• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Customer retention in the financial industry: An application of survival analysis

Dissertation
Author: Hong Zhang
Abstract:
Recently, in both marketing theory (academia) and practice (industry), the emphasis in relationship marketing has shifted to long term customer relationship management emphasizing customer retention or loyalty. This study has two main purposes: (1) to investigate the impacts of selected firm-customer interaction behavior and demographic characteristics on customer retention behavior in the financial industry, (2) to compare the results of a static method of analysis (logistic regression model) and a dynamic method (Cox's hazard method) for customer retention. The statistical analyses conducted employing two cohorts showed that the Cox model was stable across cohorts. Using Cox's hazard method, all of the independent variables were found to be significantly related to customer retention, while some of demographic factors (for example, tenure, marital status and gender, and channel usage) were not found to significantly affect customer retention in the logistic regression. The results show that with the enhancement of interaction between a bank and its customers by increasing customers' service usage, cross-buying, tenure experience, and complicated product usage, customers were more likely to stay longer with the bank. Additionally, age, education, and income were found to be positively related to customer retention, while customers with higher time pressure were less likely to defect. Single females were found to be least likely to attrite, while married customers were most likely to default or switch to an alternative. Single males were more likely to stay than married ones, but more likely to churn than single female customers. Finally, managerial implications were drawn, and limitations and potential for further study were discussed. [PUBLICATION ABSTRACT]

TABLE OF CONTENTS

Page LIST OF TABLES ..................................................................................................... vii LIST OF FIGURES ................................................................................................... viii ABSTRACT ............................................................................................................... ix CHAPTER 1 – INTRODUCTION ............................................................................ 1 Purpose of the Study ...................................................................................... 5 Significance of the Study ............................................................................... 6 Organization of the Dissertation .................................................................... 13 CHAPTER 2 – LITERATURE REVIEW ................................................................. 14 Customer Retention ....................................................................................... 14 Antecedents of Customer Retention/Loyalty ................................................. 16 Triggers of Customer Attrition/Switching ......................................... 17 Groups of Variables Affecting Customer Retention .......................... 18 Customer Demographics ........................................................ 18 Customer Behavior ................................................................ 18 Customer Perception .............................................................. 19 Macro-environment Variables ............................................... 19 Strategies of Increasing Customer Retention ..................................... 20 Satisfaction ............................................................................. 20 Switching Costs ..................................................................... 21 Hypotheses Development .............................................................................. 22 Firm-Customer Interaction Behavior Factors .................................... 23 Service Usage......................................................................... 24 Tenure or Length of Relationship .......................................... 26 Cross-Buying ......................................................................... 29 Product Complexity ............................................................... 32 Channel Usage ....................................................................... 35 Demographics .................................................................................... 40 Age ......................................................................................... 40

v Page Income.................................................................................... 43 Education ............................................................................... 45 Marital Status and Gender ..................................................... 47 Time Pressure ......................................................................... 49 CHAPTER 3 – METHODOLOGY ........................................................................... 51 Statistical Methods of Customer Retention Analysis .................................... 51 General Discussion of Survival Analysis .......................................... 52 Cox’s Proportional Hazard Model ..................................................... 54 Reasons for Choosing Cox’s Hazard Method .................................... 56 Sample............................................................................................................ 59 Variable Operationalization ........................................................................... 61 Analysis Procedures ....................................................................................... 64 Exploratory Data Analysis ................................................................. 64 Cox’s Hazard Function ...................................................................... 68 Models................................................................................................ 70 CHAPTER 4 - RESULTS .......................................................................................... 72 Effects of Firm-Customer Interaction Related Variables .............................. 72 Service Usage..................................................................................... 72 Tenure or Length of Relationship ...................................................... 74 Cross-Buying ..................................................................................... 74 Product Complexity ........................................................................... 75 Channel Usage ................................................................................... 76 Effects of Demographics ................................................................................ 76 Age ..................................................................................................... 77 Income................................................................................................ 77 Education ........................................................................................... 78 Marital Status and Gender ................................................................. 78 Time Pressure ..................................................................................... 79 Stability and Results Comparison Between Methods .................................... 79 CHAPTER 5 - CONCLUSIONS ............................................................................... 82 Discussion of General Findings and Theoretical Contributions .................... 82 Firm-Customer Interaction Related Factors ................................................... 85 Service Usage..................................................................................... 85 Tenure or Length of Relationship ...................................................... 86 Cross-Buying ..................................................................................... 87 Product Complexity ........................................................................... 88 Channel Usage ................................................................................... 89

