A type of experimental design (see “Experimental Design“) that compares two different advertising executions in which a random half of all those exposed to an ad sees one version of the ad, while the other random half sees a different version. Responses to each version — measured by a post-exposure survey or by other passive measures of advertising performance (such as click-throughs, conversions, engagement, time spent for digital ads) are compared to determine which version is stronger.
