!"#$%&'(%&)*+&,$*-.& & /012310& & & & 456*+'"-$+7&)+8&"%98&:8;8;8:8&<**%)*+-& 4*=$%)&+*'"*>*+7&)+-8&"%98&;8?8@8&:*5A*#(%-! & & B%"'*+-".*".&'(%&;#-.*+)(#& ;#-.*+)(#&C5-"%*--&4=D$$A& E(=5A.F&$G&H=$%$#"=-&(%)&C5-"%*--&IEHCJ& :84=8&"%&C5-"%*--&4.5)"*-& &
!
!D*-"-&:84=8&"%&C5-"%*--&4.5)"*-&<"%K"%9&E(#"A"(+".F&
.$&?5-.$#*+&<"G*."#*&
L(A5*&
4)*5%6/7%*08&$6/71&)*&6"#&%*98/#*:#&)9&6#0;&;#;.#(& 90;%8%0(%61&%*&:8%#*6&$#('%:#&6#0;$&)*&6"#&"#%5"6&)9&6"#& </$6);#(&4%9#6%;#&=08/#&06&0&>(%'06#&.0*?&1
"#$%&'$!
The thesis lying in front of you is the result of several months of work and positive stimulation from the people surrounding me. This thesis is also the last part of my Master of Science degree in Business Studies at the Universiteit van Amsterdam.
Besides my study Business Studies, I had the opportunity to work in the private banking business as an assistant to investment advisor. The combination of study and working in a challenging environment appeared to be a great accelerator for my development in terms of knowledge, experience and personality. I am grateful that the company facilitated me to conduct my thesis research under their supervision.
My interest for the topic developed during my time at the company when wondering what clients experienced when their client service team mutated over the years. Examples of questions that arose were, what would happen to the loyalty of the client towards a bank when he
experiences a different banker every year. Also, how do clients respond in the case of replacement of an older and more experienced investment advisor by a younger and less experienced
investment advisor? The link to Customer Lifetime Value was made when following the master course Strategic Marketing where this topic was covered. Using this relative young concept enabled me to introduce the Customer Lifetime Value topic in the company and also to address the questions concerning team composition and the effect on performance.
Despite the fact that the past period was not easy and the moment of finishing this thesis took me a while longer than expected, I would like to thank my girlfriend for her unconditional support and remarks during the process. I also would like to thank my family and especially my eldest sister, Henriëtte, for the many hours of discussion and feedback on the topic. Of course, thanks go out to my direct colleagues at the bank for their stimulation, interest and contribution to the research. It is planned to present the results of the study to the management and direct colleagues of the company. The last words of thanks are addressed to my thesis supervisor dr. ir. Mark Leenders for his constructive feedback and advise.
Timon van der Roest
2 Abstract
This study investigated the link between team member familiarity and the Customer Lifetime Value. Team member familiarity relates to the knowledge one has about his or her workplace. There is discussion in the academic world whether team member familiarity is positive to performance. On one hand researchers state that familiarity is able to breed psychological safety and improves the team coordination. These effects are positive to performance. On the other hand, researchers who found that team member familiarity is not beneficial to
performance state that familiar teams are less flexible and have a bias to discussions. The performance in this study is measured by a relative young marketing concept, Customer Lifetime Value. This concept is able to measure the value of customers by discounting the future cash flows related to the relationship with the client. The metric can be used for corporate valuation and as an informational tool. My main research question is whether team member familiarity influences the Customer Lifetime Value. To answer the research question a case study in the private banking sector was conducted. At an annex of a financial
institution active in the private banking sector, client service team changes are described for a period of four year. This enables the measurement of team familiarity. Also the Customer Lifetime Values for 178 clients are calculated who are being served by these client service teams. Regression models were developed to answer the research question.
The results from this case study revealed that team familiarity influences Customer Lifetime Value in a positive way. However, this effect is relatively small and non-significant when the control-variables on Customer Lifetime Value are introduced in the regression model namely, the assets under administration and the portfolio model. Based on the data available and the multiple regression model developed no evidence was found that the effect of team member familiarity decreases at a certain point in time. Also the data did not show that international clients ascribe more value to the relationship with the client service team than domestic clients.
!
!
!
!
(&)*$!+%!'+,-$,-!
!
