A Product Demand Forecasting Model based on Exponential Smoothing through Analysis of Consumer Requirements

Authors

  • Yeong-Hwi Ahn
  • Koo-Rack Park
  • Hwang-Rae Kim
  • Dong-Hyun Kim

Abstract

Background/Objectives: The main factors that define a company's growth and performance are the product market demand that meets consumer needs and services, and it is important for companies to exactly identify trends in market demand. The existing demand forecasting technique is limited to analyzing rapidly changing market trends and ensuring the output of results, so it is necessary to establish and operate a system based on demand forecasting that enables rapid planning of global products.
Methods/Statistical analysis: System: In order for product planners to develop the best products that fully reflect the exact needs and services of consumers, we analyzed diverse and vast big data of portal companies, collected data on products based on web crawling. That can collect and utilize the data that users want in real time, and analyzed demand in the past, used additive seasonal exponential smoothing among time series forecasting techniques to forecast future demand as well as demand patterns at the present time. We also used Nelson Rules Logic to eliminate inaccurate demand needs, and analyzed and designed the exact demands of consumers to analyze out-of-range signals for demand forecasting. Based on the above, satisfaction with product usage was calculated and used for product planning.
Findings: For companies, product planning based on market demand forecasting is very important, which is to maximize the growth and performance of the company by making the best use of limited human resources. To this end, companies are spending enormous costs and outsourcing the analysis of consumer needs and services to external marketing companies to obtain data. It can be said that launching the right product at the right time controls the company’s fate. The product demand forecasting model based on the time series forecasting technique proposed in this paper is a system for obtaining useful and accurate information for planning the best products. According to the results of a survey of the use of the proposed model for planners engaged in product planning, the satisfaction level was higher than expected in terms of system satisfaction, system efficiency, and system effectiveness.
Improvements/Applications: In order to analyze more sophisticated consumer needs and services in the future, research on product demand forecasting considering rapidly changing global environment and climate as well as consumer emotion analysis should continue with natural language processing using deep learning on customer reviews.

Downloads

Published

2020-03-26

Issue

Section

Articles