vi Page Demographic Factors ..................................................................................... 91 Age ..................................................................................................... 91 Income................................................................................................ 91 Education ........................................................................................... 92 Marital Status and Gender ................................................................. 93 Time Pressure ..................................................................................... 93 Managerial Implications ................................................................................ 94 Limitation and Future Study .......................................................................... 97 LIST OF REFERENCES ........................................................................................... 101 VITA .......................................................................................................................... 112

vii

LIST OF TABLES

Table Page 1 Sample Characteristics ................................................................................ 60 2 Variables for Customer Retention Duration Model .................................... 65 3 Model Comparison for Two Cohorts .......................................................... 73 4 Coefficients and Hazard Ratio for Customer Attrition Models .................. 80 5 Summary of Results of Hypotheses ............................................................ 83

viii

LIST OF FIGURES

Figure Page 1 Foote, Cone and Belding Involvement Matrix Grid ................................... 44 2 Survival rate chart for Cohort ..................................................................... 66 3 Survival rate chart for new households in Dec. 2000 ................................. 67

ix

ABSTRACT

Zhang, Hong. Ph.D., Purdue University, August 2008. Customer Retention in the Financial Industry: An Application of Survival Analysis. Major Professor: Richard Widdows.

Recently, in both marketing theory (academia) and practice (industry), the emphasis in relationship marketing has shifted to long term customer relationship management emphasizing customer retention or loyalty. This study has two main purposes: (1) to investigate the impacts of selected firm-customer interaction behavior and demographic characteristics on customer retention behavior in the financial industry, (2) to compare the results of a static method of analysis (logistic regression model) and a dynamic method (Cox’s hazard method) for customer retention. The statistical analyses conducted employing two cohorts showed that the Cox model was stable across cohorts. Using Cox’s hazard method, all of the independent variables were found to be significantly related to customer retention, while some of demographic factors (for example, tenure, marital status and gender, and channel usage) were not found to significantly affect customer retention in the logistic regression. The results show that with the enhancement of interaction between a bank and its customers by increasing customers' service usage, cross-buying, tenure experience, and complicated product usage, customers were more likely to stay longer with the bank. Additionally, age, education, and income were found to be positively related to

x customer retention, while customers with higher time pressure were less likely to defect. Single females were found to be least likely to attrite, while married customers were most likely to default or switch to an alternative. Single males were more likely to stay than married ones, but more likely to churn than single female customers. Finally, managerial implications were drawn, and limitations and potential for further study were discussed.

1

CHAPTER 1 – INTRODUCTION

Generally speaking, a company could increase its profits by acquiring new customers, augmenting profitability from existing customers by enhancing their one time purchase volume, and enhancing the duration of customer retention (Grant and Schlesinger 1995). In the past, most companies focused on the first two approaches. However, those strategies have been found not to be very effective and efficient in markets that are saturated. Recently, in both marketing theory (academia) and practice (industry), the emphasis in relationship marketing has shifted to long term customer relationship management (Reinartz and Kumar 2003; Al-Hawari 2006). Managers and researchers have emphasized the importance of customer retention, the dynamics of customer relationship, and customer lifetime value (Reinartz and Kumar 2003) for which customer retention is an important component (Gupta, Lehmann, and Stuart 2004). Customer retention has been suggested as an important antecedent to financial outcome (Evanschitzky and Wunderlich 2006). Compared to short-term customers, long-term oriented customers could offer substantial benefits to a company. Higher retention leads to higher profits across firms in various industries (Reichheld 1991-1992; Reichheld, Markey, and Hopton 2000). Increasing customer retention could be effective in both raising revenue and lowering costs (Keaveney and Parthasarathy 2001). On the revenue side, continuing customers have been found to buy

