Listen to the authors present their findings (source: September 2020 JM Webinar)
“No spoilers!” say many directors. Their concern is that if publications or moviegoers reveal plotlines and surprises, the public won’t want to pay for the movie. But is that concern well-founded?
Contrary to the concerns of some directors, a new study in the Journal of Marketing finds empirical evidence that spoiler reviews actually have a positive and statistically significant relationship with box office revenue. To uncover this finding, our research team examined daily box office revenues for movies released between January 2013 and December 2017 in the United States. These movies were then matched with their respective reviews collected from Internet Movie Database (IMDb), the most popular movie review platform in the United States. IMDb has a policy where its users are required to label their own reviews as spoiler reviews if they believe they give away any important plots. The published reviews are regularly monitored by the platform’s moderators and users that fail to adhere to this policy have their reviews taken down from the website and their accounts blacklisted. This policy provides us a labeled sample of either spoiler or non-spoiler reviews, which we then use to measure the degree of spoiling.
It is insufficient to measure only spoiler volume or the number of spoiler reviews for a movie. For example, two spoiler reviews could reveal very little about the plot, whereas a single spoiler review can divulge the plot in much greater depth. This motivates our team’s conceptualization of a measure for spoiler intensity, or the degree of plot uncertainty resolved by reading spoilers in movie reviews. Our measurement procedure applies a machine learning algorithm to uncover the topics from the text of both spoiler and non-spoiler reviews. Some of the topics include: “acting performance,” “cinematography,” “relationship,” and “death.” Not surprisingly, the topics of “relationship” and “death” are associated more with spoiler than with non-spoiler reviews. We aggregate the number of words related to these topics associated with spoiler reviews to calculate our measure of spoiler intensity. We consistently find that spoiler intensity has a positive and significant relationship with box office revenue. In contrast, we find little evidence for the effect of spoiler volume.
We postulate that uncertainty reduction is the driving mechanism behind this positive spoiling effect. If potential moviegoers are unsure about the quality of a movie, they are likely to benefit from the plot-related content of spoiler reviews when making their purchase decisions. Consistent with this, we find an inverted-U relationship between average ratings and spoiler intensity, which suggests that the positive spoiling effect is stronger for movies that receive moderate or mixed ratings compared to movies that receive either very high or very low ratings. This is likely because user ratings in the middle range tend to convey more ambiguous signals about movie quality compared to extreme ratings. Second, we find that the positive spoiling effect is stronger for movies that receive less advertising. Advertising can serve an informative function for consumers and is seen as a credible signal of quality in the movie industry. Less advertising should therefore lead to greater uncertainty about movie quality for potential moviegoers. Third, the positive spoiling effect is stronger for movies of limited release, which is a strategy often employed by independent and arthouse studios associated with greater uncertainty in terms of artistic quality. Finally, the positive spoiling effect declines over time, likely because consumers have greater uncertainty in the earlier periods of a movie’s life cycle.
This leads to several implications for stakeholders in the movie industry. First, online review platforms can potentially increase consumer welfare. The uncertainty-reduction mechanism suggests a spoiler-friendly review platform can help consumers make purchase decisions. Accordingly, online review platforms should maintain the availability of spoiler reviews, especially plot-intense spoiler reviews. We also recommend that review platforms keep the warning labels of spoiler reviews because of the benefit of allowing consumers to self-select into the exposure to spoilers. These spoiler-alert warnings reduce the search cost for consumers who seek to reduce movie uncertainty, while shielding consumers who wish to avoid spoiler reviews.
Second, movie studios and theaters should actively monitor the content of spoiler reviews to better forecast future box office revenue. The benefit of this monitoring activity is greater for movies with less advertising spending. Third, managers should make greater monitoring efforts in the earlier, rather than later, period of a movie’s life cycle. Finally, for movies with small advertising budgets and mixed user ratings, marketing managers should place great emphasis on stimulating online word of mouth, even when it might spoil the movie plot. However, for movies with large advertising budgets and extreme user ratings, we do not recommend encouraging consumers to generate spoiler reviews.
From: Jun Hyun (Joseph) Ryoo, Xin (Shane) Wang, and Shijie Lu, “Do Spoilers Really Spoil? Using Topic Modeling to Measure the Effect of Spoiler Reviews on Box Office Revenue,” Journal of Marketing.
Go to the Journal of Marketing
Jun Hyun (Joseph) Ryoo is Assistant Professor of Marketing, City University of Hong Kong, Hong Kong.
Xin (Shane) Wang is Associate Professor of Marketing and MBA ‘80 Faculty Fellow, Western University, Canada.
Shijie Lu is Howard J. and Geraldine F. Korth Associate Professor of Marketing, University of Notre Dame, USA.