Why Nanneh Chehras is committed to being a mentor
The senior economist knows what it means to pursue a career path like hers, and she’s determined to help others along the way.
Nanneh Chehras is a senior economist with a reputation for being an advocate. In her four years at Amazon, she’s become someone who advocates for her colleagues, mentors the next generation, and invests in women across a space where there are few. She’s also moved from data science to economics along the way, always sifting for the story within the data.
Chehras was born, raised and educated all within what she thinks of as her California triangle. Her home was in Burbank in Los Angeles County — her parents settled there after they immigrated because of the large Armenian community. From there she got her undergraduate degree at UCLA and then went to graduate school at UC Irvine. “And all those places essentially make a little triangle with about an hour drive between each point,” she says.
She accepted a job with Amazon in 2017 just before finishing her PhD. It was a data science role on Amazon’s delivery team. The team is part of Amazon’s larger Supply Chain Optimization Technologies (SCOT) organization, and her initial focus was on causal analysis of how certain factors — such as a tweak in the process or launching something new — might influence delivery time.
Chehras was drawn to data science in graduate school. Her PhD research focused on labor and education issues, but she always valued the role of data in decision making.
“I enjoyed working with novel datasets from both a curiosity perspective, but also because I often had to grow my skill set to work with them,” she said.
From data science to economics
Three months into her time at Amazon, Chehras transferred to her current job as an economist — a role she said was a better fit for her skill set. And while the title was different, in many ways her work was the same: taking unruly datasets and turning them into easy-to-explain business recommendations.
In reality the lines between data science, economics, and research science are blurred.
“In reality the lines between data science, economics, and research science are blurred,” she said. The key in all these roles, Chehras observed, is merging the time-consuming research process with changing business demands. The researcher has to be skilled enough to identify caveats in their approach and then decide how to proceed. “We need the judgement to say, ‘This gets you 80% there today’ while at other times we have to insist on longer timelines to improve the methodology,” she said.
Many of the tools she uses as part of the delivery team are fundamentals she used in grad school: different nuances but similar approaches. For example, she still uses causal econometric methodologies like difference-in-differences, propensity score matching, and more.
For those who are interested, economics holds a lot of possibilities, but Chehras cautioned the bar to entry can be high.
“For me it required a PhD,” she said. “So, it takes a while to get to the fun stuff.” To those considering a similar path, she advises careful thought about what they want their future to look like before they pursue graduate school.
Why mentoring matters
When she was finishing graduate school in 2017, the concept of economists within tech companies was still emerging. “It seemed like a black box. I wasn’t sure what I was entering,” Chehras said. Now, she’s intentional about being available to young people with the same dilemma. She’s quick to answer messages on LinkedIn and shed light on what it looks like to be an economist in tech — a field that will only continue to grow.
I think it's really important to be a mentor and I think people can always be a mentor, no matter what stage of your career you're at.
“I think it's really important to be a mentor and I think people can always be a mentor, no matter what stage of your career you're at,” Chehras said. The key, as she sees it, is to simply be available. “Listening and asking questions is often 95% of the value,” she said.
Her availability isn’t limited to young professionals and new recruits — it’s something she’s known for at Amazon. She co-leads a chapter of the Women in Engineering affinity group at Amazon that works to promote and foster community among women engineers.
Her co-lead Julie Friend, a senior business intelligence engineer, recalls being at a conference with Chehras before the two women became close friends. Over drinks, Friend mentioned that she had been in her role for a while and needed a change.
Instead of offering general statements of support, Chehras listened and asked questions about Friend’s situation and ambitions.
After the conference, Chehras asked Friend if she could propose a position change to Chehras’ supervisor. “She advocated for me when I didn’t have the guts to do it for myself. And she’s the reason I have this job I really like,” Friend said.
Connecting outside of Amazon
Chehras also works to share her experience and connect with women outside of Amazon. In 2019, she attended the Grace Hopper conference as one of more than 25,000 attendees — over 95% were women. She returned to the 2020 virtual conference as a presenter; she and a fellow Amazon senior economist, Mallory Montgomery, presented two of the causal methodologies they utilize at Amazon.
She advocated for me when I didn’t have the guts to do it for myself. And she’s the reason I have this job I really like.
“Experiments are the gold standard in causal analysis but you can imagine they’re not always feasible,” Chehras said. To demonstrate this, their talk presented a hypothetical: a new mask style for use at a hospital. How might the new mask impact coronavirus cases? It might be more effective at protection, but it might also be more uncomfortable, so fewer people wear it. “You don't know the direction of the impact, but you also can’t run an experiment because you don’t want to restrict mask usage,” Chehras explained.
Chehras and her colleague were able to demonstrate how Amazon does causal analysis in these types of situations. “For example, if Amazon launches a new program in a country our team can use synthetic control to evaluate the impact,” she observed. “We hoped that the audience would feel empowered to tackle causal questions with non-experimental methods.”
After nearly four years at Amazon, Chehras feels empowered given her flexibility to propose and explore new projects with relative ease, lots of data to work with, and colleagues who are willing to help.
“Amazon is like a playground for scientists,” she says. “The experiences and opportunities I’ve had have far exceeded my initial expectations.”