A Weighted Product-Based Decision Support System for Multi-Criteria Selection of Imported Apparel Suppliers in a Small Boutique Retailer

Abstract

Supplier selection determines product availability, quality, and cost structure in apparel retailing, yet micro-scale retailers commonly rely on the owner’s personal judgement, producing decisions that are subjective and difficult to audit. This study designs, implements, and validates a web-based decision support system that applies the Weighted Product method to rank imported apparel suppliers at Devincci Boutique, a micro retailer in North Sumatra, Indonesia. A quantitative case-study approach was used, with data obtained through structured interviews, direct observation, and a literature review. Ten suppliers were evaluated against five criteria: price (cost, 0.25), product quality (benefit, 0.25), delivery punctuality (benefit, 0.20), service (benefit, 0.15), and supplier distance (cost, 0.15), each rated on a five-point ordinal scale. The computation produced preference values ranging from 0.0717 to 0.1358, with Toko One K (A10) ranked first and Toko Mariacloset (A3) second (0.1186). A sensitivity analysis across eleven weight scenarios left the top-ranked alternative unchanged, and cross-validation against Simple Additive Weighting yielded a Spearman rank correlation of 0.976. All twelve black box test scenarios passed, and system output reproduced the manual computation to four decimal places. The study contributes a replicable and stability-tested supplier selection procedure for micro-scale imported fashion retailing.