Google Uses Search Data To Predict Box Office Hits
Written by Janet Swift   
Sunday, 16 June 2013

It should come as no surprise that popularity of movie trailers is correlated with a movie's success. However, a Google whitepaper reveals that search query patterns can predict box office performance with a surprising degree of success.

 

In a recently released white paper Quantifying Movie Magic with Google Search, two analysts from Google's Media and Entertainment division, Reggie Panaligan and Andrea Chen, explain that:

Search offers an on demand, real-time source of information that can extend engagement with potential moviegoers. By understanding how and what they are searching for, we can uncover unique insights into moviegoers’ awareness and intent.

Previous Google research had already revealed that on average, moviegoers consult 13 sources before they make a decision about what movie to see, but Panaligan and Chen have discovered that analyzing search query patterns can predict box office performance.

mm
 
The researches first establish the link between search volume and box office performance on an overall level:
mm1 400
(click in chart to enlarge)
 
The spikes in search coincides with the release of major movies and the researchers have a second graph that splits search into queries on specific movie titles and generic film-related queries:
 
mm2 400
 (click in chart to enlarge)
 
They use this chart to argue:
For film marketers, understanding these patterns can present a substantial opportunity. By adjusting search marketing strategies to these trends, marketers can either capture the attention of the “curious” moviegoer, or deepen audience engagement with a blockbuster title.
 
Using linear regression with 99 films released in 2012 the study concludes that:
  • 70% of the variation in box office performance over its opening weekend can be explained with movie-related search query volume seven days prior to the release date.

The researchers then combined search ad click volume, theater count and franchise status in a predictive model and achieve 92% accuracy in predicting box office sales, concluding:

  • If a new film has 20,000 more paid clicks than a similar film, it is expected to bring in up to $7.5M more during opening weekend receipts.

The researchers also looked at the influence of trailers perhaps the most interesting finding is:

  • 94% of variation in a film's box office performance at its opening can be explained with trailer-related title search volume 4 weeks prior to release, coupled with seasonality and franchise status.

mm3 400

(click in chart to enlarge)

 
With regard to this finding, the researchers comment:
 
The availability of content, specifically trailers, is important for moviegoers at all stages of the decision process. Earlier searches four weeks from release week for a film have the strongest link to intent ... presumably because the most ardent fans are among the first to search for specific film’s content.
 
They also point out  that as 48% of moviegoers decide what film they want to watch the day of ticket purchase it is critical for movie marketers to continue provide content that can aid in the decision to see a particular movie even after it has been released.
 
While the findings of this research are not spectacularly different from what was previously understood - an engaging trailer gets movie goers motivated to see a film - the use of big data leads to firmer conclusions than were otherwise possible.
qmm

More Information

Quantifying Movie Magic with Google Search

Related Articles

Google Flu Prediction - Beware The Media Effect

Intern at Google - Film, Fiction and Fact

The Significance Of Big Data

Twitter Can't Predict Elections Either

How the Music Flows from Place to Place

 

To be informed about new articles on I Programmer, install the I Programmer Toolbar, subscribe to the RSS feed, follow us on, Twitter, Facebook, Google+ or Linkedin,  or sign up for our weekly newsletter.

 

espbook

 

Comments




or email your comment to: comments@i-programmer.info

 

Banner


GitHub Universe AI Announcements - Copilot And Spark
30/10/2024

GitHub has announced several improvements for developers at Universe, its annual conference. Developers will get multi-model Copilot and GitHub Spark, an AI-native tool for building applications in na [ ... ]



Microsoft Open Sources Drasi
18/10/2024

Microsoft has announced the open source availability of Drasi, a data processing system designed to simplify the detection of and reaction to critical events within complex event-driven infrastructure [ ... ]


More News

 

 

Last Updated ( Sunday, 16 June 2013 )