US 12,579,552 Granted · Mar 2026

Systems and Methods for Machine Learning Modeling of Mixed Media Marketing

An ML system that measures the conversion impact of 20+ advertising channels — TV, social, search, email, and more — and optimizes budget allocation across channel-campaign combinations. Combines a multiplicative regression model for impression-to-conversion attribution with a Lagrangian constrained optimization engine, enabling marketing stakeholders to maximize sales conversions under real-world spend constraints.