Can data influence your next purchase?

eBay comes to NYU’s Center for Data Science

eBay Project Mercury Data Center

Last Wednesday, eBay’s Director of Software Development and Merchandise, Giri Iyengar, came to the Center for Data Science to discuss data science career options in the e-commerce industry.

Initially founded as an online auction site in 1995, eBay has become of the world’s leading e-commerce platforms today. With over 160 million active members and one billion items listed, eBay’s merchandise team harnesses big data and machine learning tools to boost sales.

As Iyengar explained, there are four steps to facilitating a successful sale: awareness, interest, desire, and action. Like setting up a store front, awareness involves arranging all of the items for sale in an attractive and user-friendly manner on the platform. After inspiring the undecided customer, the platform must then generate interest and desire by recommending suitable products to the customer based on their previous searches or purchases. These recommendations should eventually lead the customer towards an item that they like enough to click ‘purchase’. The customer’s purchase completes the final stage: action.

eBay NYC is the powerhouse that crafts the powerful algorithms behind this merchandising process. For example, to generate the list of ‘recommended items’ alongside the initial item (often referred to as the ‘seed item’) that the customer originally searched for, the algorithm sends the seed item’s data to the Merchandizing Back End (MBE). The MBE compares the seed item’s details with other data sets to analyze the titles, images, purchase rates, and views of other products. Those which bear the most similarity to the seed item then appear as the ‘recommended items’ for the customer.

To generate better recommendations, the platform also records its customer’s behavioral data. Which products were clicked on, viewed, and purchased? Which products were clicked on, viewed, but not purchased? Collecting and analyzing these results is crucial in helping algorithms make better recommendations every time customers use their platform.

This complex data science work is at the heart of eBay NYC, whose office is conveniently located in Manhattan. Being only a stone’s throw away from Center for Data Science, the exciting opportunity to work for Iyengar’s team is not to be missed.

by Cherrie Kwok


Originally published at cds.nyu.edu on October 19, 2016.

Show your support

Clapping shows how much you appreciated NYU Center for Data Science’s story.