Career Interview with Happiness Munedzimwe: Product Quality Analyst at Facebook Inc

What is your job title and where do you currently work?

I am a Product Quality Analyst at Facebook Inc., specializing in the Facebook Ads & Business Platform. (Data Scientist/Analyst can be distinct but are often interchangeable in some companies)

What does your firm/ organisation do?

Facebook Inc. is a social media company, with a portfolio of products aimed at meaningful connections among people across the world (Facebook, Whatsapp, Instagram, Oculus, Internet.Org).

Tell us a little bit about your career history?

I started out as an Operations Research Analyst at a start-up company called Zong Inc. which facilitated mobile payments especially around gaming and social media. The company was subsequently acquired by PayPal Inc, an online payments company. I held respective positions at PayPal as Product Analyst and Checkout Product Analytics Manager in the course of 4 years. I then moved to Walmart eCommerce as Manager of Mobile Analytics. Walmart Inc is the largest US retailer with both a physical store presence and an eCommerce presence. I supported the eCommerce side on iOS and Android as well mobile web platforms. I did that for two years before transitioning to Facebook Inc, where I am at present.

Was this always your dream career or it changed somewhere along the line?

I wanted to be a chip design engineer, but when I graduated from my first Masters in 2010 the US was emerging out of an economics depression and finding jobs was quite challenging. So I opted to try out Operations Research/Analytics.

How closely does your academic education fit in with your job?

People in my area of work typically have a technical degree of some sort. It’s only now that academic specializations in analytics and data science are starting to emerge. So I would say my educational path is not unusual.  Increasingly though the “right” path includes Computer Science and Statistics as pre-requisites.

What qualifications do you hold? In short what educational path got you where you are? Could you have made the path shorter?

  • B.S. Integrated Business & Engineering in Electrical Engineering (B.S. IBE EE)
  • Masters, Engineering Management (MEM)
  • Masters, Information and Data Science (MIDS)

Strictly speaking you don’t need a Masters to do analytics/data science if you have a solid programming and statistics background as your first degree. There is intense competition though, and a Masters/PhD qualification will often give you a higher starting salary and more career impacting responsibilities. Shortest path would be pursuing Computer Science with an AI/Machine Learning focus or Statistics/Math with a Computer Science focus, or to pick one for the first degree and do the other for the second.

What are the tasks that you do regularly in your profession?

Analytics is about statistical experiments (A/B testing), hypothesis testing, impact quantification, causal factor analysis, database engineering, data visualization, statistical programming, communication and project management. In short, we use a ton of data to influence business decisions, measure success and to drive product changes and organizational changes. Analytics/Data Science is a decision science. It’s using computer science methods to solve statistical problems with a business decision outcome. The environment for all this is SQL databases and statistical languages like Python/R/Stata/SAS (sometimes hard languages like C++/Java)—increasing Python is the dominant one. PHP is important in some areas. At the end of the day, because the results are given to business audiences, analytics involves a lot of charting and power-point style presentations.

Can you tell us some of the projects you have worked on, which you found interesting?

Most of what I have built can be described as database powered product monitoring systems and business intelligence and reporting platforms.  Some of my projects have internal engineers as the customers, such as monitoring API handoffs and service degradations or business leaders as the customers, such as using data to do market segmentation, mining customer feedback and highlighting underperforming areas.

What is it that excites you the most when you are doing your job?

As an engineer at heart what I will always enjoy the most is designing and building a self-sustaining system that answers an ongoing business need. The other type of work is ad hocs but building systems is more satisfying.

What bits do you find boring in your daily tasks?

A lot of data science is data munging. Real world data doesn’t always come in neat, consistent, easy to utilize packages. You may need to do significant cleaning up, exclusions and imputations before you can apply statistical models.

Any advice to those studying or aiming at this job or career?

Analytics/Data Science is increasingly dominated by Machine Learning/Artificial Intelligence. In the future a lot of analytics will involve building and training Neural Networks. The best preparation is to make sure you are solid in at least one of each from each of these groups: (1) C++/Java/Scala (2) Python or, to an increasingly lesser extent, R. In addition to that you need to know Statistical Methods and Machine Learning. They say a data scientist is a programmer who can do Statistics (or conversely a statistician who can program), so those are the two pillars.  Finally, you will not be effective if you don’t apply good Project Management techniques as often the people who implement data science recommendations are part of other business functions without direct accountability to the recommenders.