2 higher volumes at higher margins and increase service usage even when price increases (Reichheld 1996). On the cost side, researchers claim that the cost of recruiting a new customer is estimated to be five times more than that of retaining an existing customer (Hart, Heskett, and Sasser 1990). Therefore, improving customer retention could benefit profits of companies. In the sector of interest to this study, it has been estimated that “reducing defections by just 5% generated 85% more profits in one bank's branch system, and 50% more in an insurance brokerage” (Reichheld and Sasser 1990). An increase in customer retention is suggested to be helpful for companies to gain a competitive advantage, expand their market share, and increase employee satisfaction (Buttle and Ahmad 2002; Swailes and Dawes 1999). Loyal customers have been found to have a greater tolerance of negative customer experiences, lower price sensitivity, higher price acceptance, and a greater willingness to purchase other products. Enhancing customer retention is beneficial for acquiring new customers. Loyal customers are more likely to generate word of mouth (WOM) advertising because of their positive attitude toward the current provider. New customers could be attracted by positive WOM, which enhances revenue and market share. Reichheld (1991-1992) found that between 20% and 40% of new customers chose a bank based on a referral. Enhancing customer retention duration could help a company keep potentially profitable customers, such as students, who are not profitable while they are studying, but may be so later in their lives. Researchers claim that individuals are more likely to keep their first credit card used in college a long time after they graduate from college and generate reasonable income (Dugas 2001). Therefore, increasing student customers’ retention duration potentially benefits a company in the long run.

3 Increasing customer retention is helpful for extending market share. For some kinds of services (for example, mortgages and pensions), consumers only have limited requirements. For example, most households could only afford one household by one mortgage. Hence, if customers buy those products from one company, they could not buy them from the competitors, which might cause shifts in market share. Based on the above discussion, we could conclude that customer retention or loyalty analysis is very important for building long term customer relationships and beneficial for the long term profits of a company. Customer retention analysis could help answer the following questions: How can one evaluate the impacts of marketing activities on customer retention? What are the strengths and directions of the impacts of customers’ characteristics? How can one identify customers with higher risk of churn, and how can they be segmented? How can one predict customer retention duration? Segmenting customers and predicting customer retention duration can be used in strategic or tactical decision making. For example, it might indicate whether more resources should be allocated to serve long-term oriented customers or to retain customers who are more likely to attrite. Additionally, customer retention analysis is fundamental to evaluating customer lifetime value and the intangible value of a company (Gupta et al. 2004). Customer lifetime value is referred as a long-term view of a customer’s profitability. It is defined as the present value of the future profitability based on the customer relationship (Pfeifer, Haskins, and Conroy 2005). Customer lifetime value for a firm is the net profit or loss to the firm from a customer over the entire relationship life (Singh 2002). Customer lifetime value (CLV) is increasingly considered as a guide for a firm in

4 optimizing its marketing mix across the customer base and in decision-making toward marketing strategy (Libai, Narayandas, and Humby 2002). Besides tangible assets listed in the annual report, intangible assets (such as brand, customers, and employees) are critical to firm value, especially when considering future profitability (Gupta et al. 2004). Researchers have suggested that customer based value forms a large part of a company’s intangible value and could be treated as a proxy for firm value (Gupta et al. 2004). A firm’s customer-based value is the sum of the customer lifetime values (CLV) of its current and future customers. A customer retention forecast is one component of the formula to calculate CLV. Increased customer retention was found to have the greatest effect on customer lifetime value, followed by improved margin, reduced acquisition costs, and the discount rate (Gupta et al. 2004). Therefore, long-term customer retention projections could be very valuable for fully assessing the value of a company. Due to the saturation and fierce competition of financial markets, as a prerequisite to profitability and intangible value of a company, customer retention is very important (Veloutsou, Daskou, and Daskou 2004). Retention or attribution research has been increasingly emphasized in this context. Many studies of customer retention have been conducted in a service-wide context, such as retailing, insurance, and banking (Al-Hawari 2006; Boulding, Kalra, Staelin, and Zeithaml 1993; Ranaweera and Neely 2003; Zeithaml, Berry, and Parasuraman 1996). Service quality, customer satisfaction, trust, switching costs, pricing, and brand image have been suggested as factors impacting whether a customer will stay or switch (Bloemer, Ruyter, and Peeters 1998; Baumann, Burton, and Elliott 2005; Colgate and Lang 2001). Different Customer

