Resumo:
This study develops a hybrid method for optimized product selection in promotional campaigns for retail chains, combining the Analytic Hierarchy Process (AHP) with a rating system based on relative product participation. The method was tested in a Brazilian retail chain using 12 months of data and a portfolio of 3,270 Stock Keeping Unit (SKUs) distributed across 16 categories. Traditionally, promotional product selection relies on subjective criteria and manager experience, compromising results. The proposed approach uses AHP to hierarchize decision criteria (net sales, commercial margin, and quantity), adapting the method to handle the large volume of SKUs characteristic of retail. The methodological innovation consists of applying the AHP structure for defining criteria weights, expanding the analysis to all items through a rating system. For each product, its relative participation in each criterion is calculated and multiplied by the respective weight determined by AHP. Validation was conducted through two promotional scenarios: the first for the Easter campaign and the second for the Father”s Day campaign, both involving the selection of 25 promotional items, comparing with the traditional manager selection. The method demonstrated greater alignment with seasonal factors and a more robust decision-making process in both applications, validating its replicability across different promotional contexts. Computational implementation in KNIME enables replicability across different retail chains, contributing to increased profitability and market competitiveness.
