eBay is looking for university students to help it improve the delivery estimates it displays, which lets shoppers know approximately how long it will take for any particular item to arrive on their doorstep once purchased. It's a crucial piece of information that has the potential to set unrealistic expectations on the part of impatient buyers, especially around the holidays.
eBay put the call out as part of a competition for students interested in applying Machine Learning technology to a real-world problem. eBay described the overall challenge as follows:
"The problem statement we invite you to consider is how to build a model that can accurately predict delivery dates for items sold on eBay, given a dataset of pertinent shipping information.
"The accuracy of shipping estimates plays a significant role in providing a hassle-free and trusty customer experience.
"However, this particular area has not received enough attention within the machine learning community despite its growing importance in the new online world. We at eBay want to change that."
"The question we invite you to address is to estimate the delivery date of shipments of online purchases. The shipments come from a diverse set of sellers on eBay, ranging from people selling items from their households to large business sellers.
"The journey of a package from a seller to buyer is made up of 2 parts. The first part is the handling time, which covers the time taken by the seller to package the item until it is handed over to the carrier. The second part is the transit time, which is the time taken by the carrier to deliver the package."
Students' objective is to estimate the total number of calendar days after payment it will take to have a purchased item show up at the buyer's address.
eBay will provide competitors with a dataset consisting of 20 million randomly selected shipments from transactions on eBay. (eBay said all records would include a record identifier and anonymized identifiers for the seller, the shipping service, and the category of items.)
This is eBay's 3rd Annual University Challenge in the space of Machine Learning on an e-commerce dataset - the winning students will receive a summer internship with eBay next year.
This year's contest will be too late to impact the 2021 holiday shopping season - the deadline for submitting solutions is January 14, 2022, and eBay will notify winners of the competition on January 28th.
Regardless, there's an old saying when it comes to technology: Garbage In, Garbage Out. It was disappointing to see unrealistic delivery estimates during last year's holiday shipping crisis - and sellers paid the price when their orders arrived late - in some cases, many weeks late.
If marketplaces like eBay can't obtain and integrate accurate real-time data from shipping carriers into its system, even the best Machine Learning technology may not help.
Feel free to weigh in on the accuracy of shipping delivery estimates on eBay and other marketplaces and any ideas you may have on improving them.