Answers are grounded in this paper and cite their sources.
Purpose Marketing analytics of electronic word of mouth for consumer segmentation treat linguistic markers, such as lexical diversity, code-switching and review length, as universal indicators for consumption preferences. This assumption breaks down in Asia, where multilingual consumers use languages that carry unequal social status. Drawing on theories of indexicality (semiotics) and language as social practice, this study aims to demonstrate that in countries like the Philippines where translanguaging practices dominate, languaging behaves asymmetrically depending on the social status and symbolic capital embedded in the linguistic resources involved. Design/methodology/approach Using large-scale eWOM data from two online platforms operating in the Philippines, this paper conducts three complementary studies using a hybrid machine learning approach. This paper utilizes k-means clustering for customer segmentation and LightGBM to assess the predictive power of translanguaging features. By integrating Explainable AI (XAI) through SHAP value analysis, this paper provides model interpretability by deconstructing how variables, such as English dominance, code-switching, frequency, lexical diversity and review length can serve as asymmetric predictors of purchasing power. Findings The findings demonstrate that the predictive strength of lexical diversity and linguistic purity for premium consumption preferences applies only for English-dominant reviews, with the same linguistic features predicting more price sensitive consumption in Tagalog-dominant reviews. Originality/value By integrating sociolinguistic theory with machine learning-powered marketing analytics, this paper extends eWOM analytics beyond sentiment and brand perception toward a culturally-embedded understanding of consumers in low resource, multilingual contexts. It positions translanguaging, a dominant communicative practice in Asia, as a context-sensitive and resource-efficient mechanism for marketing segmentation in multilingual emerging markets.
No takes yet. Share an insight, caveat, or question.
Galang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: