تحلیل پیشایندها و پیامدهای استریمرهای مجازی در اکوسیستم بازاریابی دیجیتال: فناوری‌های پیشرفته هوش مصنوعی

This qualitative, exploratory study investigates the antecedents and consequences of AI-powered virtual streamers within the digital marketing ecosystem. Given the rapid proliferation of virtual streamers as emerging marketing instruments and their growing influence on consumer behavior, brand image, and sales outcomes, the research addresses a gap in comprehensive, theory-grounded analysis of this phenomenon. Data were collected through semi-structured interviews with 26 digital marketing experts, academics, and experienced users of virtual streaming platforms, selected via snowball sampling until theoretical saturation was reached. Trustworthiness was established through triangulation and member checking. Thematic analysis following the six-phase Braun and Clarke procedure yielded six overarching antecedent themes, being technological advancement and design, interaction and anthropomorphism, effects on consumer and brand behavior, challenges and threats, audience expectations, and success/failure factors. Five overarching consequence themes covering consumer behavior, brand image and loyalty, trust and authenticity, sales and marketing performance, and constraints. Findings indicate that human-like design and personalized, real-time interaction strengthen trust and engagement, while concerns over privacy, authenticity, and expectation violations constrain effectiveness. The study offers a structured antecedent–consequence framework and practical guidance for optimizing virtual-streamer strategies.
Extended Abstract:
Introduction
Virtual streamers, digital avatars controlled by real performers through motion-capture technology, have become influential figures in live-streaming commerce, particularly across East Asia, drawing large audiences through performances that merge entertainment with interactivity. Unlike traditional human streamers, virtual avatars enable interaction that is not tied to a physical body, shaping distinct audience expectations and experiences. Advances in artificial intelligence have further enhanced these avatars' capabilities, allowing more realistic personas, personalized engagement, and data-driven audience targeting, while simultaneously raising concerns about authenticity and data privacy. As virtual streamers increasingly shape purchase intentions and brand perceptions in e-commerce, understanding the interplay between AI, virtual streaming, and digital marketing becomes essential for practitioners and scholars. Despite mounting interest, most prior research has examined isolated variables, such as credibility, presence, or anthropomorphism, rather than offering an integrated account of the full chain of antecedents (from technological design to audience expectations) and consequences (from trust and loyalty to constraints and threats). This gap is especially pronounced in non-Western, developing-market contexts where adoption patterns may differ from those documented in Western or East Asian settings. The present study addresses this gap by identifying and systematically categorizing the antecedents and consequences of AI-enabled virtual streamers in the digital marketing ecosystem, guided by the question: what are the principal antecedents and consequences of virtual streamers under the influence of advanced AI technologies, and how are these factors interrelated? The study draws on expectancy violations theory, social presence theory, and the anthropomorphism model of intelligent agents as its theoretical foundation.
Methodology
This research adopted a qualitative, exploratory design grounded in interpretive philosophy and employing thematic content analysis. The study population comprised two groups: digital marketing experts, including faculty members from leading Iranian universities and senior managers of streaming and virtual-reality startups, and professional users of virtual streaming and virtual-reality platforms with direct experience interacting with AI-powered virtual streamers. Participants were selected through snowball sampling; 26 individuals were interviewed until theoretical saturation was reached, evidenced by repetitive coding after the seventeenth interview. Data were gathered via semi-structured interviews addressing success and failure factors, the influence of AI on consumer behavior, the role of human-likeness in perceived authenticity and trust, potential threats, design features affecting engagement, and differences between audience expectations of virtual versus human streamers. Trustworthiness was ensured through source triangulation across academics, industry managers, and platform users, supplemented by relevant literature, and through a member-checking process in which preliminary findings were returned to participants for validation and refinement. Reliability was strengthened by iterative pilot testing and revision of interview questions to ensure conceptual clarity and shared understanding of key terms. Coding followed Braun and Clarke's six-phase thematic analysis procedure: familiarization with the data, initial coding, searching for themes, reviewing themes, defining and naming themes, and producing the final report.
Results and Discussion
Interviews with the 26 participants (75% male, 25% female; predominantly aged 40–50 with graduate-level education) yielded 180 open codes for antecedents, organized into 30 organizing themes and six overarching themes: technological advancement and design, interaction and anthropomorphism, effects on consumer and brand behavior, challenges and threats, audience expectations, and success/failure factors. Technological advancement and human-like design were found to establish the infrastructure for engaging interactions, consistent with prior evidence that interface attractiveness and system quality drive user satisfaction, though this study additionally foregrounds perceived authenticity and expectations of life-like behavior as under-examined drivers. Anthropomorphism and emotional interaction emerged as central to building trust and purchase intention, aligning with research linking emotional and cognitive trust to purchase decisions and social presence to impulsive buying, while extending this literature by emphasizing real-time and quasi-social interaction as distinct antecedents. Personalization and engagement were shown to shape consumer behavior and brand perception, corroborating findings that different streamer types differentially affect cognitive versus emotional trust, while this study additionally identifies expectation alignment and perceived brand performance as broader outcomes. Privacy, algorithmic bias, authenticity concerns, high development costs, and audience-attrition risk were identified as major threats, echoing previous work on privacy dilemmas and authenticity skepticism in live streaming, with this study contributing two less-documented dimensions: the competitive cost of sustained innovation and the risk of losing audiences due to unmet expectations. Divergent audience expectations regarding professionalism, empathy, and continuous accessibility were found to shape adoption, in line with literature on emotional authenticity assessment, while extending it through the themes of disembodied interaction and constant availability. Interaction quality, credibility, and regional popularity emerged as decisive success factors, with expectation violation identified as a key driver of failure (an integrative emphasis less common in earlier studies). On the consequence side, analysis of 100 open codes, organized into 20 themes and five overarching categories, showed that virtual streamers enhance purchase intention, shopping experience, and preference shifts, and support behavior prediction; strengthen brand image, loyalty, engagement, and reputation, particularly through social presence; and increase trust through anthropomorphism and emotional interaction, although authenticity concerns can simultaneously erode it, revealing a dual role for authenticity as both trust enabler and trust inhibitor. Personalized interaction and streamer credibility were also linked to improved sales performance, targeting, and conversion rates, alongside evolving marketing strategies, while unresolved challenges such as reduced genuine engagement, privacy and transparency issues, and high technical costs persisted as constraints.
Conclusions
This study offers an integrated, theory-informed framework of six antecedent and five consequence themes explaining how AI-powered virtual streamers function within the digital marketing ecosystem. The findings demonstrate that human-like design, emotional and real-time interaction, and personalization jointly cultivate consumer trust, engagement, and brand loyalty, while privacy concerns, authenticity ambiguity, and unmet expectations remain significant barriers to sustained effectiveness. By tracing the full pathway from technological antecedents to marketing and behavioral consequences, this research extends prior variable-specific studies and contributes novel constructs, including expectation-driven failure, authenticity's dual trust function, and innovation-sustaining costs. Practically, brands are advised to prioritize human-like and emotionally resonant design, continuous personalized accessibility, transparent data practices, and high-quality, responsive interactions in order to fully capitalize on the marketing potential of virtual streamers while mitigating the associated risks.