Using metadata from The Movie Database (TMDb) we built a model to predict box office revenue, demonstrating shared aspects of top performing films. We highlight variables such as genre and budget as being most predictive of high revenue generation (box office success) across more than 7000 films over decades from TMDb. These analyses include multiple linear regressions. We additionally examine hypothesized predictors such as a famous actor being in a film but do not find these to be associated with film success. In addition to our predictive modeling, we also note opportunities for future work, and present key examples of when and why box office revenue will perhaps no longer be the gold standard for measuring film success, of use to production companies, marketing teams, and investors alike.