Students Score Big in Super Bowl Commercial Competition

PROVO, Utah – Mar 18, 2021 – Usually while watching TV, people tend to skip or purposefully ignore the commercials. However, once a year, there is one event where the commercials draw almost as much as attention from viewers as the actual program people watch—the Super Bowl. Since 2016, the Eccles School of Business at the University of Utah has hosted its Game Day Analytics Challenge, where student teams from various universities in Utah analyze tweets to determine the effectiveness of Super Bowl ads. This year, the BYU Marriott School of Business team comprised of information systems (IS) students took first place in the undergraduate category and received a $1,000 cash prize. 

Members of the BYU Marriott team included IS juniors Carter Beck, Parker Mecham, Benjamin Sierra, and Scott Young. Their challenge involved not only analyzing tweets but also presenting the group’s findings in a meaningful and interesting way. “The goal of the challenge was to analyze tweets from the Super Bowl Ads and create an infographic with unique and helpful insights that advertising managers could benefit from,” says Sierra, who hails from South Jordan, Utah. “Using the Twitter data set, which included 1.2 million tweets, we identified which commercials did well and which did not do well,” adds Beck, a student from Meridian, Idaho. 

After measuring the commercials’ engagement, the team compiled a thesis that visually presented key elements of a successful commercial. “Through the use of various data visualizations, we highlighted certain aspects of commercials that made the ads stand out and perform well,” says Beck. “Our thesis was that to build a winning commercial, a company needs three things: first, celebrities and preferably more than one; second, an entertaining tone as opposed to an inspiring one; and third, a brand ambassador who is willing to tweet about the commercial during the Super Bowl.”  

The skills taught by BYU Marriott IS professor Mark Keith in his IS 415 Machine Learning class helped students analyze and compile the data. "Our success as a team would not have been possible without our prior knowledge taught by the amazing professors in the IS program,” says Mecham. “In Dr. Keith’s class, we learned how to break down big sets of data and analyze them effectively. For the challenge, we implemented skills such as sentiment analysis on the data set, which helped us understand how viewers actually felt about the commercials rather than just looking at how much people tweeted about them.”

As the faculty advisor for the team during the challenge, Keith believes that the students’ problem-solving abilities were the keys to their success. “The students were incredible—they’ve only taken one class so far in Python machine learning and data analytics,” he says.  “However, I believe the reason for the team’s success had less to do with the class and more to do with their willingness to tackle unstructured business problem solving.”

“There are no ‘right answers’ you can memorize from a book on real-life data problems like these,” Keith continues. “These students had to be willing to try and possibly fail—one of the key characteristics of successful people. They found their own creative solutions using only their existing business knowledge, technology skills, and statistical reasoning. The students on this team exemplify the type of people that we hope all our students will become.”