A mechanism for modeling statistically significant propensity controls in survey lift studies is disclosed. The proposed mechanism establishes a statistically significant control model for organic viewers in order to measure attitudinal shifts and lift from the viewing population. On a high level, audience overlap is measured to find channels that are similar to the channels in an organic video campaign. Active subscribers of the identified similar channels are then used as the non-exposed group, after filtering out those who watched the organic videos. The resulting channels that share viewer audiences are filtered by channel topicality (e.g., electronics reviews, beauty tips, etc.) and channel size (subscribers within a standard deviation). Viewers are filtered by demographic, technographic and psychographic traits to align with that of the exposed groups.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Mulford, Brian; Du, Yueheng; and Ranger, Colby, "MODELING STATISTICALLY SIGNIFICANT PROPENSITY CONTROLS", Technical Disclosure Commons, (December 30, 2019)