1& M,HE;?H NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNO& 2& ABSTRACTNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNP& 3& INTRODUCTIONNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN/&
3.1& INTRODUCTION TO RESEARCH TOPIC_____________________________________________________ 5&
3.2& PROBLEM STATEMENT AND CONCEPTUAL MODEL___________________________________________ 6&
3.3& APPROACH ________________________________________________________________________ 6& 3.4& RELEVANCE _______________________________________________________________________ 7& 3.5& STRUCTURE OF THE RESEARCH_________________________________________________________ 8& 4& !QHR,S NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN1&
4.1& THE EMERGENCE OF USING TEAMS IN ORGANIZATIONS_______________________________________ 9& 4.2& TEAM COMPOSITION________________________________________________________________ 10& 4.3& THE RELATION BETWEEN TEAM FAMILIARITY AND TEAM PERFORMANCE________________________ 12&
4.4& CUSTOMER LIFETIME VALUE__________________________________________________________ 14&
4.5& ISSUES RELATING TO ETHICAL BEHAVIOUR_______________________________________________ 17& 5& !QH&ETU;U?T;<&4H?!R,NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN V1&
5.1& BANKING INDUSTRY CHARACTERISTICS_________________________________________________ 19&
5.2& THE CREDIT CRISIS: CAUSES, IMPACT, CURRENT STATE OF AFFAIRS AND EXPECTATIONS____________ 20& 5.3& PRIVATE BANKING _________________________________________________________________ 23&
6& ,H4H;,?Q&WH4TXU NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN O0&
6.1& HYPOTHESIS DEVELOPMENT__________________________________________________________ 27&
6.2& RESEARCH CONTEXT________________________________________________________________ 28& 6.3& RESEARCH STRATEGY_______________________________________________________________ 29& 6.4& DATA COLLECTION AND SAMPLE SELECTION _____________________________________________ 29& 6.5& VARIABLE MEASUREMENT ___________________________________________________________ 31&
6.5.1& Customer lifetime value calculation ________________________________________________ 31& 6.5.2& Familiarity measurement ________________________________________________________ 34& 6.5.3& Other variables ________________________________________________________________ 35& 6.5.4& Control variables ______________________________________________________________ 35&
7& ,H4B<!4NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN P0&
7.1& DESCRIPTIVE STATISTICS ____________________________________________________________ 37& 7.2& HYPOTHESIS TESTING_______________________________________________________________ 39&
7.3& CONCLUSION______________________________________________________________________ 45&
7.4& LIMITATIONS______________________________________________________________________ 46& 7.5& CONTRIBUTION TO RESEARCH AND MANAGERIAL IMPLICATIONS______________________________ 48& 8& ,HEH,HU?H4NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN /2&
!
3 Introduction
3.1 .,-#+/0'-1+,!-+!#$2$&#'3!-+41'!
Confidence is the first word that appears when thoughts about the financial world are crossing my mind. It can be seen as the most precious asset of the financial world and banks are built on it. Since the beginning of the economic crisis, as we know it today, banks had difficulty to preserve, not even to mention gain, the confidence of their clients and the general public. Everyone remembers the devastating effects of the loss of confidence of the public and business on for example Lehman Brothers, AIG, Merrill Lynch and HBOS in the autumn of 2008. In those cases the confidence of the public and business devalued very rapidly and within a very short period of time the end was near. The result; trillions of dollars were transformed or burned. Except the importance of confidence for the financial world, it is also important for the development of the relationship between the members in a team in
companies. Confidence in the capabilities of the other members will result in a more agile and capable team resulting in better performance. An environment in the financial world where teams are often used to serve clients is the private banking segment.
Private banking is a business area where high net worth and high-income private individuals are offered tailor-made financial solutions. Financial institutions active in this segment have nowadays almost USD 14.5 trillion in assets under management (Scorpio Partnership, 2009). Here again, confidence of the private banking customer in the team that serves the client and the financial institution where the money is held is vital for a successful, long-term and durable business relationship.
There are many variables that can influence the performance realized by a team. For a part they are realized by more external factors such as the economical outlook and corporate strategy. For the other part internal factors such as team composition and team member familiarity play an important role on performance. This thesis will focus on the effect of team member familiarity on team performance. The private banking business segment will form the specific research environment. Team member familiarity relates to the knowledge team members have of each other or to the position in the team they fill in and the tasks that relate
to that position. The output factor produced by the team, in the form of financial performance, will be measured with the Customer Lifetime Value (CLTV from now on). The determination of CLTV is a relatively young research domain in marketing. Pfeiffer et al defined CLTV as
‘the present value of the future cash flows attributed to the customer relationship’ (2004, p.
10).
3.2 "#+)*$5!2-&-$5$,-!&,/!'+,'$4-0&*!5+/$*!
The main goal of this research is to investigate the relationship between team member familiarity and the effect on the Customer Lifetime Value of a client and to determine whether senior management should stimulate long-term teams where the number of member changes needs to be minimized. This results in the following problem statement:
‘What is the relation between team member familiarity in client service teams on the customer lifetime value of a client at a private bank?’
The conceptual model in this thesis is the following.
3.3 644#+&'3!
Whether there is a link between team familiarity and CLTV will be studied by conducting a case study in the private banking annex of a large Dutch financial institution. At this private
bank, client service team are serving the whole package of financial needs of a portfolio of a private client. Private banking clients have an amount of EUR 1,000,000 and above free for investment opportunities in the form of savings, deposits, treasury and equity products. The annex of the private banking business unit, used for this case study, serve clients from the segments international clients, professionals & executives and institute & charity banking. This study hypothesizes that a higher levels of team member familiarity correlates positively with the CLTV.
There are four major reasons why this study has been conducted in the financial sector and in specific the private banking industry. At first, the financial sector is experiencing a turbulent era after the emergence of the credit crisis. Confidence of the public and business dropped to almost historic low levels and go together with a suspiciously attitude against people working in the financial sector. In other words, the financial industry is a hot topic. Second, many service organizations are using state of the art CRM systems describing individual customers and are thus a great source for this type of research. Third, private banking is aimed at developing long-term, durable relationships with their client base, which is useful since this study has a longitudinal aspect. The fourth and last reason is that the private banking segment often uses teams to serve their customer base. The core of the client service team can consist of a banker and an investment advisor. Besides these two team members, there are in most cases two assistants, one for the banker and one for the investment advisor.
