Understanding and forecasting consumer sequential multiscreen viewing behavior

Abstract

Nowadays, consumers often sequentially use multiple screens to watch media content, so that firms advertise on TV, websites, and apps at the same time period to reach consumers widely. However, sequential multiscreeners may repeated exposure to the same advertisement and develop negative attitudes. Utilizing real-world data, this paper employs a zero-inflated negative binomial (ZINB) model to predict consumer sequential multiscreen viewing frequency. Moreover, we quantify the net impacts of significant predictors on the sequential multiscreen viewing frequency and find that media factors (such as number of viewing devices, internet access, PC screen size, cellphone ownership, and device concentration ratio) have equivalent impacts as audience factors (including user demographics and past viewing behaviors) and the new created variable device concentration ratio has the largest impact. The accurate prediction and quantified net impacts can guide firms allocate their advertising budget more efficiently across multiple screens and avoid consumer overexposure to the same ad content.