Detecting digital voice of customer anomalies to improve product quality tracking

Abstract

Purpose: The aim of this study is to enhance the product quality management by proposing a framework for the classification of anomalies in digital voice of customer (VoC), i.e. user feedback on product/service usage gathered from online sources such as online reviews. By categorizing significant deviations in the content of digital VoC, the research seeks to provide actionable insights for quality improvement. Design/methodology/approach: The study proposes the application of topic modeling algorithms, in particular the structural topic model, to large datasets of digital VoC, enabling the identification and classification of customer feedback into distinct topics. This approach helps to systematically analyze deviations from expected feedback patterns, providing early detection of potential quality issues or shifts in customer preferences. By focusing on anomalies in digital VoC, the study offers a dynamic framework for improving product quality and enhancing customer satisfaction. Findings: The research categorizes anomalies into spike, level, trend and seasonal types, each with distinct characteristics and implications for quality management. Case studies illustrate how these anomalies can signal critical shifts in customer sentiment and behavior, highlighting the importance of targeted responses to maintain or enhance product quality. Research limitations/implications: Despite its contributions, the study has some limitations. The reliance on historical data may not hold in rapidly changing markets. Additionally, text mining techniques may miss implicit customer sentiment. Practical implications: The findings suggest that companies can enhance their quality tracking tools by digital VoC anomaly detection into their standard practices, potentially leading to more responsive and effective quality management systems. Originality/value: This paper introduces a novel framework for interpreting digital VoC anomalies within the Quality 4.0 context. By integrating text mining techniques with traditional quality tracking, it offers a novel approach for leveraging customer feedback to drive continuous improvement.