3.4 7$*$8&,'$!
This study is relevant in a practical as well in a theoretical sense. The practical relevance is expressed by the fact that knowledge generated by this research can be used by the
organization where this case study is conducted. They can use this knowledge as a handle to decide whether to stimulate job-rotation in the organization or to stimulate team cohesion. Besides CLTV is a new construct for the organization. Knowledge from the study can be used for further and more advanced development of this construct and could eventually result in implementing it as a tool to, for example, identify their most profitable clients on the long-term. The main theoretical relevance is that this study adds to current research. A new and
specific link is developed between the relative new construct of CLTV and team composition elements and more specific, team member familiarity.
3.5 9-#0'-0#$!+%!-3$!#$2$&#'3!
The research is structured as follows. In the third chapter, previous research and theories on the relations between team performance and team composition and to be more specific, team familiarity are examined. Since performance in this study will be measured by CLTV
literature in this area of research will also be examined. The fourth chapter will provide an overview of the context in which this study is conducted. This chapter will cover a brief description of the financial world, the credit crisis and the private banking sector. The fifth chapter entails the research and data collection method. The sixth chapter will be used to present the used data and the results that were derived from the study. In the last chapter the main findings and conclusions will be discussed. Also managerial implications, contribution to knowledge, limitations and suggestions for future research will be elaborated.
Figure 2 Thesis structure
T%.+$)5=."$%&
?D(6.*+&V7& T%.+$)5=."$%& Y! "%.+$)5=."$%& Y! &6+$ZA*#&-.(.*#*%.& Y! #*.D$)& Y! +*A*'(%=*&!D*$+*."=(A&
+*-*(+=D&
?D(6.*+&P7& !D*$+F& Y! .*(#&=$#6$-"."$%& Y! +*A(."$%&.*(#& =$#6$-"."$%&.$&& 6*+G$+(#(%=*& Y! =5-.$#*+&A"G*."#*& '(A5*& ?D(6.*+&[7&?(-*& -.5)F&=$%.*\.& Y! ]"%(%="(A&-*=.$+& Y! =+*)".&=+"-"-& Y! 6+"'(.*&Z(%K"%9&H#6"+"=(A&
+*-*(+=D&
?D(6.*+&/7& ,*-*(+=D&)*-"9%& Y! DF6$.D*-"-& Y! =$%.*\.& Y! -.+(.*9F& Y! )(.(&=$AA*=."$%& Y! '(+"(ZA*-&;%(AF-"-&
?D(6.*+&37& ,*-5A.-&(%)&& (%(AF-"-& Y! )*-="6."'*-& Y! =$++*A(."$%-& Y! +*9+*--"$%-&?$%=A5-"$%&
?D(6.*+&07& ?$%=A5-"$%-& Y! =$%=A5-"$%-& Y! A"#".(."$%-& Y! =$%.+"Z5."$%&.$& K%$>A*)9*& Y! "#6A"=(."$%-&G$+& #(%(9*#*%.& Y! -599*-."$%-&G$+&G5.5+*& +*-*(+=D&4
(3$+#:!
4.1 (3$!$5$#;$,'$!+%!021,;!-$&52!1,!+#;&,1<&-1+,2!
The last decades a transition occurred on the organizational level as well on the level of research. The work-structure changed from being based on the individual level to less hierarchal structures using various types of teams to get the job done. This resulted that the vast majority of the companies are using some form of teams to fulfil their corporate mission or tasks (Harrison, Mohammed, McGrath, Florey, & Vanderstoep, 2003, p. 634). Rationale for the use of teams is based on the assumption that the quality of decisions made by groups with diverse expertise will be higher than decisions that are made by individuals (Gruenfeld, Mannix, Williams, & Neale, 1996, p. 1). This emergence of using teams in organizations gained interest of the scientific world and resulted in a great deal of research on this area. A part of this research was aimed on issues relating to the composition of a team and the effects on performance. Specific interest of some researchers was aimed at the effects of team member familiarity on the team performance, which originated from the demands for greater efficiency and flexibility.
Guzzo and Dickson (1996) discussed in their article the definition of the word team. They stated that the word team is a replacement for the word group. The definition for group was derived from work from Alderfer and Hackman and was defined as; ‘a work group is
made up of individuals who see themselves and who are seen by others as a social entity, who are interdependent because of the tasks they perform as members of a group who are
embedded in one or more larger social systems and who perform tasks that affect others’
(Guzzo & Dickson, 1996, pp. 308-309).
Nowadays organizations rely on many kinds of teams for the development of products, delivering and improving services to their consumers and for the management of projects. A business environment that is getting more and more complex and uncertain causes the reliance on groups as the fundamental unit of the organization structure. The use of groups enables a more flexible approach to the dynamic outside world because decision-making is decentralized (Oh, Chung, & Labianca, 2004, p. 860). McGrath (1984) characterized different
kinds of team-types, which were gathered in three classes. These team-types are natural, quasi and concocted-groups. This research will focus on the last group since this is the most
prevalent type in organizations. Those teams have a length of a few days to multiple years.
