Market research

Market research

Goal

Find what already exists, who gets paid today, and where the wedge is for one idea in the current market. Use live web research; do not rely on memory or uncited synthesis.

Keep validation order: idea validation → market validation → product validation. This skill covers market validation research, not MVP build.

Revealed preference (paid tests, pre-orders) outweighs stated preference (interviews alone) for demand proof.

Workflow

1. Frame the idea — one-sentence job-to-be-done, ICP, geography, and solo-plus-AI constraints.
2. Resolve the research — read discovery/INDEX.md, then that folder's TOPIC.md and existing artifacts under discovery/<topic-slug>/ before contradicting them. Do not read discovery/archive/ or unrelated active researches unless the user explicitly asks. If discovery/INDEX.md is missing, create it with the active topic slug.
3. Search the open web — run multiple queries; vary keywords (problem, buyer role, category, "alternative", pricing, complaints).
4. Collect primary evidence — live pricing pages, product positioning, reviews, founder posts, listings, and forum threads.
5. Map the market — summary competitor set (see Output); full memo → competitor-landscape.
6. Identify gaps — recurring complaints, missing workflows, switching costs, and segments incumbents under-serve.
7. Size bottom-up (sketch) — segment count × adoption × price; full TAM/SAM/SOM → market-sizing skill.
8. Why now — timing, regulation, tech shift, or buyer urgency.
9. Decide — proceed, pivot wedge, or kill; name the next cheapest test.

Search playbook

Run at least three of the four families below before concluding:

Family	Example queries
Category	[category] software, [job to be done] tool
Buyer pain	[ICP] [pain] workaround, how to [job] without [incumbent]
Competition	[competitor] alternative, [competitor] pricing, best [category] for [segment]
Demand proxy	[problem] reddit, [category] site:news.ycombinator.com

Prefer current pages. Capture URL, publisher, and access date for each material claim.

Source tiers

Use the smallest set that answers the question. Prefer primary pages over roundups.

Tier	Use for	Examples
Primary	Offers, pricing, positioning, reviews, public traction	Company sites, pricing pages, App Store / Play Store, G2, Capterra, Trustpilot, Product Hunt, Indie Hackers, LinkedIn headcount, public founder posts
Market and keyword proxies	Demand and category language	Google Trends, Google search results, Reddit, Hacker News, niche forums
Idea and revenue databases	Comparable ideas and public revenue signals	IdeaProof, BigIdeasDB, AppMagic, SensorTower
Secondary	Pointers only until triangulated	News roundups, listicles, VC blogs, analyst summaries

Do not treat database labels, TAM copy, or viability scores as verified facts. Triangulate with live competitors and buyer language.

Gap analysis checklist

• At least five named solutions mapped (direct, indirect, or substitute) — summary matrix only
• Pricing captured from live pages when available
• Recurring complaints from reviews or forums (not marketing copy)
• Gap hypothesis stated as a buyer job plus segment, not "no one does this"
• Clone vs wedge decision documented — full positioning → positioning-wedge skill
• Why now articulated
• Kill criteria if no wedge or economics fail the stated goal

Output

Save to discovery/<topic-slug>/research/YYYY-MM-DD-<topic-slug>-market.md unless the user requests a brief-only pass. If the topic folder does not exist, add it under active topics and update discovery/INDEX.md.

Default sections:

• Facts
• Assumptions
• Evidence (with URLs)
• Analysis (market map summary, gap, sizing sketch, why now, risks)
• Decision
• Next experiment

Include:

• Summary competitor matrix (≤5 rows; type, audience, price, strength, weakness, evidence URL) — expand in competitor-landscape → -competitors.md
• Gap table (segment, unmet job, workaround today, wedge, confidence)
• Bottom-up sizing sketch with formula — full worksheet → market-sizing
• Sources list at the end

Escalation

• Economics and pricing depth → revenue-model skill
• Structured competitor memo and battle cards → competitor-landscape skill
• Bottom-up TAM / SAM / SOM → market-sizing skill
• Positioning and category → positioning-wedge skill
• Demand tests after market evidence → validation-experiments skill
• End-to-end orchestration → discovery-playbook or behavior-led-validation skill
• Final go/no-go → idea-scorecard skill

Do not recommend implementation until a scorecard records Strong go or Conditional go with explicit gaps to close.