Unveiling data quality dynamics in online retail: An exploration of product attributes

Abstract

Data quality is essential for effective predictive analytics and operational efficiency in e-commerce. This study investigates the quality of product attribute data such as size, colour, style, and gender classification within an online retail environment, where inaccurate or incomplete attributes can disrupt decision-making, inventory management, and customer experience. The research aims to (1) identify relevant data quality dimensions for assessing product attributes, (2) explore tools and techniques for data profiling, (3) examine the operational impact of inferior data, (4) understand how product attribute data informs business decisions, and (5) propose actionable metrics to improve data quality. Guided by a pragmatic research philosophy, a mixed-methods case study approach was employed. Quantitative analysis evaluated the completeness, consistency, and accuracy of product attribute data from a major e-commerce retailer, while qualitative insights were gathered through semi-structured interviews with key stakeholders, including procurers and retail analysts. Ethical procedures such as informed consent and data confidentiality were rigorously maintained.