Expert Data Scientist – Ad Fraud & Attribution

Protecting our advertising ecosystem through analytical depth and rigorous modeling

How do you make our customers happy?

By ensuring advertisers reach real customers and receive clear, fair value from their campaigns. You do this by analyzing large-scale behavioral data, identifying sophisticated ad fraud patterns, and improving attribution accuracy through statistical and econometric modeling.

Your analytical work directly strengthens trust in our entire marketing & advertising ecosystem.

The biggest challenge

As an Expert Data Scientist – Ad Fraud & Attribution, you set the methodological direction for fraud detection, attribution and adjacent analytical domains, providing the quantitative rigor needed to protect our platform and refine how we measure value. Fraudsters evolve quickly, and attribution signals are influenced by noise and bias. You help us stay ahead by combining hands-on analytics, statistical thinking, econometrics, and machine learning.

You will independently investigate problems, extract and analyze data using SQL/BigQuery, and translate complex patterns into insights that guide product, engineering, and business decisions.

Fraud detection and attribution modeling require understanding messy, adversarial, and biased datasets. You’ll need to:

• separate genuine customer behavior from malicious or noisy signals
• quantify uncertainties, biases, and econometric effects
• validate hypotheses with rigorous statistical reasoning
• design models robust to adversarial adaptation

This role requires someone who is both analytically independent and capable of bridging the gap across teams, who builds long-term partnerships across the product group and provides advice and influence beyond content-driven arguments.

What you will do

• Analyze platform traffic with SQL/BigQuery to uncover emerging fraud patterns
• Validate suspicious signals, estimate false positives, and measure economic impact
• Build statistical and ML-based fraud detection models
• Develop adaptive anomaly detection systems
• Investigate attack surfaces, bot behavior, and suspicious cohorts
• Improve attribution logic using statistical and econometric methods
• Model and correct biases (e.g., position bias) in attribution pipelines
• Communicate insights clearly to product, engineering, and leadership
• Proactively reach out across the organization to close analytical gaps
• Drive methodological improvements across fraud, attribution and adjacent analytical domains
• Lead AI-focused experimentation and tackle ambiguous problems requiring AI-native thinking
• Improve DS tooling and processes across teams and introduce new methods where needed

Why you can make a difference

You bring depth in statistics, econometrics, machine learning, and analytical investigation.

You are energized by exploring ambiguous or messy data, reasoning economically about value, uncertainty, and incentives, separating signal from noise and independently diving into large datasets to uncover actionable insights.

Your work directly influences platform trust, advertiser value measurement, and detection quality.

3 reasons why this is (not) for you

What helps you succeed

• Ambiguity energizes you: You get energy from solving ambiguous, high-stakes challenges and shaping analytical direction for others.
• You enjoy wrestling data (and usually win): You love diving deep into messy, adversarial data using tools like SQL, BigQuery, and ML frameworks to surface actionable insights.
• You’re a cross-team player: You communicate clearly and influence across teams—engineering, product, business—bringing people together to drive real results.

What may make this role a poor fit

• Uncertainty slows you down: You prefer stable, predictable datasets and aren’t comfortable with analytical uncertainty or behavioral noise.
• Works best with structured questions: You avoid collaborative, iterative investigation and would rather wait for someone else to define the problem fully.
• You work best in your silo: You’re not interested in influencing or guiding cross-functional teams toward the best analytical choices.

Where you'll be working

You’ll join the Reliable product group within Marketing & Advertising, collaborating closely with Engineering, Product, Data, and Analytics teams. You will be accountable for the fraud detection, attribution, ranking fairness and ad quality analytical function within the product group — translating product-group strategy into objectives, sequencing work across teams, and making trade-offs visible when the roadmap impacts adjacent products.

Perks of having a blue heart

• 29 days to recharge
• Travel costs: Public transport, car, parking & charging covered
• Pension plan: 75% premium covered
• Annual bonus based on sustainability goals