Staff Machine Learning Engineer - Acquisition Paid Search

StubHub is on a mission to redefine the live event experience on a global scale. Whether someone is looking to attend their first event or their hundredth, we’re here to delight them all the way from the moment they start looking for a ticket until they step through the gate. The same goes for our sellers. From fans selling a single ticket to the promoters of a worldwide stadium tour, we want StubHub to be the safest, most convenient way to offer a ticket to the millions of fans who browse our platform around the world.

About the Opportunity

We’re seeking a Staff Machine Learning Engineer to lead the science and systems that power our paid search marketing. You’ll design the causal measurement stack, ship models that influence bidding and budgeting in real time, and partner with marketing, data, and platform teams to drive profitable, incremental growth.

StubHub is the largest secondary ticket market in the world, generating massive amounts of consumer data that are leveraged to tackle many unique and interesting predictive and inference problems across user acquisition, product recommendations, pricing optimization, ticket fulfillment mitigation, and business forecasting. The core challenge for our marketing efforts is acquire as many new customers as possible, efficiently, and at the right time in their customer journey, making it a complex and highly impactful domain.

Location: Hybrid (3 days in office/2 days remote) – New York, NY or Los Angeles, CA    What You'll Do:

• Own causal measurement for paid search: Stand up uplift/incrementality frameworks (e.g., doubly robust learners, causal forests, DML, IVs, synthetic control, DiD, BSTS) to quantify lift beyond correlation.
• Ship production models: Build and serve models that inform bids, budgets, and query-level targeting using signals like incremental CPA, tROAS, LTV, and heterogenous treatment effects.
• Design experiments & guardrails: Architect geo/cell tests and online experiments; handle power analysis, pre-trend checks, SUTVA threats, SRM detection, and sequential monitoring.
• Integrate with ad platforms: Translate science into APIs/feeds for Google Ads, Microsoft Advertising, and SA360; validate against auction dynamics and Quality Score mechanics.
• Data & MLOps leadership: Partner with platform teams to instrument events, build reliable feature stores and ETL (batch/stream), and establish monitoring for drift, bias, leakage, and attribution sanity.
• Mentor & influence: Provide technical leadership across science, engineering, and marketing; set standards for methodology, code quality, documentation, and reproducibility.
• Tell the story: Communicate trade-offs and impact to execs and non-technical partners; make the complex understandable and actionable.

What You've Done:

• 8+ years in applied ML/causal inference (or equivalent) with direct paid search/auction experience.
• Expert in causal methods (uplift modeling, DML, IV, DiD/synth control, BSTS/Bayesian time series) and experimental design.
• Strong software engineering: Python (pandas, numpy, scikit-learn, LightGBM/XGBoost), SQL; experience with Spark and one of AWS/GCP/Azure.
• Hands-on with A/B frameworks, power analysis, and measurement diagnostics (SRM, balance, interference).
• Proven track record integrating with Google Ads/Microsoft Ads/SA360 and moving the needle on tROAS, CPA, LTV.
• Clear communicator who can mentor senior ICs and partner with product/marketing.

Nice to Have:

• Recsys, bandits/RL for bidding/budget pacing, MMM and privacy-aware attribution
• Scala/Java or microservices experience; Airflow/DBT; Kafka/PubSub; Feast or similar feature stores.
• Domain knowledge of auction theory, query taxonomy, brand vs. non-brand dynamics, and budget rebalancing

Staff-Level Capabilities 

• Technical leadership through influence rather than formal management authority 
• Strategic thinking with the ability to balance long-term technical vision with immediate organizational needs 
• Cross-functional collaborati