Connect Google Search Console to ChatGPT. For your whole team.
One Google sign-in in Loopthink, one connector in ChatGPT, and your team is reporting. The sign-in asks for the read-only scope, so sitemap writes are out of reach from the start, and the search terms real people typed stay masked until you say otherwise. Every ChatGPT user gets their own login and their own role.
Search Console has no MCP server of its own. The community ones run on one Google login or a key file that sits in every person's client config, and they carry the write methods along with the reports.
Free for 2 servers, no card · Paid plans from €39 · EU-hosted
“Which pages lost the most clicks in the last 28 days?”
Three pages are down against the 28 days before:
| Page | Clicks | Change | Position |
|---|---|---|---|
| /pricing | 412 | −118 | 8.8 |
| /blog/rollout-guide | 268 | −74 | 11.3 |
| /integrations | 197 | −53 | 7.4 |
Want the queries behind them? Those are masked for your role.
- read-only
- as s.keller
- queries masked
Google Search Console in a chat client, today.
Today a team that wants Search Console in an AI client either adds every colleague as a user on the property, or one person's Google login ends up in a shared MCP config. The first means Full or Owner rights for people who only need a chart. The second means a login that can also submit and delete sitemaps, and no record of who asked which query. And every search term a real person typed goes to the model, because nothing in that path knows to hold it back.
A silent loop through the whole setup. In Loopthink: the server overview, the New workspace dialog, the New MCP server dialog, the connection screen where Google access is granted, Google's own sign-in page, the read-only tools the grant published, and the connector page where the workspace URL is copied. Then in Claude: Settings with Connectors open, the Add custom connector dialog with the name and the URL, the page of options that follows, scrolled down and left untouched because everything on it is detected, and the connector, named after the company, switched on in a chat answering from Google Ads. A cursor moves to the button each screen is about.
- 1
Start with a workspace in Loopthink. Its servers, its members and its key are separate from every other one.
- 2
Add a server and pick Google Search Console as its connection.
- 3
Press Connect Google account.
- 4
Sign in with Google. The consent screen asks for the Google Search Console scope and nothing else.
- 5
Tick the accounts or properties your team may report on. Each one publishes its own read-only tools.
- 6
Copy the workspace connector URL.
- 7
In ChatGPT, open Settings, then Connectors, then Add custom connector.
- 8
Give it a name your colleagues will recognise, Acme Group rather than Loopthink, and paste the URL. Those two fields are the whole form.
- 9
Continue opens a page of options. Scroll down past it and change nothing: ChatGPT has already read the server, so the sign-in and the OAuth client are detected and chosen for you, and there is no client id, no secret and no request header to type. Press Add at the bottom.
- 10
Each colleague connects once with their own account, toggles it on in a chat, and asks. Every call runs as them, with their role, and lands in the audit log.
Steps checked against the vendor documentation on 17 September 2026: Search Console API reference · Search Analytics query · Users and permissions · Developer mode and MCP apps in ChatGPT · Connectors in ChatGPT
One matrix for Google Search Console. Roles across, tools down.
| Tool | SEO | Content | Management |
|---|---|---|---|
| List properties | SEO may use List properties | Content may use List properties | Management may use List properties |
| Performance per property | SEO may use Performance per property | Content may use Performance per property | Management may use Performance per property |
| Sitemaps per property | SEO may use Sitemaps per property | Content may not use Sitemaps per property | Management may not use Sitemaps per property |
| Search terms (QUERY) | SEO may use Search terms (QUERY) | Content may not use Search terms (QUERY) | Management may not use Search terms (QUERY) |
| Write actions off by default | SEO may not use Write actions | Content may not use Write actions | Management may not use Write actions |
Masked for every role: search queries, the text real people typed, until an admin enables the QUERY dimension on the connection.
You set this once. Every ChatGPT user in your team gets exactly this.
Google Search Console runs on the free tier: two MCP servers for one person, no card. The roles, the identity per person and the masking above start at €39 a month, as soon as colleagues join. See the plans
Four checks between ChatGPT and Google Search Console.
