Marketing Data Science Manager

1. Core Technical Expertise (Non-negotiable)

Causal & Predictive Analytics

• Strong in lift measurement, incrementality, and experimental design.
• Able to frame business problems using appropriate methods (e.g., Bayesian, regression, uplift modeling, time-series forecasting, optimization, casual ML).
• Ability to quantify uncertainty and defend ROI confidence intervals.
• Strong experimentation expertise: geo tests, lift tests, synthetic controls, DiD.

Marketing Mix Modeling (MMM)

• Has built MMM end-to-end and have deep understanding on marketing media optimization, not just contributed.
• Experience scaling MMM across multiple regions/products.
• Experience presenting MMM outputs to executives and adjusting the model based on business feedback.

Search / tROAS / Value-Based Bidding

• Hands-on experience building or maintaining:
• tROAS models, geo experiments, incrementality tests
• Media spend optimizers and bid strategies
• Understands channel behavior and measurement nuances.

2. Business & Stakeholder Communication

The candidate must:

• Communicate complex modeling clearly to VP-level, non-technical audiences.
• Translate model outputs into actionable marketing recommendations (e.g., where to shift $20M).
• Defend model assumptions and ROI estimates under pressure.
• Demonstrate ownership of cross-functional stakeholder conversations.

This is where many technical candidates fail. We need someone who thinks like a marketing strategist, not only a modeler.

3. Marketing Domain Knowledge

Candidate must understand:

• Media channels (search, display, paid social, TV, OLV, direct mail)
• Funnel stages, attribution issues, incrementality
• Budget allocation logic and marginal ROI curves
• Creative impact and measurement

This cannot be taught easily—must have hands-on applied experience.

4. Hands-On ML + AWS/MLOps (Good to have)

• 5+ years Python/R/STATS modeling
• Strong with SQL and large data sets
• Deployment experience using:
• AWS SageMaker, Lambda, EC2
• Terraform (nice to have)
• Comfortable owning production pipelines (MMM, optimizers, segmentation, etc.)

5. Leadership & Ownership

• Can lead projects end-to-end with minimal direction.
• Can guide analysts or engineers and review their modeling work.
• Comfortable setting standards for MMM, creative measurement, and search models.
• Can influence marketing, product, and finance partners.