CinemaCon 2017: Who Is Your Audience in 2017?

By Will Palmer, Chief Executive Officer, Movio

The theatrical business as a whole is actually swimming in data – loyalty, POS, credit card, Mobile, Web, Email, social media interactions, exit polls, cookies, IP addresses and if we combine all this together we have “Big Data.”

Frankly, all this “Big Data” is useless unless it produces actionable insights.

So, instead of standing up here and preaching the virtues of “Big Data,” we are going to attempt to make sense of all this information and demonstrate actionable insights that our industry can immediately apply to our marketing strategies and ultimately, box office success.

The Box Office Challenge

Capturing the attention of moviegoers requires a personalized and relevant approach. Until now this has been difficult at best with movie audience data limited to post-screening polls, focus groups or digital sentiment analysis.

While this provides a useful snapshot, it offers limited behavioral information, such as preferred genres, average frequency, preferred day of the week, preferred session, or the average group sizes. All of which would be highly valuable to the wider film industry to inform decisions about what to produce and how to market and distribute content.

By harnessing the power of moviegoer data, we’ll paint the picture of the American moviegoer providing full audience profiles covering demo composition and insights (age, gender, ethnicity), and behavioral patterns (frequency, admissions) while focusing on several audience-based film clusters, from Horror to Comedy.

The tool for the job is our Movio Media solution, which with the support of our exhibition partners, captures and profiles the full attendance history of over 14 million active US moviegoers, including film, actor, director, session, day, time, number of tickets, and ticket type, providing a more complete picture of the movie-going audience.  

The Movie Cluster Map

By analyzing 100 million unique box office admissions and hundreds of movies each year, we created the Audience Similarity algorithm, a measure of the similarity of the audience for each film. The key to this approach is starting with movies rather than audience demographics. We don’t care if you are an 18-year-old that loves Philomena or an 80-year-old that likes Teenage Mutant Ninja Turtles.

Each dot of the cluster map represents a particular movie and the nearest dot represents the movie with the most similar audience. This feature is used to build the ideal comparable titles list with our studio partners and this week is being released to our exhibition partners as part of Movio Cinema’s Movie Insights module.

As a case study, we conducted 30 campaigns using our proprietary Audience Similarity algorithm. Moviegoers in the target group were 9 times more likely to watch the given film than the rest of the movie-going population.

When applying Audience Similarity, genre-based patterns emerge and connections between genres become apparent:

• The Tentpoles cluster sits in the middle of the cluster map, with many genres overlapping.
• African American Comedies, Horror, and Faith-based clusters have the least overlap with other genres.

This cluster map shows a map of these fourteen clusters into two dimensions, with Horror at the bottom left, Animation at the bottom right, Critically Acclaimed Drama at the top right, and African American Comedy at the top left.

• From bottom to top increasing in age. With film examples: Moonlight appealing to older audience and Pride and Prejudice and Zombies appealing to a younger audience
• From left to right as increasing seriousness or realism. With film examples: McFarland, USA being more serious and Hot Tub Time Machine being less serious
• From top left to bottom left as an increase in appropriateness for children. With film examples: Straight Outta Compton less and Finding Dory more appropriate

If we know what movies people watched and have the permission to communicate via social media, mobile, and email, then we can apply Audience Similarity for all targeted campaigns. However, to apply this concept to media planning and a wider marketing strategy, we need to define the audience into segments that provide insight into the demographics and behavioral patterns, which can be applied to our marketing or programming.

The Target Audience

The profile we used forms a target audience of moviegoers that watch between 3 and 52 films.

One interesting point is that high-frequency moviegoers, defined as those that attend over 12 movies per year, account for 13% of the movie-going audience but 30% of all admissions. Another way to look at this is that 6-7% of the overall US population accounts for $3 billion in box office receipts.

It is also obvious that most moviegoers enjoy films from multiple clusters, and it is likely that preferences for different clusters are correlated within moviegoer segments. Therefore, in order to best characterize the audience for each cluster, we next group the set of moviegoers into ten segments based on the percentage of visits to each movie cluster. Interestingly, when segmented in this manner, as set of demographic and behavioral trends falls out for each segment in addition to the characteristic set of genre preferences.