Available theory on team composition will be covered in the next section of this chapter. The third paragraph discusses studies on the relation between team member familiarity and team performance. The fourth paragraph will discuss literature on CLTV, the performance output factor that will be used in this study. In the last paragraph hypotheses development process will be discussed.
4.2 ($&5!'+54+21-1+,!
When a new team is initially composed, the organization has to deal with the important question which members to choose. These questions are important because they will affect team performance in some way. This process relates to the research area of team composition, which concerns what team members contribute to the group in terms of skill, ability and disposition and how the combination of these attributes adds to outcomes (Matthieu, Maynard, Rapp, & Gilson, 2008, p. 433). Meta-analytical research by Stewart (2006) and Mathieu et al. (2008) provided a very rich understanding of this research topic and the relation with performance. This paragraph will highlight the key findings of these studies.
The first is the use of mean values or aggregated characteristics. These characteristics can relate to personality. For example, a study from Bell in 2007 showed that average team conscientiousness, agreeableness, extraversion, emotional stability and openness to
experience related positively to performance. Another relation can be made with
competences. A meta-analytical research by Devine and Philips concluded that the amount of cognitive ability would predict team performance. The effects will be most visible for
decision-making tasks.
The second is member diversity. This line of research suggests that individual
characteristics will not aggregate in a linear fashion and there is thus a need for assessing the fit between individual team members (Stewart, 2006, p. 33). When relating this construct to performances there are various outcomes. Certain studies support the theory that
heterogeneity is desirable for teams. Underlying arguments for this theory is that when team member heterogeneity is stimulated this will result in teams with diverse points of views and skill sets. Other studies state that teams where homogeneity between members is stimulated will experience less conflict. The meta-analytic study by Stewart founds that heterogeneity has weak relationships with performance (Stewart, 2006, p. 42).
The third is demographics. In the literature concerning team composition, demographic variables are clustered under surface-level diversity and comprehend for example age, gender, race and marital status (Harrison, Price, Gavin, & Florey, 2002). When evaluating the research in this area Harrison, et al. (2002) came to the conclusion that team members will have a less positive attitude with those who do not resemble themselves. They thereby tend to commit to fewer social connections. Another study found that diversity in age and tenure could be beneficial to performance while other studies state that they are
unfavourable for team performance (Matthieu, Maynard, Rapp, & Gilson, 2008, p. 438). Noteworthy here is that the study by Harrison, et al. (2002) revealed that a demographic surface level variable, such as age, interacted with time so that these effects were eliminated as teams get to know each other better.
The fourth relates to team size. In this case there are also different effects on teams. Certain studies revealed that larger teams are harder to coordinate while other studies found that larger teams are more effective. One argument behind the last finding is that larger teams are better able to obtain resources that might be useful for completing difficult tasks in
complex and uncertain environments (Stewart, 2006, p. 33). In general, the meta-analytical study by Stewart founds that project teams should be smaller than production teams.
There is not an overall clear consensus in the findings of the different studies on how individual team composition variables affect performance. However there is evidence that individual ability and disposition relate positive to team performance. Stewart states: ‘Who is
included in the team matters’ (Stewart, 2006, p. 44).
Harrison, et al. described the options companies have to compose a team based on earlier research (2003, pp. 636 – 637). The first option is to choose members who are experienced with the task or project. The second option is to choose members on the basis of how well they are able to work together and are thus selected based on their personal characteristics. The third option is to choose members based on the fact that they do not know each other. Choosing this last option organization suppose that team member familiarity will build over
time. Harrison, et al. stated that the knowledge concerning the last option is little because not much is known about the impact of interpersonal experience on team performance (Harrison, Mohammed, McGrath, Florey, & Vanderstoep, 2003). This issue relates to team member familiarity, which will be discussed in the next paragraph.
4.3 (3$!#$*&-1+,!)$-=$$,!-$&5!%&51*1-:!&,/!-$&5!4$#%+#5&,'$!
Team member familiarity is a broad construct and relates to the knowledge a team member has about the workplace to which he or she has been appointed. This workplace is defined by Goodman and Leyden as a ‘unique configuration of people, machinery, physical environment,
performance strategies and jobs’ (Goodman & Leyden, 1991, p. 1978). These unique
configurations are the source of performance differences and it is thus essential to become familiar with the unique configuration, enabling effective performance. Team member familiarity in the context of this study is related to knowledge about other team members, the client portfolio and team members role specific tasks. Once a team is composed familiarity will increase positively related to time (Watson, Michaelsen, & Sharp, 1991).
