Arshavir Blackwell, PhD

How Marketing on Facebook Correlates with Domestic Box Office Returns

Another great CitizenNet post on Facebook and media returns—this time, for box office. The referenced paper is second in a series examining the relationship between online social behavior and offline purchase behavior.

Key Findings:

  • Facebook behavior is indeed correlated with box office returns.
  • The most correlated metrics are the number of people who have liked a movie’s Page, the click-through rate (CTR) of the movie’s content, and the Unlike rate of the Page.
  • More than 75% of all impressions of non-franchise film content on Facebook are sponsored, indicating that organic traffic is not enough to drive awareness
  • Understanding audiences is paramount to driving quality advertising

Many attempts at predicting box office – this is not one of them

Hollywood’s dramatized depiction of the business of statistics in Money Ball is an apt description of what studios need to contend with every opening weekend. Much like minor league scouts in baseball, studios routinely make bets on ticket sales, in excess of $100M on a film, with only statistics to ease nerves all the way up to the all-important opening weekend. With consumers’ increasing desire for IMAX and dazzling 3D action, the pressure to perform continues to increase.


Today, more than half a billion tweets and 2.7 billion Facebook® Likes are generated everyday. At CitizenNet, we are big believers in the power of social. Not only is the volume of social data huge, but in a world of device fragmentation, social identity may also become more robust than cookies and other browser-based behavior.


Many have looked at using social data to predict box office success. Similar to doctors’ recommendations on what food you should eat, the studies seem to conflict each other.


Back in 2010, an HP report showed how Twitter (much smaller in volume then, compared to today) could predict box office. Since then, newer reports have refuted that claim, with other researchers saying that Wikipedia or even human-scored sentiment analysis is the right approach. Rather than predicting box office success, we strive to:


Determine if social behavior, and Facebook in particular, correlates with box office sales Discover the main factors that drive such correlation Understand how paid marketing plays a role Armed with the knowledge of what tends to correlate with success, marketers can be smarter about what metrics are important and how to affect them.