Market Research by Desire

Reverse-engineer market opportunities from human desires through automated 3-phase multi-agent analysis.

Overview

Instead of starting from “I have an idea” or “I have technology,” this plugin starts from “인간 욕망” (human desires) and systematically discovers market opportunities, competitive gaps, and revenue models.

Core Philosophy: 욕망 → 시장 구조 → 경쟁 분석 → 수익 모델

Architecture

graph TD
    A[User Trigger] --> B[3-Round Interview]
    B --> C{Phase 1: PARALLEL}
    C --> D[desire-cartographer]
    C --> E[market-trend-researcher]
    D --> F[desire-map.json]
    E --> G[market-trends.json]
    F --> H{Phase 2: SEQUENTIAL}
    G --> H
    H --> I[competitive-scanner]
    I --> J[competitive-landscape.json]
    J --> K[gap-opportunity-analyzer]
    K --> L[gap-analysis.json]
    L --> M{Phase 3: SOLO}
    M --> N[revenue-model-architect]
    N --> O[revenue-models.json]
    O --> P[Generate 3 Final Documents]
    P --> Q[market-analysis.md]
    P --> R[competitive-analysis.md]
    P --> S[revenue-model-draft.md]

Installation

Part of the claude-ai-engineering plugin collection. No separate installation needed if the repo is already linked.

Usage

Trigger Phrases

English:

• “market research by desire”
• “desire research”
• /market-research-by-desire

Korean:

• “욕망 기반 시장조사”
• “욕망 리서치”
• “욕망에서 시장 찾기”

Example Session

You: 욕망 기반 시장조사 시작
Claude: [Presents 5 desire categories]
You: 성장과성취
Claude: [Presents sub-categories like 전문성개발, 커리어성장, etc.]
You: 전문성개발
Claude: [Asks context questions: target market, solo-dev preference, etc.]
You: Korea, solo-dev preferred, bootstrap budget, tech industry
Claude: [Launches 5 agents in 3 phases, generates 3 documents in 12-18 min]

Output Location

~/.market-research-by-desire/projects/{timestamp-slug}/
├── interview-responses.json
├── artifacts/
│   ├── desire-map.json
│   ├── market-trends.json
│   ├── competitive-landscape.json
│   ├── gap-analysis.json
│   └── revenue-models.json
├── market-analysis.md
├── competitive-analysis.md
├── revenue-model-draft.md
└── README.md

Agents

Agent	Phase	Role	Model
desire-cartographer	1 (Parallel)	Map desires → market structure, generate search terms	sonnet
market-trend-researcher	1 (Parallel)	WebSearch for TAM/SAM/SOM, trends, growth drivers	sonnet
competitive-scanner	2 (Sequential)	WebSearch competitors, pricing, features	sonnet
gap-opportunity-analyzer	2 (Sequential)	Synthesize all data → gaps & positioning	sonnet
revenue-model-architect	3 (Solo)	Design 3-5 revenue models with unit economics	sonnet

Configuration

Edit config/settings.yaml:

output:
  language: "ko"  # or "en"
  base_directory: "~/.market-research-by-desire/projects"

web_search:
  max_results_per_query: 20
  enable_web_fetch: true

market_research:
  default_target_market: "Korea"
  som_market_share_y3: 0.030  # 3% market share by year 3

solo_dev:
  disqualify_two_sided_marketplace: true  # Filter out high-ops models

Key Differentiators

1. Desire-first framework: 5-category taxonomy (생존과안전, 성장과성취, 연결과소속, 자유와통제, 즐거움과자극)
2. Desire intersection opportunities: Unique cross-desire market gaps (e.g., 성취 + 연결 = community learning)
3. Solo-dev feasibility scoring: Automatic filtering for solo entrepreneurs
4. Korean market specialization: 통계청, KOSIS, Naver/Daum search optimization
5. 3-document deliverable: Market analysis, competitive analysis, revenue model draft

Troubleshooting

Issue	Cause	Solution
No market data found	Emerging/niche market	Plugin generates estimates using proxy markets
Few/no competitors	Blue ocean market	Report includes “market education risk” section
WebSearch quota exceeded	Too many queries	Set enable_web_fetch: false in config
Agent timeout	Complex market	Increase timeout in config/settings.yaml

Performance

• Execution time: 12-18 minutes
• Cost estimate: $1-2 USD (varies with WebSearch volume)
• Output quality: Sonnet model for all agents ensures high-quality analysis

Next Steps After Running

1. Validate market size: Cross-reference TAM/SAM with additional sources
2. Interview 20 target users: Confirm desire intensity and pain points
3. Build MVP for top gap: Focus on highest-scoring gap from gap-analysis.json
4. Test revenue model: Run pre-sales or landing page experiment
5. Use Business Avengers plugin: For full product development pipeline

Related Skills

• business-avengers: Follow-up for MVP planning and execution
• planning-interview: For detailed product feature planning
• ai-digest: For ongoing market trend monitoring

License

MIT

Author

Jay Kim (https://github.com/JayKim88)