5 Relationship Management strategies have been used to retain customers and build loyalty. Generally, those strategies involve creating loyalty programs, selling more products or services to existing customers, improving customer service quality and customer satisfaction, developing consumer trust, and increasing customer switching cost (Fitzgibbon and White 2005). Customer retention and loyalty were usually considered as synonymous by practitioners and academic researchers (Al-Hawari 2006; Boulding et al. 1993; Ranaweera and Neely 2003; Zeithaml et al. 1996), while customer retention has been used to measure customer behavioral loyalty. In the current research, those two terms are treated as transferable constructs. There are two analytical or statistical approaches used by prior customer retention research. The first one is static and short-term customer attrition or retention analysis. It is usually conducted with a forecast window of less than one year and used to identify customer segments, and set marketing campaigns. The object is to reduce attrition or increase customer loyalty. The second approach is dynamic. Long-term retention forecasting is conducted to calculate customer lifetime value and guide long- term business strategy, while aiding short-term marketing campaigns.

Purpose of the Study

The primary purpose of this study is to investigate the impacts of demographics and customer behavior factors on customer retention based on data from a consumer bank. Cox’s method, one kind of survival analysis, was used to build a probabilistic model. Additionally, this study compares the results of a static method (logistic regression model) and a dynamic method (Cox’s hazard method), and examines model

6 stability according to different cohorts of customers. The following questions guide this study: 1) In the financial industry context, what are the differences in age, education, income, marital status, and gender between short-term customers and long-term customers? 2) In the financial industry context, does time pressure of a household impact its attrition? 3) In the financial industry context, do customers with more interaction and experience with a bank stay longer? Interactions comprise using multiple channels, holding longer tenure with the bank, using more complex products from a bank, and using more services.

Significance of the Study

“The importance of customer loyalty in the service industry cannot be overstated” (Lim 2005). Many companies treat increasing retention rate as one of the most powerful weapons for reaching financial goals and improving potential firm value. Even though these benefits and the importance of customer retention are also recognized by financial institutions, many have not fully employed the concept in practice. This study will make several contributes to academic research and practice in financial industry. 1) The current study provides a dynamic methodology for predicting customer retention, which not only predicts whether a customer will stay or attrite at

7 one point of time, but also when he/she will attrite. It complements the existing customer retention or loyalty research in the following aspects: a) This study fully investigates loyal customers’ past behavior and demographics. Prior research only tested the effects of some past customer behaviors, such as specific product ownership, inter- purchase time, or balance, on customer retention (Van den Poel and Lariviere 2004). The factors in this study include customers’ channel choice, product usage, tenure of relationship, service usage, cross- buying behavior, and demographics. This not only contributes to academic customer loyalty research, but also is very helpful in guiding banks to segment customers and set up tailored marketing strategies. Unlike general retail consumers, financial consumers purchase new products or services at specific times in their lifecycle, reflecting their needs at these stages. For example, young consumers in college are more likely to use education loans as financial aids to accomplish their degree. After finding a job and starting to buy their own house, they need mortgage, home equity lines or home equity loans. Later, they may think more about their pension and annuity for retirement. Therefore, customers’ past behavior may well provide an indication of future purchasing patterns (Harrison and Ansell 2002). b) The current study compensates for biases in survey analysis. Most loyalty studies are based on data collected by questionnaires which may include some biases. Firstly, responses in questionnaires might

