As companies increasingly prioritise the optimisation of business communication to achieve better returns on their advertising investments, and with a growing eagerness to leverage artificial intelligence (AI) for business goals, this paper evaluates the suitability of AI solutions for industry professionals aiming to optimise advertising campaigns. Using a sample of eight static advertisements of varying complexity, the study conducts a series of eye-tracking evaluations with human participants—the ‘gold standard’ for advertising evaluations—alongside AI-supported automated software solutions to generate various metrics for informing design recommendations for promotional content. Following these evaluations, 40 interviews with industry professionals are conducted to assess the comparability of conclusions drawn from the metrics regarding the optimisation of advertising campaigns. The investigation reveals that while AI-supported metrics often appear similar on the surface, this similarity is superficial, which could be particularly misleading when implemented in industry. The inconsistent quality and frequency of misleading results of AI-supported metrics present real concerns that could lead to unnecessary, incorrect, and costly changes to the design of promotional content in the fast-paced environment of advertising practice.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

AI Meets Advertising: Evaluating AI-Supported Automated Eye-Tracking Solutions to Optimise Advertising Campaigns

  • Dennis Olsen

摘要

As companies increasingly prioritise the optimisation of business communication to achieve better returns on their advertising investments, and with a growing eagerness to leverage artificial intelligence (AI) for business goals, this paper evaluates the suitability of AI solutions for industry professionals aiming to optimise advertising campaigns. Using a sample of eight static advertisements of varying complexity, the study conducts a series of eye-tracking evaluations with human participants—the ‘gold standard’ for advertising evaluations—alongside AI-supported automated software solutions to generate various metrics for informing design recommendations for promotional content. Following these evaluations, 40 interviews with industry professionals are conducted to assess the comparability of conclusions drawn from the metrics regarding the optimisation of advertising campaigns. The investigation reveals that while AI-supported metrics often appear similar on the surface, this similarity is superficial, which could be particularly misleading when implemented in industry. The inconsistent quality and frequency of misleading results of AI-supported metrics present real concerns that could lead to unnecessary, incorrect, and costly changes to the design of promotional content in the fast-paced environment of advertising practice.