Customer engagement (CE) within social media has emerged as a focal point in marketing research due to its significant impact on brand and firm-related outcomes. However, effective measurement methodologies for CE behaviours remain underexplored in this context. Building upon the existing concept of CE behaviour proposed by Brodie et al. (J Serv Res 14(3):252–271, 2013), and the CE cycle theory, this paper introduces an operational definition that supports an entropy-based CE behaviour measurement framework, offering a nuanced reflection of social media dynamics. The proposed measurement approach, among the initial attempts to integrate the multi-criteria decision-making (MCDM) weighting method into brand marketing analytics, ensures a reliable and rigorous assessment of CE performance with brands. Additionally, a range of analytical methods, including regression modelling using dummy variables, sentiment analysis, and hierarchical clustering algorithm, were synthesized to assist the objective evaluation of brands’ impacts on CE behaviour outcomes. A peer group of six luxury brands were chosen to analyse their impacts on engagement performance independently. Key findings from this application demonstrate the approach’s ability to compare the engagement performance of different brands, offering a significant tool to differentiate engagement behaviour levels while providing actionable insights for brand marketing initiatives.
Several scholars have utilized hierarchical network Data Envelopment Analysis modeling techniques to assess the performance of complex structures. However, there has been limited consideration given to the integration of a peer-appraisal setting within a self-evaluation hierarchical context. This aims to enhance discriminatory power and mitigate the issue of unrealistic weighting scheme. To this end, our study extends the single-stage hierarchical additive self-evaluation model of Kao (Omega 51:121–127, 2015. https://doi.org/10.1016/j.omega.2014.09.008), by integrating the well-established cross-efficiency method. An original combination of a maxmin secondary goal model and the Criteria Importance Through Inter-criteria Correlation (CRITIC) method is proposed, to expand the basic hierarchical self-evaluation model. The maxmin model addresses the issue of the non-unique optimal multipliers obtained from the self-evaluation model, ensuring a more realistic weight scheme. The CRITIC method, that tackles the aggregation problem by objectively determining weights of criteria, rewards the minority and is conducive to a fairer evaluation. Results indicate that the proposed approach is more likely to obtain a unique efficiency and ranking score for the units under consideration. This study entails a numerical experimentation aimed at evaluating the efficiency of a set of 20 universities while validating the applicability of our proposed approach. To conclude, the practical applications of this methodological framework could encompass assessing services within the higher education sector or fostering sustainable development across various operations within a hierarchical structure.
20 November 2024 in Socio-Economic Planning Sciences
In the digital era, with the growing importance of social media platforms, this study advances customer engagement (CE) research by moving from ex-post measurement to ex-ante prediction. While recent social media CE literature has proposed sophisticated measures of CE, most indices are based on public metrics, leaving most of the work descriptive and retrospective in nature. However, marketing practitioners require predictive capabilities to optimise content and timing decisions before posting social media content. Responding to that challenge, this paper extends the CE measurement approach into a predictive modelling framework. The innovation lies in utilising supervised topic modelling (STM), specifically Anchored CorEx, to generate probabilistic indicators of content levers that are systematically integrated into predictive pipelines. The probabilistic lever features (Entertainment, Trendiness), combined with time and brand metadata, are incorporated into five predictive families: Random Forest, XGBoost, Keras MLP, OLS, and KNN. Results on an authentic luxury brand Twitter dataset (2017-2022) show that the proposed model yields satisfying predictive performance, improving both accuracy and rank validity. Notably, the predictive results prove the validity of the entropy-computed behavioral CE approach in extracting valuable information from social media contexts, leveraging ex-ante marketing strategic decision-making for practitioners.
30 January 2026 in Socio-Economic Planning Sciences