Senior Data Scientist, Ads Experimentation

“The front page of the internet," Reddit brings over 430 million people together each month through their common interests, inviting them to share, vote, comment, and create across thousands of communities. 

We are looking for a Data Scientist to work on our experimentation platform, which has been used by our Product and Engineering teams to run hundreds of experiments in order to ensure that we fully understand our users and create the best possible experiences, as well as to rollout thousands of features in a safe, appropriate way. In 2020, we have begun to grow the experiments team and are quickly turning the platform into a world-class framework, capable of even greater scale and effectiveness. 

As the Data Scientist working on Ads Experimentation at Reddit, you will be collaborating closely with our Data Science team (Ads, Experimentation) and you’ll be pivotal in driving company-wide innovation and rigor, shaping our metrics strategy, and building a strong experimentation culture. The opportunity for impact will be great as you enhance the success and understanding of our products across our diverse and growing user base.

How You’ll Have Impact:

Reddit’s mission is to bring community and belonging to everyone in the world, and you will be pivotal in making that happen. As the 6th largest site on the internet Reddit is still a small/mid-sized company, so you won’t be stuck making small incremental gains on enterprise systems. You will create organization-wide capabilities to run more advanced experiments and create publishable quality research, leverage complex testing and optimization algorithms to grow revenue, and produce more robust results faster to help teams increase their velocity and effectiveness.

What You’ll Learn:

You will partner closely with a world-class org of Engineering, ML, Data Science, Data Engineering, and Analytics practitioners who work together to create a variety of capabilities for Reddit to be more successful in its decision-making and product creation. In particular, you’ll be exposed to and drive adoption of a variety of new techniques in modeling, experimentation, statistical frameworks, and causal inference.

What You’ll Do:

• Create and implement the roadmap for the stats engine of Reddit’s experiment platform. Represent use cases that are specific to ads and enable ads experimentation at scale through the platform.
• Apply statistical fundamentals and enable various statistical tests and measurements in our experimentation platform. 
• Set the standard for experiment design for advertising teams 
• Be the centralized owner for all experiments within Ads. You will sign off experiment specs, readouts and launch decisions by ensuring scientific/statistical correctness in measurements
• Build a framework and platform for running multiple marketplace experiments simultaneously.
• Perform meta-analysis on experiments to understand macro trends in our ads ecosystem
• Create templates/processes and ensure consistency in how different sub-groups within Ads run experiments.

Who You Might Be:

• 5+ years of experience in quantitative or data science roles, preferably for a consumer-facing service/app
• Familiarity with statistical analysis and programming languages (e.g., R / Python)
• Proficiency with relational database (e.g., SQL)
• Proficient in applied statistics and causal inference in an industry setting, including experience conducting and analyzing experiments
• Experience making data-guided recommendations about courses of action, and communicating these to stakeholders with various levels of technical knowledge
• Passionate about statistical fundamentals and translating complex business problems into tractable statistical problems
• Results-oriented with a strong customer and business focus
• Entrepreneurial and self-directed, demonstrated ability to innovate and bias toward action in fast-paced environments
• Ability to communicate and discuss complex topics with technical and non-technical audiences.
• Ability to tackle ambiguous and undefined problems