algo-seo-content
從關鍵字研究到站內最佳化的內容 SEO 策略
中文翻譯
---
name: "algo-seo-content"
description: "執行內容 SEO 策略,從關鍵字研究一路到內容規劃、撰寫和站內最佳化。當使用者需要建立 SEO 最佳化的內容、做關鍵字研究、找出內容缺口、或提升現有內容的排名時使用,即使他們說的是「內容策略」、「關鍵字研究」或「這個主題要怎麼排上去」也一樣。"
metadata:
category: "WP-35 SEO 演算法"
tags: ["seo", "content-strategy", "keyword-research", "on-page-seo"]
---
# 內容 SEO 策略(Content SEO Strategy)
## 概述
內容 SEO 是有系統地建立並最佳化內容,使其符合搜尋意圖並自然排名的過程。流程:關鍵字研究 → 意圖對應 → 內容建立 → 站內最佳化 → 成效監控。成功取決於意圖吻合,不是關鍵字密度。
## 何時使用
**觸發條件:**
- 要規劃新內容來抓自然搜尋流量
- 要最佳化現有表現不佳的內容
- 要做關鍵字研究和內容缺口分析
**何時不要用:**
- 問題出在技術面(頁面速度、可爬取性)時,用技術 SEO
- 問題出在站外(反向連結、權威度)時,用反向連結分析
## 演算法
```
鐵律:內容必須符合「搜尋意圖」
針對錯誤意圖、但最佳化得很完美的頁面,「不會」有排名。
四種意圖類型:
1. 資訊型:想學習(「how to」、「what is」)
2. 導航型:想找特定網站(「github login」)
3. 商業型:比較選項(「best CRM 2025」)
4. 交易型:想買/想做(「buy iPhone 16 case」)
動筆前,先看 SERP 結果來判斷實際的意圖。
```
### 階段 1:輸入驗證
定義目標主題/利基。從腦力激盪、競爭對手分析和工具(Ahrefs、SEMrush、Google Keyword Planner)蒐集種子關鍵字。
**關卡:** 種子關鍵字清單,附上搜尋量和難度估計。
### 階段 2:核心演算法
1. **關鍵字分群**:依意圖和主題把相關關鍵字分組
2. **SERP 分析**:檢查每一群的前 10 名結果,找出意圖、內容形式和深度
3. **內容缺口分析**:找出競爭對手有排名、你卻沒有的關鍵字
4. **內容 brief**:定義:目標關鍵字、意圖、形式(指南/清單/比較)、字數基準、從 SERP 分析得出的必要子主題
5. **站內最佳化**:標題標籤(關鍵字放前面)、meta description、H1/H2 結構、內部連結、圖片 alt 文字
### 階段 3:驗證
檢查:標題包含主要關鍵字、意圖符合 SERP、涵蓋所有關鍵子主題、有連到相關內容的內部連結。
**關卡:** 內容符合已辨識的意圖,且涵蓋從 SERP 得出的子主題。
### 階段 4:輸出
回傳內容 brief 和最佳化檢查清單。
## 輸出格式
```json
{
"content_brief": {"primary_keyword": "...", "intent": "informational", "format": "how-to guide", "target_word_count": 2000, "subtopics": ["...", "..."]},
"optimization": {"title": "...", "meta_description": "...", "h2_structure": ["...", "..."], "internal_links": 5},
"metadata": {"search_volume": 2400, "keyword_difficulty": 35}
}
```
## 範例
### 輸入輸出範例
**輸入:** 主題「email marketing automation」,目標市場:台灣中小企業
**預期:** 主要關鍵字:「email行銷自動化」、意圖:商業調查型、形式:比較指南、2500 字
### 邊界情況
| 輸入 | 預期 | 原因 |
|-------|----------|-----|
| 搜尋量為零的關鍵字 | 想想它是新興的還是根本不存在 | 如果主題相關,可能還是值得鎖定 |
| 新網站遇到 KD > 80 | 先鎖定長尾變體 | 新網站在高 KD 的詞上沒辦法競爭 |
| 混合意圖的 SERP | 建立符合主導意圖的內容 | 不要想在一個頁面滿足所有意圖 |
## 常見陷阱
- **關鍵字密度已死**:沒有所謂最佳的關鍵字密度。自然地寫。硬塞關鍵字會傷害可讀性,還可能觸發垃圾訊號。
- **搜尋量 ≠ 流量**:搜尋量 1 萬的關鍵字排第 1,不會帶來 1 萬次造訪。點擊率會隨 SERP 功能(廣告、精選摘要、PAA)而不同。
- **內容新鮮度**:有些查詢需要新內容(例如「best laptop 2025」)。過時的內容,就算曾經排第 1,排名也會掉。
- **自我蠶食(cannibalization)**:多個頁面鎖定同一個關鍵字,會互相競爭。一個強的頁面勝過三個平庸的。
- **E-E-A-T**:對 YMYL 主題(健康、金融),Google 要求展現經驗、專業、權威和信任。匿名內容排不上去。
## 參考資料
- 關鍵字分群的方法,見 `references/keyword-clustering.md`
- 內容 brief 的範本,見 `references/content-brief-template.md`英文原文
---
name: "algo-seo-content"
description: "Execute content SEO strategy from keyword research through content planning, writing, and on-page optimization. Use this skill when the user needs to create SEO-optimized content, perform keyword research, identify content gaps, or improve existing content rankings — even if they say 'content strategy', 'keyword research', or 'how to rank for this topic'."
