A Predictive Model of Seasonal Clothing Demand with Weather Factors

New normal weather, more significant variations in temperature and precipitation, and more extreme weather events have resulted in sluggish sales in the clothing industry. Since the industry is a highly fragmented global value chain and procurement of clothing products requires long lead times, weather forecasts are essential information for product planning. Thus, this study aims to develop a model of merchandising strategy for retailers using weather factors within a season. As a place-specific study, we acquire the data from Goggle Trend and weather observation sites from October 2008 to January 2019. Temperature, precipitation, wind speed, snowfall, snow depth, and wind chill were analyzed to find influential factors in seasonal clothing demand. Exploratory data analysis, Pearson correlation analysis, cross-correlation analysis, and Genialized Linear Mixed Model (GLMM) are conducted. Wind chill and the month of the year are significant predictors of seasonal clothing demand. Since the wind chill changes lead to the demand for seasonal clothing one day ahead, this study uses the one-day-lagged windchill data (WindChill_lag1). GLMM model separates the linear relationship between WindChill_lag1 and monthly consumer demand in winter seasons with random effects for multiple years. In addition, the model shows that there are unmeasured characteristics of consumer demand which are indicated through intercept covariance between years. This study provides meaningful information to the clothing industry, which can modify their merchandising plan at the right time according to a weather change.