MCP Server for Google Analytics 4

MCP Server for Google Analytics 4

By HosakaKeigo GitHub

-

Overview

what is MCP Server for Google Analytics 4?

MCP Server for Google Analytics 4 is a server designed to facilitate the retrieval and analysis of Google Analytics 4 data, providing various tools to fetch metrics and generate reports.

how to use MCP Server for Google Analytics 4?

To use the MCP Server, clone the repository, install the necessary dependencies, set up your credentials in the .env file, and run the server. Integration with Claude for Desktop is also supported for enhanced functionality.

key features of MCP Server for Google Analytics 4?

  • Fetches page views, active users, events, and user behavior metrics.
  • Provides metadata and resources for GA4 properties.
  • Includes prompts for data analysis and report generation.

use cases of MCP Server for Google Analytics 4?

  1. Analyzing website traffic and user engagement metrics.
  2. Generating reports for marketing performance.
  3. Monitoring user behavior to improve website design.

FAQ from MCP Server for Google Analytics 4?

  • What are the prerequisites for using the server?

You need a Google Analytics 4 property, a Google Cloud service account, Node.js 20 or higher, and the pnpm package manager.

  • How do I troubleshoot authentication errors?

Check your service account key and permissions to ensure they are correctly set up.

  • Is the server fully tested?

No, the server is a work in progress and not yet fully tested. Use at your own discretion.

Content

MCP Server for Google Analytics 4

NOTE

This is a work in progress. The server is not yet fully tested. Use at your own discretion.

Key Features

This server provides the following MCP functionalities:

Tools

  • get-page-views: Fetches page view metrics for a specified date range.
  • get-active-users: Fetches active user metrics for a specified date range.
  • get-events: Fetches event metrics for a specified date range.
  • get-user-behavior: Fetches user behavior metrics such as session duration and bounce rate.

Resources

  • ga4://property-info: Metadata for the GA4 property including available dimensions, metrics and comparisons.

The followings are more lightweight resources.

  • ga4://dimensions: List of available GA4 dimensions.
  • ga4://filters-help: Help documentation on using GA4 filters.
  • ga4://metrics: List of available GA4 metrics.

Prompts

  • analyze-data: Data analysis assistant.
  • create-report: Report generation template.
  • select-dimensions: Assists in selecting appropriate dimensions for analysis goals.

Setup

Prerequisites

  • A Google Analytics 4 property.
  • A Google Cloud service account with access to the GA4 API.
  • Node.js 20 or higher.
  • pnpm package manager.

Installation

# Clone the repository
git clone https://github.com/yourusername/mcp-server-ga4.git
cd mcp-server-ga4

# Install dependencies using pnpm
pnpm install

# Create the .env file
cp .env.example .env

# Edit the .env file and set your credentials
# GOOGLE_CLIENT_EMAIL=your-service-account-email@example.iam.gserviceaccount.com
# GOOGLE_PRIVATE_KEY=your-private-key
# GA_PROPERTY_ID=your-ga4-property-id

Auth

If GOOGLE_CLIENT_EMAIL and GOOGLE_PRIVATE_KEY are set, the server will authenticate with the service account. If not, it will use the default credentials.

Make sure your project enables the Google Analytics API and that your service account has the necessary permissions.(https://www.googleapis.com/auth/analytics.readonly

Here's how to set up Google Application Default Credentials:

  1. Install the Google Cloud SDK.
  2. Run the following command to authenticate:
gcloud auth application-default login --scopes=openid,https://www.googleapis.com/auth/userinfo.email,https://www.googleapis.com/auth/cloud-platform,https://www.googleapis.com/auth/analytics.readonly

# If you want to set a quota project
gcloud auth application-default set-quota-project <my-quota-project>

Build and Run

# Build the server
pnpm run build

# Run the server
pnpm start

Usage

Integration with Claude for Desktop

Add the following to your claude_desktop_config.json file:

{
  "mcpServers": {
    "ga4": {
      "command": "node",
      "args": ["/path/to/mcp-server-ga4/dist/index.js"],
      "env": {
        "GOOGLE_PROJECT_ID": "your-project-id",
        "GOOGLE_CLIENT_EMAIL": "your-service-account-email@example.iam.gserviceaccount.com",
        "GOOGLE_PRIVATE_KEY": "your-private-key",
        "GA_PROPERTY_ID": "your-ga4-property-id"
      }
    }
  }
}

If you want to use the default credentials, set GOOGLE_APPLICATION_CREDENTIALS in your environment:

{
  "mcpServers": {
    "ga4": {
      "command": "node",
      "args": ["/path/to/mcp-server-ga4/dist/index.js"],
      "env": {
        "GOOGLE_PROJECT_ID": "your-project-id",
        "GA_PROPERTY_ID": "your-ga4-property-id",
        "GOOGLE_APPLICATION_CREDENTIALS": "/path/to/your/credentials.json"
      }
    }
  }
}

Filtering

You can use the filter parameter with all tools to narrow down results:

{
  "dimension": [
    {
      "name": "pagePath",
      "stringEquals": "/blog/article-1"
    }
  ],
  "metric": [
    {
      "name": "screenPageViews",
      "greaterThan": 100
    }
  ],
  "operator": "AND"
}

Pagination

When dealing with large amounts of data, use the limit and offset parameters:

// First page (items 1-50)
get-page-views with limit=50, offset=0

// Second page (items 51-100)
get-page-views with limit=50, offset=50

Troubleshooting

  • Authentication errors: Check your service account key and permissions.
  • No data showing: Verify the date range and property ID are correct.
  • Connection errors: Check your network settings and firewall.

Testing

Test using the MCP Inspector:

pnpm dlx @modelcontextprotocol/inspector node /path/to/mcp-server-ga4/dist/index.js

(Note: pnpm dlx is the equivalent of npx for executing packages)

License

MIT

Thanks

No tools information available.
No content found.