The Effects of Online User Reviews on Movie Box-Office Performance: Accounting for Sequential Rollout and Aggregation Across Local Markets
Our objective in this paper is to measure the impact of national online user reviews (valence, volume and variance) on Designated Market Area (DMA) level local geographic box-office performance of movies. We account for three complications with analyses that use national level aggregate box-office data – (i) aggregation across heterogeneous markets (“spatial aggregation”); (ii) serial correlation due to sequential release of movies (“endogenous rollout”); and (iii) serial correlation due to other unobserved components that could affect inferences regarding the impact of user reviews. We use daily box-office ticket sales data for 148 movies released in the U.S. during a 16-month period (out of the 874 movies released) along with user review data from the Yahoo! Movies website. The analysis also controls for other possible box-office drivers. Our identification strategy rests on our ability to identify plausible instruments for user ratings by exploiting the sequential release of movies across markets – since user reviews can only come from markets where the movie has previously been released in, exogenous variables from previous markets would be appropriate instruments in subsequent markets.
In contrast with previous studies that have found that the main driver of box-office performance is the volume of reviews, we find that it is the valence that seems to matter and not the volume. Further, ignoring the endogenous rollout decision does not seem to have a big impact on the results from our DMA-level analysis. When we carry out our analysis with aggregated national data, we obtain the same results as those from previous studies, i.e., that volume matters but not the valence. Using various market level controls in the national data model, we attempt to identify the source of this difference.
By conducting our empirical analysis at the DMA level and accounting for pre-release advertising, we are able to classify DMA’s based on their responsiveness to firm initiated marketing effort (advertising) and consumer generated marketing (online word-of-mouth). A unique feature of our study is that it allows marketing managers to assess a DMA’s responsiveness along these two dimensions. The substantive insights can help studios and distributors evaluate their future product rollout strategies. While our empirical analysis is conducted using motion picture industry data, our approach to addressing the endogeneity of reviews is generalizable to other industry settings where products are sequentially rolled out.
Keywords: Online Word of Mouth, Sequential New Product Release, Endogeneity, Instrumental Variables, Generalized Method of Moments, Motion Pictures