Forward Deployed Marketing Data Scientist
Job description
About the Role
Hightouch seeks a Forward Deployed Marketing Data Scientist to serve as a critical bridge between advanced AI Decisioning systems and both customer stakeholders and internal engineering teams. The primary responsibility is to ensure that AI-driven marketing campaigns generate measurable, compounding value for clients. This position requires a unique blend of technical depth and communication skill, as the role holder will diagnose model performance, refine machine learning parameters, measure campaign incrementality, explore complex customer data, and translate intricate model behaviors into clear, actionable insights for marketing and executive audiences.
Marketing teams at Hightouch utilize AI Decisioning to move away from static, weeks-long campaign planning toward dynamic, real-time 1:1 customer engagement that adapts to individual preferences. The Forward Deployed Marketing Data Scientist is essential for verifying that these AI agents function optimally and for helping customers understand the underlying reasons for campaign performance. Approximately 30% of the role involves direct interaction with clients, while 70% is dedicated to deep analytical and modeling tasks. No two days will be identical, yet the core responsibilities will remain consistent and impactful.
Key Responsibilities
The day-to-day work of this role centers on analysis, diagnosis, and strategic insight. You will own the diagnostic processes that clarify why AI Decisioning successfully improves marketing performance across diverse customer environments. This involves explaining performance lift through counterfactual analysis, incrementality breakdowns, and detailed cohort examination. You will actively debug performance issues, iterate on reward functions, and ensure that agent recommendations align precisely with stated customer goals. A further responsibility involves investigating experiment configurations, including send volumes, reachability factors, and channel limitations, to formulate concrete, actionable recommendations.
A significant portion of the role is spent working directly with data in notebooks and customer data warehouses. You will pull historical data to conduct exploratory analysis using Polars and Pandas within Jupyter notebook environments. This work includes modifying and enhancing customer feature matrices to enable deeper personalization and more accurate responses to real-time signals. When insights are not readily apparent in user interfaces, you will perform deeper warehouse-level SQL analysis to uncover underlying trends.
Beyond technical analysis, you will be responsible for building lightweight tooling that creates efficiency and enables scalability across multiple customer engagements. This includes creating reusable templates, notebooks, scripts, and workflows that standardize performance analysis across the client base. You will also identify systemic gaps within current methodologies and influence the future direction of ML reporting and introspection capabilities.
Communication is a core function of this position. You will regularly present model insights and strategic recommendations to marketers, analysts, and executives. This requires explaining complex ML concepts in accessible terms, such as how the decision engine manages cold start scenarios, utilizes message transfer learning, and balances exploration versus exploitation. You will partner closely with Solutions Consultants to identify new opportunities for campaign uplift and transform these insights into repeatable, structured campaign frameworks.
Compensation and Location
-
Location: Remote (North America)
-
Engagement: Full-time
-
Compensation: $130,000 to $180,000 USD per year
Requirements
To succeed in this role, you must possess a strong ability to perform deep exploratory data analysis using Python, specifically with Polars and Pandas libraries within Jupyter notebooks. Proficiency in writing and interpreting complex SQL queries against customer data warehouses is essential for answering nuanced marketing questions. A high-level understanding of core ML modeling concepts is required, including cold start challenges, the exploration versus exploitation dilemma, and message transfer learning principles. You must be adept at explaining model behavior through rigorous counterfactual analysis, incrementality testing, and cohort breakdowns to provide clear reasoning behind campaign results.
Nice to Have
Prior experience working with composable Customer Data Platforms and other marketing data platforms that effectively connect analytics with activation is highly valued.
Skills & Tools
- Composable CDP
- AI Decisioning
- Polars
- Pandas
- Jupyter notebooks
- SQL
Practical Notes
Please What you'll do