8 be impacted by subjects’ emotion, social identity, and data collector’s idiosyncrasies (researchers), which limit the reliability and validity of survey responses. As an instance, Mazursky and Geva (1989) conducted a study regarding the relationship between satisfaction and future purchase intention. They measured satisfaction and intention on the same subjects twice by using the same survey. However, the positive impact of satisfaction was found not to be significant based on the second measurement of intention conducted two weeks later. In the current study, we investigated customer retention by using real customer behavior and demographics data from a bank and randomly selecting the sample of customers to avoid subjective effects. c) The current study investigates factors impacting actual customer retention behavior rather than retention attitude or intention. Most existing studies explored factors affecting customers’ propensity to switch or stay, which is a common shortcoming in customer loyalty studies (Baumann et al. 2005; Evanschitzky and Wunderlich 2006).Such studies focusing on propensity have some biases. Firstly, as mentioned before, there is measurement bias for those studies, since researchers usually conducted analysis based on survey data. Intention is usually measured by five or seven-point scales, which causes information to be lost as a result of not capturing the extreme variation in the construct and the

9 within range variation in the construct (Seiders, Voss, Grewal, and Godfrey 2005). Secondly, attitude and behavioral intention were often found to be poor predictors of actual behavior (Baumann et. al. 2005; Chakravarty, Feinberg, and Rhee 2004; Uncles, Dowling, and Hammond 2003). A strong attitude towards a brand may provide only a weak prediction of customers’ repurchase behavior (Uncles et al. 2003), since the relationship between intention and actual behavior is moderated by various variables (Seiders et al. 2005). Thus, management strategies and inferences based on the studies of behavioral intention could be misleading and generate costly service mistakes (Seiders et al. 2005). Thirdly, most of the attitude information used for analysis could not be found in the data warehouse of industries and can not be collected for all customers, since not every customer is willing to answer survey questions. Therefore, management of industries could not determine the propensity to stay or switch for each customer. The current study identifies factors impacting customer retention among data usually available for each customer in an actual data warehouse. The results could be directly used for conducting marketing strategies. d) The model developed in the current study also contributes to the body of customer retention or loyalty by capturing customer information dynamically and over time. In the traditional database marketing literature, logistic regression, neural networks, decision tree, and other similar methods are conducted to predict customers’ response as a

10 function of customer or marketing related factors (Bloemer et al. 1998; Homburg and Giering 2001; Wallace, Giese, and Johnson 2004; Veloutsou et al. 2004). Those response measures are usually discrete dichotomous measures, such as purchase or not, respond or not, and so forth. The models are static models. Based on those methods, researchers could get a "snapshot" of customers at various stages of their relationship, but are unable to follow individual customers over time. However, firm- customer relationship is a dynamic entity. The duration of a relationship will change over time due to firm-customer interactions during the whole service procedure, or due to external factors such as changes in the customer’s situation. A model based on cross sectional data to predict a discrete, dichotomous measure such as buy/no buy, respond/did not respond, and the like could not catch these dynamics. As a survival analysis approach, Cox’ s proportional hazard method used in this study includes time-varying variables by using longitudinal data (Weerahandi and Moitra 1995). It could provide better forecasting and prediction by conducting analysis based on actual long term dynamic customer behavior. 2) The second way in which the current study contributes to knowledge is that it provides an approach to forecast customer retention rate, which is fundamental to evaluating customer lifetime value and firm value (Gupta et al. 2004). Additionally, strengths and directions of the relationship between

11 factors and retention duration could guide managers’ strategic and tactical decision making. As the most important intangible asset of a company, estimating customer lifetime value could provide a more accurate understanding of the potential value of a company. Recently, interest in customer lifetime value (CLV) research has increased (Gupta et al. 2004). To calculate individual CLV, the important component is forecasting the time to customer attrition. Gupta et al. 2004 identified three components needed for predicting customer lifetime value: the retention rate of different time periods, the profitability or margin of the all customers, and discount rate. Among those three components, customer retention was found to be the most important for customer lifetime value, even more important than margin, acquisition costs, and discount rate (Gupta et al. 2004). Discount rate and profitability data (margin and costs) for each customer usually can be obtained from financial departments of a company. Therefore, forecasting customer retention becomes the last piece for estimating customer lifetime value or residual customer lifetime value which would be brought to firms by existing customers. After forecasting long-term customer retention based on the current study, one could fully evaluate customer lifetime value and assess the potential value of companies. C LV models are mainly strategic models that could provide a guide for resource allocation decisions for the entire customer base or segments (Gupta et al. 2004). As one type of customer profitability model, CLV models have a limited capability to guide individual customer level marketing decisions (Libai

