An Interview with Steph Pedron

What did you study in your PhD?
I focused on American Politics and Methodology with a minor in Political Psychology. My substantive research specialties were public opinion, race and ethnicity politics, and political behavior. I was especially interested in the unconventional aspects of American political attitudes like conspiratorial thinking and the tendency for individuals to embrace political positions that conflict with their own collective interests.

What is your current position and where do you work?
I’m a Research Data Scientist at Google. I work for the Google Ads organization on the Ads Insight and Measurement team. My team specializes in multi-touch attribution and privacy conversion modeling.

What does your work involve?
On a high-level, I augment multi-touch attribution models used across Google, making sure our suite of measurement tools stay reliable despite large-scale privacy constraints so users can better understand what actually drives ad performance. In practice, that translates into a pretty wide range of work. My day-to-day usually involves juggling multiple projects, from developing and validating new metrics for Google products like YouTube and Search to designing large-scale experiments across countries to test new features. A large part of my time is spent collaborating with engineers and other research and product data scientists, coding, and working with large datasets.

What do you enjoy most about your work?
The people. I’m surrounded by brilliant colleagues who are genuinely collaborative. There’s a steep learning curve because of the number of internal tools and processes involved, but people are very willing to answer questions, share context, and help you ramp up rather than expecting you to figure everything out alone. It creates an environment where you’re constantly learning. I’m the type that gets bored easily, so I also enjoy the pace and variety of the work. The problems tend to be large-scale and challenging, so there’s always something new to dig into. And, of course, the benefits aren’t bad either.

How did you get into this line of work?
I always knew I wanted to transition into industry after my PhD, but the direction became much clearer toward the latter half of my time in the program. While writing my dissertation, I had a lot more time to pursue other things I thought were cool. I started making my own R packages and spending more time on collaborative projects outside of my direct research interests. I found that I enjoy the implementation side of research way more than the theory. That eventually pulled me toward data science and measurement work in tech.

What transferable skills from your graduate studies are you using in your current career?
I would say my methods training (especially the skills I developed one-on-one with certain professors), my ability to quickly and independently learn new things, and my experience managing multiple projects simultaneously. That combo has been useful in fast-moving environments.

How has your background in political science influenced your work?
Political science gave me a strong foundation in thinking about causality and systems with many interacting variables, which in turn shaped how I approach problems involving interpretation under uncertainty and making defensible claims from imperfect data. That way of thinking has carried over directly into my current work in measurement, where a key part of the job is understanding what you can and can’t conclude from the data.

Any advice for those looking to enter your field?
Expand your toolkit early. R is valuable, but Python and SQL are must have skills for research DS. For both R and Python, you should be comfortable writing functions on the fly and iterating on your code based on feedback from reviewers, as this is often part of the live interview process.

After building a foundation in statistics through concepts like regression, hypothesis testing, etc., I recommend focusing on causal inference, machine learning, and communication skills. A chunk of the job is explaining concepts like p-values, bias, or bootstrapping in a way that non-technical colleagues can understand. A good rule of thumb is that if you can’t explain a concept clearly to someone outside your field in 2-3 sentences, it’s worth revisiting until you understand it more deeply yourself. If it’s hard to figure out on your own, you should generally know when to ask for help.

It also helps to have a visible body of work on a website to demonstrate practical experience. Your website should be non-technical for the most part as well because recruiters (who are typically the first ones to see it) aren’t technically trained. In several cases, my website was enough to skip me through the usual recruiter phone call screening step. Finally, while in graduate school, internships, contract work, or other applied research opportunities are valuable both for networking and for gaining practical experience.

Anything else you’d like to share?
The best advice I ever got was to remember how to talk like a normal person. Simplicity and being casually relatable matter just as much as technical depth. I’ve also found that perfection isn’t the goal in most industry settings. “Good enough” is often the right standard. What matters is getting things out, learning from the feedback and incorporating it, then moving onto the next project.