With the prevalence of social media platforms, the way customers engage with brands has greatly been altered. Choosing appropriate social media marketing strategies to stimulate customer engagement in different forms is an important issue. In order to better understand customer behaviours in the social media marketing context, we draw on the Stimulus-Organism-Response theory, and conceptualise and characterise marketing posts from six dimensions to get various features as stimuli, which induce or activate customers’ cognitive and affective states to varying levels, and ultimately lead to different behaviour responses. Machine learning algorithms are applied to the customer engagement behaviour choice prediction when facing marketing posts. It is proved that the post features designed by humans can be used to get good predictions, while the best results are achieved by combining the human-designed features with the high-dimensional features automatically extracted from post texts by the BERT model. Our research provides insights for firms to effectively conduct social media marketing design and customer engagement behaviour choice prediction.

随着社交媒体平台的流行，客户与品牌互动的方式发生了巨大变化。选择适当的社交媒体营销策略以刺激不同形式的客户参与是一个重要问题。为了更好地理解社交媒体营销背景下的客户行为，我们借鉴了刺激-有机体-反应理论，从六个维度对营销帖子进行概念化和刻画，得到各种特征作为刺激，诱导或激活客户的认知和情感。状态不同，最终导致不同的行为反应。当面对营销职位时，机器学习算法被应用于客户参与行为选择预测。事实证明，人类设计的帖子特征可以用来得到好的预测，而最好的结果是通过将人工设计的特征与 BERT 模型从帖子文本中自动提取的高维特征相结合来实现的。我们的研究为公司有效地进行社交媒体营销设计和客户参与行为选择预测提供了见解。