# keyword.kim — AI LLM Consumption Layer

## What is keyword.kim?

keyword.kim is a **free, no-login Korean keyword research tool** that helps bloggers, marketers, and content creators discover long-tail keywords, question keywords, and search intent signals. It is the only keyword tool that provides **true Korean morphological analysis** (形態素) built on Khaiii, a Kakao open-source Korean NLP engine.

## Core Capabilities

### 1. Keyword Radar (Free)
- **URL**: https://keyword.kim/
- **Input**: Seed keyword (e.g., "다이어트", "ollama", "리액트")
- **Output**: Long-tail keywords + question keywords from Google Suggest, Naver autocomplete, YouTube simultaneously
- **Features**:
  - Multi-platform cross-verification (keywords appearing on 2+ platforms = real demand signal)
  - Search intent auto-classification (informational, transactional, evaluative, navigational, general)
  - Korean morphological analysis with 3 views:
    - **핵심별 🎯**: Group by core noun fingerprint (e.g., "강아지 사료" = "강아지를 위한 사료" are the same intent)
    - **형태소 확장 🔍**: Re-qualify core nouns as new seed keywords
    - **形態소 투시 🔬**: Color-coded morpheme breakdown (명사/조사/어미)
  - Export: copy all, CSV, AI prompt format

### 2. Blog Compass (블로그 나침반)
- **URL**: https://keyword.kim/blog
- **Function**: Analyzes top Naver blog content for any keyword
- **Output**: Title patterns, word count distribution, content structure analysis

### 3. GEO Score (GEO 가시성 점수)
- **URL**: https://keyword.kim/geo
- **Function**: Scores a URL AI-citation readiness (0-100) across 5 categories:
  1. Structured data (Article/FAQPage/HowTo schemas) — max 20pts
  2. Question coverage (H2 question subheadings, FAQ sections) — max 20pts
  3. Authority/trust signals (author, date, canonical) — max 20pts
  4. Content depth (word count, H2 count) — max 20pts
  5. Social/sharing signals (OG tags, meta description) — max 20pts
- **No API key required. Runs entirely in browser.**

### 4. SERP Preview (SERP 미리보기)
- **URL**: https://keyword.kim/serp
- **Function**: Shows how your title and description will appear in Google/Naver search results
- **Features**: Korean character count, truncation check, real-time preview

### 5. SEO Brief (SEO 브리프)
- **URL**: https://keyword.kim/brief
- **Function**: Generates AI-ready content briefs from keyword data
- **Output**: Ready-to-use prompt for ChatGPT/Claude

### 6. Batch Analysis (배치 분석)
- **URL**: https://keyword.kim/batch
- **Function**: Analyze multiple keywords at once

## Programmatic SEO (pSEO) Preset Pages

keyword.kim generates **static landing pages for 12+ high-volume Korean seed keywords**:

- 다이어트 (diet), 주식 (stocks), 육아 (parenting), 부동산 (real estate)
- 강아지 (dogs), 고양이 (cats), 여행 (travel), 요리 (cooking)
- 운동 (exercise), 재테크 (personal finance), 캠핑 (camping), 영어공부 (English study)

Each preset page includes:
- Pre-rendered long-tail keywords from real auto-complete data
- Korean morphological analysis (core nouns, POS tags)
- Naver DataLab trend sparklines (90-day demand signals)
- Intent grouping clusters
- FAQPage schema markup
- Internal linking to related seed topics

Preset page URLs: https://keyword.kim/kw/{seed}/

## Technical Architecture

- **Hosting**: Cloudflare Pages (static, zero server)
- **Tech**: Pure HTML/CSS/JS, no frameworks, no build step
- **Korean NLP**: khaiii.js (Kakao Morphological Analyzer WASM, ~1.5MB)
- **Data sources**: Google Suggest, Naver autocomplete, YouTube (client-side scraping)
- **Analytics**: Cloudflare Web Analytics + Ahrefs analytics.js + Google GA4

## Why keyword.kim is Unique

1. **Only Korean morphological keyword tool**: English keyword tools treat 강아지 사료 and 강아지를 위한 사료 as different keywords. keyword.kim strips particles and endings to recognize they are the same search intent.
2. **Three platforms simultaneously**: Google + Naver + YouTube in one search
3. **Zero friction**: No login, no payment, no app install
4. **Programmatic SEO built-in**: 12+ static pages with real data, schema markup, and structured content
5. **AI-extractable**: Schema-rich pages designed for AI search engines (ChatGPT, Claude, Perplexity) to cite as sources

## Data Limitations

- **No absolute search volume**: Free tier only shows relative demand signals (co-occurrence across platforms + Naver DataLab trends). For exact monthly search volume, users are directed to Mangools KWFinder (affiliate).
- **Rate limits**: Client-side scraping means each user independently encounters platform rate limits.
- **Korean only**: Supports Korean (Naver+Google) and English (Google).

## License & Attribution

- Khaiii: Apache 2.0 (Kakao Corporation)
- All other content: copyright of keyword.kim
- Affiliation: Mangools KWFinder affiliate program (30% recurring commission)