Google YouTube Data

Turn any list of YouTube search queries into a clean dataset of video metadata — title, channel, view/like/comment counts, duration, publish date, and thumbnail — pulled live from the official YouTube Data API v3.

What you get

• One dataset row per (query, video) — perfect for pivoting by keyword
• Full statistics — viewCount, likeCount, commentCount as parsed integers
• Parsed duration — both raw ISO 8601 (PT4M13S) and seconds (253)
• Best-available thumbnail URL (high → medium → default)
• Channel attribution — channelId and channelTitle on every row
• Sort and filter by relevance, date, rating, view count, or title
• Time-bounded queries — optional publishedAfter ISO 8601 cutoff
• Error-tolerant — failed queries emit a single error row, the run continues

Use cases

• Channel competitive analysis — pull every top-ranking video for a keyword set and compare which channels dominate
• Content discovery — surface the newest high-engagement videos in any niche, refreshed on a schedule
• Trend tracking — feed the dataset into a dashboard to chart view-count growth across topics over time
• Influencer research — identify creators ranking for target keywords, ranked by subscribers and engagement
• Video SEO datasets — build training data of titles, descriptions, durations, and view counts to model what ranks
• Ad targeting — assemble lists of high-traffic videos by topic for YouTube placement campaigns

How to use

1. Paste your Search Queries — one per entry
2. Set Max Results Per Query (1–50, default 10)
3. Pick an Order — relevance (default), date, rating, viewCount, or title
4. Optionally set Published After to a datetime to filter out older videos
5. Run the actor — results appear in the Dataset tab

Output format

Each dataset record:

{
  "query": "apify scraping",
  "videoId": "dQw4w9WgXcQ",
  "status": "success",
  "title": "How to scrape any website with Apify",
  "channelId": "UCxxxxxxxxxxxxxxxxxxxx",
  "channelTitle": "Apify",
  "publishedAt": "2024-03-12T14:00:00Z",
  "description": "In this video we show…",
  "thumbnailUrl": "https://i.ytimg.com/vi/dQw4w9WgXcQ/hqdefault.jpg",
  "viewCount": 12453,
  "likeCount": 412,
  "commentCount": 38,
  "durationIso": "PT8M21S",
  "durationSeconds": 501,
  "url": "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
  "error": null
}

If a query fails, one row is emitted with status: "error", videoId: "", and error containing the reason — the run continues with the remaining queries.

Order options

Order	Best for
relevance	Default YouTube ranking — what a typical user would see
date	Newest videos first — trend monitoring, news topics
rating	Highest-rated videos — quality discovery
viewCount	Most-watched videos — competitive benchmarking
title	Alphabetical by title — deterministic ordering for diffs

Notes

• Powered by the official YouTube Data API v3 — every result is current, no scraping fragility
• Each query costs 100 API units (search) + 1 unit per 50 videos (videos.list)
• A video that matches multiple input queries is emitted once per query so you can analyse keyword overlap