BACK TO ALL BLOGS

Hive + Bain Media Lab: Who Was Seen and What Was Said During the 2020 Academy Awards

Next-day analysis highlights trends in measured exposure during Hollywood’s biggest night

Hollywood’s biggest night, the Academy Awards, wrapped up this year’s awards season in style. Red carpet fashion started the night and Parasite stole headlines after the Korean-language film claimed four awards including Best Picture.

While awards are permanent markers of achievement, exposure is a broader prize shared by winners, nominees, performers, and presenters. Hive’s Celebrity Model, used by agencies to measure endorsement value and by media companies to enrich metadata in their video libraries, measured the screen time earned by the stars during the 2020 Oscars.

Bong Joon Ho – who took the stage as a winner four times for Parasite – earned the most time on screen during last night’s telecast of the Academy Awards, according to data from Hive’s Celebrity Model (see Figure 1). The remainder of the top 10 was made up of winners (Joaquin Phoenix, Brad Pitt, Laura Dern), presenters (Steve Martin, Chris Rock, Kristin Wiig, Maya Rudolph), and some celebs wearing multiple hats (Elton John as a winner and performer; Taika Waititi as a winner and presenter).

Source: Hive Celebrity Model
Source: Hive Celebrity Model

Much was said leading up to the event about another host-less award show short on diversity. A diverse mix of presenters and performers aimed to compensate for nominees that skewed white and male. However, while these themes stole headlines leading up to the event and were scattered across acceptance speeches during the night, most of what was said during the show was relatively consistent year-over-year (see Figure 2).

Source: Hive Speech-to-Text Model
Source: Hive Speech-to-Text Model

Hive’s Speech-to-Text model, with commercial uses including transcription of audio and monitoring of brand mentions in TV, radio, and digital video, was used to track mentions and keywords from within the Oscars’ telecast. Insights from what was said across award presentations and acceptance speeches included:

  • Thanks were given more than 120 times and love was expressed more than 55 times – mostly to thematic groups including The Academy, parents, partners, children, and God, as well as casts and crew
  • Statements on diversity and inclusion – spanning gender, race, and sexual orientation – were sprinkled throughout the night and were material in aggregate
    • Women, plural, were referenced as a group more than 3 times as often as men, most notably differentiated by messages of strength and unity (“all women are superheroes”)
    • The presence of Black Panther and BlacKkKlansman in the 2019 Oscars drove more significant conversation on race during last year’s telecast, which was less in frequent this year although still referenced across multiple speeches (e.g., Matthew A. Cherry and Karin Rupert Toliver), award presentations (e.g., Chris Rock and Steve Martin), and performances (e.g., Janelle Monae)
  • References to current events were scattered across the awards show, reflecting topics that impacted society over the past year including climate change and the environment, politics, and the death of Kobe Bryant
  • For the second year in a row, Netflix earned the highest count of mentions by award recipients among media companies – even with just 2 of its 24 nominations resulting in wins

About our models:

Hive’s Celebrity Model is trained to identify more than 80,000 public figures in media content and uniquely leverages Hive’s distributed workforce of more than 1.5 million registered contributors to efficiently optimize the precision and recall of low confidence results. Commercial uses of the model include measurement of endorsement value by agencies and enrichment of metadata in media companies’ video libraries.

Hive’s Speech-to-Text Model parses and transcribes speech data from video and audio content, and can be accessed via an API or on device. The model is trained by tens of thousands of hours of paired audio and speech data. Commercial uses of the model include transcription of audio and monitoring of brand mentions in TV, radio, and digital video.

Kevin Guo is the cofounder and CEO of Hive and is based in San Francisco. Dan Calpin is President of Hive Media and a Senior Advisor with Bain & Company based in Los Angeles; he was a founding partner of Bain Media Lab. Laura Beaudin is a Bain partner in San Francisco and leads Bain’s Global Marketing Excellence practice. Andre James is a Bain partner in Los Angeles and leads Bain’s Global Media & Entertainment practice; he was a founding partner of Bain Media Lab.