Media Viewership in the Connected World: A Big Data Case Study


U.S. consumers are adding time to their media day and making time to connect with their favorite content, no matter where it exists (Nielsen 2014). But how they’re consuming media is ever-changing thanks to the continued proliferation of technological devices, 24/7 availability of the media content, ease-of-access, and economics. Whether streaming or satellite, over-the-air or over-the-top, understanding how consumers are consuming media is more important than ever, particularly for companies providing these services since advertising is their major source of revenue. For researchers, this consumption ecosystem has given rise to big datasets consisting of millions and millions of viewing records to mine thru in order to discover trends, viewing patterns, and relationships. In this study, we are attempting to do just that.Using a very large viewing dataset (48 million records) coming from a large frame (1.2 million) of connected devices (ex: Roku media player), we explored the Big Data characteristics of volume, variety, and velocity in streaming data. We performed a variety of analysis ranging from time series analysis to demonstrate growth trajectory of the connected devices and their households, to conducting spatial analysis to uncover the geographical spread of these devices and their households. Preliminary results showed the presence of interesting viewing trends between weekday/weekly, time of the day, and content types (i.e. genre), and an unbalanced geographic distribution. This study will provide valuable practical insight to researchers with respect to the appropriate ways of utilizing large volumes of digital media viewership data for viewership estimation methodologies. Also, for the production agency on optimizing its content production and lastly for the ad agency on improving its ad campaign strategies.

Recommended Citation:

Zhang, M., Rao, K., Fan, J., & Ion, C. (2016). Media Viewership in the Connected World: A Big Data Case Study. Paper presented at the Midwest Association for Public Opinion Research, Chicago, IL.

Attached Documents:

  • MAPOR 2016 Program (see page #35 for the mention)
  • For a copy of this presentation, please send me a comment with your email address in the box below.
  1. No comments yet.

  1. No trackbacks yet.