Students in Michael Ratajczyk’s data analysis and business modeling course reviewed notes and discussed tactics as they waited for their turn in front of the judges.
The students, primarily sophomores and juniors, had been given the task of analyzing data from a fictional business, Cardinal Hardware, which utilizes multiple suppliers and uses their own fleet of trucks to save on shipping costs. In return for saving on shipping and overhead, the suppliers give Cardinal Hardware a rebate on their orders. Based on analysis of 10 years of data, teams were asked to forecast whether it would be more profitable for the business to continue to use a fixed rate rebate program or switch to a variable rate rebate program.
Regardless of their findings, students were judged on their ability to present their results in a clear, professional—and understandable—way.
Just to add to their anxiety a bit, their judges included former business alumni professionals, as well as area business professionals (including judges from 3M, Fastenal, Chrysler Winona, and Winona Radio).
“The data analytics course is one of four required in the business intelligence and analytics major,” Ratajczyk said. “While mathematics and business acumen are key ingredients in a successful graduate, it is of greater importance that graduates can communicate well with a variety of audiences. In the field, they will work with people who have a varying degree of data literacy.”
His students, he said, were given a limited set of instructions, a couple of data sets, and a deadline. The students needed to find additional data sets from the government and economic agencies and condense more than two weeks of data science modeling into a 15-minute presentation. Based on techniques used in class, they could utilize decision trees, logarithmic, or linear regression.
Dixon Irwin, a sophomore Business Intelligence and Analytics major, and his group ultimately decided that Cardinal Hardware should continue with a fixed rebate program in 2017.
“First were created a price model that would help us forecast for 2017,” Irwin said. “We considered Gross Domestic Product, CPI Consumer Price Index, and inflation in our forecasting method. Our next task was to use a percent change qualifier to determine how low our percent could drop before the company would have to revise their contracts. My group and I found that, with our model, sticking with the company’s original fixed rebate program would be more profitable for 2017. Along with that, we also found that if the company were to use our variable rebate program from the start they would have seen more profits.”
Junior Tara Nagy, a junior Business Intelligence and Analytics and Finance double major, said her team also ultimately recommended using a fixed variable rate in 2017.
“We audited the financial quarterly reports of Cardinal Hardware and their shipping orders. Along with this, we calculated the fixed rebate amount based off a fixed cost for all suppliers for the years 2007-2016. Following this, we had to calculate the variable rebate cost of the suppliers based off a percent change qualifier per quarter from 2007-2016. After calculating the rebates for fixed and variable, we had to forecast what the rebates will be in 2017. What my group found was that the variable rebate was better over the time span of 2007-2016 compared to the fixed rebate, but the fixed rebate was better when forecasting 2017 compared to the variable rebate. To find this, we used a weighted average over the last three months to forecast the diesel fuel prices and used a percent change qualifier to determine if it was sensible to change the rebate percent per quarter or if having the fixed would be better. Overall we found that a fixed rebate percent gave the company more profit.”
Both students hope one day to get a job in business in the business intelligence and analytics field. For Irwin, this exercise helped solidify that he had chosen the correct career path. “This experience was a great confidence booster for me,” he said. “I always knew I like doing this kind of work, but the communication part of the job was the big question mark for me. If you ask my group members, before the presentation I was extremely nervous, but once it came my turn to present, I effectively communicated our data to our audience. Knowing that I can do this and do a respectable job was the confidence I needed.”
Nagy knows she’s headed into a field that has a shortage of job candidates with the necessary skills in data science and analytics.
“There is such a high demand everywhere for someone who knows how to look at data, manipulate it, visualize it, and ultimately present it in a way that allows people to make decisions based on these findings,” she said. “These classes will make me more valuable as a job candidate because the projects that we do are from the real world.”
Irwin agrees that he’s learning skills that are applicable in any career path he chooses. “I’m not only learning valuable skills that can help a company make better decisions, but I am also learning skills that can help me communicate better. The projects we work on are very similar to possible situations we could see on the job, so it’s beneficial to get this ‘real world’ experience.”
Photo caption: Michael Amelio ’17 presents his findings to the judges.