Evaluating the influence of consumer preference in embroidery industry — An integrated forecasting model

Abstract

The development of embroidery art has been in Taiwan for centuries. From handmade production to digital manufacturing, embroidery products have been commercialized into consumer market as part of modern fashion. As a result, consumer preference proves decisive on purchasing embroidery products but makes demand forecasting in embroidery industry more difficult compared to general consumer goods industry. To improve forecasting performance, this study presents an integrated model to mitigate forecasting error. The model evaluates consumer preference by brand choice models, makes time-series forecasts by ARIMA, and integrates the outcome from one with the other. The empirical result reveals that the presented model outperforms ARIMA. It benefits related industry by pointing out a possible direction of consumer taste and providing more accurate forecasts. It may be instrumental to embroidery manufacturers' lower inventory, higher customer satisfaction, and maximum profit.