---
name: youtube-to-website
description: Work out which YouTube videos send visitors to the user's website, whether those visitors convert, and which videos earn their keep. Use when someone asks whether YouTube drives traffic or sales, which videos bring website visitors, or whether their YouTube effort is worth it. Needs the Caybl.ai connector with YouTube and Google Analytics connected.
---

# YouTube to website

Uses the Caybl.ai connector. You need both `youtube_` tools and Google Analytics (`run_report`). If either is missing, say which one to connect at https://caybl.ai/dashboard and stop.

## 1. Website side (Google Analytics)

Call `run_report` for the last 90 days:
- dimensions `sessionSource`, `landingPage`; metrics `sessions`, `keyEvents`, `totalRevenue` (if e-commerce)
- filter `sessionSource` contains `youtube` (this covers youtube.com, m.youtube.com and youtu.be)

Also get the same metrics without the filter, to know YouTube's share of all sessions.

## 2. YouTube side

- `youtube_top_videos` for the same 90 days (limit 20).
- Match landing pages to videos: a video usually links to one page in its description. Ask the user if a match isn't obvious, and don't guess.
- If the site uses UTM tags on video links (`utm_source=youtube`, `utm_content=<video>`), use `sessionManualAdContent` or `sessionCampaignName` to attribute exactly.

## 3. Report

1. One sentence: how many sessions and key events YouTube sent in 90 days, and its share of the site total.
2. A table: video (or landing page) · YouTube views · website sessions · key events · conversion rate.
3. Which videos earn their keep (website value per 1,000 views), and which get views but send nobody.
4. Two or three actions, for example adding a clear link and pinned comment to high-view videos that send no traffic, or tagging description links with UTMs so next month's numbers are exact.

Be clear about what's measured and what's matched by hand. Without UTMs, YouTube referrals can't be tied to a specific video for certain.