Staff Data Scientist

Google is seeking a Staff Data Scientist to join the Ads Metrics team onsite in Zürich, ZH. The Ads Metrics group is the core data science engine behind Google Search Ads, building the frameworks, metrics, and libraries that power the next generation of advertising products. This is a role for someone who loves turning data into decisions, designing rigorous experiments, and communicating findings to leadership.

Based in Zürich, the position sits on site and partners with software engineers, product managers, researchers, and analysts to push the boundaries of experiment design, causal inference, and time-series analysis for launches that matter to billions of users. The work sits at the heart of Google Ads, a multi-billion-dollar global business that relies on AI-powered models and Generative AI capabilities such as Gemini to deliver value to advertisers and users alike.

Turning questions into rigorous experiments in ads metrics

In this role, you’ll translate business questions into robust experiments, evaluation metrics, and mathematical frameworks, collaborating closely with cross-functional partners to frame the right problems and the right solutions. You’ll architect end-to-end analysis pipelines that convert ambiguous questions into rigorous analyses, and you’ll choose the analytical approaches that best illuminate the path forward.

You’ll design and assess complex experiments or models, such as causal inference approaches or hierarchical structures, to tackle questions in retrieval and ranking, often with limited precedent. Translating system telemetry into statistically sound evidence is a core part of the job, ensuring research-grade analysis that can unblock billion-dollar launches.

What you’ll need to bring and what would help you stand out

The minimum qualifications mix formal education with hands-on practice. A master’s degree in statistics, data science, mathematics, physics, economics, operations research, engineering, or a closely related quantitative field, or an equivalent level of experience, is required. You’ll also need either eight years of analytics experience solving product or business problems with coding in Python, R, or SQL and querying databases, or six years if you hold a PhD.

• A master’s or equivalent in a quantitative field (statistics, data science, mathematics, physics, economics, operations research, engineering) or comparable practical experience
• Eight years of analytics work solving product or business problems with programming in Python, R, or SQL, or six years with a PhD

• PhD in a quantitative discipline such as statistics, operations research, economics, computational biology, computer science, mathematics, physics, or engineering
• Experience with statistical software (Python, R), database languages (SQL), and agentic development tools like Gemini or Antigravity
• Proven ability to perform statistical data analysis and to design experiments
• Deep grounding in modern statistical theory, including experiment design, regression, causal inference, sampling methods, time-series analysis, and hierarchical modeling
• Strong track record of scientific communication and the ability to distill investigative findings for executive stakeholders
• Solid data intuition and business sense, with experience analyzing data to solve business problems in complex, fast-moving, and ambiguous environments

Tips for standing out as a candidate

Lead your resume with the must-have items: show explicit experience solving product or business problems, and highlight eight years of analytics work or six years with a PhD, plus concrete examples where you used Python, R, or SQL to drive insights.

Demonstrate the must-have skills with real projects. Include end-to-end pipelines you built, your approach to translating business questions into experiments, and any work involving causal inference, time-series, or hierarchical models. If you collaborated with engineers, PMs, researchers, or analysts, spell out what you delivered and how your work influenced launches.

Prepare for interviews by talking through how you would translate a business question into an experimental design, the metrics you would use, and how you would validate findings with telemetry data. Be ready to discuss your approach to data integrity and how you ensure results are robust enough to support high-stakes decisions.

Ask a smart question about how theAds Metrics team collaborates with Google Research and DeepMind, and how generative AI tools like Gemini are actually used in practice to improve measurements, models, or evaluation. A good question will show you’re thinking about both the technical and organizational side of big, fast-moving advertising problems.