Publication Type

PhD Dissertation

Version

publishedVersion

Publication Date

3-2025

Abstract

As digital platforms increasingly rely on unstructured data such as images, audio, and video to engage users, understanding how these sensory elements influence consumer behavior has become essential. This dissertation investigates the role of visual and auditory cues in shaping user engagement and decision-making in digital environments. Specifically, it examines how image distinctiveness and vocal tone characteristics affect consumer attention, interaction, and commitment across different online contexts.

The first essay explores the impact of image distinctiveness on consumer search behavior within a Southeast Asian real estate platform. Grounded in the Theory of Visual Attention (TVA), this study develops a novel metric for image distinctiveness based on informativeness and contextual similarity. Using validated measures and real-world clickstream data, the findings demonstrate that higher image distinctiveness significantly increases click-through rates (CTR), particularly among new users and those navigating platform-initiated channels.

The second essay investigates the influence of anchors’ vocal tone characteristics on viewer engagement in live-streaming video platforms. Employing advanced audio mining techniques, this study analyzes over 1,000 hours of live-streaming data from a major gaming platform. Results reveal that specific vocal tones—such as expressive energy, cognitive tones, and stress—affect both short-term interactions (chat activity) and long-term engagement (subscriptions). The effects vary according to channel popularity and viewer sentiment, offering nuanced insights for anchors and platform managers seeking to optimize engagement strategies.

Collectively, this dissertation contributes to the sensory marketing and digital engagement literature by demonstrating how visual and auditory stimuli can be strategically leveraged to enhance consumer engagement and platform performance. Methodologically, it introduces novel metrics and robust empirical strategies to address endogeneity concerns, offering practical implications for digital marketers, platform designers, and content creators.

Keywords

Unstructured data, online platforms, audio analysis, live streaming, text analysis, image distinctiveness, visual attention

Degree Awarded

PhD in Business (Marketing)

Discipline

Marketing

Supervisor(s)

CHANDUKALA, Sandeep Rao

First Page

1

Last Page

112

Publisher

Singapore Management University

City or Country

Singapore

Copyright Owner and License

Author

Download

Included in

Marketing Commons

Share

COinS