Senior Data Scientist - AdTech / MarTech

We are looking for a Senior Data Scientist to join our outcome data pod. You will evaluate third-party consumer, business, panel and outcome data, and turn validated data into features for a household-level intelligence layer that powers audience and measurement use cases.

This is an analysis-led role. You will test what a data provider s data actually delivers, using real samples, rather than relying on the provider s reported numbers. You need a solid grasp of how the consumer and business data ecosystem works, including identity resolution, data licensing and audience data. You will work closely with Data Engineering and report findings that are sourced, reproducible and clear to non-technical stakeholders.

Key Responsibilities:

• Evaluate third-party outcome, panel and video data against real data samples. Measure match rate, label precision, geographic distribution, deduplication and bot quality, and viewership precision.
• Design and run match-rate testing to independently validate provider-reported figures, using both deterministic and probabilistic matching approaches.
• Map provider-specific schemas and data dictionaries into a shared feature taxonomy.
• Build feature definitions and data-quality checks for a feature store supporting a household embedding model of roughly 100 features.
• Document data-quality findings in a standard scorecard format, with a clear evidence trail that others can reproduce.
• Identify privacy and consent gaps in third-party data, such as unclear consent for ML training, and escalate them for legal review.
• Partner with Data Engineering to define the validated schema each data pipeline should deliver.
• Present findings and recommendations to business and technical stakeholders. Qualifications

Must Have Skills

• 6+ years of experience in data science or advanced analytics, ideally in adtech, martech, data brokerage or programmatic/real-time platforms.
• Statistical analysis of sample and panel data: match-rate testing, deduplication and bot detection, and geographic (DMA-level) distribution analysis.
• Strong Python (pandas, scikit-learn or equivalent) and SQL.
• Working knowledge of identity resolution: hashed PII, deterministic vs. probabilistic matching, and MAID, HEM and household-ID concepts.
• Experience matching and joining disparate datasets, including third-party consumer and business data.
• Experience evaluating third-party or panel data quality: data dictionary review, sample audits, provenance and overlap checks.
• Feature engineering for downstream ML and embedding use.
• Clear written communication of quantitative findings to non-technical stakeholders.

Good to Have

• Experience with feature stores or embedding-based customer or household representations.
• Exposure to marketing and media measurement, telecom or advertising audience data.
• Familiarity with data clean-room platforms (Snowflake Data Clean Rooms, LiveRamp, Habu, InfoSum).
• Experience with Snowflake or Databricks.
• Working familiarity with CCPA and consent frameworks, enough to recognize gaps rather than adjudicate them.