agentsclimarketplace

Validation experiments

Skill eltntawy/idea-discovery-playbook/skills/validation-experiments

Designs smoke tests, fake doors, concierge and Wizard of Oz MVPs, pre-sales, and design-partner experiments with ILI metrics, success and fail thresholds. Use when the user asks to test demand, run a smoke test, pretotype, validate an idea, or design an experiment.From its SKILL.md

Install
npx -y skills add eltntawy/idea-discovery-playbook --skill validation-experiments

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

4.3 KB, 717 tokens by cl100k_base, as published. Nobody here has run it

Validation Experiments

Link each experiment to one row in the assumption-map ledger. Map results to evidence levels in idea-scorecard (Intent = 3; payment = 5).

Loop

  1. Assume — state the riskiest assumption
  2. Design — choose the smallest test that can falsify it
  3. Test — run the experiment
  4. Measure — collect quantitative data (ILI, CTR, conversion, payment)
  5. Learn — extract insights
  6. Decide — proceed, pivot, or kill

Technique picker

TechniqueBest forNot for
Fake doorDemand for a specific offer/feature; priced CTAProving retention or full workflow
Smoke test / landingMessage–market fit, segment + offer at scaleDeep workflow discovery
Concierge MVPLearning what to build; pains and language (5–10 customers)Scalable unit economics
Wizard of OzTesting UX and workflow without backendFirst read on abstract demand
Pre-sale / design partnerWillingness to pay and delivery commitmentAnonymous traffic only

Recommended sequence: fake door (priced) → concierge (5–10) → Wizard of Oz at small scale → build only if assumptions falsify the kill path.

ILI (Initial Level of Interest)

From pretotyping: ILI = actions_taken / opportunities_offered

  • Opportunities: visitors who saw the offer (or segment reached)
  • Actions: paid pre-orders, deposits, qualified signups with stated intent, or explicit purchase clicks per your threshold

Record ILI with segment, offer, and date. Label as proxy — not audited revenue.

Experiment plan template

FieldValue
Riskiest assumption
Linked assumption ID(from assumption-map)
Methodfake door / smoke test / concierge MVP / Wizard of Oz / pre-sale / design partner
Audience
Duration
Price shown?yes / no — must be yes for WTP tests
Success threshold
Fail threshold
Evidence quality target(scorecard level 3–5)
ILI targetoptional numeric

Ethical fake door

After the CTA, be honest: the product is not built yet. Offer waitlist, founder call, or refund policy — do not imply shipping software that does not exist.

Benchmarks (calibrate per segment)

  • Landing visitor → signup: 5–15% can be healthy for warm/niche traffic; <1% on cold paid traffic is often a kill signal for that offer.
  • Pre-sales and design-partner time commitments beat email-only signups.
  • Do not count views, likes, or unqualified clicks as validation.

Output

Save plans and results to discovery/<topic-slug>/experiments/YYYY-MM-DD-<topic-slug>-experiment.md.

Default sections: Facts, Assumptions, Evidence, Analysis, Decision, Next experiment.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 326,835. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.