Our paper titled "Expectations and beyond: The nexus of AI instrumentality and brand credibility in voice assistant retention using extended expectation-confirmation model" has just been published in the Journal of Consumer Behaviour (ABDC A/ SCOPUS/ WOS/ AJG 2).
My co-authors are Ezlika Ghazali and Na Kai Lun (Universiti Malaya).
Abstract:
AI-based voice assistants (AIVA) are capable of interpreting human speech and responding with useful information, aiding with tasks, and controlling other devices. The usage of these AIVAs has grown significantly worldwide. Despite this growth, studies on user behavior related to continued usage intention of AIVAs and effects on the long-term commercial sustainability of brands, remain low. What is less understood is the potential of AI instrumentality attributes and brand credibility components in provoking shifts in post-use behavior of AIVAs. This study proposes a model which expands on the Expectation-Confirmation Model for user continuance behavior of AIVAs, by integrating user's technology related traits, AI instrumentality attributes and brand credibility. To verify the research hypotheses, the study employed partial least square—structural equation modelling, based on 281 validated responses of a survey. This study highlights the significance of Optimism, Innovativeness, and Discomfort for post-adoption confirmation. Higher post-adoption confirmation is strongly associated with perceived intelligence, anthropomorphism, information quality, and system quality. Anthropomorphism and information quality are key factors for brand expertise, while anthropomorphism and system quality are significant for brand trustworthiness. The study confirms that brand expertise and trustworthiness lead to post-use satisfaction and use continuance intention. Understanding the antecedents of satisfaction and continuance intention extends the existing literature on AIVAs and provides valuable insights for academics and practitioners alike. Some implications for researchers and managers are discussed.
Here are the key findings:
🧠Users' Tendencies and Expectations: Users with higher optimism, innovativeness, and lower discomfort have confirmed expectations after using AIVAs, enhancing perceptions of AI-based factors.
📊 Demographics and Perception: Younger, tech-savvy users tend to be less skeptical, not significantly correlating insecurity with confirmation (79.4% being 34 or younger).
👥 Anthropomorphism's Role: Perceived human-like qualities in AIVAs foster connection and familiarity, driving brand trustworthiness and expertise.
ℹ️ Information Quality and Brand Expertise: Accurate and complete information shapes brand expertise but not trustworthiness.
🎛️ System Quality's Impact: A user-friendly system enhances trust in AIVAs, but may not contribute to brand expertise perceptions.
🧩 Intelligence Not a Major Driver: Intelligence is valuable but not a primary factor in brand credibility; focus on anthropomorphism, system quality, and information quality is more critical.
🎗️ Theoretical and Managerial Implications: Insights into factors that influence the adoption and usage of AIVAs, with strategies for product developers and marketers to enhance user confirmation, satisfaction, and continued use.
