mcp-google-ads and adloop
These are **competitors** — both provide MCP servers that expose Google Ads data to Claude/AI assistants, but A offers broader campaign analysis while B specifically integrates GA4 feedback loops, making them alternative solutions for the same core use case of AI-driven ad platform interaction.
About mcp-google-ads
cohnen/mcp-google-ads
An MCP tool that connects Google Ads with Claude AI/Cursor and others, allowing you to analyze your advertising data through natural language conversations. This integration gives you access to campaign information, performance metrics, keyword analytics, and ad management—all through simple chat with Claude, Cursor or Windsurf.
Implements FastMCP server architecture with OAuth 2.0 and service account authentication, exposing tools like `execute_gaql_query` and `run_gaql` for direct Google Ads Query Language access alongside pre-built performance analytics. Supports automatic token refresh, multiple authentication methods, and integrates with Claude, Cursor, and Windsurf via the Model Context Protocol standard.
About adloop
kLOsk/adloop
MCP server connecting Google Ads + GA4 into one AI-driven feedback loop inside your IDE
Implements 38 MCP-compatible tools across GA4 and Google Ads APIs with automatic cross-referencing — joining ad clicks to GA4 sessions to isolate conversion gaps from GDPR consent rejection vs. tracking breakage. Built on Python 3.11+ with draft-preview-confirm workflows for write operations and safety guardrails (e.g., preventing Broad Match + Manual CPC combinations that waste budget). Integrates directly into Claude, Cursor, and other MCP-compatible IDEs via stdio transport, enabling real-time campaign diagnostics and optimization without leaving your editor.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work