Comcast Data Analytics Problem
I participated in the Fall 2017 Data Analytics Challenge. My team member, Malik Donald, and I chose to work on the Comcast problem. Throughout my courses in the MIS program, I have had a particular interest in the Data Science and Data Analytics courses. I believe it is a very powerful skill to be able to use and manipulate data in ways that allow us to do business better and to understand our customer’s better. One of the most fascinating parts about these challenges is that contestants get an opportunity to work with real data that has been generated from large corporations, such as Comcast.
For our analysis, we compared box office gross returns to the amount allocated to the film’s budget. In order to do this, we first cleaned the data getting rid of any box office revenues that contained non-U.S. currencies because our prompt only mentions U.S Dollars. This still left us with movies generating revenues in the form of U.S dollars. After doing further research online about what is considered to be a “small budget film”, we were able to make an educated assumption that this would be any film whose budget was less than $2.5 million. This led us to separating our data on films with low and high budgets. Our graphic illustrates that there is a trend in box office success of genres. Further we found that these successful genres differ depending on the size of the budget. For small budget films, Horror and Drama perform exceptionally well according to a sum of the films’ box office gross. For large budget films, Horror and Family films are less successful in terms of box office gross. Our packed bubbled graphic demonstrates this through its size and color. The color hue shows the budget ranged from small to large budget films, while the size of the bubble displays the sum of the box office gross under the films’ primary genres. We believe this analysis provides useful insight when predicting which movies will be a hit or a bust.
Check out our final submission here: Fall 2017 Data Analytics Challenge Submission