Supermetrics MCP Server

Supermetrics MCP is the official, vendor-hosted Model Context Protocol server from Supermetrics. It exposes the Supermetrics Data API to MCP-compatible clients like Claude, ChatGPT, Microsoft Copilot, Gemini Enterprise, n8n, and Power Automate so AI agents can query live marketing data without manual exports or per-platform integrations. The server provides a single OAuth-authenticated endpoint that fans out to 170+ underlying data sources including Google Ads, Meta Ads, LinkedIn Ads, TikTok Ads, Microsoft Ads, GA4, HubSpot, Salesforce, Shopify, and many more.

The server exposes a small, opinionated toolset built around a discovery-then-query flow. Agents first call data source and account discovery to find what is connected, then call field discovery to learn available metrics and dimensions, then submit asynchronous data queries that return a schedule_id used to fetch results. Helper tools for the current UTC date, authenticated user info, and a health check round out the surface. Queries support relative date ranges (e.g. last_7_days), custom date ranges, period comparisons, and a filter expression language with operators for equality, ranges, contains, and regex.

Because it is provider-hosted, there is no code to clone or container to run: setup is just adding the remote MCP URL to a client and authorizing via OAuth. Access is included with any Supermetrics plan that includes Data API access, and a 14-day free trial is available without a credit card.

Tools

Prerequisites

• A Supermetrics account on a plan that includes Data API access (a 14-day trial is available at mcp.supermetrics.com, no credit card required).
• At least one connected data source in Supermetrics (Google Ads, Meta, GA4, etc.).
• An MCP client that supports remote/HTTP MCP servers and OAuth (Claude Desktop, Claude.ai, ChatGPT, Cursor, VS Code, n8n, Power Automate, etc.).

Server URL

https://mcp.supermetrics.com/mcp

Authentication

OAuth 2.0 per user. The first time you connect, your MCP client will open a browser window to authorize against your Supermetrics workspace. Each user only sees the data sources and accounts they are authorized for. API key auth is also available for server-to-server workflows.

Claude Desktop / Cursor / generic MCP config

Add the remote server to your client's MCP config:

{
  "mcpServers": {
    "supermetrics": {
      "url": "https://mcp.supermetrics.com/mcp"
    }
  }
}

If your client does not yet support remote MCP servers natively, you can bridge with mcp-remote:

{
  "mcpServers": {
    "supermetrics": {
      "command": "npx",
      "args": ["mcp-remote", "https://mcp.supermetrics.com/mcp"]
    }
  }
}

Claude.ai connector

Visit the Supermetrics connector page in the Claude directory and click Connect to link your Supermetrics account, no JSON required.

Microsoft Power Platform

A SuperMCP custom connector is published for Power Automate, Power Apps, Copilot Studio, and Azure Logic Apps. Authentication is OAuth 2.0 and the throttling limit is 100 calls per connection per 60 seconds.

• Pull a unified weekly performance summary across Google Ads, Meta, LinkedIn, and TikTok (spend, revenue, ROAS, CPA, top campaigns) without exporting CSVs.
• Let an agent answer ad hoc questions like "Which GA4 channels drove the most signups last month?" by discovering data sources, fields, and running a live query.
• Build automated pacing and anomaly alerts in n8n or Power Automate that query Supermetrics on a schedule and surface campaigns drifting from target ROAS or budget.
• Power a marketing copilot that compares period over period performance (last 7 days vs prior 7 days, or vs the same period last year) directly inside Claude or Copilot.
• Centralize cross-platform reporting for a marketing team where each user only sees data sources they're authorized for via OAuth scoping.

• "List my connected Supermetrics data sources, then show me last 7 days of spend, clicks, and conversions for all Google Ads accounts."
• "For our Meta Ads account, find the top 5 campaigns by ROAS over the last 30 days and compare them to the prior 30 days."
• "Pull GA4 sessions and conversions by channel grouping for last month, filtered to country == United States."
• "List available metrics for LinkedIn Ads, then build a query for cost per lead by campaign for last_14_days_inc."
• "What were our total ad spend and revenue across Google Ads, Meta, and TikTok last quarter, with percentage change vs the previous quarter?"

• Official, vendor-hosted server with OAuth 2.0, so there is no code to maintain and per-user data access is enforced by Supermetrics.
• Single endpoint fans out to 170+ marketing, analytics, CRM, and ecommerce data sources, which is far broader than most single-platform MCP servers.
• Well-structured discovery-then-query flow with schema introspection (data_source_discovery, field_discovery) makes it agent-friendly and reduces hallucinated field names.
• Works across Claude, ChatGPT, Copilot, Gemini Enterprise, n8n, and Microsoft Power Platform via a published custom connector.

• Requires a paid Supermetrics plan that includes Data API access (beyond the 14-day trial); not free for ongoing use.
• Read-oriented: the toolset is built for querying and reporting, not for editing campaigns or budgets inside the source platforms.
• Asynchronous query model (data_query returns a schedule_id, then get_async_query_results) adds a step compared to synchronous APIs and can complicate simple agent flows.

• Google Analytics / Google Ads MCP servers for direct, per-platform access to a single source without a Supermetrics subscription.
• Windsor.ai or Improvado, similar marketing data pipeline vendors that have begun shipping MCP or AI-agent integrations.
• Building a custom MCP server on top of the Supermetrics Data API or a warehouse (BigQuery, Snowflake) using a generic SQL MCP server.