Three Essays on Marketing Analytics in the Age of Generative AI
Hinweis: Auch kostenlose digitale Publikationen wie Studien, Dissertationen und Zeitschriftenartikel müssen zuerst in den Warenkorb gelegt werden. Nach Eingabe Ihrer Kontaktdaten können Sie die Dokumente direkt herunterladen und erhalten zusätzlich einen Download-Link per E-Mail.
Author: Christopher Marc Schraml
Language: English (nur als PDF-Datei erhältlich)
This dissertation advances marketing analytics through three essays that explore how unstructured video data and structured survey data on customer journeys can enhance marketing decisions and performance. The essays focus on automated video analytics and established marketing analytics approaches to better understand and influence consumer behavior.
Essay I closes a methodological gap in marketing research by presenting the first systematic literature review on automated video analytics. The literature review outlines extractable features across text, audio, and visual modalities and the methods and tools used to extract them. Further, this essay introduces a scalable, cost-effective zero-shot video analysis method based on multimodal large language models, demonstrating that ChatGPT-4o can achieve human-level coding accuracy.
Essay II investigates how product presentation styles in shoppable video clips affect consumer behavior on e-commerce sites. Two experiments and a field study reveal that videos depicting influencers using (vs. displaying) products significantly boost purchase intentions and add-to-cart rates, especially for experience products, strong brands, and in the upper stages of the purchase funnel. The study applies the method from Essay I to code the product presentation styles in thousands of videos automatically and links it to behavioral metrics.
Essay III analyzes how the socialness of the pre-purchase customer journey influences customer inspiration and purchase behavior. Analyzing thousands of customer journeys, the study finds that higher socialness of the pre-purchase phase significantly boosts customer inspiration, but only when the touchpoints used are primarily online. Customers who feel more inspired through social online touchpoints spend more and make unplanned purchases more frequently.
Together, the essays show how generative Artificial Intelligence enhances video analytics (Essay I), improves marketing performance (Essay II), and how traditional marketing analytics supports the design of customer journeys in order to shape early-stage customer behavior (Essay III). This dissertation offers actionable insights for improving video marketing, influencer strategies and customer journey design to increase marketing performance and drive business success.
Essay I closes a methodological gap in marketing research by presenting the first systematic literature review on automated video analytics. The literature review outlines extractable features across text, audio, and visual modalities and the methods and tools used to extract them. Further, this essay introduces a scalable, cost-effective zero-shot video analysis method based on multimodal large language models, demonstrating that ChatGPT-4o can achieve human-level coding accuracy.
Essay II investigates how product presentation styles in shoppable video clips affect consumer behavior on e-commerce sites. Two experiments and a field study reveal that videos depicting influencers using (vs. displaying) products significantly boost purchase intentions and add-to-cart rates, especially for experience products, strong brands, and in the upper stages of the purchase funnel. The study applies the method from Essay I to code the product presentation styles in thousands of videos automatically and links it to behavioral metrics.
Essay III analyzes how the socialness of the pre-purchase customer journey influences customer inspiration and purchase behavior. Analyzing thousands of customer journeys, the study finds that higher socialness of the pre-purchase phase significantly boosts customer inspiration, but only when the touchpoints used are primarily online. Customers who feel more inspired through social online touchpoints spend more and make unplanned purchases more frequently.
Together, the essays show how generative Artificial Intelligence enhances video analytics (Essay I), improves marketing performance (Essay II), and how traditional marketing analytics supports the design of customer journeys in order to shape early-stage customer behavior (Essay III). This dissertation offers actionable insights for improving video marketing, influencer strategies and customer journey design to increase marketing performance and drive business success.