Like many of the Tech giants, Facebook, leverages data to the fullest extent in all parts of its business. This provides many on-going opportunities for those looking to get hired into a Data Scientist role at Facebook! This blog post will take you through the role, required skills, the interview process, a few example questions, and some tips to help you start prepping for your interview.
Data Scientist roles are embedded throughout Facebook, affording many opportunities. Below are a few types of data scientists Facebook hires for:
Financial Data Scientists - Folks in this role ensure that Facebook is spending and making money for the business according to plan. Financial Data Scientists help build models to understand what revenue and spending will look like in the years to come.
Product Data Scientists - These data scientists work on understanding how users interact with the different Facebook products and provide analysis to help improve products and user experience.
Community Operations and Business Integrity Data Scientists - Folks here ensure Facebook’s content-focused products (such as Ads, Facebook Posts, Marketplace Posts) have content compliant with Facebook’s policies. These teams work heavily with Engineering to build/improve machine learning models which help proactively detect non-compliant content.
An average Facebook Data Scientist typically has 3-6 years of experience in a quantitative role. Other qualifications include:
Degree in a technical program such as Mathematics, Statistics, Computer Science, Physics, Engineering, Operations Research, Economics, Engineering
Experience with SQL
Experience with applied statistics
Experience with quantitative analysis
Experience with relational databases and business intelligence tools
Experience with scripting languages (e.g. Python, Java)
Experience with stats software such as Numpy/Pandas (Colab Notebooks are a great way to practice this skill)
Experience working with cross-functional teams
These skills vary from role to role, so be sure to read into the specifics for a particular role.
The typical interview process at Facebook follows the steps below:
Initial screening by a recruiter (~30 minutes). This is more of a formality to ensure the candidate fits the criteria for the role. The recruiter will ask about your past experiences, your interests, and will provide you more details about the role.
Initial phone interview with a member of the current team (45 minutes). This interview will again typically cover past experiences, but with more of a focus on technical skills and problem solving ability. It’s important to use stories / past experiences that highlight your technical expertise here.
Onsite interview. These interviews will be with members of the current team (it's likely one of interviews will be with the hiring manager) and cross-functional team members. The interviews will take about 1/2 a day - interviewing with 3-4 people for about 45 minutes each. These interviews can be structured in many different ways depending on what team you’re interviewing for (your recruiter will provide you with the specifics when you reach this stage), but generally expect 1-2 technical interviews and the remaining be case-like/behavioral interviews.
Example Facebook Data Scientist Interview Questions
What are the challenges you faced in a previous project and how did you resolve it?
Write pseudo code for a dot product of two sparse vectors, ensuring it’s time and space efficient.
How would you measure the effectiveness of Facebook Lite?
How would you measure the effectiveness of Facebook Marketplace?
How do you work with cross-functional teams? Can you give me an example?
What is a F-test? What is a T-test? How are these different?
How does maximum likelihood relate to OLS? Under what conditions are they the same?
If you are the owner of a flower shop, how would you target your customers online? (who? when? where? how?)
How would you measure customer check out rates? How would you go about doing this?
How would you measure how much users liked videos?
Explain confidence intervals to someone with no statistics background.
Tips to help prepare for your interview
Come prepared with 6-7 stories that cover the following:
A time you disagree with a team member, how did you deal with this?
A time where a client had unreasonable demands, how did you respond?
A time where you took on a leadership role formally or informally.
A time where you took a risk, what made the risk notable?
A time where you worked on a particularly difficult project. What made the project difficult?
An example of a time you were leading a group and it wasn’t going well, how did you turn it around?
Interested in practicing for data scientist or analyst interviews?
We send 3 practice questions each week to thousands of data scientists and analysts preparing for interviews (or just keeping their skills sharp). You can sign up to receive the questions for free on our home page.