Henry Mei wasn’t sure what to expect when he interviewed for a data science role at one of San Francisco’s largest tech companies. But he walked away overwhelmed and disheartened, an experience echoed by many data scientists early in their careers.
“The experience was brutally disorienting,” Mei admits. The problem wasn’t so much that he was interacting with tech professionals who’d set out to confuse or mislead him, but rather that the massive scope of this competitive interdisciplinary field left his head spinning.
And yet, Mei found his perfect fit at GoodRx, where he currently serves as senior manager of data science and advanced analytics. We asked this thriving young professional how he prepared for the data science job he wanted, what it was like to navigate interviews at the most daunting companies in his field, and what wisdom from marketing specialist Guy Kawasaki he swears by.
GoodRx is an American healthcare company that operates a telemedicine platform and a free-to-use website and mobile app. These track prescription drug prices in the United States and provide free drug coupons for discounts on medications.
Tell us a little bit about your first experience interviewing for a data science role.
My first interview for a data science role was mostly a fluke. As an engineering major, I had already interviewed for software engineering roles but decided to accelerate my academic career instead. It wasn’t until I became a struggling graduate student that I realized my eclectic background working on disparate research projects that were becoming increasingly more data-driven — and my love for the collaborative spirit within hackathons — was more of a strength than a weakness. Feeling burned out at school, I applied to a few roles in data science and machine learning.
My first interview was with one of the larger tech businesses in San Francisco, and while the format was similar to other engineering-centric interviews (phone screen with Jack, tech assessment with Jane, onsite with James’ product managers and Jessica’s data folk — okay, I’m running out of names that start with J, but you get the idea...), the experience was brutally disorienting.
Data science, as it turns out, is this nebulous term that can mean just about anything and can vary even within a company. Being prepared to know a little about everything means that you end up knowing a lot about nothing.
What is the most important thing you do to prepare for a data science interview, and why?
In my opinion, the single most important thing outside of getting quality experience is to know what you are looking for in a data career — and know whether or not it is compatible with how “data science” is defined within the role and company that you’re interested in. An objectively “right” definition does not exist, beyond what is right for you. Having sat on both sides of the proverbial table, it’s surprising how often there’s a mismatch in expectations, wherein some consider data science to be statistical modeling or machine learning and others believe it involves building and maintaining production data systems, running A/B tests, and building dashboards. To others still, being in data science means being a fluent storyteller to non-technical stakeholders.
What advice do you have for someone preparing for a data science interview at your company?
At GoodRx, we value curiosity, and that means curating diversity in all senses of that word. Like many businesses, we strive for efficiency by using a standardized and primarily AWS-based tool stack, but we try to prioritize aptitude over skills. While we enjoy nerding out on the details of a particular algorithm or maybe a special stopping criteria for Bayesian A/B testing, being able to take a step back and reason out what a solution to a business need looks like is a cornerstone of what successful data scientists at GoodRx do every day. Systems thinking is more important to us than knowing a textbook’s worth of algorithms by heart.
While our data scientists embed into different areas of our business on a roughly quarterly basis, one thing remains constant: we favor rapid iteration, because learning happens by doing. If a series of small steps can get us to the same place in the same time or less, we overwhelmingly prefer that over a complicated solution.
As Guy Kawasaki put it, simply, “The hardest thing about getting started, is getting started.” Indeed, every small bit of progress we make means another American can feel empowered by access to affordable and convenient healthcare.
