Feb 8, 2024

Can AI predict the success of a TV series just by analysing the script?

StoryFit & Kouo have set up a strategic partnership and research study to investigate how audience ratings are predictable from scripts. The results are groundbreaking and promise to change the way content gets produced.

Can AI predict the success of a TV series just by analysing the script?
Photo by Bastian Riccardi / Unsplash

Since Netflix set the bar extremely high and started a streaming platform race to the top, with every new show produced, millions of dollars and the growth of the platforms are at stake. So understanding what makes scripted content engaging is more vital than ever. The challenge is that even series nearly identical in talent, themes, genres, and production value will still be rated very differently by audiences.

As StoryFit and Kouo prove with their study, the success of series and films depends less on marketing, cast and production values alone and more on hitting the right moments of emotional intensity for audiences.

Now, with the help of StoryFit (predictive script analysis) and Kouo (live emotion analytics), producers, studios & streaming platforms can understand what those moments are, predict a script’s success before it goes into production and measure real-time audience reactions to produced content, reducing guesswork and the time it takes to fine-tune those notorious final cuts.

Deciphering Audience Reactions & Making Better Shows

The focus of the study was to prove the predictability of audience ratings from scripts and produced content. To this end StoryFit & Kouo analysed the pilot episodes of two Crime-Drama-Mystery series, Killing Eve and Clarice. Seemingly identical shows in themes and genre, yet only one made it to global success.

StoryFit’s AI engine ran on the publicly available version of the pilot scripts, applying hundreds of proprietary models to measure character and story elements and simulating audience responses to produce powerful predictive insights.

StoryFit scene-by-scene analysis of the emotional intensity of Clarice and Killing Eve.

Kouo’s AI ran on the biological data of participants who remotely shared their Apple Watch data while watching the finished episodes. Analysing human biometric feedback during the viewing process to understand the viewers’ emotional responses, Kouo can pinpoint salient moments of interest and areas for improvement.

Kouo’s analysis of viewer emotional states while watching the finished episodes.

Despite using different modalities at different stages of the content’s life cycle, StoryFit and Kouo identified the same differentiator. Audiences found Clarice to be 50% less emotive than Killing Eve, reflected in the higher ratings and better viewership performance.

While the two seem, at face value, quite similar, based on the kurtosis and the skewness of the scoring distribution for TV series (See the analysis of rating distributions for IMDb scores), Clarice scores in the bottom percentile (25%), while Killing Eve scores in the top 10% of titles.

Both modalities were also able to highlight moments of peak audience interest, flag potential areas of improvement and successfully predict audience rating of the content.

The capabilities StoryFit and Kouo represent are vital for studios, streaming platforms and production companies wishing to grow their audiences and slate of entertaining and engaging content. The two deployed together perfectly cover the entire life cycle of every production, from script writing to test screenings. What is most, as both employ advanced AI engines, they are fit for a world where attitudes towards content are constantly changing and evolving.

What does this mean for scripted content?

The significant overlap in StoryFit’s data and Kouo’s measured audience engagement levels prove to be a groundbreaking feat in the field of predictive analytics, and the two modalities capable of predicting audience reception will revolutionise how content gets produced.

And no, AI will not replace talented writers and creatives. Instead, it will ensure that productions spend less time, money and resources producing shows that ultimately flop.

Get in touch with us to learn more about our work and the study, or read our white paper here.