The partnership has led to new study from SU applying its machine learning to Comscore’s TV Essentials second-by-second television viewing data, programme ratings, and to Comscore’s Exact Commercial Ratings data to identify key attributes and patterns that predict commercial viewership.

Julia Johnston, senior vice-president, regional and local agency sales at Comscore, said: “Every day, consumers are inundated with thousands of advertisements across the fragmented media landscape. More than ever, marketers need data that explains what messages are working and what makes viewers tune out or change the channel.”

One insight from the research suggests that the number of ads in an ad break during linear TV maybe a factor. An analysis of the audience viewership of The Big Bang Theory from September 2016 – February 2018 isolated several factors driving viewing declines. The combination of programme originality/rerun, ad duration, number of ads in a break, and number of breaks in a programme had more than 80% accuracy in predicting ad viewing declines.

The number of ads in a break (pod) and the originality of a first-run programme are the most important factor in maintaining ad viewing.

Table: Relative importance of each attribute, extracted from trained neural network

Source: Comscore