docs-as-marketing
について
このスキルは、開発者が技術文書を効果的なマーケティングツールに変え、ユーザーを惹きつけてコンバージョンにつなげるための支援をします。検索最適化されたドキュメント、コンバージョン重視のクイックスタート、開発者中心の情報アーキテクチャに関する戦略を提供します。ドキュメント主導の成長を実装する必要がある場合や、APIリファレンスを最適化して採用率を向上させたい場合にご活用ください。
クイックインストール
Claude Code
推奨npx skills add jonathimer/devmarketing-skills -a claude-code/plugin add https://github.com/jonathimer/devmarketing-skillsgit clone https://github.com/jonathimer/devmarketing-skills.git ~/.claude/skills/docs-as-marketingこのコマンドをClaude Codeにコピー&ペーストしてスキルをインストールします
ドキュメント
Documentation as Marketing
Documentation is often a developer's first meaningful interaction with your product. Great docs don't just explain—they market. They reduce friction, build trust, and turn curious visitors into active users who recommend your product to others.
Overview
Developer documentation serves multiple marketing functions:
- Acquisition: Docs rank in search and attract developers actively seeking solutions
- Activation: Well-structured quickstarts reduce time-to-value
- Retention: Comprehensive references keep developers building
- Referral: Developers share docs they love, not marketing pages
This skill covers the intersection of technical writing and developer marketing—creating documentation that serves both education and conversion goals.
Before You Start
Review the developer-audience-context skill to understand your target developers:
- What problems are they searching for solutions to?
- What's their technical sophistication level?
- What frameworks and languages do they use?
- Where do they currently look for answers?
Your documentation strategy should directly address these audience insights.
Information Architecture That Converts
The Four Types of Documentation
Structure your docs around the four types developers need:
| Type | Purpose | Marketing Function |
|---|---|---|
| Tutorials | Learning-oriented, step-by-step | Builds confidence, shows product value |
| How-to Guides | Task-oriented, problem-solving | Demonstrates capability breadth |
| Reference | Information-oriented, accurate | Proves product depth and reliability |
| Explanation | Understanding-oriented, conceptual | Establishes thought leadership |
Navigation That Reduces Bounce
Good Navigation Structure:
Getting Started
├── Quickstart (< 5 min)
├── Installation
└── Core Concepts
Guides
├── Authentication
├── [Most Common Use Case]
├── [Second Most Common Use Case]
└── ...
API Reference
├── Overview
├── Authentication
├── Endpoints (alphabetical or logical grouping)
└── SDKs
Resources
├── Examples
├── Changelog
└── Support
Bad Navigation Structure:
Documentation
├── Chapter 1: Introduction
├── Chapter 2: Getting Started
├── Chapter 3: Advanced Topics
├── Appendix A
└── API (link to separate site)
Information Hierarchy
Every documentation page should follow this hierarchy:
- What is this? (1 sentence)
- Why would I use it? (1-2 sentences)
- How do I use it? (the bulk of the page)
- What's next? (clear next steps)
Quickstart Optimization
Your quickstart is your most important conversion page. Optimize ruthlessly.
The 5-Minute Rule
Developers should reach a meaningful success moment within 5 minutes. If your quickstart takes longer, you're losing developers.
Measure and optimize:
- Time from page load to first successful API call
- Drop-off points in the quickstart flow
- Completion rate
Quickstart Structure
# Quickstart
Get your first [meaningful result] in under 5 minutes.
## Prerequisites
- [Specific version] of [language/tool]
- [Account/API key] (link to signup)
## Step 1: Install
[Single command, copy-paste ready]
## Step 2: Configure
[Minimal configuration, explain what each part does]
## Step 3: Run
[The payoff—show them it works]
## What You Built
[Explain what just happened and why it matters]
## Next Steps
- [Immediate next tutorial]
- [Reference docs for what they just used]
- [Community/support link]
Good vs. Bad Quickstarts
Good Quickstart:
# Send Your First Message
Send an SMS in under 5 minutes.
## Prerequisites
- Node.js 16 or higher
- A Twilio account ([sign up free](link))
## Install the SDK
```bash
npm install twilio
Send a Message
Create send-sms.js:
const twilio = require('twilio');
const client = twilio('YOUR_ACCOUNT_SID', 'YOUR_AUTH_TOKEN');
client.messages.create({
body: 'Hello from my app!',
to: '+15551234567',
from: '+15559876543'
}).then(message => console.log(`Sent: ${message.sid}`));
Run it:
node send-sms.js
You should see: Sent: SM1234...
What Just Happened
You authenticated with your API credentials and sent an SMS...
**Bad Quickstart:**
```markdown
# Getting Started
Welcome to our platform! Before we begin, let's discuss
the architecture of our messaging system...
[500 words of background]
## Installation
First, ensure you have the correct version of Node.js.
You can check this by running...
[200 words on version checking]
You'll also need to configure your environment variables.
Create a .env file and add the following variables...
[Complex configuration with 10+ variables]
API Reference Best Practices
Every Endpoint Needs
- One-sentence description of what it does
- Authentication requirements clearly stated
- Request format with all parameters documented
- Response format with example
- Error responses with common causes
- Copy-paste example that actually works
Copy-Paste Code That Works
Critical: Example code must work when copied. Test it.
Good Example:
## Create a User
Creates a new user in your organization.
### Request
```bash
curl -X POST https://api.example.com/v1/users \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"email": "[email protected]",
"name": "Jane Developer"
}'
Response
{
"id": "usr_123abc",
"email": "[email protected]",
"name": "Jane Developer",
"created_at": "2024-01-15T10:30:00Z"
}
Errors
| Code | Meaning |
|---|---|
| 400 | Invalid email format |
| 409 | Email already exists |
| 401 | Invalid or missing API key |
**Bad Example:**
```markdown
## POST /users
Parameters:
- email (string)
- name (string)
- org_id (string, optional)
- role (enum, optional)
- metadata (object, optional)
- ...
