What is Mcp Ga4

mcp-ga4 is a Model Context Protocol (MCP) server designed to retrieve information from Google Analytics 4 (GA4) cloud server. It integrates seamlessly with Claude desktop, providing valuable business intelligence insights based on user data.

Use cases

Use cases for mcp-ga4 include generating detailed reports for marketing campaigns, analyzing user engagement in e-commerce platforms, comparing performance metrics over different periods, and providing real-time insights for decision-making.

How to use

To use mcp-ga4, first initialize the project by cloning the repository and navigating to the project folder. Then, create a virtual environment and activate it. Install the necessary dependencies and set up your Google account credentials by downloading the JSON file and configuring the environment variables accordingly.

Key features

Key features of mcp-ga4 include tools to retrieve GA4 reports based on specific metrics and dimensions, real-time reporting, comparative reporting for different time periods, ranking reports based on specified metrics, and listing all available GA4 properties and accounts.

Where to use

mcp-ga4 can be used in various fields where data analytics and business intelligence are crucial, including marketing, e-commerce, web development, and any organization that utilizes Google Analytics for tracking user behavior and performance metrics.

MCP Server For Google Analytics 4

This is a Model Context Protocol (MCP) server with tools to retrieve information from Google analytics 4 cloud server. It integrates easily with claude desktop and gives helpful business intelligence feedback based on your data

Setup Steps

1. Initialize the project

   launch CMD where you want to keep your project, and clone the repo, enter project folder

    git clone https://github.com/smacient/mcp-ga4.git
    cd mcp-ga4

2. Create virtual environment and activate it

    uv venv
    .venv\Scripts\activate

3. Install dependencies:

    uv sync

4. Setup environment

• set up your goggle account credentials and download the json file. rename it credentials.json

• create a .env file inside the localserver folder

• add the correct path to your google credentials in the env file

  GOOGLE_APPLICATION_CREDENTIALS="C:\\Users\\path to\\credentials.json"

Features

The server implements the following features:

Tools

• get_report: retrieve your GA4 report based on required metrics, dimension and timeframe and give a detailed summary and insight based on the data
• get_realtime_report: retrieve your GA4 report realtime report based on required metrics, dimension and timeframe and give a detailed summary and insight based on the data
• compare_report_metrics: retrieve your GA4 report for two periods based on required metrics, dimension and give a comparative report and insight based on the data
• get_report_with_order: retrieve your GA4 report report based on required metrics, dimension and timeframe, rank it based on specified metrics and give a detailed summary and insight based on the data
• list_all_properties: retrieve your GA4 properties available in the current user
• list_all_accounts: retrieve your GA4 account information, which it uses to get other information

Running the Server

To run on terminal

To run the server with the MCP Inspector for development:

uv run mcp dev localserver/ga4_server.py

To run the with openai client:

update the .env

  GOOGLE_APPLICATION_CREDENTIALS="C:\\Users\\`path to`\\credentials.json"
  OPENAI_API_KEY="add your openai api key"

run command

uv run localserver/openai_client.py

To run the with gemini client:

update the .env

  GOOGLE_APPLICATION_CREDENTIALS="C:\\Users\\`path to`\\credentials.json"
  GEMINI_API_KEY="add your gemini api key"

run command

uv run localserver/gemini_client.py

To run on Claude desktop

To install the server in Claude desktop app:

uv run mcp install localserver/ga4_server.py

Set up the configuration properly

• go to claude desktop.
• click settings, go to developer and click on edit config
• open the config file with sublime text
• add the env section as the one shown below.
• edit the claude configuration file to look like the one below. add the paths needed. remove path to and add the correct path

{
  "mcpServers": {
    "Google Analytics 4": {
      "command": "uv",
      "args": [
        "run",
        "--with",
        "mcp[cli]",
        "mcp",
        "run",
        "C:\\Users\\`path to`\\Gemini-Agent-MCP\\src\\gemini_server_sample.py"
      ],
      "env": {
        "VIRTUAL_ENV": "C:\\Users\\`path to`\\Gemini-Agent-MCPmcp_ga4\\.venv",
        "PATH": "C:\\Users\\`path to`\\Gemini-Agent-MCP\\.venv\\Scripts;${PATH}",
        "GOOGLE_APPLICATION_CREDENTIALS": "C:\\Users\\`path to`\\credentials.json"
      }
    }
  }
}

End claude in task manager and restart it. Your tools will be visible on the chat interface.

Ask Claude to get your GA4 information and accept the popup permit to use the tools required.