12 et al. 2002). Compared to other long term retention models which usually use “average retention rates, or sample-based switching probabilities and customer perceptions to calculate CLV of an average customers or the expected value of all customers” (Libai et al. 2002), the survival model developed in this study could complement the CLV model and give guidance to individual customer level marketing decision. 3) This study has several managerial implications. The current model provides a way to understand which group of customers managers need to focus on, how to segment customers, and how to affect customer retention. Generally speaking, based on the model developed in the current study, managers could partition their customer base into behaviorally and demographically homogeneous groups and then evaluate customers in terms of their estimated risk of attrition. This could be used in strategic and tactical decisions. The strategic decisions for a company could be, for example, identifying the characteristics of its customers who might not attrite in the short run and developing marketing strategies for targeting those customers in the long run. The tactical decisions might be, for example, how to allocate short-term resources among marketing activities and factors, which means deciding on which customers to serve and how to allocate their resources more effectively.

13 Organization of the Dissertation

This dissertation is composed of five chapters. The first contains the overview of the importance of customer retention research, the objectives, and significance of this study. Chapter two provides a review of existing literature in the area of customer satisfaction, switching cost, and customer retention. Hypotheses are also developed in this chapter. Chapter three describes the methods of survival analysis which was employed to investigate the impacts of behavioral and demographic factors on customer retention. It also includes sample and data description. Chapter four presents the results of statistical analysis. Chapter five consists of a discussion of the findings and provides conclusions, academic contributions, and managerial recommendations. The limitations and suggestions for future research are also included in this chapter.

14

CHAPTER 2 – LITERATURE REVIEW

This chapter contains a review of literature. The review is conducted in order to develop a probabilistic model for customer retention with respect to financial product consumers. The chapter contains a general review of the concept of customer retention, the factors and triggers for switching, and strategies impacting customer retention. Additionally, the nature of the relationship between customers’ demographics (age, education, income, marital status and gender, and time pressure) and customer retention, and the impacts of firm-customer interaction factors (service usage, tenure, cross-sell, channel usage, and product complexity) on retention are explored. Hypotheses are developed based on the literature review.

Customer Retention

The research regarding customer loyalty or retention can be traced back to the 1920s (Copeland 1923). Researchers defined customer loyalty as a biased (nonrandom) repeat purchase of a specific brand over time by a consumer (Jacoby and Kyner 1973) and operationalized customer loyalty in three ways. First, customer loyalty was measured as a form of repeat purchasing of one specific product or service. Second, loyalty was measured by the proportion of purchases devoted to a given brand (Brody and Cunningham 1968). Third, some researchers used the probability of purchase as a measurement of loyalty (Farley 1964). Day (1969) criticized the measurement of

15 customer loyalty based only on behavior, which could not distinguish true loyalty from spurious loyalty, which referred to customers’ consistent purchasing due to no alternatives available or inertia. He advocated adding an attitudinal dimension to behavioral loyalty. Later, Jacoby and Chestnut (1978) claimed that belief, affect, and intention should be examined in customer loyalty analysis. In 1999, Oliver defined loyalty as a commitment to repurchase a preferred product or service consistently in the future. He introduced a detailed framework of loyalty comprising four consecutive phases. He suggested that different aspects of loyalty could not emerge simultaneously, but one by one over time. Customers firstly become cognitively loyal, and then affectively loyal, followed by conatively loyal, and finally behaviorally loyal (Oliver 1999). Cognitive loyalty is the first stage, which refers to the existence of beliefs of preference in one brand (Oliver 1999). Cognitive loyalty is impacted by consumers’ experience and price equity. Customers with cognitive loyalty are more likely to switch when they believe an alternative to have a better cost-benefit ratio (Evanschitzky and Wunderlich 2006). Affective loyalty reflects a favorable attitude or liking based on satisfied usage (Oliver 1999). Oliver (1980) mentioned that, consumers form expectations of product or service performance prior to purchase. Then, they compare the expectations with actual performance based on subsequent purchase and usage. When perceptions exceed expectations, customer satisfaction develops. Satisfaction with past experience was suggested as an affective process and also a seed for customer loyalty (Oliver 1997; Bloemer and Kasper 1995).

16 Conative loyalty was defined as attitudinal loyalty accompanied by an intention to a behavior, such as repurchasing (Oliver 1999). Service failure has a negative impact on conative loyalty. When facing frequent service failures, customers are more likely to switch to an alternative (Oliver 1999). Action loyalty relates to the conversion of intentions to action with a willingness to overcome impediments to such action (Oliver 1999). Not all intentions could be transformed into actual action. Only if the customers’ other three loyalties were converted to actual repurchase behavior, would the current provider have direct financial benefits. In the current study, because of the objectives of the study, we used actual customer behavior to measure retention. Customer retention refers to a customer’s decision to continue business with a particular service firm (Stewart 1998). Customer retention is operationalized as the customer still appearing in the firm’s database of customer accounts in the next time period. Should the customer not appear, s/he is regarded as having defected.

Antecedents of Customer Retention/Loyalty

In existing studies, several factors were found to impact customer attrition/switching behavior, customer retention attitude and intention, and customer loyalty (Al-Hawari 2006; Boulding et al. 1993; Ranaweera and Neely 2003; Roos, Edvardsson, and Gustafsson 2004; Zeithaml et al. 1996). Those factors could be classified into categories based on different criteria, or “triggers”.

17 Triggers of Customer Attrition/Switching Researchers claimed that customers’ switching behavior could be impacted by three triggers as follow: situational triggers, influential triggers, and reactional triggers (Roos et al. 2004). Situational triggers are defined as changes in a customer’s situation outside of the customer relationship (Roos et. al. 2004). Those changes are related to customers’ own lives and not necessarily associated with service providers. Those might result from changes in customers’ demographics, work situation, financial circumstances, usage of spare time, changes in mobility (car, local means of conveyance), and product expertise (Roos et. al. 2004). For example, due to aging, financial industry customers might no longer need some products or services, such as student loans or mortgage. Those customers may defect or switch to other providers with more appropriate product categories. Influential triggers are factors related to the competitive situation (Roos et. al. 2004). The typical situation is one in which a new service provider tries to penetrate a market. Besides competitors’ efforts to increase their market share, company’s attributes could also impact customers’ switching behavior. Bank customers might also be influenced by other customers, or some other factors. Reactional trigger refers to critical incidents of deterioration, such as a decline in performance (Roos et al. 2004). Those incidents might redirect a customer to evaluate the current product or service, which puts customers on a switching path. For example, reactional triggers might be core service failures, failed service encounters, or customers’ perceived decline in service quality (Keaveney 1995).

Full document contains 127 pages
Abstract: Recently, in both marketing theory (academia) and practice (industry), the emphasis in relationship marketing has shifted to long term customer relationship management emphasizing customer retention or loyalty. This study has two main purposes: (1) to investigate the impacts of selected firm-customer interaction behavior and demographic characteristics on customer retention behavior in the financial industry, (2) to compare the results of a static method of analysis (logistic regression model) and a dynamic method (Cox's hazard method) for customer retention. The statistical analyses conducted employing two cohorts showed that the Cox model was stable across cohorts. Using Cox's hazard method, all of the independent variables were found to be significantly related to customer retention, while some of demographic factors (for example, tenure, marital status and gender, and channel usage) were not found to significantly affect customer retention in the logistic regression. The results show that with the enhancement of interaction between a bank and its customers by increasing customers' service usage, cross-buying, tenure experience, and complicated product usage, customers were more likely to stay longer with the bank. Additionally, age, education, and income were found to be positively related to customer retention, while customers with higher time pressure were less likely to defect. Single females were found to be least likely to attrite, while married customers were most likely to default or switch to an alternative. Single males were more likely to stay than married ones, but more likely to churn than single female customers. Finally, managerial implications were drawn, and limitations and potential for further study were discussed. [PUBLICATION ABSTRACT]