Cracking Data Science Careers: Advice from a Data Scientist

I had a great chat with Dawn Choo, a Data Scientist at ClassDojo, about the data science career. We delved into the role of a product data scientist, the transferable skills you can leverage to break into the field, and Dawn's inspiring journey to landing her dream job at Facebook (Meta)!

What is a Data Scientist?

Jean: What is a product data scientist? The term seems to encompass a wide range of responsibilities.

Dawn: You're absolutely right. "Product data scientist" can definitely be a loaded term! It varies a lot between companies and even between teams within the same company. But here's a general idea:

• Product focus: A product data scientist works on a specific product, whether it's something broad like Instagram Stories or something more granular like stickers within that feature.
• Product team: You'll collaborate with product managers, designers, engineers, and other stakeholders to bring data-driven insights to the product development process.
• Impact across the board: Your goal is to help the product team with everything from strategy and user research to understanding the business model and maximizing return on investment (ROI).

Jean: In your previous role as a business analyst, how did your work differ from what you do now as a data scientist?

Dawn: That's a great question, especially since there are so many data-related roles in tech. As a business analyst, my role felt more like a support function. I was embedded in an account management team, providing data and building tools to make their work more efficient.

For example, they might ask me to pull reports on click-through rates for specific advertising campaigns.

Now, as a product data scientist working in education, I'm focused on helping the product team make strategic decisions. Here are some questions I might explore:

• Do we have a good product-market fit with our current user base?
• Are there untapped markets with high potential revenue?
• Where should we invest our resources for the best ROI?

Jean: It sounds like data science skills are transferable across different areas. Can you elaborate on that?

Dawn: Absolutely! One of the best things about data science is the transferability of skills. You can start in one area, like business analysis, and then transition to another, like data science, if your interests shift.

For example, many data scientists come from backgrounds in data engineering, product management, or even other areas. Everyone is figuring out their path based on their current priorities and interests.

Jean: How much statistics and machine learning do I need to become a data scientist?

Dawn: It depends on your focus. Product data scientists, like myself, leverage statistics but focus more on applying them to business problems. We might be asked about linear or logistic regression, but the emphasis is on practical application, not deep theory. 

Machine learning expertise varies, but most product data science interviews won't delve into super complex algorithms.

‍

Dawn Cracks the Facebook Interview Code

Jean: You mentioned getting rejected by Facebook four times before landing the offer. What was your turning point?

Dawn: It all came down to targeted improvement. After each rejection, I identified my weaknesses. Product sense and translating statistics into actionable business insights were my biggest gaps.

To bridge the product sense gap:

• Shadowed product managers: I learned their thought processes through one-on-one interactions.
• Managed personal projects: Applied product strategy principles to my own endeavors.
• Studied for product manager interviews: Gained a deeper understanding of product strategy.

For statistics communication:

• Worked with account managers: Learned about client-side experimentation.
• Studied and practiced: I honed my ability to explain statistical concepts in a business context.
• Developed frameworks: Structured approaches helped me answer interview questions confidently and communicate effectively.

Jean: That's impressive! This reminds me of software engineering interviews. You can learn to interview well, even if you lack specific experience.

‍

Breaking into Data Science

Jean: What advice would you give aspiring data scientists?

Dawn: Here are four key areas to focus on:

1. Coding Skills: Master SQL first, and consider expanding to Python.
2. Experimentation & Statistics: Understand core concepts like hypothesis testing and statistical distributions.
3. Product Sense: Learn how to translate data into actionable insights for product teams and practice articulating it. Read "Cracking the Product Manager Interview" and talk to product managers if possible.
4. Behavioral Interviewing: Be prepared to answer behavioral questions. You should be 100% prepared for this no matter what interview you go to because you CAN 100% prepare for it. 

‍

Summary:

• Product Data Scientists typically focus on specific products. They work collaboratively within the product team, and use data to guide product strategy.
• Skills gained in business analysis or other data roles can be a stepping stone to data science. 
• Product data science interviews emphasize applying statistics to business problems, not deep theory or complex algorithms. 
• When faced with interview rejections, identify weaknesses and target improvement through learning and practice.
• To build “product sense, shadow product managers, manage work projects, and study for product manager interviews. 
• Similar to software engineering interviews, data science interview skills can be learned and honed even if you lack industry experience. 

‍