Does it matter how I stream? Comparative analysis of livestreaming marketing formats on Amazon Live

The COVID-19 pandemic accelerated the shift towards livestreaming commerce, particularly in China. Livestreaming marketing, where sellers demonstrate products via live video, has seen explosive growth, generating billions in sales. While offering significant opportunities, sellers face the challenge of selecting the most effective product demonstration format to maximize viewer engagement and sales. This study addresses this gap by investigating the relationships between social capital acquisition, social endorsement, and consumer engagement across three distinct livestreaming formats: interview, tutorial, and behind-the-scenes. The research is grounded in social capital and signaling theories, which posit that social interactions and endorsements influence consumer behavior. The study aims to determine which format most effectively facilitates social capital acquisition and endorsement, ultimately leading to increased consumer engagement.

Existing research has explored livestreaming in various contexts, including social media and gaming, but has largely overlooked the differences in product demonstration formats within livestreaming marketing. Studies have examined consumer motivations for participating in livestreaming commerce and the psychological aspects of livestreaming in business-to-consumer e-commerce. However, a critical gap exists in understanding how different livestreaming formats affect consumer engagement. This study addresses this gap by examining the three main formats identified by Bous (2019a): interview, tutorial, and behind-the-scenes. The study leverages social capital theory, focusing on factors like physical proximity, social proximity, familiarity, parasocial interaction, and self-disclosure, and signaling theory to explain how social endorsement (follower and viewer endorsement) impacts engagement.

Data was collected from 766 sellers on Amazon Live using a survey questionnaire distributed via email and social media. The questionnaire included sections on demographic information and constructs related to social capital acquisition, social endorsement, and consumer engagement. These constructs were measured using a seven-point Likert scale. The data was then categorized into three groups based on the dominant livestreaming format used by each seller (interview, tutorial, and behind-the-scenes). Partial Least Squares Structural Equation Modeling (PLS-SEM) was employed to analyze the data, specifically using a cross-case analysis to compare the three livestreaming formats. This method was chosen for its suitability for newer marketing concepts and its robustness with limited sample sizes and skewed data. The measurement model's reliability and validity were assessed using various metrics, including confirmatory tetrad analysis, factor loadings, composite reliability (CR), average variance extracted (AVE), and rho_A. Discriminant validity was established using the Fornell-Larcker criterion, and collinearity was assessed using variance inflation factor (VIF) and tolerance levels. Finally, Henseler's bootstrap-based multigroup analysis (MGA) was used to compare the differences in path coefficients across the three formats.

The PLS-SEM analysis revealed that social capital acquisition and social endorsement positively influenced consumer engagement across all three livestreaming formats, supporting the hypotheses (H1a, H2a, H3a). However, the strength of these relationships varied significantly across formats (H1b, H2b, H3b). Specifically, the tutorial format showed the strongest direct effect of social capital on consumer engagement, while the interview format had the strongest direct effect of social endorsement on consumer engagement. The behind-the-scenes format fell between the two. Social capital significantly explained variance in social endorsement across all formats. R-squared values indicated substantial explained variance in consumer engagement (88.1%, 73.2%, and 79.1% for interview, tutorial, and behind-the-scenes formats, respectively) and social endorsement (68.7%, 59.8%, and 62.9%, respectively). The multigroup analysis confirmed statistically significant differences in the effects of social capital and social endorsement on consumer engagement between the tutorial and behind-the-scenes formats, and between the interview and tutorial formats. The differences were less pronounced between the interview and behind-the-scenes formats for the relationship between social capital and consumer engagement.

The findings highlight that while the motivation for choosing a livestreaming format is generally to increase consumer engagement, the impact of each format differs significantly. The interview format, despite being less frequently used, demonstrates a stronger reliance on social capital and endorsement to drive engagement compared to the tutorial and behind-the-scenes formats. This suggests that the interaction and personalization offered by the interview format may be highly effective. The results support the importance of social capital in driving consumer engagement in livestreaming, but also show the significant role of social endorsement. This study offers a nuanced perspective compared to previous research that focused primarily on consumer behavior. The findings underscore the importance of considering both social capital and social endorsement when selecting a livestreaming format, and that optimal strategy depends on the format chosen.

This study makes several key contributions. First, it provides a seller-centric perspective on livestreaming marketing, focusing on format selection and its impact on engagement. Second, it offers a comparative analysis of three commonly used livestreaming formats, revealing significant variations in their effectiveness. Third, it highlights the crucial interplay of social capital and social endorsement in driving consumer engagement. Future research could explore the impact of seller demographics (e.g., gender), product type, and different platforms on the relationships identified. Expanding the study to other livestreaming platforms would enhance generalizability.

The study's reliance on data from a single e-commerce platform (Amazon Live) might limit the generalizability of the findings. While the sample size is relatively large, other factors not included in the model (e.g., seller gender, specific product characteristics, platform-specific features) could influence consumer engagement. The cross-sectional nature of the data limits causal inferences.