MS Business Analytics Students Examine Long-Held Parcel Delivery Problem

In the wake of the COVID-19 pandemic, the demand for home delivery services skyrocketed. Package delivery hasn’t been novel across the U.S. since the Pony Express was inducted in the middle of the 19th century, but it still has its limits. Namely, it is limited by last-mile delivery logistics.

The last mile is the final step in a supply chain process that begins at a warehouse where a package is selected, sent to a distribution hub, and then transported to a final destination. That very last mile of the parcel’s journey comprises more than 50 percent of the delivery cost.

“In fact, for a lot of delivery, it's just impractical,” said Alan Montgomery, Professor of Marketing and Head of Master of Science, Business Analytics (MSBA) program. “If the delivery is not large enough, the shipping costs just take up too much of a percentage of the transaction, which is why we still go to the grocery store to pick up our groceries and why we still go to Walmart to pick up our bundle of goods, put it in our car, bring it home—because we're solving the last-mile delivery problem.”

Montgomery organized a team of MSBA students who aimed to provide a fresh perspective on the last mile delivery dilemma for educational project sponsor Tata Consultancy Services (TCS), an information technology services and consulting company. The TCS Research & Innovation Group previously hosted the TCS Innovation Forum, which included Carnegie Mellon University faculty as speakers. Sponsoring a student project focused on a supply chain challenge seemed like a fitting next step for the developing partnership between the school and company.

 “We identified last-mile delivery as a priority challenge, especially with the heightened awareness of multiple same-household deliveries and the rapidly increasing demand and variability that resulted from consumer response to the pandemic,” explained Matt Lekstutis, Global Managing Director, Supply Chain as a Service at TCS. “We knew that we would be challenged to obtain data to support the hypotheses raised regarding the challenge and that the project would be like peeling layers from an onion in search of the root cause and possible solutions towards the ‘new normal’ for parcel delivery operations.”

To begin their project, the student team—which included four MSBA candidates Kristina Schiffhauer, Sivanagahari Devarapalli, Stefanie Montgomery, and Timothy Marshall—made two hypotheses:

1. Parcel consolidation would reduce last-mile delivery costs.
2. There are incentives to motivate customers to choose delayed parcel delivery.

To test their predictions, they conducted surveys, analyzed data, and reviewed research. The team also built predictive models to determine consumer demands and the estimated delivery cost. Next, they wove their findings together to shape insights around how to optimize and reduce delivery cost; how to incentivize customers; and, what an appealing incentive entails.

TCS was delighted by the results, which can be viewed on the MSBA demo site. “The students not only developed supportable hypotheses based on test data creation and consumer survey research but exceeded the expectations in developing a scientific approach to optimizing home delivery,” said Lekstutis. “An unexpected result was the identification of consumer sentiment that actually would accept less than ideal delivery service in order to reduce home delivery frequency that further helps improve carrier capacity utilization.”

Montgomery summarized: “It turns out that people are actually quite willing to have their deliveries consolidated and they actually don't really need compensation, what they want is more control over [delivery] dates.”

According to the surveys, consumers would also appreciate earning a gift card in order to receive packages on an assigned weekday. Ultimately, very little incentive was needed. These results suggest that a delivery company can route its system more efficiently by taking fewer trips through a neighborhood, which significantly reduces the cost of last-mile delivery. Additionally, the results signal positive environmental effects.

“One of the benefits is less pollution, less driving, potentially even less packaging if you can consolidate packages together,” said Montgomery. “There's a green benefit of being able to have a smaller number of packages or a reduced carbon footprint that's associated with the delivery.”

“The students brought a fresh lens to the problem that has been analyzed across the industry for several years,” said Lekstutis. “Additionally, the student qualifications, skills, competency in data science, and business experience were invaluable. Another benefit to leveraging university students is the expertise and guidance the university faculty as project advisors bring to the project.”

Montgomery noted, “I think the greatest impact for the sponsor is that they were able to explore an idea that they may not have had the resources or time to examine in-house.” With many years of project advising to his credit, he feels these projects are great for answering the questions, “‘Wouldn't it be great if you had time to do this?” The key ingredients for setting up a great project, according to Montgomery, are an important problem and a data set that's available and compatible.