How MCP servers connect GA4, Google Ads and more directly to AI

MCP servers in marketing: analyse marketing data directly in Claude

MCP Server connects marketing data from GA4, Search Console, Google Ads and Meta Ads with an AI chat for real-time analysis.

Ralph Grundmann · Last updated on 4 November 2026

So far, every reporting process has been the same: exporting data from GA4, downloading from Google Ads, pulling from Meta, merging into a table, preparing, and interpreting. This takes time that no one has and produces reports that hardly anyone reads.

Now that AI assistants such as Claude can connect directly to external data sources via the Model Context Protocol (MCP), this is changing fundamentally. Instead of exporting data and evaluating it manually, you simply ask a question and receive a complete analysis in return.

The protocol that connects AI with your data

What is MCP anyway?

MCP stands for Model Context Protocol and is an open standard that allows AI assistants to access external data sources directly. The AI assistant queries the data itself via the MCP server, without copying or CSV uploads.

Specifically, this means: Claude, the AI assistant from Anthropic, can communicate directly with GA4, the Google Search Console, Google Ads or the Meta Marketing API via a configured MCP server.

MCP Server connects GA4, Search Console, Google Ads and Meta Ads with Claude for real-time data queries.

The result is a conversational partner that knows your marketing data.

From data query to finished analysis in seconds

What this means in practice

A concrete example: You open Claude, type “How did my organic traffic develop in April?” and receive a structured analysis within seconds, including sessions, top landing pages, keyword trends and a comparison to the previous month.

Exports, pivot tables and preparation are no longer needed.

The same works across channels. When GA4, Search Console, Google Ads, Meta and affiliate data all converge in one system, 360° analyses become possible with far less manual work:

Organic traffic declines. It becomes immediately visible whether Paid Search is compensating for it
Meta campaign performs strongly. You can see immediately if there is a halo effect on organic visibility
Affiliate revenue increases, directly linked to traffic from all other channels

Connections that today disappear in data silos become visible.

The current status of the MCP integrations

Which channels can be connected

For many marketing platforms, MCP servers already exist:

GA4: easiest setup, ready to use immediately
Google Search Console: rankings, impressions, CTR directly queryable
Google Ads: campaign performance, keywords, ROAS
Meta Ads: Facebook and Instagram in one query
Microsoft Advertising: campaign data can be integrated
Pinterest Ads: campaign data readily available
Awin: publisher and affiliate data can be queried

All channels in one system, queryable with a single question.

Faster decisions, less effort

What this means for your business

When you can access a complete channel analysis at any time without waiting days for a report, it changes the way you make decisions. What used to take hours now takes minutes.

In concrete terms: more frequent insights, faster reactions to changes in the market and a marketing team that works with data.

What MCP does not provide

Where the limits are

MCP is not a substitute for strategy. The quality of the analysis directly depends on the quality of the questions asked and how well the underlying data is maintained. Garbage in, garbage out applies here just as it does with any other analysis tool.

For productive use, a well-maintained data structure and a GDPR-compliant infrastructure are also required. This is no small task, but it is a one-time effort that pays off in the long run.

The first step is easier than you think

How to get started now

The easiest entry point is GA4. The connection to Claude is set up in under 15 minutes, without any programming knowledge, without any infrastructure of your own. This is enough to get a feel for how AI-supported data analysis works in practice.
The next step is a central system that brings together all your channels and makes them available to your entire team. We will demonstrate what this could look like further down.

Step by step

How to set up GA4 in Claude

The setup runs via Claude Desktop (the free desktop app from Anthropic). Important: download the .exe directly from claude.ai/download, as the Windows Store version does not reliably support local MCP servers.

Install Node.js

Claude Desktop starts the connection to the GA4 MCP server via npx (mcp-remote package), so you need Node.js. Download it from nodejs.org; the LTS version is sufficient. After installation, check in the terminal:

node --version

If a version number appears, everything is correct.

Install Claude Desktop

Download and install directly from claude.ai/download. Avoid the link from the Windows Store, as the Store version currently does not reliably support local MCP servers.

Open config file

In Claude Desktop: Settings → Developer → Edit Config. The file is located at:

C:\Users\[Username]\AppData\Roaming\Claude\claude_desktop_config.json

On a Mac, it is located at ~/Library/Application Support/Claude/claude_desktop_config.json.

Edit configuration file

Insert the following block:

{
  "mcpServers": {
    "ga4-mcp-server": {
      "command": "C:\\Program Files\\nodejs\\npx.cmd",
      "args": [
        "-y",
        "mcp-remote",
        "https://mcp-ga.stape.ai/mcp"
      ]
    }
  }
}

On a Mac, "npx" is enough as the command.

Save configuration file

Save, completely close Claude Desktop and restart.

Under Settings → Developer, ga4-mcp-server should appear with the status Running.

Connect your account and ask a question

The first time you call it, a browser window will automatically open for Google authentication. Select the Google account that has access to the GA4 property and confirm access.

In a new chat, simply ask: Which GA4 properties do I have?

Or get started directly: Show me the sessions of the last 30 days broken down by channel.

More than GA4

The next step: all channels in one system

GA4 is the easy entry point. The real value arises when all marketing channels (Search Console, Google Ads, Meta, Microsoft Advertising, Pinterest, Awin and Sistrix) converge in a central system that your entire team can access.

Such a system has to be built around your customer structure, your channels and your internal workflows. The technology for this is available. What is needed is someone to set it up properly.

We at Rheinwunder build exactly that: If you want to know what a central MCP reporting system could look like for your company, get in touch with us. You can read what we offer on the page MCP server for marketing data.