metadata:
category: "WP-35 SEO 演算法"
tags: ["seo", "content-strategy", "keyword-research", "on-page-seo"]
---
# Content SEO Strategy
## Overview
Content SEO is the systematic process of creating and optimizing content to match search intent and rank organically. The pipeline: keyword research → intent mapping → content creation → on-page optimization → performance monitoring. Success depends on intent match, not keyword density.
## When to Use
**Trigger conditions:**
- Planning new content to capture organic search traffic
- Optimizing existing underperforming content
- Conducting keyword research and content gap analysis
**When NOT to use:**
- When the issue is technical (page speed, crawlability) — use technical SEO
- When the issue is off-page (backlinks, authority) — use backlink analysis
## Algorithm
```
IRON LAW: Content Must Match SEARCH INTENT
A perfectly optimized page targeting the wrong intent will NOT rank.
Four intent types:
1. Informational — wants to learn ("how to", "what is")
2. Navigational — wants a specific site ("github login")
3. Commercial — comparing options ("best CRM 2025")
4. Transactional — wants to buy/do ("buy iPhone 16 case")
Check SERP results to determine actual intent before writing.
```
### Phase 1: Input Validation
Define target topic/niche. Gather seed keywords from brainstorming, competitor analysis, and tools (Ahrefs, SEMrush, Google Keyword Planner).
**Gate:** Seed keyword list with search volume and difficulty estimates.
### Phase 2: Core Algorithm
1. **Keyword clustering**: Group related keywords by intent and topic
2. **SERP analysis**: Check top 10 results for each cluster — identify intent, content format, and depth
3. **Content gap analysis**: Find keywords competitors rank for that you don't
4. **Content brief**: Define: target keyword, intent, format (guide/list/comparison), word count benchmark, required subtopics from SERP analysis
5. **On-page optimization**: Title tag (keyword front-loaded), meta description, H1/H2 structure, internal links, image alt text
### Phase 3: Verification
Check: title contains primary keyword, intent matches SERP, all key subtopics covered, internal links to related content.
**Gate:** Content matches identified intent and covers SERP-derived subtopics.
### Phase 4: Output
Return content brief and optimization checklist.
## Output Format
```json
{
"content_brief": {"primary_keyword": "...", "intent": "informational", "format": "how-to guide", "target_word_count": 2000, "subtopics": ["...", "..."]},
"optimization": {"title": "...", "meta_description": "...", "h2_structure": ["...", "..."], "internal_links": 5},
"metadata": {"search_volume": 2400, "keyword_difficulty": 35}
}
```
## Examples
### Sample I/O
**Input:** Topic "email marketing automation", target market: Taiwan SMBs
**Expected:** Primary KW: "email行銷自動化", intent: commercial investigation, format: comparison guide, 2500 words
### Edge Cases
| Input | Expected | Why |
|-------|----------|-----|
| Zero search volume keyword | Consider if it's emerging or nonexistent | May be worth targeting if topically relevant |
| KD > 80 for new site | Target long-tail variants first | New sites can't compete on high-KD terms |
| Mixed intent SERP | Create content matching dominant intent | Don't try to serve all intents in one page |
## Gotchas
- **Keyword density is dead**: There's no optimal keyword density. Write naturally. Forced keyword insertion hurts readability and may trigger spam signals.
- **Search volume ≠ traffic**: A #1 ranking for a 10K volume keyword won't bring 10K visits. Click-through rates vary by SERP features (ads, featured snippets, PAA).
- **Content freshness**: Some queries demand fresh content (e.g., "best laptop 2025"). Outdated content drops rankings even if it once ranked #1.
- **Cannibalization**: Multiple pages targeting the same keyword compete with each other. One strong page outperforms three mediocre ones.
- **E-E-A-T**: For YMYL topics (health, finance), Google requires demonstrated Experience, Expertise, Authority, and Trust. Anonymous content won't rank.
## References
- For keyword clustering methodology, see `references/keyword-clustering.md`
- For content brief templates, see `references/content-brief-template.md`