Students collaborate during the first-ever WSB hackathon. PHOTO: PAUL L. NEWBY II
Thousands of data points on shortages and overages—customer purchase orders, trucks, load diagrams, and damaged products.
Far from a theoretical exercise, these real-life issues were part of the Wisconsin School of Business’s first-ever hackathon, a collaboration with Kimberly-Clark and IBM that gave students a real-world supply chain problem to tackle using company data.
During the one-day session, 15 Wisconsin MBA students and six engineering students at the University of Wisconsin–Madison were tasked with examining Kimberly-Clark data, looking for possible solutions to bring the company’s aggregate loss closer to a zero balance.
“Hacking is taking something and making it better through some kind of effort,” says Jake Dean, director of the Grainger Center for Supply Chain Management. “You’ve got this huge data set—picture it as an Excel spreadsheet that is probably 20 columns wide and 700,000 rows long. Now how do you take something like that and hack through it and get it to give you useful information?”
The event grew out of the applied learning curriculum course Dean teaches each Friday during the semester and demonstrates to students that the things they learn in classrooms Monday through Thursday “actually go on in real life,” Dean says.
Kimberly-Clark’s Stacy Wilcox, a reverse logistics consultant and graduate of UW–Madison’s civil and environmental engineering program, said the supply-chain issue is a “hot topic” at the company right now.
When customer loads are either short, damaged, or have more product than ordered, the company has to spend resources to resolve invoice claims from customers, Wilcox says. The problem could stem from a customer using a code incorrectly that results in short or damaged claims, as some customer sites aren’t fully digitized in their accounting and inventory. Or, the origin might be a shipping and distribution issue, such as the way a truck is loaded, that is causing damaged products. Most likely, it’s a combination of things, which is why the company is reaching out for student perspectives.
IBM's Bob Warpinski coaches students during the hackathon at WSB. PHOTO: PAUL L. NEWBY II
“We’re not seeing a whole lot of consistency with the trends we’ve looked at,” Wilcox says, “and we want to know if anyone else can see what we’re not seeing. It’s a big challenge.”
Bud Kane, business partner in information technology with Kimberly-Clark, says the problem initially landed in the domain of the company’s distribution center, but it is more complicated than that and spans more than one department. That’s beneficial for students to know at the outset.
“It’s really easy to go right to one solution when you may not have the root cause. That’s probably a good thing for them to find out right away.”
Another good experience for students: watching such a well-known company work to continuously improve when an issue arises.
“I think it can be eye-opening for students to look in the real world and understand real-world problems and how they can personally get involved and make a difference,” Kane says.
Hackathon participants also learned not just how to understand the data but also how to read for what isn’t in the data, Dean says. For example, a destination could have the most claims, but it could simultaneously have the most shipments without claims because it’s such a high-volume location.
IBM’s Watson Analytics platform, provided exclusively for groups to use during the hackathon, helps. The analytics tool takes what can seem like endless Excel data and processes it into visualizations and insights that users can easily read to quickly discover patterns and meaning in the data. With guided data discovery, automated predictive analytics and cognitive capabilities such as natural language dialogue, the students were able to interact with data conversationally to get answers they understand.
With prior supply chain experience at Amazon, Piyush Kumar (MBA ‘18) is already familiar with examining data sets for trends and insights, knowing where to sink the time into a particular set of data and when to pass over.
“Some of the places we are able to find a pattern, and some of the places we are not,” he says, scanning some data columns on how trucks are stacked with Kimberly-Clark consumer tissue and personal care products.
A hackathon participant examines the data set. PHOTO: PAUL L. NEWBY II
Dean says the cross-disciplinary component—joining the engineering students with the WSB students—provided more balance in terms of the skill set for the day. The setup also was designed to mirror common work settings—where those with more business knowledge and responsibilities often pair with more technically oriented colleagues to solve business problems.
Beyond the problem-solving challenge, the hackathon strengthened partnerships and gave students a chance to shine in front of potential employers.
Kimberly-Clark employs multiple graduates across all of its departments, Kane says, so events like the hackathon “help us keep an eye out for talent for either future careers or internships.”
During a debriefing session at the end of the day, students shared the insights they were able to obtain with the IBM and Kimberly-Clark teams. The back-and-forth exchange resulted in additional learning on both sides—with students gaining an appreciation for the challenges of gathering and cleansing data, and with Kimberly-Clark learning of some useful signals in the data that they hadn’t uncovered before.
IBM’s Bob Warpinski, a client executive with the company who helped bring the partners together initially, says he can see the hackathon happening again in the future.
“Hopefully, this becomes sort of a living, breathing, ongoing relationship,” Warpinski says.
For students, Dean sees the hackathon as extremely valuable and far more interactive than having a speaker come to the class or discussing a case study.
Learning how to handle complex sets of data is a “key supply chain skill, if not the perfect supply chain skill,” Dean says.
“This was hands getting dirty and coming up with conclusions based on the data.”