Decisions, Decisions: How One Class is Using Science to Predict Craft Beer Sales

On the surface, craft beer and decision and system sciences studies may seem unrelated, but to Ron Kilmberg, Ph.D., professor in Saint Joseph’s Haub School of Business, they go hand in hand. 

For the last few years, Klimberg has partnered with software developers at Branded Aware to host a course-long competition in which students predict the market trends of the craft beer industry. Using weekly syndicated data from Nielsen, a global marketing research firm, student teams select the six-pack of craft beer they believe will have the highest percentage increase in weekly sales; the teams who are most accurate vie for ultimate bragging rights and a top spot on the class’s digital leaderboard. 

The competition begins with a “pre-season,” during which students apply various forecasting techniques to calculate historical sales of craft beers. This provides them with a testing ground for the methodology they will later use to predict live weekly sales. Then, the competition begins. 

For the remainder of the semester, student teams track current data in order to select the six-pack of craft beer (among 200 leading craft beers) they believe will have the highest sales week-over-week in the country. At the end of the project, students present their methodology and are evaluated on how closely they predicted the numbers and how they adjusted and learned from a fluctuating market.

Klimberg and his students agree that this gamified version of experiential learning helps them build skills that will prepare them for their careers. This process of learning by doing requires teamwork, critical thinking and a holistic worldview of the market — all of which are part of the decision and science system curriculum. 

According to Klimberg, “Teams that do well are the ones that explore other techniques — such as using seasonal data, weather, recent news in supply chains and multiple sources of market data — and work closely together. It shows teamwork and their ability to adjust to change in the market.” 

He says that the challenges in supply chain exacerbated by the COVID-19 pandemic have only made the course more interesting this semester. 

“Sales actually went up last year. Seasonal trends affect the predictions more than the pandemic did overall. This is a key concept that students need to know in the field of analytics,” he says. “You need to balance the analytics with human intuition. In some ways, it’s like predicting a stock or whether the batter will get a hit or not.”

Joe Risi ’23, a junior business intelligence and analytics major from Doylestown, Pennsylvania, is currently enrolled in the class and is hoping to finish in first place with his team. Risi says that his team is using a combination of three different methods of analytics each week — kept under close wraps until they present their methodology — to predict trends of the beers and says it has not been easy.

“We’ve been off on some picks,” he says. “It is very difficult to predict some of the different variables. There is a lot more to predicting and some of the analysis than you’d expect.” 

Risi has a co-op lined up with Johnson & Johnson next semester and hopes to work in sports analytics one day. However, he says that this course has application to nearly any career. 

“Everything is trending toward data, not just the business world, but society as a whole. So this is a great way to jumpstart a career,” says Risi. 

While gamification is not a new concept in education, using it in the decision and science systems course is. “Our program continues to try to lead in terms of the education of our students,” says Klimberg. “Our online masters analytics program was one of the first in the world and our undergraduate analytics programming was the first one in this area; this gamified experiential learning project is just another example of how Saint Joseph’s leads the way.”