YouTube Most Replayed Scraper - Find the Hottest Parts of Any Video 🔥 ## Overview This Apify Actor, YouTube Most Replayed Scraper, helps you quickly identify and extract data about the most replayed sections of any YouTube video. Instead of watching entire videos, you can pinpoint the most engaging and interesting moments, saving time and gaining valuable insights. This is perfect for content creators, researchers, marketers, or anyone who wants to quickly understand what parts of a video resonate most with viewers. Also known as Youtube Most Watched or Rewatched, heatmap, most viewed parts... - For more in-depth information about the "Most Replayed" feature in YouTube videos, see the references at the end of this article. ## Features 🚀 * Extracts "Most Replayed" Data: Gets the visibleTimeRangeStartMillis, visibleTimeRangeEndMillis, and decorationTimeMillis values for each "Most Replayed" segment in a video. * Retrieves Heatmap Data: Provides the intensityScoreNormalized for every segment of the video. * Easy Input: Just provide the YouTube video URL(s). * Fast and Efficient: Optimized for speed to get you the data you need quickly. * Structured Output: Returns data in a clean, easy-to-use JSON format. * Gets additional Video Information: Also extracts, if found: * Channel Owner * Title * Video ID * View Count * Likes * Comments * Date and relative date ## How It Works 🤔 YouTube's "Most Replayed" feature displays a graph above the video's progress bar, highlighting the sections that viewers have rewatched most often. This actor scrapes the underlying data that powers this graph. The data is normalized, meaning the values represent the relative replay frequency within that specific video. A value of 1 (or 100%) represents the absolute peak – the most replayed part of that video. Other values show how frequently those sections were replayed compared to that peak. ### Input ⚙️ The actor takes a simple input: * url (string or array of strings): The URL(s) of the YouTube video(s) you want to analyze. You can provide a single URL or an array of multiple URLs for bulk processing. Example Input (single URL): json { "url": "https://www.youtube.com/watch?v=3CVHLAg55sQ" } Example Input (multiple URLs): json { "url": [ "https://www.youtube.com/watch?v=3CVHLAg55sQ", "https://www.youtube.com/watch?v=nnFFmgtCKOI", "https://www.youtube.com/watch?v=yworB2ySUUc" ] } ### Output 🎁 The actor returns a JSON array. Each object in the array represents a scraped video and contains the following data: * channelOwner: (string) The name of the YouTube channel that uploaded the video. * title: (string) The title of the video. * titleJson: (string) The title in JSON format. * videoId: (string) The unique YouTube video ID. * viewCount: (string) The number of views, formatted as a string (e.g., "1,112,246 views"). * likes: (string) The number of likes, formatted as a string (e.g., "45K"). * comments: (string) The number of comments, formatted as a string (e.g., "810"). * dateText: (string) The upload date (e.g., "May 24, 2023"). * relativeDate: (string) A relative representation of the upload date (e.g., "1 year ago"). * mostReplayed: (array) An array of objects, each representing a "Most Replayed" segment. Each object has these properties: * visibleTimeRangeStartMillis: (number) The start time (in milliseconds) of the replayed segment. * visibleTimeRangeEndMillis: (number) The end time (in milliseconds) of the replayed segment. * decorationTimeMillis: (number) The time (in milliseconds) at which the "Most Replayed" visual cue (label) is displayed. This usually falls within the visibleTimeRangeStartMillis and visibleTimeRangeEndMillis. * heatSeek: (array): Contains and array of objects, each represents a segment of the video: * startMillis: (number) start time in milliseconds. * durationMillis: (number) duration of the segment in milliseconds. * intensityScoreNormalized (number): Normalized values represented by "intensity scores". Each point in time of the video has a score between 0 and 1. Example Output: json [ { "channelOwner": "FunSizeProject", "title": "This Picture Frame Transformed with LED Lights Will Blow Your Mind! The Ultimate Home Decor Hack!", "titleJson": "This Picture Frame Transformed with LED Lights Will Blow Your Mind! The Ultimate Home Decor Hack!", "videoId": "3CVHLAg55sQ", "viewCount": "1,112,246 views", "likes": "", "comments": "810", "dateText": "May 24, 2023", "relativeDate": "1 year ago", "mostReplayed": [ { "visibleTimeRangeStartMillis": 24750, "visibleTimeRangeEndMillis": 28050, "decorationTimeMillis": 26400 }, { "visibleTimeRangeStartMillis": 41250, "visibleTimeRangeEndMillis": 46200, "decorationTimeMillis": 42900 }, { "visibleTimeRangeStartMillis": 70950, "visibleTimeRangeEndMillis": 80850, "decorationTimeMillis": 72600 } ], "heatSeek": [ { "startMillis": "0", "durationMillis": "1650", "intensityScoreNormalized": 1 }, { "startMillis": "1650", "durationMillis": "1650", "intensityScoreNormalized": 0.24851088071229727 }, { "startMillis": "3300", "durationMillis": "1650", "intensityScoreNormalized": 0.2878596089839822 }, { "startMillis": "4950", "durationMillis": "1650", "intensityScoreNormalized": 0.278429379104963 }, { "startMillis": "6600", "durationMillis": "1650", "intensityScoreNormalized": 0.25421004861285773 }, { "startMillis": "132000", "durationMillis": "1650", "intensityScoreNormalized": 0 }, { "startMillis": "133650", "durationMillis": "1650", "intensityScoreNormalized": 0.0299962236605618 }, { "startMillis": "135300", "durationMillis": "1650", "intensityScoreNormalized": 0.011892619926647401 } ] } ] ### Interpreting the intensityScoreNormalized The intensityScoreNormalized values in the heatSeek array are key to understanding the relative replay frequency. * 1 (or 100%): The most replayed point in that specific video. * 0.1 (or 10%): This section was replayed 10% as often as the most replayed section. * 0.00001 (or 0.001%): This section was very rarely replayed compared to the most replayed section (0.001% of the peak replay frequency). Remember, these values are relative within the same video. A 0.1 on one video's graph doesn't necessarily mean the same thing as a 0.1 on another video's graph in terms of absolute replay counts. ## Use Cases 💡 * Content Creators: Identify the most engaging parts of your videos to understand what resonates with your audience. Use this information to create more of what your viewers love! * Marketers: Analyze competitor videos to see which segments get the most attention. Find opportunities to create more compelling content. * Researchers: Study viewer behavior and identify patterns in video engagement. * Video Editors: Quickly find key moments in long videos to create highlight reels or shorter clips. * Anyone!: Save time by jumping directly to the most interesting parts of any YouTube video. ## Important Notes ⚠️ * The video must have at least 50K views to contain a heatmap graph * Not all videos have heatmap graphs. * The "Most Replayed" feature is typically available for longer-form videos, but not always for videos with manually defined chapters. * YouTube may change its website structure, which could affect the actor's functionality. If you encounter any issues, please report them. ## Get Started Today! 🎉 Use the YouTube Most Replayed Scraper to unlock valuable insights from YouTube videos and make data-driven decisions! ## References: I. Websites: * TubeBuddy Blog * Snapmuse Blog * Tubeloop Homepage * TubeRanker Blog * Reddit * Google answers * YouTube video explainer Further In depth references: II. Research Papers (Google Scholar Links): * EnTube Dataset * XGBoost YouTube Popularity * Three Metrics for YouTube Engagement * Zhou & Qiao (Temporal Segmentation) * Kuehne et al. (Temporal Segmentation) * Mr. HiSum (Highlight Detection) * Understanding the YouTube Algorithm * Auditing Radicalization Pathways * Predicting Most Replayed Data Some other Related topics: III. General Searches (No Direct Links): * "youtube audio normalization" (Google Search) * "Loudness normalization LUFS" (Google Search)