Returns a user object.
Language-Specific Examples
Provide examples in languages your developers actually use:
- cURL (universal, always include)
- JavaScript/Node.js
- Python
- Go
- Ruby
- PHP
- Your most-used SDK languages
Search Optimization for Docs
Docs That Rank
Developer documentation can capture high-intent search traffic.
Target Query Types:
- Problem queries: "how to send sms from node.js"
- Comparison queries: "[your product] vs [competitor]"
- Integration queries: "integrate [your product] with [popular tool]"
- Error queries: "[specific error message]"
SEO Fundamentals for Docs
Page Titles:
Good: "Send SMS with Node.js | Twilio Docs"
Bad: "Documentation - Messaging - SMS - Send"
Meta Descriptions:
Good: "Learn how to send SMS messages using Node.js and the
Twilio API. Includes code examples and troubleshooting tips."
Bad: "This page contains documentation for the SMS sending
functionality of our messaging product."
URL Structure:
Good: /docs/sms/send-messages/nodejs
Bad: /docs/section/3/page/27?lang=nodejs
Internal Linking
Create a documentation web, not documentation silos:
- Link related concepts
- Link from reference to tutorials
- Link from tutorials to reference
- Cross-link between SDK docs
Measuring Documentation Effectiveness
Key Metrics
| Metric | What It Tells You |
|---|---|
| Time on quickstart | Engagement (but also confusion) |
| Quickstart completion rate | Conversion effectiveness |
| Search → signup rate | Docs as acquisition channel |
| Support ticket deflection | Docs comprehensiveness |
| Page ratings/feedback | Content quality |
| Internal search queries | Content gaps |
Feedback Loops
Implement:
- "Was this helpful?" on every page
- Internal search analytics (what are people searching for?)
- Support ticket analysis (what questions do docs fail to answer?)
- Developer interviews (what's confusing? What's missing?)
Common Documentation Anti-Patterns
The "Wall of Text"
Problem: Pages with no code, no structure, no visual breaks Fix: Lead with code, use headers liberally, break up paragraphs
The "Assumed Knowledge" Trap
Problem: Assuming developers know your terminology Fix: Define terms on first use, link to glossary
The "Everything Page"
Problem: One page trying to cover all use cases Fix: Separate pages for distinct tasks, link between them
The "Outdated Quickstart"
Problem: Quickstart code that no longer works Fix: Automated testing of documentation code samples
The "Hidden Prerequisites"
Problem: Discovering requirements mid-tutorial Fix: All prerequisites at the top, with version numbers
Tools
Documentation Platforms
- GitBook: Good for smaller teams, nice defaults
- ReadMe: Interactive API docs, metrics built-in
- Mintlify: Modern, fast, good DX
- Docusaurus: Flexible, self-hosted, React-based
- Notion: Quick to set up, limited customization
Code Sample Testing
- Doctest: Python code in docs
- mdx-js: JSX in markdown
- Custom CI: Run code samples as tests
Search and Analytics
- Algolia DocSearch: Free for open source, powerful
- Google Analytics: Basic traffic metrics
- FullStory/Hotjar: Session recording, heatmaps
- Internal search analytics: What are devs searching for?
Related Skills
- api-onboarding: Optimize the complete first API call experience
- sdk-dx: Create SDKs that make your docs simpler
- developer-sandbox: Interactive environments that complement docs
- technical-content-strategy: Broader content strategy including docs
- developer-audience-context: Understanding who you're writing for
GitHub リポジトリ
関連スキル
content-collections
メタこのスキルは、Content Collections(Markdown/MDXファイルを型安全なデータコレクションに変換するTypeScriptファーストのツール)の本番環境でテストされた設定を提供します。Zodバリデーションによる型安全性を実現し、ブログ、ドキュメントサイト、コンテンツ重視のVite + Reactアプリケーション構築時にご利用ください。Viteプラグインの設定、MDXコンパイルから、デプロイ最適化、スキーマバリデーションまで、すべてを網羅しています。
polymarket
メタこのスキルは、開発者がPolymarket予測市場プラットフォームを活用したアプリケーション構築を可能にします。API統合による取引や市場データの取得に加え、WebSocketを介したリアルタイムデータストリーミングにより、ライブ取引や市場活動を監視できます。取引戦略の実装や、ライブ市場更新を処理するツールの作成にご利用ください。
creating-opencode-plugins
メタこのスキルは、開発者がコマンド、ファイル、LSP操作など25種類以上のイベントタイプにフックするOpenCodeプラグインを作成することを支援します。JavaScript/TypeScriptモジュール向けに、プラグイン構造、イベントAPI仕様、および実装パターンを提供します。カスタムイベント駆動ロジックでOpenCode AIアシスタントのライフサイクルをインターセプト、監視、または拡張する必要がある場合にご利用ください。
sglang
メタSGLangは、高性能なLLMサービングフレームワークであり、RadixAttentionプレフィックスキャッシュを活用したJSON、正規表現、エージェントワークフロー向けの高速で構造化された生成を特長とします。特にプレフィックスが繰り返されるタスクにおいて、大幅に高速な推論を実現し、複雑な構造化出力やマルチターン対話に最適です。制約付きデコードが必要な場合や、広範なプレフィックス共有を伴うアプリケーションを構築する場合は、vLLMなどの代替案ではなくSGLangを選択してください。
