YouTube is one of the largest and longest-standing video social media platforms in the world. As of 2023, YouTube has around 2.6 billion active users, 3.7 million video uploads per day by creators worldwide, and over 122 million people accessing the site each day. YouTube is a one-stop place for people, where they can find news updates, educational content, tutorials, movies, vlogs, and many more. Throughout the millions of videos uploaded each day, it is fascinating to think about how some videos go viral and generate high engagement while others struggle to get views. This made us wonder: what factors contribute to the success or popularity of a YouTube video? In this project, we will analyze YouTube data related to likes, shares, saves, comments, video category, length, and posting time in relation to how well they perform or trend. We will also analyze the most subscribed channels to observe how well their videos perform and their engagement rate across multiple platforms.
Our project aims to visualize the factors that contribute to the popularity of videos on YouTube and observe the consumption patterns of users. We will achieve this by comparing and contrasting data, identifying patterns, and observing how these factors evolve and relate to each other. This is an interesting story to tell because it will allow content creators to understand their audience and tailor their content to increase engagement. By analyzing the unique characteristics of the videos that have generated high engagement, we can understand the formula for creating viral videos and replicating the success. In addition to that, knowing what makes a YouTube video popular will improve trend forecasting, as content creators will be able to stay ahead of the curve and produce relevant content that caters to their audiences.
This project will also look into how YouTube’s algorithm affects how well videos perform and how the users operate. Understanding the platform’s algorithm is imperative for creators who are trying to increase their visibility on the surface, and reach of their content. Furthermore, we noticed how it is essential for content creators to ride the waves of the current trend — to understand and adapt to the ever-changing landscape of YouTube and digital media. Amid political turmoil, environmental catastrophes, and the not-long-past-global pandemics, the ability of content creators to engage with their audience and remain relevant has become increasingly important.
This study's purpose stems from YouTube being the preeminent platform for entertainment and learning in this contemporary world. This project will examine what makes a YouTube video viral, the engagement of content creators as well as the impact of events on YouTube video categories, to identify the trends and patterns, alteration in focus, and uncover mechanisms and their resulting impacts on the content consumption patterns of users and the role of YouTube’s algorithm in influencing how well videos perform.
-
Where did you download the data (e.g., a web URL)?
https://www.kaggle.com/datasets/rsrishav/youtube-trending-video-dataset -
How was the data collected or generated? Make sure to explain who collected the data (not necessarily the same people that host the data), and who or what the data is about?
The data is collected using the YouTube API and is collected by Rishav Sharma. This data shows the trending Youtube videos each day, along with information that can be useful to gain insights such as its view count, likes, dislikes, tags used as well as its channel name. -
How many observations (rows) are in your data?
197590 rows -
How many features (columns) are in the data?
16 columns
-
Where did you download the data (e.g., a web URL)?
https://www.kaggle.com/datasets/themrityunjaypathak/most-subscribed-1000-youtube-channels -
How was the data collected or generated? Make sure to explain who collected the data (not necessarily the same people that host the data), and who or >hat the data is about?
The data is collected using the BeautifulSoup python package and is collected by Mrityunjay Pathak. This data shows the 1000 most subscribed youtube channels as of January 2023. The dataset also includes information about the channel such as the category of the channel, ranking, and the total views and subscribers. -
How many observations (rows) are in your data?
1000 rows -
How many features (columns) are in the data?
7 columns
- The article also mentions and gives analytics on other apps such as Facebook, Snapchat and Instagram, whereas we will only focus on Youtube.
- Both our idea and the article presents how the pandemic has affected the consumption of Youtube.
- The article talks about the revenue associated with YouTubers with most subscribers, while we won’t be covering the income they make.
- Both our idea and the article explains how the consumption of YouTube videos progressed throughout the years.
- The article also goes in-depth into what day gives out the most trending videos per category, whereas we will not focus on that aspect.
- Both our idea and the article include visualizations of the top categories, the tags that make the video trending, and how long it takes for a video to become trending.
- The article includes an analysis of whether a title of a video could also impact the number of views, however, we won’t be delving deep into that part.
- Both our idea and the article will include a section that includes what time is best to post a video to get the highest number of views.
- The article analyzes the YouTube trend in Indonesia, whereas we are only going to spotlight the US region.
- Both our idea and the article will analyze the number of likes, dislikes, and comments of the video. We will also analyze the mean, minimum value, maximum value and standard deviation of those fields.
By Charity Joy Njotorahardjo, Eugene Alexander Wongso, Steven Wilbert Heng.