Do online searches influence sales or merely predict them? The case of motion pictures

Purpose: This paper aims to examine whether a film’s search volume causes its ticket sales in different stages of its lifecycle. Design/methodology/approach: This study tests the causality between searches and sales by using an instrumental variable approach. This study exploits the ideas that consumers’ perception of a product’s consumption risk is correlated with their search efforts and consumers use multiple information sources to infer a product’s consumption risk. As an instrument for a focal film’s search volume, this paper uses review disagreement for past movies related to the focal film. This paper incorporates the ideas in a model of weekly online search volume and revenues and apply it to a movie data set. Findings: Films’ search volume influences their revenues only until the opening week. A 10% increase in opening-week search volume generates a 7.4% increase in opening-week revenue, while the same increase in pre-launch search volume generates a 4.1% increase. Although searches are not an influencer of sales from the second week on, the random forest models and cross-validation studies find that weekly search volume is a strong predictor of weekly revenues in this period. Research limitations/implications: Testing the findings in other product categories is important for generalizing the findings. Practical implications: This study suggests different usage values for online searches, depending on a film’s lifecycle stages. Furthermore, given that review disagreement has a positive influence on opening-week revenue through searches, managers should encourage diverse opinions about their films until the opening week to increase sales through searches. Originality/value: Regarding the role of online search, previous studies have maintained the perspective that online search is a predictor of sales. This is the first study that finds causality between searches and sales for films.