Google Analytics & Search Console Improvement Skill

中文文档

An AI agent skill that performs goal-driven website analysis — collecting data from Google Search Console and GA4, auditing SEO/GEO readiness, and generating prioritized improvement plans with data visualization.

What It Does

Give your AI agent a website URL, and it will:

1. Understand your website goals — visit the site, define the intended user journey
2. Collect search & analytics data — via GSC and GA4 APIs (or manual CSV export)
3. Analyze where users diverge — map search queries and user behavior against your goals
4. Audit the live site — SEO metadata, GEO (AI search) readiness, performance, security
5. Generate a prioritized report — P0-P3 action items with charts, execution roadmap

Core Principle

Goal → Data → Gap → Action

Every analysis starts from what your site wants users to do. The skill finds where reality diverges from intention and tells you exactly what to fix.

Installation

npx skills add morvanzhou/data-driven-product

Quick Start

Tell your AI agent:

Analyze example.com using the data-driven-product skill

The agent will guide you through setup and run the full analysis automatically.

Data Collection Modes

Mode	Setup	Best For
API auto-collect (recommended)	Google Cloud Service Account (~10 min first time)	Full analysis with all data
Manual CSV export	None	Quick analysis with exported data
Browser audit only	None	Technical SEO/GEO audit without analytics data

API Setup (Mode A)

1. Create a Google Cloud project, enable Search Console API, Analytics Data API, and PageSpeed Insights API
2. Create a Service Account and download the JSON key
3. Grant access: add the SA email as viewer in both GSC and GA4
4. Tell the agent your key file path, GSC site URL, and GA4 Property ID

Tip: GSC has two property types — Domain (sc-domain:example.com) and URL-prefix (https://example.com). Using the wrong format causes a 403 error.

See references/gsc-api-guide.md for detailed setup instructions.

Analysis Workflow

Phase 0  →  Website reconnaissance & goal definition
Phase 1  →  Data collection (API / CSV / browser-only)
Phase 2  →  Search performance analysis (GSC)
Phase 3  →  User behavior analysis (GA4)
Phase 3b →  Funnel exploration (optional, custom events)
Phase 4  →  Live site audit (performance, SEO, security)
Phase 5  →  Source code review (optional)
Phase 5b →  SEO & GEO optimization checklist
Phase 6  →  Goal-aligned improvement report with charts

Output

The skill generates a full set of analysis reports and charts in .skills-data/data-driven-product/:

• analysis/improvement-report.md — Final report with executive summary, goal achievement status, P0-P3 prioritized actions, and execution roadmap
• analysis/ — Phase-by-phase detailed reports (search analysis, behavior analysis, funnel analysis, site audit, SEO/GEO checklist)
• charts/ — Data visualization charts (PNG) embedded in reports

Companion Skills

Skill	Use Case
seo-geo	Implement SEO/GEO recommendations from the report
agent-browser	Browser automation for site auditing
frontend-design	Implement frontend/UX improvements

License

MIT