Search Console

    Google Search Console MCP: the two realistic setups

    Getting clicks, impressions and average position into Claude or ChatGPT is straightforward. The choice is whether you want to run the plumbing yourself or have it run for you — both are legitimate, and the free one is genuinely free.

    Self-host an open-source Search Console MCP

    Several community servers wrap the Search Console API. You create a Google Cloud project, enable the Search Console API, make a service account, grant it access to your property and run the server locally or on your own host.

    Good: Free, and the data is first-party and complete.

    Cost: It is developer work, it breaks when credentials rotate, and a local server cannot answer anything while your machine is off.

    Connect Search Console to LemRank and use the LemRank MCP

    LemRank imports your Search Console performance data alongside its own daily rank checks. You authorise Google once in the app; the MCP server then answers from both.

    Good: No service account, no hosting, and search data sits next to tracked positions so the assistant can compare what you rank for with what you actually get clicks on.

    Cost: It is part of a paid product, and Search Console history only backfills as far as Google exposes it.

    Worth knowing first

    What Search Console cannot tell you

    • Search Console only reports queries you already appear for. It cannot tell you about keywords you do not rank for yet — that needs a rank tracker or a keyword tool.
    • Data is delayed by roughly two days, so today's numbers are never in it.
    • Query rows are sampled and filtered for privacy, so the totals will not match the sum of the query list.

    This is why most people end up running a Search Console MCP alongside a rank tracker: one tells you what is already working, the other tells you where you sit on terms you are chasing. We compare both kinds here.

    Install

    Connect the LemRank MCP server

    1. Create an API key in LemRank: sign in, open REST API Documentation (/api-docs) and use the API Keys tab. The key is shown once and stored only as a hash.
    2. Add the server below to your assistant and paste the key in place of YOUR_LEMRANK_API_KEY.
    3. Ask it something like “which of my keywords dropped this week?”
    Endpoint
    https://lemrank.com/api/public/mcp
    Claude Desktop, Cursor, Windsurf and other clients that read a JSON config
    {
      "mcpServers": {
        "lemrank": {
          "url": "https://lemrank.com/api/public/mcp",
          "headers": {
            "Authorization": "Bearer YOUR_LEMRANK_API_KEY"
          }
        }
      }
    }
    Claude Code
    claude mcp add --transport http lemrank https://lemrank.com/api/public/mcp \
      --header "Authorization: Bearer YOUR_LEMRANK_API_KEY"

    In ChatGPT, add it as a remote connector using the same endpoint and put the key in an Authorization: Bearer header. Any client that speaks Streamable HTTP MCP will work — the server holds no session state. The same data is available over REST.

    Search data and rank data, one connection

    Authorise Search Console inside LemRank and your assistant can compare impressions with where you actually rank. Free trial, no card.