Davis Balaba is a seasoned data professional currently working as Head of data science at Meta’s New Product Experimentation division. There he leads efforts to develop and test new product concepts across multiple verticals that could help build and foster community globally. Previously Davis has worked on AR and VR products for Facebook’s Building8 and as a growth marketing analyst implementing data-driven strategies to engage active users and resurrect churned users of the company.
Together with Tina Huang, he is the author of 365 Data Science’s latest course Data-Driven Business Growth dedicated to the implementation of business development strategies based on data. We caught up with Davis to talk about his data science journey, his work at Meta, and what to expect from his new course.
I'm excited to deep dive with you on your work. Can you give us a bit of background about yourself?
I was born in Uganda. I eventually left for high school at the United World College of the Atlantic in Wales, UK. After that, I relocated to Spain and eventually the US where I earned a Ph.D. in aerospace engineering and an MBA.
My work has spanned multiple verticals ranging from manufacturing through banking and insurance to High Tech eventually.
What is the most exciting part of what you’re doing right now?
The most exciting part of my current job is I get to explore a complex design space of new app ideas that range from fun stuff like collaborative music creation to products that aim to foster equity in the world. This is a rare combination within one job.
Can you think of a specific data science project you are particularly proud of because of the effect it had on the company or society at large?
The Facebook ad algorithm is widely revered. However, at one point in my work, we had the challenge of rapidly identifying optimal ad configurations out of a very large-scale ad catalog, O(10^5). I borrowed an algorithm from my aerospace training used to identify optimal configurations out of large dimensional spaces to try to solve this problem. Unlike aircraft, ad performance decays with time so the best ad today is not necessarily the best one tomorrow. After some customization, we were able to develop a framework that improved cost per conversion by 30% - 50%.
What will we learn from your new course created in collaboration with Tina Huang, and why is it important?
Every company wants to achieve its topline goals. Every company says and sometimes believes they are data-driven. Many data transformation projects fail because either leadership is too scared to invest a lot, or the top-down push starts off with very large requests.
This course is aimed at helping companies take a low-risk strategy to data transformation by gradually developing proof points of the impact that can be delivered through data and then using this impact to invest a commensurate amount.
Additionally, we want companies to develop good data hygiene where “Do you have the data to back this up? or What is your hypothesis?” are part of everyday decision-making conversations. While it is not always possible to get historical data to size and justify a new initiative, it is always possible to be data-driven. Where data is not available, a hypothesis and a plan to validate or invalidate it should be articulated upfront. A company with good data hygiene has few or no cases of teams fighting over their share of impact because each team’s contribution is well-measured and defensible. Your data team will let out a collective woosah when this state is attained as they are usually caught in the middle when this debate arises.
And with that, we conclude our interview with the inspiring Davis Balaba. We thank him for his time and look forward to more insights from him in the future. If our talk with Davis has got you excited about the enormous possibilities of data science for business, check out our new Data-Driven Business Growth course where he and Tina share data strategies from their FAANG professional experience. Sign up below to access the first few lectures for free!
