Google MCP servers connect Claude Code to your own Google data: Search Console, Analytics, Business Profile, Merchant Center and Trends, so you can query it in plain English instead of exporting spreadsheets. This page maps every surface, which servers exist for each, and the stack I actually run on this site every day.
The distinction that makes this page necessary: most "SEO MCP" tools ship market data (keywords, rankings, backlinks) from a third-party API. That is useful, and open-source options like open-seo cover it well. But none of them can read YOUR Search Console or YOUR Analytics, because that requires OAuth into your own Google account. First-party access is a separate class of tool, and it is the half where the real decisions live.
What is a Google MCP server?
A Google MCP server is a small program that translates between an AI client like Claude Code and one Google API, using the Model Context Protocol. You authorise it once against your Google account, and from then on the AI can call tools like "run a Search Console query" or "pull GA4 sessions by landing page" as part of a conversation. The data stays between your machine and Google; there is no third-party platform in the middle.
MCP (Model Context Protocol) is the open standard Anthropic released for connecting AI clients to external systems, and it is now supported by Claude Code, Claude Desktop, Cursor and most agent frameworks. One protocol, many servers, which is what makes a per-product stack practical.
Which Google surfaces have MCP servers?
Five Google surfaces matter for search work, and their MCP coverage is uneven. Here is the honest state of each:
| Surface | Official server? | Community options | My take |
|---|---|---|---|
| Google Analytics (GA4) | Yes (googleanalytics/google-analytics-mcp) | Several | The only surface Google itself ships a server for. Start official. Full guide. |
| Search Console | No | 7+ mature options | The richest community field. The one I use most, daily. Full guide and comparison. |
| Google Trends | No (no official API) | Few | I built my own on pytrends and use it for demand-direction on every content decision. Free download below. |
| Business Profile | No | Sparse | The API exists but access approval is a hurdle. Watch this space; sparse tooling today. |
| Merchant Center | No | Sparse | Content API access makes this buildable for ecommerce teams, but off-the-shelf options are thin. |
How do the first-party and market-data halves fit together?
They answer different questions, and you want both connected to the same conversation. Your Search Console tells you what YOUR pages earn. A market-data tool tells you what the rest of the SERP is doing. Neither substitutes for the other.
The cleanest open-source market-data option right now is open-seo (MIT licensed), which runs on DataForSEO and exposes keywords, rank tracking, backlinks and audits over MCP. I treat it as the right-hand column of the diagram above: the market half. The first-party servers on this page are the left-hand column. Claude Code with both connected can answer "which of my striking-distance queries has rising market demand" in one pass, which neither half can answer alone.
What does this look like in real use?
In real use, the first-party stack turns analysis that took an afternoon into a conversation. Three examples from this site in the past fortnight, all run through MCP tools in Claude Code:
- Caught a vanity-metric mirage. Impressions tripled overnight in June while clicks stayed flat. Pulling 90 days of daily Search Console data through MCP showed machine-generated operator queries entering at position 4 to 20: about 96,000 impressions and roughly zero clicks. A dashboard read would have called it growth.
- Stopped a title rewrite that would have cost two page-1 rankings. A full query-profile pull on one tool page showed its title already held position 9.75 for its exact-match term before we traded it away for a 60-impression variant.
- Ran a weekly trends page refresh from live winners-losers data in minutes, operator noise filtered, instead of hand-exporting CSVs.
The pattern across all three: the win is not the data access, it is that the analysis happens where the follow-up questions are. That is also true of the demand-direction layer, which is where Trends comes in.
Get the Google Trends MCP I use (free)
Google Trends has no official API, so I built a small MCP server on pytrends and Google's trending RSS, and I use it on every content decision. Four tools: interest over time (trajectory), related queries (the rising list is the content-gap goldmine), interest by region, and live trending searches. No API key, free, Python.
Download the Google Trends MCP (free)
Install: unzip, pip install -r requirements.txt, add server.py to your MCP client config, and ask Claude "is demand for [topic] rising or dying?". The reading rules that stop the common misreads (discount the last two data points, filter co-occurrence noise, empty does not mean no demand) are in the README.
FAQ
Do I need to know how to code to use Google MCP servers?
No. Most servers install with one command (npx or uvx) plus a one-time Google authorisation in the browser. The setup cost is roughly 15 minutes per surface, once.
Is it safe to connect Claude to my Google account?
The mainstream servers run locally on your machine and talk directly to Google's APIs, so your data does not pass through a third-party platform. Use read-only scopes where offered, and prefer open-source servers whose code you can inspect.
What is the difference between a Google MCP server and open-seo?
open-seo is a market-data platform: keywords, rankings, backlinks and audits from DataForSEO, exposed over MCP. Google MCP servers are first-party: they read your own Search Console, Analytics and other account data via OAuth. The two are complementary halves of one stack.
Which Google MCP server should I set up first?
Search Console, because it carries the decisions (what ranks, what gets clicked, what is decaying) and the community tooling is mature. Analytics second, using Google's official server. Trends third for demand direction.
Sources
- Anthropic, Model Context Protocol: https://modelcontextprotocol.io/
- Google Analytics official MCP server: https://github.com/googleanalytics/google-analytics-mcp
- open-seo (MIT): https://github.com/every-app/open-seo
Soaring Above Search
Weekly AI search insights from the front line. One newsletter. Six sections. Everything that actually moved this week, with a practitioner's take.
Chief of Staff at StudioHawk, Australia's largest dedicated SEO agency. Specialising in AI search visibility, technical SEO, and organic growth strategy. Leading a team of 120+ across Melbourne, Sydney, London, and the US. Book a free consultation →