The Impact of Abnormal Event to Customer Value －A Research of Time Series Model in An Aggregate View
Date Issued
2007
Date
2007
Author(s)
Zeng, Siang-Hao
DOI
zh-TW
Abstract
The most valuable marketing people are who can help their companies improve customer value most. From strategic perspective, to improve customer value can be done in two ways. On the one hand is to improve the price customers will pay; on the other hand is to lower the costs. In addition by constructing model and consumer behavior analysis, marketing people can improve the price customers willing to pay in terms of more precise marketing activities, marketing people can also respond actively to the situation of customer leaving by observing the transaction data of customers.
There are two reasons will result in customers leaving. One is that disappear naturally, and the other is that results from abnormal events deriving from the competitors or environment. Thus, analyze and evaluate these events which will make impact to the customer value of the company, caution the manager for the damage as well as make marketing strategy to respond the impact is an evitable task for marketing people in the future.
The case company in this thesis is an on-line company. This research use the Chow-test of breakpoint test of time-series method and product life cycle theory to prove that there was an abnormal event happened in this company. After comparing the coefficient of determination, RMSE and MAPE, this research then find out a ARIMA(2,1,1and5) first-step step-function intervention model which is better to model and analyse the event. Finally, the impact resulted from the event caused customer value loss of N.T. 45484290 dollars to the company.
Subjects
異常事件
顧客價值
時間序列
介入模型
Abnormal event
Customer value
Time series
Intervention model
SDGs
Type
thesis