Modeling and forecasting box office revenues of blockbusters

By Philip Hans Franses, Professor of Applied Econometrics
Carolyn Ridsdale

With a simple model the Professor describes weekly box office revenues for 40 top movies in the USA. The model allows for two groups of adopters. One starts to adopt before the launch, while the second group starts adopting right in the first week of release. It turns out that the two types of adopters are about equal in size. Also, marketing efforts before launch do not have much predictive value for the total amount of revenues in the end, but these efforts do have a positive effect on how long a movie can be viewed in cinemas.

'Pre-launch marketing activities are important and influential'

Look at the figure below which presents the cumulative revenues in the USA (in millions of USD) for Star Wars: Episode VII, The Force Awakens. The horizontal axis gives the weeks that the movie was shown in cinemas, and the vertical axis concerns millions of USD. Clearly, there is no visible S-shaped pattern, which you would perhaps have expected.

It could now be that the pattern in the revenues associates not with a single group of adopters, but two such groups. The first adopters could adopt already before the launch, due to marketing and trailers, but of course these adopters can only actually go to a movie at or from the time of launch. The second group of adopters could start to adopt at the time of the launch. If we describe each of the two types of adoptions using a logistic function, then it turns out that summing the two functions allows for describing the pattern as in the above figure.

The model is applied for 40 all times top movies in the USA for many years (data collected in May 2020), and these include Avatar, Frozen, The Hunger Games, and Titanic, see the next figure, where the data were retrieved from

'a good predictor for total final revenues is the amount of revenues in the first week'

Analysing the estimates of the key parameters in the 40 models reveals that the revenues from the two types of adopters are about equal in size. This means that social and other media apparently play a very important role in arousing preferences for movies even though the movies have not yet been launched. Prelaunch marketing activities are important and influential. It is likely the case that this becomes even more so in the future.

Furthermore, adoption by those who go to the movies when these are in the cinema is faster than for the pre-launch group. We also learn that earlier adoption by the first generation does not associate with a larger total amount of revenues. But more trailers and advertising does associate with longer playtime in cinemas. Hence, more marketing efforts prior to launch make movies to last longer in cinemas. And finally, a good predictor for total final revenues is the amount of revenues in the first week. In fact, a simple forecasting scheme is roughly 300 million USD plus the revenues in the first week, at least for blockbusters

Erasmus School of Economics

Philip Hans Franses

Philip Hans Franses is Professor of Applied Econometrics at Erasmus School of Economics. From 2006 to 2019 he served as its dean. His research interests concern the development and application of econometric methods for relevant and interesting problems in marketing, finance, and macroeconomics.

More information

This item is part of Backbone Magazine 2023. The magazine can be found in E-building or Theil-building for free. Additionally, a digital copy is available here. Backbone is the corporate magazine of Erasmus School of Economics. Since 2014, it is published once a year. The magazine highlights successful and interesting alumni, covers the latest economic trends and research, and reports on news, events, student and alumni accomplishments.

Compare @count study programme

  • @title

    • Duration: @duration
Compare study programmes