The effects of familiarity on performance and team outcome have been studied in research and the outcomes are not consistent with each other. On one hand there are studies that identified positive relations between team member familiarity and the team performance. A study in the coal mining industry under 26 teams revealed that lower level of familiarity led to lower levels of productivity, therefore the researcher advocates for familiarity enhancement (Goodman & Leyden, 1991). In a longitudinal study by Watson, et al. where groups were studied spending more than 30 hours in decision making tasks they found that team members over time became more familiar to each other. In this way they were better able to identify the specific needed information and the evaluation of the groups’ input which enables better decisions (Watson, Michaelsen, & Sharp, 1991, p. 807). Jehn and Shah focussed in their studies on the performance of friendships versus acquaintances (Jehn & Shah, 1997; Shah & Jehn, 1993). They showed that friendships, because of the higher level of familiarity,
performed better than acquaintances. In a recently published case study at a software
company the researchers revealed that team familiarity has a positive effect on adherence and quality (Huckmann, Staats, & Upton, 2009).
On the other hand there are studies that identified negative relations between team member familiarity and team outcomes. In an experiment, formal interventions were exposed on familiar teams. This resulted in lower performance because of lower flexibility caused by the fact that familiar groups have norms and routines (Okhuysen, 2001, p. 795). In a study by Petersen and Thompson teams of friends and teams of strangers were compared when they were negotiating. It seems that teams of strangers achieved better results than the teams of friends (1997, p. 379). A study under students revealed that teams with task and team experience displayed a greater bias toward discussions resulting in poor results in descision making (Kim, 1997, p. 172).
Besides a negative or positive side there are also studies which indicated a mixed effect. For example, a study on group composition revealed that familiar groups were better equipped in the psychological sense to resolve problems in opposition to stranger groups who have different perspectives and know different facts. They however lack the personal ties and interpersonal knowledge to communicate effectively (Gruenfeld, Mannix, Williams, & Neale, 1996, pp. 10-11). Research by Katz (1982) revealed a mixed effect on the positive effects of team familiarity. He found that team familiarity in the first period is developing positively and can develop negative in the second period. The turning point in that specific research occurred after about three to five years. As a reason for this pattern of team familiarity development Katz stated that groups that are together for a great amount of time would be secluded from information sources.
There are various explanations for the positive relationship between team familiarity and team performance. Huckmann, et al. divided these reasons into two subgroups,
coordination and willingness to engage in a relationship (2009, pp. 87-88). Familiarity is able to improve the coordination to act as a group in a well-organized way. Also related to
coordination is the growth of team human capital because of shared experiences of team members. When looking at the willingness to engage in a relationship the following
explanations are derived. The first is that familiarity breed’s psychological safety, which will positively influence performance. Second, the shared experiences of team members leads to trust which results in more information sharing and an improved quality of the information.
This paragraph reviewed available literature on how team familiarity relates to team
performance. The next paragraph will review the literature on the area of Customer Lifetime Value
4.4 >02-+5$#!*1%$-15$!8&*0$!
In the last decennia more and more firms are concerned with the relationships they built with clients. Before that phase, firms businesses with customers were treated as discrete activities. In this period there was a focus on costs, product lines and competition. Over time firms realized the importance of loyalty and they switched from product-centric approach to a customer-centric approach (Jain & Singh, 2002, p. 35). Product-centric firms will treat products as assets and are focussing on increasing the profitability of each product. The customer-centric approach on the other hand embraces customers as assets and is focussed on acquiring and retaining customers. The competitive advantage in this case is built on the customer database of the firm. It is this state of mind when Customer Relationship Management (CRM from now on) emerged.
CRM is defined as the process that aims at the retention and development of relationships with individual clients so value is created for the company as well for the customer (Verhoef, 2004, p. 81). Panda adds to this definition that CRM must enable firms to terminate relationships with customers and stakeholders when necessary (2002, p. 63). Gupta, et al. suggested three reasons why CRM became more important (2006, pp. 139-140). The first reason is that there is an increasing pressure in companies to make marketing
accountable. Marketing metrics such as for example brand awareness, attitudes and market share are not useful enough to calculate return on marketing investments. The second reason is that financial metrics are also not appropriate because research demonstrated that not all customers are equally profitable. With this in mind it is reasonable to ‘eliminate’ those less profitable customers. The third reason is related to the emergence of advanced information technology that enabled companies to gather huge amounts of customer data on different aspects such as transactions and demographics.
CRM provided management with tools for decision-making and evaluation on which customers to engage in a relationship and to monitor the, potential, value of those customers. Because one must know whether or not a relationship is lucrative and thus one must be able to quantify the relationship. One tool to quantify the relationship is the CLTV, which aims at defining the total net income a company can expect from a customer. Pfeiffer, et al.,
discovered inconsistencies in the use of the term CLTV and proposed a definition based on their literature study. They defined CLTV as the present value of the future cash flows attributed to the customer relationship (2004, p. 10).
CLTV can be an informational tool to management in several ways. The first is from Jain and Singh (2002) stating that CLTV can help, as stated earlier, to quantify the
relationship of the firm with their customer base enabling them to make more informed decisions. CLTV will also help the firm to understand which customers are the most lucrative and in this way it is a metric for marketing resource allocation (Jain & Singh, 2002, p. 36). Second, a stream of study on CLTV demonstrated the link with firm value (Bauer,
Hammerschmidt, & Braehler, 2003). Gupta, et al. set the tone on this area with their study of five companies in several industries (2004). For three of the five firms a clear link was demonstrated between valuations on basis of the CLTV to the valuation of shareholders. Only for the Internet based firms the link was missing which could have been caused by the
turbulent developments in that specific market. A study on CLTV in the European retail bank industry recognized the need for customer valuation based on the growing number of merger and acquisitions (Haenlein, Kaplan, & Beeser, 2007). The number of mergers and acquisitions in the retail bank environment rose in the period before the economic crisis because banks wanted to expand their scale of operations. This resulted in questions about customer
valuation. The researchers developed a CLTV model to fit in the retail bank environment and validated the use by testing it on 6.2 million datasets. The possible applications of this tool in the financial service industry are numerous (2007, p. 12). For example the derived figures could be used to group customers based on CLTV and a specific CRM strategy could then be developed and applied on this group. Second, setting acquisition allowances by using the model. And third by optimizing the client base by for example looking for unprofitable customers and decide whether the relationship needs to be terminated or that the relationship should be enhanced. Venkatesan and Kumar (2004) also derive similar results of the success of CLTV in their study. They revealed that customers selected on basis of their individual CLTV, provided higher profits than customers who are selected on basis of other metrics.
The concept of CLTV can be compared with the net present value calculation of an
investment project. Instead the expected future cash flows generated by the customer is used to determine the present value of the stream of cash flows (Verhoef, 2004).
The basic formula for calculating CLTV is stated below.
!
CLTVi,0 = profiti,t (1+ d)t
t= 0
"
#
(1)The way CLTV can be calculated is numerous and the models range from basic formulas to complex mathematic models (Berger & Nasr, 1998). Verhoef stated that almost all models contain three basic components (2004, pp. 83-90).
The first component used to calculate the CLTV is profit. Profit is composed of proceeds minus costs. Proceeds are divided in purchasing-related profits and non-purchasing-related profits. The first are profits that originate from transaction with the company. These purchasing related profits can further be distinguished in three dimensions; length, depth and width. Length relates to the time that a client keeps making transactions with the company, depth relates to the volume of the transactions and width relates to the number of services or products the client uses from the company. Costs are categorized in client specific and non-client specific costs. Client specific costs are costs that can be assigned to individual non-clients and can vary per client. Non-client specific costs cannot be assigned to individual clients. Those costs are for example marketing and IT related activities.
The second component concerns the discount ratio. This ratio describes the cost of capital while it also can be used as a measure for uncertainty or risk in the future. Literature does not give a lot of attention to this important component. Most of the time the costs of capital or the Weighted Average Costs of Capital is used.
The third and last component is the time frame. The basic formula assumes that the profit should be calculated to the infinite. It is however impossible to foresee the
developments on this time frame. It therefore seems to be more reasonable to use a limited timeframe. Rust, Zeitalm and Lemon argue to use the planning period of the company, which in most cases lies between three and five years.
Question remains how accurate the measurement of the CLTV is. This must be questioned because prediction is involved in the calculation of CLTV and historical data is not a very good predictor of future value. There are several studies that put this issue to the test. The first study tests the use of CLTV to identify the future most profitable customers (Malthouse & Blattberg, 2005). Detecting these customers is one of the most important goals of CLTV. Their research revealed that of the top 20% customers based on CLTV, approximately 55% is misclassified. Of the bottom 80% ‘normal’ customers based on CLTV, 15% is misclassified and should deserve special attention (Malthouse & Blattberg, 2005, p. 10). Another study by Gupta, et al. that is already discussed, indicated that the level of accuracy for CLTV
calculation could depend on the type of market in which a company is active (2004). Concluding you could state that CLTV calculation can be accurate in markets where the behaviour of customers is predictable.
4.5 .220$2!#$*&-1,;!-+!$-31'&*!)$3&81+0#!
A point of attention has to made regarding to ethical behaviour and the relationship with marketing and the financial services industry, the specific research environment of this study. Despite decades of research on topics relating to ethics and ethical behaviour the academic world has not been able to develop a universally accepted definition of ethic or ethical
behaviour. Regarding business ethics, Robin tried to develop an applicable concept definition and defined ethical behaviour as a goal to treat stakeholders with fairness and respect in their naturally occurring exchanges with business (2009, p. 148). Briefly, the ethical issues relating to financial industry, the use of CRM systems and client segmentation are reviewed.
Ethical behaviour is playing a central role in the financial industry and especially in the current economic situation in which the public doubts the ethical correctness of this sector. Ethical sales behaviour in this industry is important to built long-term relationships with customers (Wray, Palmer, & Bejou, 1994). Sales people in financial organizations are the most visible elements of the organization for the client and often form the only direct contact with the organization. Ethical correct behaviour of this person, or sales team, is thus critical for the relationship. Problem here is that the stress on these person is higher because they are responsible for generating revenues. In a study on ethical behaviour in the financial
industry, the effect of ethical behaviour of the sales person on the client has been tested (Román, 2003). Here unethical behaviour was defined as short-run salesperson’s conduct that enables him to gain at the expense of the customer. The study showed that ethical sales behaviour led to increased trust, loyalty and satisfaction of the customer (2003, p. 927).
The emergence of customer information systems led to questions about the ethical correctness of collecting massive amounts of customer data that are privacy sensitive in a lot of cases. Mason summarizes the issues on this area in the P-A-P-A model that stands for: privacy, accuracy, property and accessibility (Boyce, 2002, p. 114). These subgroups of area’s need to be evaluated when using or designing a CRM system where the privacy aspects should gain the most attention. Reason for this is that privacy is valued in all societies (Boyce, 2002).
