Predictive Engagement is EDO’s proprietary behavioral measure of consumers’ brand engagement in the moments following a TV ad airing. Engagement is a leading indicator of market share, and an indication of consumer discovery, interest, and intent.
When TV viewers see content that interests or excites them, they pick up a second device and to get more information. EDO captures this consumer response to TV advertising by measuring these engagements in real-time, analyzing the granular data that precisely attribute consumers’ online behavior and content consumption to individual TV ad airings.
EDO’s data modeling approach provides a comprehensive view across linear and streaming television, showing marketers what works to drive consumers to action and what doesn’t. With a unique combination of intent signals (Search and Site Visitation) on a per exposure basis, EDO accomplishes one-to-one matching.
First, EDO captures Convergent TV ad occurrence data via Smart TV or First-Party data sharing. Then, we capture outcomes data — individual-level data for search and site visit activity for all brands. Next we match occurrences to outcomes in an encrypted, privacy-safe method. Finally - an advanced data model is executed to identify incremental search and site activity and contribution of audience, creative, media.
EDO uses a Propensity Model and a Lift Model to attribute credit to individual ad exposures. For a given ad exposure, its attributed credit represents the incremental impact it had in driving behavior. EDO’s Propensity Model measures the baseline probability that a behavior happens without any ad exposure. Because this probability is not incremental, it is separated and is not attributed to any individual ad exposure. EDO’s Lift Model then uses algorithmic attribution to attribute credit to individual ad exposures. Different ad exposures will receive different credit based on their estimated impact on driving incremental behaviors.