Emotional Congruence at Scale: Using Databricks AI to Align Ads and Content

TV advertisers know context matters, but most contextual signals stop at coarse attributes like genre or keywords. What about emotion? What we watch shapes how we feel—and how we feel influences how we engage with brands.At NBCUniversal, we built an Emotional Congruence engine to quantify whether alignment between these emotional signals drives intent and conversion—and how to scale it using Databricks.Using foundation models served on Databricks, we extract and summarize unstructured signals from content and ads, tagging each with a standardized emotional taxonomy across themes, tones and personal values. Governed end-to-end with MLflow 3.0, the system is orchestrated for rigorous testing and deployment.Testing shows emotional alignment drives measurable lift—often in the double digits. As a result, it is now embedded within NBCUniversal’s broader contextual advertising offering.