Real but deceptive: effectiveness of personalized synthetic advertising for event promotion

Purpose AI’s effectiveness in generating advertising content engaging marginalized sport consumer populations remains underexplored. As an initial empirical effort addressing this gap, this study aims to examine consumer response to event promotion ads that were synthetically modified for a femvertising purpose (i.e. showcasing female-dominant crowds in a male-dominated event) as well as the effect of disclaimers revealing the use of AI and its true intent. Design/methodology/approach A between-subjects experimental study, consisting of three conditions featuring ads (human-created vs synthetic vs synthetic-with-disclaimer) promoting a major professional golf event, was conducted with 175 female participants representative of the target audience of synthetic femvertising. Multivariate analysis of covariance (MANCOVA) was performed to test the hypotheses. Findings When sport involvement was controlled for, synthetic ads were less effective than those created by humans in terms of the target audience’s perceived quality, realism of content, attitudes toward both the ad and the event, interest and intention to attend the event. The negative effect of synthetic femvertising is attributed to the distrust resulting from seeing a false reality, which can be mitigated by including the disclaimer, leading to conative responses (interest and intention to attend) equivalent to those elicited by human-created ads. Originality/value While generative AI enabled sport marketers to develop personalized synthetic advertising, our findings provide valuable implications for future research and practices in sport marketing toward underrepresented consumer populations through cautious and transparent application of generative AI.