Immersive Fairness in E-Commerce: How Can a Bias-Aware WebXR Pipeline Transform Regional Delivery Analytics?

E-commerce delivery systems often exhibit regional and socioeconomic disparities in delays, cancellations, and refunds that are not evident in aggregate dashboards, especially for intersectional demographic and geographic cohorts. Existing work primarily emphasises sales and marketing analytics, leaving delivery fairness underexplored and exposing the shortcomings of traditional two-dimensional tools in uncovering multidimensional bias across geography, income, and order attributes.This paper introduces a bias-aware delivery analytics pipeline that integrates intersectional fairness metrics with WebXR-based immersive visualisation, using the Brazilian Olist dataset as a real-world case study. The pipeline fuses relational order, review, and geolocation data with fairness metrics derived from an extended version of Fairlearn’s MetricFrame to capture disparities among overlapping groups. The delivery performance metrics, such as delay and cancellation rates, are associated with three-dimensional representations—bar height, colour scales of equal perceptual difference, and dynamic particle orbitals—in a browser- native WebXR environment that is accessible from both desktop and VR. An interactive governance layer equipped with immersive tools enables the establishment of an audit trail through confidence intervals that can be interacted with, causal flow overlays, and snapshots of the metric, all of which are designed to help stakeholders identify when, where, and for whom unfairness has occurred. The initial results reveal a considerable range of disparities in worst-case scenarios, where orders with far-off delivery times and low prices had delivery times and cancellation rates two to three times longer than those of central, high-priced orders. A planned comparative user study will estimate how the WebXR system impacts the accuracy, time-to-insight, and confidence in bias detection compared to standard dashboards, positioning immersive analytics as a practical and helpful tool for more equitable, data-driven logistics decisions.