Jeff Parks
Marketing measurement: MMM, incrementality, forecasting & causal inference, and evaluation systems for applied AI. Selected work, experience, and contact.
10+ yrs$500M+ media spendMMM, incrementality, causal inference
About
I build the causal infrastructure behind commercial decisions across consumer and B2B lines, and I lead the teams that ship it into production. Ten years turning behavioral signal into GTM action: marketing mix modeling, audience and account segmentation, production propensity and recommendation models run at consumer scale.
Recent work spans geo-experiment design, synthetic-control validation, and self-serve tooling so marketing partners can answer routine questions without a data scientist in the loop.
Selected Work
Revenue forecasting / backtesting harness for multi-channel paid media — a rolling backtest notebook plus a Streamlit diagnostics viewer.
Time-series forecasting harness for future media-cost assumptions — an essential MMM input, same proof-of-concept family as above.
CUNY capstone identifying systemic constraints limiting X/Twitter Community Notes interventions at platform scale during the 2024 US election.
A git workflow assistant that integrates an LLM — automated commit messages, PR-review suggestions, and README generation from commit history.
Adstock transformation modeling for time-decay effects of media on consumer response, with a flexible half-life parameter.
EleutherAI Summer of Open AI Research — a knowledge-graph pipeline extracting concepts and interventions from 11,800 AI-safety papers. One of the project's top contributors.
Experience
Leading a 5-person data science & measurement team building ML-powered revenue forecasting and cross-channel performance measurement for a top retail brand.
Advanced measurement and agentic solutions for clients in the media and higher education sectors.
Led a 4-person consumer analytics team; owned model deployment and evaluation across enterprise personalization channels.
Led analytics transformation for a $500M+ portfolio (Lilly, TikTok, Lululemon); built production ML audience models and data infrastructure from the ground up.
Turned around a $100M data science engagement for L'Oréal USA across 30+ brands on a distributed Spark/Databricks stack.
Built and scaled an analytics practice supporting $50M+ in annual media spend across healthcare, retail, and technology clients.
Founded and scaled the agency's first analytics practice from zero to $5M+ in annual revenue.