1. Avid Action Fans: 40% Tentpoles, 15% Mature Thrillers
2. Event Tentpole Fans: 70% Tentpoles, +20% Opening weekend
3. Young Action / Dystopian Fans: 28% Dystopian, 25% Tentpoles
4. Mature Action Fans: 40% Mature Thrillers, 13% Tentpoles
5. Avid Families Female: 40% Animation, 11% Tentpoles
6. Avid Families Male: 40% Tentpoles, 35% Animation
7. Event Families: 75% Animation, -15% Opening weekend
8. Drama Fans: 27% Critically Acclaimed Drama + Drama, 13% Mature Thrillers, 12% Female Comedy, -20% Opening weekend
9. Horror Fans: 33% Horror, 15% Tentpoles
10. Comedy Fans: 25% Male Comedy, 17% Tentpoles

To put all this into action and demonstrate the effectiveness of combining content, demographic, and behavioral information for audience identification, we will attempt to predict the audience for upcoming films across different genres.

Horror Cluster

60% of the Horror audience is aged 18-30, representing an increase of 80% from average. 35% are High-Frequency moviegoers, over-indexing by 10%. Hispanic moviegoers make up 26% of the Horror audience, over-indexing by 18%. 34% of this audience attends Night sessions (after 9pm), representing an increase of 46% from the average.

Horror Fans make up 39% of the Horror cluster audience. This segment comprises 53% Female, 56% aged 18-30, 24% Hispanic, 19% African-American, 70% Evening and Night, with 35% after 9 pm. Only 6% of members make up 39% of the Horror visits.

Horror fans are fanatical, however, we have anecdotal evidence that increasing frequency of Horror releases cannibalizes itself and may not lead to extra box office. It appears the optimum period between Horror releases is circa 5 weeks.

The audience of Annabelle 2, releasing on August 11, 2017, is likely to be young, highly Hispanic and African American. It will comprise high-frequency moviegoers that also enjoy Male and African American Comedies. Evening and late sessions will dominate, with slightly higher than average attendance on opening night/weekend.

The segments that this film will appeal to are:

• Horror Fans: 40-45% of the audience
• Comedy Fans: 13% of the audience, over-indexing by 40% on Horror films

Mature Thrillers Cluster

One of the key findings of ‘The 50-plus Moviegoer, An Industry Segment That Should Not Be Ignored’, the study AARP commissioned from Movio, and published in March, highlighted that the 50+ moviegoer has helped establish an emerging genre referred to as ‘Mature Thrillers’. This new category includes action-driven films starring veteran male actors.

The study found that 51% of the moviegoers who saw Jack Reacher: Never Go Back, Taken 3, The November Man and A Walk Amongst The Tombstones, were over age 50 compared with 32% for the total American cinema audience. The research suggests this phenomenon is because fans are aging alongside their favorite action stars such as Liam Neeson, Kevin Costner, and Tom Cruise. Interestingly Matt Damon and Ben Affleck have just graduated into this category.

47% of the Mature Thriller audience is aged over 50, over-indexing by 65%. 40% are High-Frequency moviegoers, representing an increase of 40% from average.

The Mature Thriller cluster audience comprises two segments: Avid Action Fans and Mature Action Fans, respectively 25% and 28%. 19% of Avid Action Fans are high-frequency moviegoers, 64% Male, 62% buying less than 2 admissions per visit. 67% of Mature Action Fans are over 50, 70% Caucasian, 55% Male, 45% visit in the afternoon, 21% are high-frequency moviegoers. 24% of moviegoers make up 53% of Mature Thriller visits.

The audience of American Made, releasing on September 29, 2017, is likely to skew towards a Male, Caucasian, older, higher frequency audience that primarily purchases tickets in singles or pairs, and is interested primarily in other Action or Tentpole films.

Segments that this film will appeal to:

• Avid Action fans: 30% of the audience
• Mature Action Fans: 30% of the audience

Tentpoles Cluster

Clearly, most moviegoers watch Tentpoles, hence the term 4-quadrant films, so more eyeballs seeing the marketing has to be a priority. However, through our research (‘Breaking the Blockbuster Code’) into audience evolution, we mapped how the audience profile changed throughout the release of a film. What was surprising in that there was a pattern and that it was predictable.

So, what have we learned?

Our data-driven strategy should start with the movie taste and not demographics.

To be relevant, marketing now needs to target people based on what they do and what they like instead of what they are.
Let’s change the narrative from “Big Data” to “Actionable Insights.”