Preliminary Framework for Verifying the Syntactic Quality of Digital Voice-of-Customer for Product Quality Monitoring

Voice-of-Customer (VoC) data, such as online customer reviews and unsolicited customer feedback, has become a valuable complement to traditional quality information in manufacturing contexts and within the broader paradigm of Quality 4.0. In particular, digital VoC provides post-market evidence on product quality, supporting the monitoring of field performance and the early detection of customer-perceived defects. However, the reliability of digital VoC-based monitoring depends critically on the quality of the underlying data. When digital VoC datasets are noisy, incomplete, inconsistent, or manipulated, even advanced analytics may generate misleading signals, reflecting the well-known “garbage in, garbage out” effect. This paper addresses the problem of digital VoC data reliability for product quality monitoring by grounding the discussion in established data-quality principles, with particular reference to ISO 8000. The standard distinguishes three complementary perspectives on data quality: syntactic, semantic, and pragmatic, which together determine whether data can support reliable analysis and decision-making. Building on this conceptual structure, the paper develops a preliminary framework for assessing the syntactic quality of digital VoC datasets. Syntactic quality concerns the formal correctness of data, independently of their meaning or usefulness. The proposed framework operationalizes syntactic quality through a set of subdimensions and associated metrics, providing a foundation for assessing the reliability of VoC datasets before performing downstream analytics in manufacturing quality monitoring. While the broader problem of VoC reliability includes semantic and pragmatic considerations, this work focuses specifically on the syntactic dimension as a necessary first step toward a comprehensive reliability assessment framework for digital VoC.