Customer Engagement Prediction on Social Media: A Graph Neural Network Method

With the rapid prevalence and massive user growth of social media platforms, efficiently targeting potential customers on these platforms has grown in importance for companies. Enhancing the likelihood that a social media user will engage with brand posts holds profound implications for online marketing strategy design. However, predicting customer engagement on social media comes with its own set of challenges. In this work, we design a graph neural network model called the graph neural network with attention mechanism for customer engagement (GACE) to predict customer engagement (like/comment/share) of brand posts. We exploit large-scale content consumption information from the perspective of heterogeneous networks and learn latent customer representation by developing a graph neural network model. We examine GACE using a large-scale Facebook data set, and the comprehensive results show significant performance improvement over state-of-the-art baselines. Furthermore, we conduct an interpretability analysis, which sheds some light on the explanation of the proposed model. To illustrate the practical significance of our work, we provide examples to quantify the economic value of improved predictive power using a cost-revenue analysis in the context of targeted marketing.

History: Olivia Liu Sheng, Senior Editor; Huimin Zhao, Associate Editor.

Supplemental Material: The online appendix is available at https://doi.org/10.1287/isre.2021.0281.