YouTube has changed the way video content is stored, transmitted, and viewed in the world. Yesterday's television is today's YouTube. YouTube has put the ability to publish video content to the world at the fingertips of anyone with an internet connection. The fact that viewing videos on YouTube is free and could be viewed at the convenience of your time generates a humongous amount of traffic. The statistics are staggering:
- 6 out of 10 people prefer online video platforms to live TV
- YouTube is the third most visited site in the world after Google and Facebook
- 300 hours of videos are uploaded on YouTube every hour
- 25 billion hours of video is watched each month
This explains the reason that YouTube was acquired by Google in 2006 for $1.65 billion. One big difference between TV and YouTube is in the ability to measure precisely the impact of video content in terms of traffic and preference. TV content popularity is generally measured using representative sampling metrics and then extrapolating them to understand the number of people that watched a program, while, every view on YouTube videos is recorded and aggregated along with opinions (comments), likes, and dislikes that allow us to understand the impact of video content much more precisely. Most companies today maintain their channels on YouTube the way they have pages on Facebook or Twitter. Like advertising on television, companies put ads and videos of their products and services on YouTube to promote various products or services.
YouTube provides channel and video performance (engagement on views) and audience metrics (demographics) to channel owners, but it is also possible, as a third-party, to analyze, to an extent, the impact of video content by using data from the YouTube API. This has advantages for a media analyst in understanding how different content works and how people react to it. Eventually, it makes an interesting way to be creative using data science.
In this chapter, we will cover the following topics:
- YouTube and campaigns
- Video performance analytics using Pandas
- Video comments analytics using sentiment and emotions
- Combining structured and unstructured data results for measuring campaign effectiveness