Measuring the impact of assortment change in retail stores using the synthetic control method

Abstract

This study offers an innovative method for evaluating the effects of assortment changes on retail store revenue, employing the Synthetic Control Method (SCM) with Lasso regression. In response to the limitations of randomized control trials for large-scale interventions, we propose a unique approach using Lasso regression to build synthetic control groups from comparable stores in competitor retailers. This innovative technique overcomes the obstacle of absent control data within the same retailer. Our analysis reveals a remarkable 3.7% increase in retailer revenue after implementing the assortment changes. To confirm the validity of our findings, we conducted placebo studies, solidifying the positive impact of these adjustments. These results advocate embracing the SCM with Lasso regression as a reliable tool for measuring the effects of interventions in the business world, especially when controlled experiments are not feasible. This method empowers retailers to assess the effectiveness of assortment optimization strategies and make data-driven decisions about future changes.