Last year, a friend told Minghui He about an open internship on Amazon Search’s research team. It was exactly the type of opportunity she had been looking for: A chance to explore e-commerce research questions in the US.
Minghui was working on her PhD in the financial services analytics program (FSAN) at the University of Delaware, having just completed a year-long apprenticeship with a third-party seller on the Chinese online shopping platform, Taobao. During the apprenticeship, she monitored and analyzed key performance metrics in influencer live streams, looking for factors that translated into viewer engagement and purchases.
Minghui earned her bachelor’s degree in mathematics in China, where she grew up, and had always wanted to see more of the world — a desire that partially inspired her to pursue an advanced degree. But the Amazon internship represented an opportunity to put her skills into practice.
“One of my motivations in getting an advanced degree was to go abroad to America, because it is home to some of the biggest tech and e-commerce companies in the world,” Minghui said. “I also wanted to use my mathematics knowledge to understand customer behaviors.”
Her pursuit of an advanced degree provided opportunities to see the world, but Minghui knew her passions existed more in practice than theory.
“As a PhD, your output is research papers,” she said. “But I wanted to apply those theories to an actual product.”
Keen to apply her data analysis and machine-learning skills at an American retailer, Minghui applied for the Amazon internship and got it, joining the Search research team in June 2020.
Using data to understand customer behavior
During her internship, Minghui worked with quantitative user researchers, combining their data — derived from surveys and other methods — with qualitative research insights. The research team focuses on understanding what happens as a customer searches for a product on the Amazon Store. What are they looking for? Do they find what they want? What influences the buying decisions that take place after the search results are provided?
Specifically, she built machine-learning models to predict when customers might hesitate on, or abandon, a purchase. She also applied natural language techniques on customers’ search queries to infer their search intent, then combined it with their actual behaviors and transactions to understand their shopping paths.
The role was a natural extension of Minghui’s PhD interest. For her dissertation, she used machine learning methods to analyze major brands’ posts on social media, correlating their image attributes, such as color and content topics, to followers’ engagement.
Beyond data analysis
Alex Thayer, head of the Search research team, chose Minghui for the internship. Thayer had been seeking someone who could conduct applied science with data to enhance the Amazon search experience for customers. But Thayer also sought an individual capable of thinking beyond the numbers when analyzing customer behavior and valued Minghui’s experience studying people’s actions on Taobao and on social media.
When trying to understand people and their habits, Thayer said, you start with small-scale qualitative research and create hypotheses about people’s behavior and motivation. “But at some point, you must scale up those initial hypotheses. To do that, ideally you need someone with research science expertise to expand your ability to look for insights across tens of thousands or millions of customer signals.”
“Given Amazon’s scale, we have a lot of data. So in a sense, it’s not a challenge of looking through data and finding patterns,” Thayer continued. “It’s more about knowing where to start when you look at a giant dataset.”
Knowing where to start is a form of working backwards from the customer — a key component of Amazon’s customer obsession leadership principle — and Minghui was primed to do just that. With family and friends still in China, she pays close attention to e-commerce trends there as well as in the US. During her Taobao vendor apprenticeship, she worked on understanding customer behavior and sales conversion rates related to influencer live streams — a digital, younger-skewing version of video-driven shopping platforms like the Home Shopping Network or QVC.
Inspiring others with initiative on ideas
The friend who referred her to the internship, Lan Ma, is a data scientist in Alexa Shopping. Often during their conversations, Minghui was brimming with questions related to the US retail landscape.
Thayer said Minghui’s curiosity and initiative were part of what made her a successful intern.
“She is fearless,” he said. “The best interns don’t know whether something is a hard problem or a simple problem — that doesn’t really matter. It’s more about, ‘You asked me to go check this out. I’m going to give it a shot and come back with what I’ve learned.’”
Minghui did such a great job that Thayer extended her summer internship for one month and then hired her full time. After completing her PhD, she joined Thayer’s team at Amazon in March. As a research scientist on a team comprising mostly ethnographers, one of the pluses of Minghui’s role is interacting with user researchers, product managers, and designers. That kind of cross-disciplinary work enhances her learning experience.
Minghui inspired the team to realize that their ideas matter too.
Recently, Minghui mentioned to Thayer that she had a few research ideas she wanted to pursue. Thayer encouraged her to bring them to the whole team for discussion, so she wrote up a summary of the different concepts she had in mind. The process of doing that spurred her colleagues to think about what they wanted to propose, too — and Minghui’s brief ended up becoming a template for her team.
“She inspired the team to realize that their ideas matter too,” Thayer said. For her part, Minghui said this openness to new thinking is one of the qualities she appreciates most about working at Amazon.
Additionally, because Minghui’s internship occurred during the coronavirus pandemic, she and Thayer have yet to meet in person. Still based in Delaware for now, Minghui plans to eventually move to Palo Alto, Calif., where the rest of the team is based.
Her advice to other Amazon interns interested in pursuing a full-time role at the company: First, talk a lot with your manager about your ideas and goals — their experience, she said, can help you shape informal questions into real projects.
And the other piece of advice? “Think a lot,” she said. “And think big.”
