Abstract  

Siddharth Sinha is the CEO of an apparel retailer WE SELL STYLE (WSS). The retail chain was set up in 2008, housing more than 100 brands. In 2015, they operated over 200 stores in all four regions of the country. They primarily focused on providing good quality fashion at a remarkably low price.

Markdown planning has been an important aspect of the apparel business. It is important to understand that the demand for fashion apparel is seasonal – affected by current fashion, variations in the seasons, festivals and hence difficult to estimate. An apparel retailer could go off target – either by overestimating or underestimating the demand, with overestimating being prevalent.

The ordering–manufacturing–stocking cycle is easily a 6-month cycle before the selling actually starts; with an expectation to improve sales year on year, the procurement team buys more, making an increase in the variety of colors and styles to offer more to the consumer. However, not all styles sell as expected, leaving higher than expected stocked inventory, which requires an impetus to sell. The impetus in the industry comes in the form of ‘‘end of season markdown (EOSS)’’.

Decision on the percentage of markdown for EOSS is one of the most critical tasks for an apparel retailer. This activity starts months ahead of the EOSS. The product team and the planning team sit together and come up with an EOSS plan at the style level. In the decision process, procurement and planning team use their domain expertise and judge the performance of style using metrics such as rate of sales, full price sell-through, inventory left, and more. The key decision is to quantify the degree of non-performance of styles that did not sell as forecasted and by how much to markdown for the EOSS.  

Learning Objective 

This case can be used either in the Quantitative Methods or in the Business Analytics/Data Science course of MBA and executive MBA programs. The case is ideal for teaching markdown business strategy used by retailers and introductory/advanced-level concepts in clustering, forecasting, and optimization. The case helps the instructor to demonstrate applications of hierarchical clustering, partition around medoids, time series forecasting using regression, and non-linear optimization.