- 01
Identify
OIDC/SSO against your own IdP, including Microsoft Entra. The call runs as the person who made it, never as a shared service account.
- 02
Authorize
Per-role, per-tool, read or write. Anything not explicitly granted never shows up in the tool list in the first place.
- 03
Mask
Field-level masking and pseudonymization applied on your side, before the result travels. The model never sees the raw field. Composio, AnythingMCP and both provider tunnels have no field-level masking at all.
- 04
Log
Who asked, which tool, which records, which fields were masked. One audit trail across every AI client, exportable for review.
What people actually ask ChatGPT about Google Search Console.
- SEO
“Which pages lost the most clicks in the last 28 days compared with the 28 before?”
- Content
“How did the blog perform last month, by page and country?”
- Management
“Clicks and impressions this month versus last month, by country?”
Each question runs as the person who asked it, with that person's role. A Content colleague asking the SEO question gets the tools their own role allows, not an error, not more.
Four ways to put Google Search Console into ChatGPT.
| Native Google Search Console connector | Build it yourself | Enterprise gateways1 | Loopthink | |
|---|---|---|---|---|
| Per-user identity | One personal login, all its rights | Only if you build it | Yes | Yes, every call runs as the person who asked |
| Tool-level roles | No | Your code | Yes | Yes, write off by default |
| Field masking | No | Your code | Not documented | Yes, before the model sees the data |
| Published price | Included in your Google Search Console plan | Your engineers' time | Contact sales | Free, paid from €39 |
| EU company | No | Whatever you are | Mostly US | Yes, Germany |
Native Google Search Console connector: No official Google MCP server for Search Console. Community servers run on one OAuth login or a key file in the client config, sitemap writes included, and the search terms pass through untouched.
1 MintMCP, MCP Manager, Pomerium
Coming from one of them? Composio alternative · MintMCP alternative · MCP Manager alternative · AnythingMCP alternative
Also for
Connect another system to ChatGPT
Before you connect.
- Does every team member need their own Google Search Console account?
- No. One Google sign-in in Loopthink covers the whole team, and it is the only Google account involved. Your people sign in to Loopthink with their own identity, every call is logged under their name, and the seats in Google Search Console stay exactly where they are today.
- Who has to add the connector in ChatGPT?
- One person with the right to add connectors in your ChatGPT workspace, once. After that every team member enables it for themselves. Nobody pastes a key, nobody edits a config file.
- Where does the Search Console key live?
- There is none, and that is the point. A Google sign-in grants Loopthink read access, which is stored encrypted in the EU, is never returned by the API and never reaches ChatGPT or the people using the connector. You revoke it any time at myaccount.google.com/permissions, and nobody has to rotate or paste anything.
- Can somebody submit or delete a sitemap from the chat?
- No, and not because a switch is off. The sign-in asks for the read-only Search Console scope, so Google itself refuses the write methods, and the connection publishes no write tools either. Both limits would have to fail at once.
- What does the model see?
- The result of the tool call, after masking. Fields you mark as personal, such as emails, names or card details, are masked before the answer travels to ChatGPT. The model gets the numbers, not the people.
- Is there an audit trail?
- Yes. Who asked, which tool, which parameters, which fields were masked, when. One log across every client your team uses, exportable for review.
- What does it cost?
- Free for one person and two MCP servers, forever, no card. Starter is €39 per month and adds roles, an identity of their own for every colleague, field masking and a 30-day audit. Team is €99 and adds your own APIs and SQL databases, your own domain and your own identity provider. Prices are on the pricing page, no call needed.
- We already run our own MCP server for Google Search Console. Does it still work?
- Yes. Point it at Loopthink and it inherits the same login, roles, masking and audit. Nothing on the server has to change.
Google Search Console in ChatGPT, with rules.
One key in Loopthink, one login per person, one matrix that decides who may do what.
Free for 2 servers, no card · Paid plans from €39 · EU-hosted