Risky Exploration: Steering Product Performance with Introduction Decisions

Abstract

The fast fashion business model is founded on the ability to quickly learn about market dynamics to improve product design. We empirically examine the effect of learning on product launch decisions and sales performance. In line with common knowledge, we confirm that late-season launches sell faster. We also identify local exploration, and past obtained rewards as two novel levers of the firm's launch process that can affect individual product performance. Utilizing daily product-level data from a major apparel firm between 2021 and 2022, we find that these three variables are positively associated with product performance. Although the effect of product introduction is strong and consistent across product categories and seasons, we observe heterogeneous effects for the other two variables. Most importantly, our work provides evidence that past obtained performance is a strong indicator of success in forthcoming new launches.

Keywords: Innovation, new product introduction, exploration, fashion retailing, experimentation process