We began by reviewing the technical and management literature by leading academic researchers on recommendation engines and on the reasons underlying the success of particular films and music.i Next, we interviewed representatives of 18 organizations involved in predicting or recommending cultural products. They included software companies that offer some form of recommendation system for movies, cable television, music, books and online shopping, as well as the smaller number that provide predictions to the cultural-product creation and distribution industry. We conducted a more extensive analysis of the phenomenon in the movie industry, where we interviewed movie studio representatives, movie distributors and movie theater companies about their use of predictive models.