보유기간 구조상의 동적 고객관리 모델

서지정보

· 발행기관 : 한국마케팅학회
· 수록지 정보 : 마케팅연구 / 25권 / 1호 / 195 ~ 215페이지
· 저자명 : 김건하

초록

Most of researches on CRM (Customer Relationship Management) and CLV (Customer Lifetime Value) have been done by service providing companies especially in subscription service industry.
And their topics of researches on CRM and CLV are concentrated on retention, making their customers heavy users, and leading their cash flows into good conditions when their markets go into saturation stage.
What about the management of customers in consumer-durables market? The basic concepts and strategies of CRM in service market may be the same as those of CRM in consumer-durables market.
But, some characteristics of consumer-durables market make it difficult for marketers to apply the present CRM models of services market to customer population management of consumer-durables market.
Consumer-durables have long product-life-cycles relative to contract-periods of subscription services.
Revenues of service providers depend on contract, and cash flows resulted from service usages during contract period without churning. Revenues of consumer-durables are composed of sales of products,repairs or maintenances, and replacements without switching to the other makers. And it is true that the difficulties of data acquisition of consumer-durables market hinder the development of aggregatelevel and disaggregate-level models of customer population management in consumer-durables market.
Our study proposes a mathematical model that can analyze the dynamics of customer population at cohort level in consumer-durables market. The dynamics of customer population include the rate of customer population increase, the customer population distribution at stable state , the distribution of replacement rates, and the sensitivity analysis and elasticity analysis of the population matrix.
We adopt the concept of well-known population matrix models in ecology, redefine the terminologies for replacement behavior of consumer durables, propose a matrix model for customer population management, and test its validity and reliability by simulating it under the typical conditions of numerical experiments.
Let's think about customers as a population or a cohort; a cohort is the set of all individuals born,hatched, or recruited into population during a defined time interval. And we can classify the population of customers into several stages by duration time; how long they have their consumer durables in use.
Once a transition matrix has been estimated from trade-in data, the basic information of interest to the marketer are the rate of customer population increase (the dominant eigenvalue, λ), the stable state customer population distribution (the right eigenvector), and the distribution of replacement rates (the left eigenvector).
But, although these figures may be of interest in and of themselves, the dynamics they might change is often of interest also. Marketers may want to know how the dominant eigenvalue changes in response to a management decision; the timing of new product launch or the allocation of promotion budget to some stages of population. The goal is to understand the effect of marketing management decision on the rate of population growth. Thus, we propose sensitivity analysis and elasticity analysis of the matrix.
It is hard to acquire the trade-in data of customers' replacement behaviors in consumer durables. So, we instead design numerical experiments and simulations that test our proposed model. Through the numerical experiments and simulations, we estimate and analyze some figures including eigenvalue,eigenvectors, sensitivities, and elasticities. These figures will allude us something strategic and intrinsic information on customer management that the previous researches of customers-CRM, and retention rate or churn probability of customers-cannot provide us with.
We find some marketing implications from the simulation. First, customer population management is very important to both the retaining of customer population and the increase/decrease of customer population. We can identify and deduct the objective standards of replacement rates and survival rates for the desirable customer population distribution state on the condition that the size and the types of initial customer population distribution are specified. Second, it is supported by our study that marketers pour their efforts to grasp early adopters or opinion leaders in early stages of PLC when a company launches a new product. Third, it is the most efficient to population growth to stimulate R0 (the replacement behaviors, switching-in flows, and category starters of the starting stage cohort customers) than the replacement behaviors of other cohorts. Fourth, to maintain the early stage cohort customers is more important for population growth under the suppositions that replacement rate follows to bath tub curve and survival rate inverse bath tub curve.
Our study has some meanings in several points. One of them is the challenge for mathematical model development for the dynamic customer population management in consumer durables market. And it is a trial of the interdisciplinary study between mathematical ecology and customer population management of marketing. Fortunately we see some possibilities of the dynamic customer population management in numerical experiments and simulations of our proposed model developments.

참고자료

· 없음