Data Scientist

Confidential; United States; On-site; USD 120,000 - 190,000; Full time

Job summary

Confidential is seeking a Data Scientist focused on causal inference and marketing measurement to join our Data Science team. You will design models to measure causal impact of marketing activities on brand performance, using MMM, attribution, and Bayesian methods.

The role emphasizes communicating insights to marketing and brand leaders, partnering with Product and Engineering to translate modeling outputs into scalable capabilities. Experience with PyMC/Stan is favored.

Qualifications

• Applied experience or strong academic research in causal inference and marketing measurement.
• Strong grounding in probability theory and experimental design.
• Hands-on Bayesian modeling including priors and posterior evaluation.
• Ability to translate modeling results into actionable insights for non-technical stakeholders.

Responsibilities

• Design, build, and validate causal models measuring the impact of marketing and brand activities.
• Apply MMM, attribution, and other causal inference techniques.
• Quantify uncertainty with credible intervals and communicate findings to stakeholders.
• Collaborate with Product and Engineering to operationalize modeling outputs in products.

Skills

Causal inference

Marketing measurement

Bayesian modeling

Python

Data visualization

Communication

Collaboration

Education

MS/PhD in Statistics/Data Science/Econometrics

Tools

PyMC

Stan

Spark

PySpark

Job description

We are seeking a Data Scientist to join our Data Science team, with a specific focus on causal inference and marketing measurement. This role sits at the intersection of statistical modeling and brand strategy, helping our customers understand what drives brand and marketing performance - not just what correlates with it. The ideal candidate is deeply curious about why marketing works, is comfortable working in Bayesian frameworks, and wants to apply rigorous causal methods to real marketing and brand data. We're looking for someone who is genuinely passionate about brands and marketing, with an intellectual curiosity about what drives consumer behavior - not just someone applying models to whatever dataset happens to be in front of them.

We welcome applicants at two levels of experience:

• Early career: Recent graduate (MS or PhD) with a research or thesis focused on causal inference, econometrics, or Bayesian statistics, eager to apply that training to marketing problems.
• Experienced: 2-5 years of hands-on experience applying causal inference methods - such as marketing mix modeling (MMM), media/channel attribution, or econometric modeling -within marketing, advertising, brand, or consumer analytics.
• Causal & Marketing Measurement Modeling: Design, build, and validate models that measure the causal impact of marketing and brand activities on business outcomes - including marketing mix models (MMM), incrementality testing, and other causal inference approaches (e.g., difference-in-differences, synthetic control, instrumental variables, Bayesian structural time series).
• Bayesian Modeling: Develop and refine Bayesian models (e.g., in PyMC, Stan, or similar probabilistic programming frameworks) to quantify uncertainty, incorporate prior domain knowledge, and produce credible, decision-ready estimates of marketing effectiveness.
• Model Validation & Iteration: Rigorously test model assumptions, perform sensitivity analysis, and iterate on modeling approaches as new data and marketing channels emerge.
• Business Insights and Communication: Serve as the translator between statistical rigor and marketing strategy. Communicate assumptions, causal findings, and their business implications clearly to both technical and non-technical stakeholders, including marketing and brand leaders.
• Cross-Functional Collaboration: Partner with Product and Engineering teams to help translate data science solutions into scalable, agentic product features - working closely with those teams as they operationalize and "agentify" modeling outputs into automated, product-facing workflows.

Required Qualifications

• Domain-Relevant Experience: Applied experience (or strong academic research) in causal inference and/or marketing measurement - e.g., MMM, attribution modeling, incrementality testing, or econometrics applied to marketing/brand/advertising data. Experience in unrelated domains (healthcare, education, life sciences, etc.) without a marketing/causal inference component is not a fit for this role.
• Educational Foundation: Bachelor's, Master's, or PhD in quantitative fields such as Statistics, Data Science, Economics, Econometrics, Applied Mathematics, or a related quantitative discipline, with strong grounding in causal inference, probability theory, and experimental design.
• Bayesian Fluency: Genuine, hands-on experience with Bayesian modeling - not just familiarity with the term. Comfortable specifying priors, working with posterior distributions, and evaluating model fit and uncertainty.
• Problem-solving and Ownership: Ability to independently scope and execute causal modeling projects that answer real marketing questions, and to communicate the limitations and assumptions of those models honestly.
• Communication Skills: Strong written and verbal communication skills, with a demonstrated ability to translate technical modeling results - assumptions, uncertainty, causal claims - into clear, actionable insights for non-technical stakeholders such as marketing and brand leaders. Comfortable telling a data-driven story, not just presenting output.
• Passion for Brands & Consumer Behavior: A genuine interest in brands, marketing, and advertising, paired with intellectual curiosity about what drives consumer decision-making. Candidates should be motivated by the marketing questions themselves, not just the modeling techniques.
• Programming & Modeling Tools: Strong Python skills specifically for statistical/causal modeling - proficiency with PyMC, Stan, or a comparable probabilistic programming framework is required. General-purpose Python experience (pandas, numpy, scikit-learn, etc.) is also expected.
• Statistical & Causal Methods: Deep, applied knowledge of marketing mix modeling, attribution, and causal inference techniques (e.g., Bayesian structural time series, difference-in-differences, synthetic control, instrumental variables, uplift modeling).
• Data Visualization: Ability to build clear, decision-ready visualizations and reports that communicate model outputs (e.g., channel contribution, ROI curves, credible intervals) to marketing stakeholders.
• Nice to Have: Exposure to marketing/advertising data structures such as media spend, impressions, brand tracking surveys. Deep database engineering, complex query optimization, or Big Data pipeline experience (Spark/PySpark) is not a requirement for this role.

Note: This role is distinct from our Data Engineering and ML Engineering positions. We are not looking for candidates whose primary strength is building data pipelines or general-purpose ML infrastructure - this role is focused on causal modeling and marketing measurement.