Todd Ferris is a principal data scientist for Hershey Chocolates. He can track a cocoa bean from harvest to chocolate bar on a store shelf. Here are some excerpts from The Wall Street Journal (March 29, 2019) interview that you might use in class when you cover our new chapter, Module G, Applying Analytics to Big Data in Operations Management:
Our team goes after complex problems across the supply chain. Can we see our products from sourcing a cocoa bean in Western Africa all the way to manufacturing, shipping and getting it to the customers? It’s very difficult to keep track of all that data.
We are always fighting this bullwhip effect, the phenomenon of small changes at one end of the supply chain creating huge issues once you get back to manufacturing. If customer demand varies by 100 chocolate bars at retail, by the time that information gets back to us at manufacturing that signal may be 1,000 bars. This creates a lot of inefficiencies.
We use a programming language called R, a counterpart to Python. You’ll do your modeling inside one of those programs; we’ll use SQL to access, manipulate and filter data before we bring it into analytical tools.
You can do procurement, like forecasting the health of crops or their availability. I’ve worked on manpower analysis—how we shape our manpower in our manufacturing plants in the best manner possible. We just worked on the best way to route peanuts from our suppliers to our plants. On one side we are forecasting crops and on the other we’re at the store level trying to determine if we have too much inventory or too little. We try to predict what’s going to happen and then make sense of how we need to respond.
Classroom discussion questions:
- Describe the “bullwhip effect” and its importance in OM.
- How would you describe “data analytics” to an executive and explain its role?