a daily news desk
Tools

Webflow adds Gemini to its MCP server, exposing AEO Analytics to Google's agent

The October 15 update lets Gemini users query which prompts surface a brand, which LLM crawlers are hitting pages, and which A/B variants are winning — inside the chat window.

By Iris Holloway·Tools & frameworks·

Webflow turned on a Gemini connector for its MCP server on October 15, making Google’s agent the fifth chat interface, after Claude, ChatGPT, Codex, and Slack, that can read a Webflow site’s analytics, A/B results, and AEO data in a conversation.

The headline capability isn’t the integration itself. It’s what Gemini can now ask about. Webflow’s changelog says agents can pull “the pages LLM crawlers request most, plus your tracked prompts and how answer engines responded.” That makes a mainstream general-purpose assistant into a diagnostic surface for AI-search visibility, the question of whether answer engines are routing buyers to your brand or to a competitor’s. Until now, that read-path has mostly lived inside specialist AEO tools. See our earlier piece on AI agents as the invisible first step in a buyer’s journey for why that routing question has become a growth variable rather than an SEO curiosity.

The deployment details are deliberately low-friction. Install is free, doesn’t consume Webflow AI credits, and inherits existing roles and permissions, with AI activity attributed in Webflow’s standard logs. Enterprise plan support is still in development. The Gemini connector marketplace itself is open only to approved partners, which is part of why the launch reads as a positioning move for both companies.

“Teams move faster when people can contribute from the tools they already use,” said Ben Haefele, Webflow’s chief product officer, framing the connector less as a new product than as expanded access to an existing one.

Context matters. Inriver expanded its MCP servers to expose governed product data to outside agents the same week, Validity shipped an MCP connector for email intelligence, and MarTech’s running list catalogued more than a dozen MCP or agentic releases on October 15 alone. The MCP wave is now the default distribution model for martech data, with the chat window as the new dashboard.

Sources