signal against the pass bar.

What if an assumption gets invalidated?

That's a win. An invalidated crux for £50 beats a failed launch after six months. The map shows what to test next — or tells you to pivot the idea while it's still cheap to do so.

Who reviews the experiments?

Daniel Johnson and We Scale Startups — hands-on GTM work with founder-led B2B startups. In the pilot, every assumption map and experiment design gets operator review before you spend time or money testing the wrong thing.

Is this for SaaS only?

No. The method works for any idea with a customer and a price: SaaS, services, marketplaces, content businesses, physical products. The experiments adapt — smoke pages and outreach for B2B, ads and waitlists for consumer.

What happens after I apply?

You'll get an email within a couple of days. If there's a fit, you submit your ideas and get your first assumption map the same week. The pilot is free while we refine the process — in exchange we want honest feedback.

Your assumption map takes 2 minutes

Paste your idea. Get the riskiest assumptions, experiment ideas, and next steps — before you build.

Get my assumption map