Also the use of the CLTV method raises questions about ethical responsibility. Several applications of this instrument urge to assess some ethical topics. For example one of the practical applications of CLTV is to decide which clients to serve and which relations should be ended based on segmentation (Boyce, 2002, pp. 113-114). This form of segmentation based on customer valuation is very common these days and can be seen in the insurance and telecommunication sector. Large corporations using these segmentation techniques have a huge impact on clients because in the end they can restrain the access on certain goods and services. Main concern here is that customer valuation techniques such as CLTV could in the end restrain the access of people to basic needs and services.
This section discussed current literature on the area of team member familiarity, customer lifetime value and some ethical concerns were highlighted. In the next section the specific context of this case study is introduced.
5
(3$!%1,&,'1&*!2$'-+#!
This chapter will cover the current state of affairs in the banking environment. First,
characteristics of the banking environment will be described. Second, the effects and causes of the phenomenon that caused a major impact on the financial world will be described, the credit crisis. The third and last part will describe the sub sector private banking, in which this study is conducted, together with the key players and expectations for the future.
5.1 ?&,@1,;!1,/02-#:!'3&#&'-$#12-1'2!
The primary activities of a bank are borrowing and lending money. Besides these two primary activities banks incorporated more financial services and products over time. These banking services range from transactional services, advisory services, asset management and
proprietary trading to services like clearing and settlement. Regulation of the banking industry varies widely across the globe. For example, the United States is known for their relatively light regulations while China, with their communistic background, experiences a heavily regulated banking environment.
Due to the crisis, which will be discussed later, profits of the world’s top 1000 banks dropped with 85.3% from USD 780 bln. in 2007 to USD 115 bln. in 2008. Such an effect was also seen in the return on capital that decreased to 2.69% coming from 20% (Lambe, 2009). Despite the increasing competition from banks established in the emerging economies of this world it are still the banks from the Western economies that are leading the markets.
From a global perspective the banks that stayed at the basics of banking can be concerned to be the winners of the credit crisis. This resulted that among others Asian banks are gaining more influence. For example Chinese banks like the Industrial Bank of China and the China Construction Bank are now to be reckoned to the top of the industry. One reason for this shift can be found in the more regulated banking environment in that area of the world. Another reason is the economic growth in the emerging markets and the success of the government stimulus. The World Bank still expects an economic growth in China of 7.2% in 2009. An overview of the largest banks in the world in assets can be found on the next page.
Table 1 The Top 1000 Banks 2009. Source: The Banker
Top 15 by Tier 1 Capital ($M) Top15 by total assets ($M)
Bank Country
Tier 1
capital Bank Country
Total assets
1. JPMorgan Chase&Co US 136,104 1. Royal Bank of Schotland UK 3,500,950
2. Bank of America Corp US 120,814 2. Deutsche Bank Germany 3,065,307
3. Citigroup US 118,758 3. Barclays Bank UK 2,992,682
4. Royal Bank of Scotland UK 101,818 4. BNP Paribas France 2,888,728
5. HSBC Holdings UK 95,336 5. HSBC Holdings UK 2,418,033
6. Wells Fargo&Co US 86,397 6. Credit Agricole Group France 2,239,370
7. Mitsubishi UFJ Fin.Group Japan 77,218 7. JPMorgan Chase&Co US 2,175,052
8. ICBC China 74,701 8. Mitsubishi UFJ Fin. Group Japan 2,025,830
9. Credit Agricole Group France 71,681 9. Citigroup US 1,938,470
10. Santander Central Hispano Spain 65,267 10. UBS Switzerland 1,894,423
11. Bank of China China 64,961 11. ING Bank Netherlands 1,853,393
12. China Construction Bk Corp China 63,113 12. Bank of America US 1,817,943
13. Goldman Sachs US 62,637 13. Société Générale France 1,572,721
14. BNP Paribas France 58,175 14. Mizuho Financial Group Japan 1,494,960
15. Barclays Bank UK 54,300 15. Santander Central Hispano Spain 1,460,866
When the above table is projected in the light of profits related to the Tier 1 capital assets the results of the crisis is even more visible. We will only see two banks from the United States and one from Europe in the top 25. For example the average return on assets for US banks is 0.05% and is 3.5% for banks from Nigeria (The Banker, 2009).
5.2 (3$!'#$/1-!'#1212A!'&02$2B!154&'-B!'0##$,-!2-&-$!+%!&%%&1#2!&,/! $C4$'-&-1+,2!
The future of the banking industry will not be the same as before the emergence of the credit crisis. It is therefore important to give a quick overview of the events that lead to the financial crisis and the effects they caused. The paragraph ends with expectations for the future.
