The Propensity Algorithm™ is at the heart of Movio EQ's AI approach. By drawing on all of the individual and collective moviegoing data stored in your Loyalty database, it identifies the members who are most likely to be interested in a movie.
Important: The Propensity Algorithm™ is not an attendance or box-office forecaster, and the audience numbers are not a measure of how well or badly a movie is doing. The algorithm is a recommender system that predicts moviegoer taste and estimates the right audience for a piece of content based on your members' movie-going history.
You can use this to your advantage when you define a campaign audience or use the dynamic-content options in the campaign or journey message editor.
Marketing benefits
- Connect moviegoers with their ideal movie.
- Easily assess the likely interest in each movie.
- Create complex, granular segmentation with a high degree of confidence.
- Identify the best content to feature in your emails.
- More effectively focus marketing time and effort based on the likely interest in 'Coming soon' movies.
Effect of showtimes on propensity score
EQ retrieves showtimes from movieXchange or, in some cases, International Showtimes. These are taken into account when calculating a member's likely interest in a movie.
Note: If you had MX Showtimes (MXS) enabled in Classic, you will have it in EQ. If you didn't have it enabled in Classic, or you've never used Classic, contact us to enable it.
If any of the following are true, there will be no impact on the propensity score:
- A movie doesn't yet have any showtimes at any of your cinemas.
- A member hasn't defined any preferred cinemas and doesn't have a most-frequented cinema (based on transactions).
- A movie has at least one showtime at one of the member's preferred cinemas or their most-frequented cinema.
The only time the propensity score will be affected is if a movie has showtimes but not at a member's preferred or most-frequented cinema. In that case, the propensity score will drop to the Less likely category.
When there aren't many transactions for a movie
Initially, when there aren’t many transactions for a movie, EQ uses algorithmically generated comp (comparable movie) titles to generate propensity numbers for the movie. The comps are generated by comparing the metadata of the current movie (for example, genre, censor rating, and synopsis) with that of previously released movies, as well as taking previous moviegoer behaviour into account.
When a movie’s ‘watched’ count exceeds 5% of your active member base, the Propensity Algorithm stops using comps and calculates the numbers using the new transactions and behaviour.
See also: