YouTube Analytics MCP server for your AI

YouTube Analytics answers targeted questions on demand: pick a report, a date range and filters, and the rows come back in the same call. YouTube Data manages the channel itself, with videos, playlists, comments and captions. YouTube Reporting schedules daily bulk CSV exports for archives and warehouses. Analytics rows are keyed by bare video and playlist ids, so pair this connector with YouTube Data when you want titles next to the numbers.

YouTube finalises a day's analytics two to three days late, so the end of a range that finishes near today is incomplete rather than a real drop. The report tools default to the last 28 days ending yesterday, and channel_summary works with no arguments at all. When there is no data for a range, the response simply has no rows. Pick an end date three or more days back when you compare periods.

Not on an ordinary creator channel. YouTube only returns revenue and ad-performance metrics on content owner reports, so revenue_over_time and ad_performance return an error on a channel connection whatever permissions were granted. That is YouTube's restriction, not a setup mistake, and earnings stay in YouTube Studio. The Revenue dropped trigger has the same limit. Everything else, from watch time to traffic sources and audience demographics, works for any channel you own.

Exactly one. After you authorize, PopMCP lists the channels the Google account owns, and you choose which to connect and which workspace each belongs in. Every report is then pinned to that channel, and the content owner parameter is stripped, so the connection cannot widen to other channels. If the list is empty, that Google account has no channel: reconnect, and for a Brand Account pick it on Google's account chooser.

Three read-only scopes and no write scope: yt-analytics.readonly for reports, yt-analytics-monetary.readonly for the revenue view, and youtube.readonly, which Google requires alongside analytics for reports and group items. Nothing on the consent screen asks to manage your channel. One setup step matters: if your project's Audience says Testing, publish the app before connecting, because Google expires Testing tokens after 7 days and the connection would stop.

Eight thresholds, each with a level you choose. They fire when daily views drop against the previous period, watch time falls, subscriber losses spike in a day, one video's views take off, views from YouTube search drop, likes, comments and shares fall, the split between Shorts and long-form moves past your line, or estimated revenue drops, which only works on content owner accounts. Each hands the numbers to whatever runs next.

No. PopMCP hosts it and gives you a Streamable HTTP MCP URL. There is nothing to deploy, no callback URL to register, and no infrastructure to keep running or patch.

Any client that supports a remote Streamable HTTP MCP server: Claude Desktop, Claude Web, Claude Code, ChatGPT custom connectors, Codex and the Codex CLI, OpenCode and Z.ai's ZCode, as well as editors such as Cursor, Windsurf and Zed. Your client signs in to PopMCP with OAuth; the provider credential never leaves the server.

Connect YouTube Analytics in PopMCP first, then run claude mcp add --transport http popmcp https://app.popmcp.com/mcp in a terminal. Inside Claude Code, run /mcp and sign in to PopMCP in the browser window it opens, ticking YouTube Analytics on the consent screen. The YouTube Analytics tools are then available in that session, and no YouTube Analytics credential is stored on your machine.

Loading 40 tool definitions into every conversation burns context before the work starts. PopMCP loads a curated 26-tool set covering the everyday workflows and keeps the rest searchable, so your AI pulls a specialist tool only when a task needs it. Every tool stays searchable, and on Scale the owner can load the full catalog for a teammate's connection.

A tool is something your AI calls, and there are 40 here. A trigger is something PopMCP watches for you: 8 YouTube Analytics events that can start work on the PopMCP side. Triggers never enter your AI client's context, and none of them ask you to paste a callback URL into YouTube Analytics.