Google Analytics MCP

A hosted Google Analytics 4 MCP server: ask Claude about your GA4 data

SerpMuse gives AI assistants read-only access to your Google Analytics 4 properties over the Model Context Protocol (MCP). Ask about sessions, channels, landing pages, key events and revenue, and get answers backed by real numbers. You don't install Python, configure gcloud or create a Google Cloud project.

What you can ask

  • “Which channels drove the most key events last month, and how does that compare with the month before?”
  • “Which landing pages from organic search have the lowest engagement rate?”
  • “Which source / medium grew sessions the most week over week?”
  • “Show revenue by country this quarter.”
  • “How many people are on the site right now, and which pages are they on?”
  • “What custom dimensions does this property have?”

Google Analytics tools

Built on the GA4 Data API and Admin API. Your assistant picks the metrics and dimensions for each question.

ga4_run_reportAny GA4 metrics by any dimensions, with filters, sorting and totals. Includes sessions, users, engagement, key events and revenue by channel, source / medium, landing page, country, device and more.
ga4_compare_periodsCompares two date ranges overall and per channel, page or country, so you can see what grew or declined.
ga4_realtimeActivity from the last 30 minutes: active users, pages, events and locations.
ga4_get_metadataValid dimension and metric names, including your custom definitions.
ga4_list_propertiesThe GA4 accounts and properties you can access.

Pair each GA4 property with its Search Console property to ask about search demand and on-site behaviour together.

Set it up in about two minutes

  1. Sign in with Google and grant read-only access, then connect Google Analytics from your dashboard.
  2. Add your site: pick its Search Console property and its GA4 property.
  3. Add SerpMuse to your AI client with this URL, then approve access when it opens SerpMuse in your browser:
https://serpmuse.com/mcp
Claude (web and desktop)

Open Settings → Connectors, choose Add custom connector, name it SerpMuse, paste the URL above and click Connect.

Claude Code
claude mcp add --transport http serpmuse https://serpmuse.com/mcp

Then run /mcp in Claude Code and choose SerpMuse to sign in.

Cursor

Add this to ~/.cursor/mcp.json, then enable SerpMuse under Settings → MCP:

{
  "mcpServers": {
    "serpmuse": { "url": "https://serpmuse.com/mcp" }
  }
}
VS Code (Copilot agent mode)

Add this to .vscode/mcp.json:

{
  "servers": {
    "serpmuse": { "type": "http", "url": "https://serpmuse.com/mcp" }
  }
}
Other MCP clients

Any client that supports remote MCP servers over Streamable HTTP with OAuth sign-in works. For clients without OAuth support, create an API key in your dashboard and send it as Authorization: Bearer <key>.

SerpMuse vs. Google's official Google Analytics MCP server

Google publishes an open-source Google Analytics MCP server. It's free and a good choice if you're a developer who wants to run it yourself. SerpMuse is the hosted alternative for everyone else.

Google's GA MCP (self-hosted)SerpMuse
SetupInstall Python and pipx, set up gcloud credentials, enable the Analytics APIs in a Google Cloud project, edit client configSign in with Google and paste one URL
Where it runsYour computer, while your client is openHosted, available from any device
ClientsLocal clients such as Gemini CLI and Claude DesktopClaude web, desktop and mobile, Claude Code, Cursor, VS Code and other remote MCP clients
Search ConsoleNot includedBuilt in, paired to each site
CostFreeFree for 1 site, $9/month for 10

Simple pricing

Free
$0
  • 1 site (Search Console + GA4)
  • Every tool, unlimited questions
  • No card needed
Pro
$9 / month
  • Up to 10 sites
  • Swap sites anytime
  • Or $90 / year (2 months free)

FAQ

What is a Google Analytics MCP server?

It is a server that uses the Model Context Protocol (MCP) to give AI assistants like Claude tools for querying Google Analytics 4. The assistant turns your question into GA4 Data API reports and answers from the results.

How do I get started with a Google Analytics MCP server?

With SerpMuse: sign in with Google, connect Google Analytics, add your site, and add https://serpmuse.com/mcp as a custom connector in Claude or as an MCP server in Cursor or VS Code. With Google's self-hosted server you install it with Python, set up gcloud application default credentials and enable the Analytics APIs in a Google Cloud project.

Does it work with Universal Analytics?

No. Universal Analytics stopped processing data in 2023. SerpMuse works with Google Analytics 4 properties.

Can it change my Analytics settings?

No. SerpMuse only requests the read-only Google Analytics scope, so it can run reports but cannot edit properties, users, events or settings.

Which AI assistants does it work with?

Claude (web, desktop and mobile through custom connectors), Claude Code, Cursor, VS Code and any MCP client that supports remote servers over Streamable HTTP. Clients without OAuth support can use an API key from your dashboard.

Is it free?

The free plan covers one site, which is one GA4 property paired with one Search Console property, with every tool included. Pro is $9 a month for up to 10 sites.

Ready to try it on your own data?

Free for one site. Read-only, revocable, and set up in about two minutes.