Hollywood’s Wary Embrace of Big Data

In recent years the movie business has used social data to connect with audiences and stepped up its reliance on quantitative data to forecast box office revenues. However, if data represented a person, that individual may get a seat at L.A.’s trendiest restaurant, but would still be seated in the back room. That was the gist of a Tribeca Film Festival Industry Talks panel on Tuesday in New York.

“There are three countervailing forces at play that we need to balance, namely the artistic creative side, technological advances and commercial considerations”, said Jason Kassin, co-founder and CEO of Film Track, a rights management company.

“Navigating the world with data points is different than it was five years ago”, added Eugene Hernandez, Film Society of Lincoln Center‘s director of digital strategy. The biggest change is the use of sentiment analysis to monitor audience reactions, though the benefits appear mixed:

  • Sentiment-based date is broadly used: “Big data has become socialized”, said Bill Livek, vice chairman and CEO of entertainment measurement company Rentrak. Their customers include not only big studios, but also independent studios and distributors across the country.
  • Social media monitoring yields massive, but imprecise data: Sentiment analysis measures movie reviews, ratings and audience comments. As Stacy Spikes, CEO and co-founder of theatrical subscription service MoviePass noted, “Going to the movies now is a communal experience”. Nevertheless, social media data isn’t projectable, the panelists cautioned.
  • Sentiment analysis can point to the right direction, according to Christina Warren, Mashable’s senior tech analyst. “But since monitoring is mostly done by machine, it’s best to use the tool to help target audiences and markets”, she explained. Livek concurred, adding, “A social media database can drive certain activities, but not content creation.”

The panel also tackled the quantitative version of data, which has “big economic implications for the film industry”, noted Livek. “Data helps reduce costs, which can in turn be invested in content”. The following represent various ways the entertainment industry uses this data:

  • Worldwide data helps predict U.S. box office sales: Movies often open globally, in cities located in time zones hours ahead of the U.S. International data therefore serves as a preview of domestic releases. Movie studios have time to react and adjust marketing plans accordingly.
  • Data is used for distribution and promotion decisions: By taking a holistic approach to movie data, Film Track figures out the best movie distribution options, such as HBO, Kassin said. They also review actor-specific data to decide who to feature more prominently in preview ads.
  • Data serves to prove or dispel industry stereotypes: Hollywood has many preconceived notions, such as film length, Kassin observed. But data may show filmgoers’ affinity for other than standard two-hour offerings. Livek agreed, saying video-on-demand has greatly impacted viewing timeframes. For example, Netflix’s House of Cards series allows a binge viewing option.

“There are too many intangibles” to count on data to make completely accurate projections, Kassin said. As Spikes added,” Many other things impact movies, such as filmmakers’ vision and passion”. And actors who behave badly end up being distractions that hinder or fuel box office success. (Speaking of which, we’ll see how Reese Witherspoon’s new movie fares after her infamous arrest this week).

Livek closed by saying, “Some things you just can’t research. You need to combine information with art and common sense. Data mainly reacts to historical trends. Why do you think there are so many sequels?”