YouTube Research Methods, Tools, and Analytics

I have recently published a chapter on YouTube research methods in The SAGE Handbook of Social Media Research Methods.

Social media use has become an integral part of the lives of billions across the globe. As a modern media powerhouse, YouTube is the world’s most popular video-sharing site (Alexa, 2021). The largest market for YouTube is the United States (14.8%), followed by India (8.1%). Over 70% of the views on YouTube are from mobile devices (Cooper, 2019). YouTube content is diverse, emanating from a broad range of audiences, in a number of languages, and in a variety of genres (Khan, 2017). Videos are one of the most engaging media, and such a high level of activity on this social media platform necessitates a thorough analysis of the content posted in the form of videos and comments. We define social media analytics research as the use of varied tools and techniques to make sense of online activities and conversations by obtaining, refining, analyzing, and visualizing social media data. Such analyses can include both quantitative and qualitative research techniques with the aim of understanding online user behavior, needs, perceptions, and challenges.

In this chapter, we shed light on the various tools for YouTube data collection. These include the more recent social media analytics techniques that often employ ‘big data’. In addition, we discuss seminal research centered on YouTube, highlight limitations, and sum up the various practices to effectively conduct YouTube research. Furthermore, we offer two case studies to illuminate the latest research on the platform and the methods employed to answer the research questions. In discussing the significance of big data, artificial intelligence (AI), and machine learning, we conclude by advocating the need for collaboration across varied academic disciplines.

Researchers from various scientific disciplines such as communication, business, sociology, and informatics have applied numerous research methods to examine and gain insights into different perspectives surrounding the platform itself and its users. These include the need to explore users and communities to understand their motivations for posting and watching, and in and predicting video virality. A study by Snelson et al. (2012) identified seven key areas that needed attention in YouTube research: (1) users, groups, and communities; (2) teaching/learning; (3) social and political impact; (4) video creation/production; (5) legal/ethical concerns; (6) media management; and (7) commercial interests (Snelson et al., 2012). We have adapted and refined these categories to offer the following five major areas that need to be explored in YouTube research:

1. User motivations for engagement
2. Social, political, educational, and economic impacts of video content
3. Impacts of user engagement (views, likes, dislikes, comments, and shares) on user perceptions and sense-making
4. Formation of groups and communities
5. Ethical and legal challenges of video sharing.

Question 1: What type of YouTube videos are users posting?

Question 2: What kind of comments are YouTube users posting underneath videos?

Question 3: Why are people viewing certain YouTube videos?

Question 4: How do people derive meaning from reading comments?

Question 5: Which videos attract the highest views (passive engagement)?

Question 6: Which videos attract the highest active engagement from users in the form of likes, dislikes, comments, and shares?

Question 7: What are the roles of YouTube algorithms and the recommender systems in suggesting videos to users?

There is a range of options available for researchers in seeking answers to various questions within the five major research areas outlined above. Before discussing the various data collection and analysis techniques in YouTube research, we first want to shed light on some of the key questions that have been frequently posed by researchers in the realm of YouTube research (Das et al., 2019; Jiménez and Vozmediano, 2020; Lee and Yoon, 2020; Moon and Lee, 2020; Porreca et al., 2020; Zhang et al., 2017).