« Back to Posts


I Hate, Therefore I Watch – Using Emotions To Predict The Next Episode’s Viewership

Has your hatred of a certain character on “The Bachelorette” ever made you tune in to next week’s episode? Does the insanity of a plot point on “Empire” make you all but certain to show up next week to see what happens next? And do you talk about these things on Twitter? If you said yes to all three of these statements, you are not alone and Canvs has the research to prove that you’re driving ratings.

In a new study released today by Canvs, the emotional measurement firm (and BRaVe Ventures backed venture) found that emotions can be used to predict growth in ratings for a show’s next episode. A 1% rise in viewers expressing hatred on a reality program leads to a 0.7% growth of gross ratings for adults 18 to 49 in live plus same day viewership on the next episode. This study was conducted between January 2014 and June 2015 and according to Canvs’ CEO Jared Feldman, is the single largest viewership study to date utilizing Twitter data. The study was conducted internally by Canvs and is based on their belief that emotions drive behavior.

Going into the study, Feldman had a few assumptions that turned out to be wrong. For one, he assumed that “funny” would be the main driver of ratings growth for comedies. That turned out to be not the case, as the most demonstrable emotion was beauty, with 1% yielding .3% growth and love having 1% yield .1% growth. Beauty and love here served as commentary on the physical attractiveness of the actors, which Feldman found “comedies play to more freely than dramas or reality according to the data.”

These findings have lead to the creation of Canvs’ new Canvs Viewership Probability, or CVP. This rating system uses the qualitative, emotional reactions  expressed about TV shows on Twitter to generate a well-calibrated probability as to whether viewership will go up or down for the next episode. With this new development, Canvs is placing a bet that they can finally answer the question “how does social effect ratings?” And if the results hold up, this could be massive for the industry as the question that has been asked since the start of social TV may finally have an answer.

unnamed