YouTube Goldmine Scraper – High‑Volume Video & Channel Intelligence

Turn YouTube into a structured, analytics‑ready data source.
This scraper gives you high‑volume, low‑cost access to:

• Channels, videos, playlists, and search results
• Normalized metrics (views, likes, subscribers, dates)
• Optional comments and subtitles for deeper AI / analytics

All without using the official YouTube Data API or worrying about quota limits.

---

What this YouTube Scraper does

Use this actor as a YouTube API alternative to collect public data at scale:

• 🔍 Discover content & competitors

  • Scrape videos by search term, channel URL, playlist URL, or direct video URL.
  • Track trends, niches, and top‑performing videos in any keyword space.

• 📊 Get clean, normalized metrics

  • View counts and subscriber counts normalized to integers (K/M/B handled).
  • Publish dates normalized to ISO format where possible.
  • Ready for dashboards, BI tools, or machine learning pipelines.

• 💬 Understand your audience (optional)

  • Fetch the first N top‑level comments per video.
  • Analyze sentiment, FAQs, objections, and topic ideas.

• 🎧 Leverage subtitles for AI (optional)

  • Download subtitles / transcripts (when available) in text form.
  • Feed to LLMs for summarization, keyword extraction, or script generation.

• 💰 Optimized for low cost per result

  • Carefully tuned to avoid heavy browser work for every video.
  • Hard limits on scroll depth and per‑page items to keep resource usage low.
  • Ideal for pay‑per‑result pricing and bulk data pulls.

---

What data you can scrape from YouTube

Each run can return multiple item types, depending on your input:

1. Direct video details (type: "video")

For video URLs you get rich detail:

• videoId
• url
• title
• duration
• channelName
• channelUrl
• viewsText (original)
• viewsCount (normalized integer)
• likesText (original)
• likesCount (normalized integer)
• subscribersText (original)
• subscribersCount (normalized integer)
• publishedText
• publishedDateIso (normalized when possible)
• thumbnailUrl
• hashtags (array)
• description (optional)
• comments (optional array of basic comments)
• subtitles (optional transcript object)
• scrapedAt (ISO timestamp)

2. Channel videos (type: "channel_video")

For channel /videos pages, you get listing‑level data per video:

• videoId
• url
• title
• duration
• channelName
• channelUrl
• viewsText, viewsCount
• subscribersText, subscribersCount
• publishedText, publishedDateIso
• scrapedAt

3. Playlist videos (type: "playlist_video")

For playlist URLs:

• playlistUrl
• playlistTitle
• playlistOwnerName
• playlistOwnerUrl
• videoId
• url
• title
• channelName
• duration
• viewsText, viewsCount
• publishedText, publishedDateIso
• scrapedAt

4. Search results (type: "search_video")

For search pages or search terms:

• videoId
• url
• title
• channelName
• duration
• viewsText, viewsCount
• publishedText, publishedDateIso
• searchQuery
• scrapedAt

This structure makes it easy to join datasets (e.g., treat videoId as a primary key) and run analytics in SQL, Python, or any BI tool.

---

Typical use cases

1. Market & competitor intelligence

• Track view growth, likes, and uploads for competing channels.
• Monitor new videos in your niche by keyword or hashtag.
• Identify high‑engagement topics before they peak.

2. Content & script generation with AI

• Scrape thousands of subtitles and feed them to LLMs.
• Auto‑generate video summaries, show notes, and blog posts.
• Build knowledge bases or Q&A bots from YouTube content.

3. Lead generation & influencer discovery

• Find channels that:
  • Match certain keywords.
  • Exceed a subscriber/view threshold.
• Use social links and metadata to build outreach lists.

4. Sentiment & comment analysis

• Collect a sample of comments for many videos in bulk.
• Run sentiment analysis or topic clustering.
• Validate product ideas and track community feedback.

---

Input: how to configure the scraper

You can use either the Apify UI or JSON input.

Main fields

• startUrls (array)

  • YouTube URLs to start from.
  • Accepts:
    • Channel pages: https://www.youtube.com/@ChannelName, https://www.youtube.com/channel/ID
    • Channel videos tab: https://www.youtube.com/@ChannelName/videos
    • Playlists: https://www.youtube.com/playlist?list=...
    • Video URLs: https://www.youtube.com/watch?v=VIDEO_ID, https://youtu.be/VIDEO_ID
    • Search result URLs: https://www.youtube.com/results?search_query=keyword

• searchKeywords (string, optional)

  • Keywords as you’d type into the YouTube search bar.
  • Example: "ai automation tutorials"
  • The actor will open the search results page and extract videos.

• maxVideosPerChannel (integer, default 100)

  • Maximum number of videos scraped per channel.
  • Higher values = more data but more compute.

• maxVideosPerPlaylist (integer, default 100)

  • Maximum videos taken from each playlist URL.

• maxResultsPerSearch (integer, default 100)

  • Maximum search result videos per search term / search URL.

• maxCommentsPerVideo (integer, default 0)

  • 0 → comments disabled (cheapest mode).
  • >0 → scrape up to N top‑level comments per direct video URL.
  • Recommended range: 10–50 for cost‑effective insights.

• includeSubtitles (boolean, default false)

  • false → skip subtitles completely (cheapest).
  • true → try to fetch the first available subtitle track.

• includeDescription (boolean, default true)

  • When true, returns full video description for direct video URLs.

• proxyConfiguration (object)

  • Configure Apify proxy or your own proxies.
  • For best stability on YouTube, use datacenter proxies.

Sample JSON input

{
  "startUrls": [
    { "url": "https://www.youtube.com/@LofiGirl/videos" },
    { "url": "https://www.youtube.com/playlist?list=PL6i60qoDQhQGaGbbg-4aSwXJvxOqO6o5e" }
  ],
  "searchKeywords": "ai automation",
  "maxVideosPerChannel": 80,
  "maxVideosPerPlaylist": 80,
  "maxResultsPerSearch": 80,
  "maxCommentsPerVideo": 0,
  "includeSubtitles": false,
  "includeDescription": true,
  "proxyConfiguration": {
    "useApifyProxy": true,
    "groups": ["SHADER"]
  }
}