Files
2nd/10_Wiki/Topics/AI_and_ML/AI-Overviews-and-SGE.md
T
koriweb d8a80f6272 chore(wiki): dangling 링크 canonical 정규화 (768파일/1200건)
이름만 다른(표기 변형) [[위키링크]]를 대상 문서의 canonical 제목으로 치환해
끊겼던 1,200개 링크를 연결. 제목/파일명 정규화 일치만 적용하고 별칭 매칭은
과병합 위험으로 제외(애매성 가드). 원본은 _link_reconcile_backup/ 에 백업.
도구: Datacollect/scripts/link_reconcile_apply.mjs

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-08 12:24:15 +09:00

223 lines
5.3 KiB
Markdown

---
id: wiki-2026-0508-ai-overviews-and-sge
title: AI Overviews and SGE (Search Generative Experience)
category: 10_Wiki/Topics
status: verified
canonical_id: self
aliases: [SGE, AI Overviews, Google AI Search, generative search, zero-click search]
duplicate_of: none
source_trust_level: B
confidence_score: 0.85
verification_status: conceptual
tags: [sge, ai-overviews, google-search, seo, aeo, citation, zero-click, structured-data]
raw_sources: []
last_reinforced: 2026-05-09
github_commit: pending
inferred_by: Claude Opus 4.7 (manual cleanup 2026-05-09)
---
# AI Overviews and SGE
## 📌 한 줄 통찰
> **Google 의 search 의 매 query 의 AI-generated answer + cited source**. 매 brand 의 매 click 의 lose, 매 citation 의 gain. **Direct answer + structured data + Core Web Vitals 의 critical**.
## 📖 핵심
### Evolution
1. **PageRank** (1998): link 의 popularity.
2. **Knowledge Graph** (2012): structured entity.
3. **BERT / RankBrain** (2018+): semantic.
4. **SGE** (2023+): generative answer.
5. **AI Overviews** (2024+): permanent UI.
### 매 component
- **AI-generated answer**: top of SERP.
- **Cited source**: 매 link.
- **Follow-up question**: chat-like.
- **Traditional results**: below.
### Optimization 의 framework
#### Visual Hierarchy
- 매 H1 / H2 / H3 의 clear.
- 매 1 question / heading.
- 매 short paragraph.
#### Direct Answer
- 매 H2 (question) 직후 = 1-2 sentence answer.
- 매 expansion = 다음 paragraph.
- "Inverted pyramid" (journalism).
#### Schema.org
- FAQPage / HowTo / Article.
- Product / Recipe (specific).
- 매 entity 의 explicit.
#### Core Web Vitals
- LCP < 2.5s.
- INP < 200ms.
- CLS < 0.1.
→ 매 SGE 의 selection signal.
#### E-E-A-T
- **Experience**: 매 first-hand.
- **Expertise**: credential / qualification.
- **Authoritativeness**: 매 domain authority.
- **Trustworthiness**: secure, accurate.
→ Google 의 quality rater guideline.
### 매 query type 의 SGE behavior
#### Informational ("How to X")
- AI Overview 가 dominant.
- 매 step-by-step format.
#### Transactional ("buy X")
- AI Overview 가 less.
- 매 product result.
#### Navigational ("Tesla home page")
- AI Overview 거의 X.
- 매 direct site.
#### YMYL (Your Money Your Life)
- 매 strict E-E-A-T.
- 매 medical / financial / legal.
- 매 conservative AI Overview.
### Zero-click 의 reality
- 매 user 의 click 의 ↓.
- 매 publisher 의 traffic 의 hurt.
- 매 brand visibility 의 citation 의 trade-off.
### 매 measurement
#### Search Console
- 매 query 의 AI Overview 의 inclusion.
- 매 impression / position.
#### 3rd-party
- Semrush, Ahrefs, BrightEdge.
- 매 SGE-specific tracking.
#### Self-monitor
- 매 important keyword 의 manual check.
- 매 weekly / monthly.
## 💻 Code
### FAQ schema (top citation chance)
```html
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [
{
"@type": "Question",
"name": "How does AI Overview work?",
"acceptedAnswer": {
"@type": "Answer",
"text": "AI Overview generates a summary answer at the top of search results, citing source URLs."
}
}
]
}
</script>
```
### HowTo schema
```html
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "HowTo",
"name": "How to bake bread",
"step": [
{ "@type": "HowToStep", "name": "Mix flour", "text": "..." },
{ "@type": "HowToStep", "name": "Knead", "text": "..." }
]
}
</script>
```
### 매 page structure (AI-friendly)
```html
<article>
<h1>Topic</h1>
<p>Direct answer in 1-2 sentence (lead).</p>
<h2>What is X?</h2>
<p>Brief definition + key points.</p>
<h2>Why is X important?</h2>
<p>Context + benefit.</p>
<h2>How to use X?</h2>
<ol>
<li>Step 1</li>
<li>Step 2</li>
</ol>
<h2>Common mistakes</h2>
<ul>
<li>Mistake 1</li>
</ul>
</article>
```
### Author bio (E-E-A-T)
```html
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "Person",
"name": "Jane Doe",
"jobTitle": "Senior AI Researcher",
"alumniOf": { "@type": "Organization", "name": "MIT" },
"url": "https://janedoe.com"
}
</script>
```
## 🤔 결정 기준
| Query type | Strategy |
|---|---|
| Informational | FAQ + HowTo schema |
| Product | Product schema + reviews |
| YMYL | E-E-A-T strict |
| Local | LocalBusiness schema |
| News | Article + freshness |
**기본값**: Direct answer + FAQ schema + Core Web Vitals + E-E-A-T author info.
## 🔗 Graph
- 부모: [[SEO]]
- 변형: [[AI-Answer-Engine-Optimization]] · [[Generative-Engine-Optimization]]
- 응용: [[Core Web Vitals Optimization (INP, LCP, CLS)|Core-Web-Vitals]]
- Adjacent: [[Zero-Click-Search]] · [[Knowledge Graph|Knowledge-Graph]]
## 🤖 LLM 활용
**언제**: Content website 의 SGE traffic 의 capture.
**언제 X**: 매 internal app. 매 paid-only content.
## ❌ 안티패턴
- **JS-only render**: bot blind.
- **Fake schema**: penalty.
- **Click-bait + AI 의 misalign**: low citation.
- **No author info**: low E-E-A-T.
## 🧪 검증 / 중복
- Verified (concept).
- 신뢰도 B (Google Search Central, Schema.org).
- Related: [[AI-Answer-Engine-Optimization]] (overlap).
## 🕓 Changelog
| 날짜 | 변경 |
|---|---|
| 2026-05-08 | Phase 1 |
| 2026-05-09 | Manual cleanup — schema code + E-E-A-T + 결정 |