算法推荐与直播营销对消费者“冲动购买”与“信任转移”的机制研究

A Study on the Mechanisms through Which Algorithmic Recommendation and Livestream Marketing Drive Consumer “Impulse Buying” and “Trust Transfer”

摘要: 数字经济浪潮下，电子商务加速从“人找货”的搜索式逻辑向“货找人”的发现式逻辑转型，算法精准分发与直播社交赋能成为驱动这一变革的核心力量。为揭示算法推荐与直播营销场景下消费者冲动购买意愿的形成机制，本研究基于S-O-R理论与信任转移理论，将算法精准度、直播互动性、主播专业性作为外部营销刺激，认知信任与情感信任作为心理机体中介变量，构建“营销刺激–双信任机制–冲动购买意愿”理论模型。通过347份有效问卷数据，采用结构方程模型(SEM)与Bootstrap中介效应检验开展实证分析。结果显示：1) 算法精准度、直播互动性、主播专业性均对认知信任和情感信任具有显著正向影响，其中主播专业性对认知信任的驱动效应最强(β = 0.402, p < 0.001)，直播互动性对情感信任的驱动效应最强(β = 0.375, p < 0.001)；2) 认知信任对冲动购买意愿存在显著正向驱动作用(β = 0.283, p < 0.001)，情感信任的直接影响不显著(p = 0.244)；3) 认知信任在营销刺激与冲动购买意愿间发挥显著中介作用，情感信任的单独中介效应与“认知→情感”链式中介效应均未得到验证。本研究拓展了S-O-R理论在算法直播融合场景的应用边界，明晰了双信任机制的差异化作用，为电商平台、直播主播及品牌商家优化营销策略、提升消费转化效率提供实证支撑。

Abstract: In the landscape of the digital economy, e-commerce is rapidly transitioning from a search-based logic where consumers seek products to a discovery-based logic where products find consumers. Algorithmic precision distribution and livestream social empowerment constitute the core forces driving this shift. To uncover the formation mechanism of consumer impulse buying intention within algorithmic recommendation and livestream marketing contexts, this study draws upon stimulus-organism-response (S-O-R) theory and trust transfer theory. Algorithmic precision, live interactivity, and streamer expertise are cast as external marketing stimuli, while cognitive trust and emotional trust serve as psychological organism mediators, yielding a theoretical model that links marketing stimuli to impulse buying intention through dual trust mechanisms. Empirical analyses are carried out on 347 valid questionnaire responses using structural equation modeling (SEM) and bootstrap mediation tests. The findings reveal several patterns. First, algorithmic precision, live interactivity, and streamer expertise all exert significant positive effects on both cognitive trust and emotional trust, with streamer expertise producing the strongest influence on cognitive trust (β = 0.402, p < 0.001) and live interactivity producing the strongest influence on emotional trust (β = 0.375, p < 0.001). Second, cognitive trust significantly promotes impulse buying intention (β = 0.283, p < 0.001), whereas the direct effect of emotional trust is not significant (p = 0.244). Third, cognitive trust significantly mediates the relationships between the marketing stimuli and impulse buying intention. Neither the independent mediating effect of emotional trust nor the chained cognitive-to-emotional trust mediation receives empirical support. This study extends the application boundary of S-O-R theory to the integrated algorithmic livestream commerce setting, clarifies the differentiated roles of the dual trust mechanisms, and provides empirical evidence for e-commerce platforms, livestream hosts, and brand merchants seeking to optimize marketing strategies and improve consumer conversion efficiency.