Showcase of Undergraduate Research and Creative Endeavors (SOURCE)

Building a Time Series Regression Model to Predict the Cost of a 30-second Super Bowl Commercial

The Super Bowl is broadcast in over 180 countries and is one of the most-watched television events in the United States. Its popularity has only grown since the first Super Bowl in 1967, and over the years it has become much more than just a game of American football. It is an important cultural event that attracts millions of viewers and provides one of the most powerful platforms to reach audiences and capture their attention. From a business standpoint, it is an opportunity to promote products and services through commercials, create a lasting impression on consumers, and build product recognition and consumer appeal. This research project aims to explore the ratings and viewership trends of Super Bowl games from 1967 to 2020 and to understand the factors that affect the cost of a 30-second Super Bowl commercial. How does the popularity of the teams affect viewership? Does the success of a franchise affect the popularity of the game? Do the household ratings and share exhibit a stronger correlation with the popularity of the halftime show or with the teams playing in the game? Does the location of the game affect the total viewership around the country? These are some examples of the questions we address in our study using the exploratory data analysis . In the modeling portion of the study we build a times series regression model and employ other forecasting techniques to predict the cost of a Super Bowl commercial for the next several years.