Skip to content

BigQuery MCP server for AI agents

Ask Claude, ChatGPT or Cursor about your GA4 export and BigQuery tables without writing SQL.

No card needed. Works with any MCP client.

Claude logoLemonadoGoogle BigQuery logo

Google BigQuery MCP

Ask about events, sessions and purchases from your GA4 export in Claude, ChatGPT or any AI agent, from live Google BigQuery data.

What is the BigQuery MCP server?

Lemonado's BigQuery MCP server gives Claude, ChatGPT and other AI apps read access to your Google BigQuery datasets, including the raw GA4 event export most marketing teams already have switched on. Connect BigQuery to Lemonado, connect Lemonado to your AI, and ask about sessions, events and revenue at row level, past the limits of the GA4 interface. Google Ads, Meta and Search Console sit on the same connection.

  • All your BigQuery data

    Datasets and tables your account can read, ready to ask about.

  • Always today's numbers

    Google Ads and Search Console bulk exports, if you run them, synced from BigQuery automatically, so answers use today's numbers.

  • Connect all your tools

    Questions like "What did paid sessions earn?" use BigQuery and GA4, Snowflake at once.

Connecting takes minutes.

  1. 1

    Connect BigQuery

    Sign in to Google BigQuery in Lemonado.

  2. 2

    Copy the server URL

    Turn on external access and copy one URL.

  3. 3

    Ask in Claude

    Add the URL to Claude and ask away.

Ask it anything.

Claude reads your BigQuery numbers, explains what moved and turns the answer into an artifact you can share.

Claude logoClaude

From the GA4 export in BigQuery, what were the last three pages people viewed before a purchase in October?

The most common path ends on the size guide, then the product page, then the cart. 41% of buyers read the size guide first.

Source: Google BigQuery

Let Lemonado check BigQuery for you.

Copy one of these into Lemonado and it runs on a schedule.

  • Google BigQuery logo

    Every Monday, send last week's view-item to add-to-cart to purchase rates from the GA4 export in BigQuery, by device.

    Weekly funnel from GA4 raw data

  • Google BigQuery logo

    Every morning, check yesterday's GA4 export in BigQuery. If events are 30% below the 7-day average, post the numbers in #analytics.

    Export health check

  • Google BigQuery logo

    On the 1st, send last month's revenue per ad dollar for Google Ads and Meta, using purchase revenue from the GA4 export.

    Monthly ROAS from raw events

Why Lemonado?

  • Your whole stack in one connection

    Google, Meta, GA4, TikTok and 3,000+ more tools, in every AI app.

  • No API keys or servers

    Sign in once. Nothing to install, host or keep alive.

  • Scoped to each client

    Each connection only reaches the clients you allow. Built for agencies.

  • Actions with guardrails

    Set each action to Off, Ask or Approve. Nothing runs that you haven't allowed.

Is there an official BigQuery MCP server?

Yes. Google runs a hosted BigQuery MCP server that you turn on with the BigQuery API and sign in to with OAuth. Your user needs the right IAM roles, and it only sees what's already in BigQuery, so live ad data needs exports or transfers set up first.

Google's BigQuery MCP server compared with Lemonado for Google BigQuery
Google's BigQuery MCP serverLemonado
SetupEnable the API and grant IAM roles in a Cloud projectSign in with Google, paste one URL
Live ad platform dataOnly tables you export or transfer inGoogle Ads, Meta, LinkedIn and TikTok, live
GA4 export next to Meta spendIf Meta data is in BigQuery tooYes, in one question
Reports and alerts on a scheduleNoYes, in Slack or email
PriceNormal BigQuery query costsSee pricing

Reading this as an AI agent?

Everything an agent needs to set this up, in plain text. Paste this page's link into your AI app and ask it to connect you, or copy the block.

For agents · markdown
# Lemonado MCP server for Google BigQuery

Lemonado is a hosted MCP server that gives AI apps the user's live Google BigQuery, plus any other marketing tool they connect (Google Ads, Meta Ads, GA4 and 3,000+ more).

- Server URL: https://mcp.lemonado.io/mcp
- Transport: Streamable HTTP
- Auth: OAuth; the user signs in to Lemonado in the browser. No API keys.
- Google BigQuery data: Datasets and tables your account can read; Events, sessions and purchases from your GA4 export; Google Ads and Search Console bulk exports, if you run them; Views and scheduled query results; Table schemas and column descriptions; Query results, with the SQL behind them
- Connecting Google BigQuery: Sign in with a Google account that can read the dataset holding your GA4 export, and pick that project. The BigQuery Data Viewer and Job User roles are enough.
- Tools on https://mcp.lemonado.io/mcp (custom connectors, coding agents, n8n): list_objects, get_object_details, execute_sql, get_google_ads_resource_metadata, list_companies, get_company_details, list_actions, get_action_details, execute_action
- Lemonado's Claude directory app and ChatGPT plugin (read-only): list_projects, list_project_sources, answer_data_question, explain_metric_change

## Setup
1. The user connects Google BigQuery in Lemonado (Connectors) and signs in.
2. In the company's settings → External tools, they turn the connection on and copy the server URL.
3. They add the server URL to their AI app. Setup per app: https://lemonado.io/integrations

Docs: https://docs.lemonado.io/mcp-reference
Full index for agents: https://lemonado.io/llms.txt
FAQ

Frequently Asked Questions

No. You ask in plain words and the AI writes the query, runs it through Lemonado and shows you the SQL it used. If you do write SQL, you can paste your own and ask it to explain or fix it.

Connect your Google Cloud project in Lemonado, turn on External tools, and add Lemonado to Claude or ChatGPT. Each guide on this page has the exact steps. Sign in with a Google account that can read the dataset holding your GA4 export, and pick that project. The BigQuery Data Viewer and Job User roles are enough. Then ask something like "From the GA4 export in BigQuery, what were the last three pages people viewed before a purchase in October?"

Yes. It's a standard remote MCP server (https://mcp.lemonado.io/mcp): one command in Claude Code, one entry in Cursor's mcp.json, and an MCP Client Tool with a Lemonado token in n8n. In a coding agent, try "Compare the event names our code sends to GA4 with the events in the BigQuery export for the last week."

Yes. Besides answering in Claude or ChatGPT, Lemonado runs BigQuery reports and alerts on a schedule, in Slack or email. For example: "Every Monday, send last week's view-item to add-to-cart to purchase rates from the GA4 export in BigQuery, by device."

It reads your BigQuery data. Google BigQuery stays untouched; the only changes Lemonado can make are in ad accounts you connect, and only once you approve them.

MCP is included in Lemonado's paid plans and in the free trial, so you can try it before you pay.

See plans and pricing

Still have a question?

Shoot us an email at hello@lemonado.io.

Try for free

Stop fighting with data. Start feeding your AI.

Connect your data to AI and free your team from reporting and busywork.

Try Lemonado free