The financial turbulence started in August 2007 when a shortage of liquidity in the interbank money market came to light. The central banks of the Europe as well of the United States had to support the financial industry by providing funds. The main causes can be found in:
1. Sub-prime mortgages and securitization of loans
The start of a financial crisis threatened when the burst of the Internet bubble occurred in 2000. The FED, central bank of the United States, lowered the interest from 6% to 1.75% in an attempt to increase the trust in the market and to make sure that
companies and consumers could borrow money at a low price. High risks were taken by banks in the United States who provided people, despite the fact they did not generate enough income, with mortgages against very low interest levels. At the same time, house prices in the United States showed a strong increase in the period 2001 – 2005. Consumers with sufficient income for one house grabbed the opportunity to buy a bigger house or even a second house for speculative reasons.
Banks were aware of the risks related to these mortgages and clustered these loans into financial instruments. Investors in these packages received the cash flow of the original underlying loans while the banks received fees from selling the
instrument. This is the so-called securitization process. In this way banks were able to shift a part of their credit risk to other financial institutions.
2. The deregulation and globalization process
The financial world has expanded in a high pace in the last decades. This expansion originated for a part from the globalization process and for a part due to deregulation in the financial world. For example, the financial system in Europe was not subjected to a very large exposure to abroad financial systems. This exposure did occur when the European Central Bank and the Euro were established. Suddenly, the banks were able to look for new opportunities and to improve their result by looking across borders.
Deregulation and the effects are most evident in the United States. The process made it possible to develop and market new products such as the above mentioned securitization and credit default swaps. The largest deregulatory evolvement started in 1999 with the Gramm-Leach-Bliley Act. This Act removed barriers to competition between the traditional banks, investment banks and the insurance companies. From then on it was possible to participate in all three markets.
In both cases an eagerness for growth and returns appeared which caused higher risk taking. The higher risk taking of banks is recognized by society and the academic world. Noteworthy is that in a article from 2003 about CLTV in the financial world, Panda remarked the effects of higher competition because of deregulation; ‘In today’s deregulated world, members of
financial services industry are continuously forced to seek new ways to gain on their competitors and to outdo one another in terms of effectively reaching to retail customers demands for increasing sophisticated financial products and services’ (Panda, 2002, p. 164).
Before the shortage of liquidity in het interbank money market, a series of events happened that caused the start of the crisis. For years the interest rate in the United States was low but several causes resulted in an increasing interest rate. This caused that the monthly overheads of American citizen increased because their houses were funded with mortgages with a variable interest rate. Many American citizens could not bear the expenses, which effectuated in selling the house, or returning the house as a security to the bank. House prices tumbled and the crisis started. Together with the tumbling house prices, the stock prices also dropped. Banks experienced increasingly more trouble to fulfil their short-term liabilities and some of them even collapsed. This triggered suspicion among banks, corporations and consumers, which is a death sentence in the financial world.
At the moment of writing this study the global economy is recovering. The contraction of the economy that was reigning for three quarters was more or less controlled in the second quarter of 2009. This recovery can be ascribed to the governmental and fiscal stimulus and recovery plans that were developed and executed when the effects of the credit crunch came visible. These stimulus measures were able to boost the economy partly because of their proportions. Remarkable here is that the massive USD 787 billion stimulus program by the Obama
administration has only been executed for USD 85 billion by 21 August 2009. A great part of the stimulus is thus yet to come. Problem of the government stimulus programs is that the factors that stimulate the current economy have a temporary nature. For a fundamental and sustainable recovery major changes are necessary (Jones Lang LaSalle, 2009). First, the United States will have to focus on export and at the same time the Asian countries will have to focus more on their domestic markets. Second, the debt of governments that executed massive recovery programs will have to be scaled down. Despite the fact that the economy is recovering it cannot be confirmed with hard evidence, for the moment, that this recovery is sustainable.
5.3 "#18&-$!)&,@1,;!!
Private banking is a business area where high net worth and high-income private individuals are offered tailor-made financial advisory, investment and management services on a
comprehensive and long-term basis (Foehn, 2004). Services offered by private banks are investment advising, asset management, financial planning and alternative investments. The major part of the earnings originates from investment transactions and interest margin. The earnings from the various sources are correlated with the economic situation. In the current economic situation the proceedings from interest margin is significant lower because the spread is low. Also the proceedings from investment transactions are low because customers are less willing to invest in uncertain markets.
Foehn states that the target group is often described in terms of their wealth and the way they attracted this wealth (2004). He argues that variables such as sensitivity to price and service should also be introduced since competition in the private banking world increases. This increased competition is caused by a larger number of providers such as new established private banking business units, Internet brokers and insurance companies. The larger players who are active on the private banking market are, besides developing their home market, also developing the international markets and attract international clients. There are various reasons why a client is willing to bank in a foreign country. These reasons can vary from tax benefits, historic ties to a certain bank to political instability in the country of origin. The relationship with a foreign client is more complex than with most local clients. This
complexity is higher because of for example communication, expectations, culture and time zones. The question why a foreign customer wants to open a bank account can partly be explained by the country-of-origin effect. The country-of-origin effect can be described as a trigger to the global evaluation of quality, performance and specific service attributes. It is an extrinsic factor that is hardly controlled by a company and is used by the customer to reduce the risk in the decision-making process (Michaelis, Woisetschlager, Backhaus, & Ahlert, 2008, p. 407). In the case of private banking business, the team that is servicing a client can also be an important trigger to choose and stay with a certain foreign bank. This idea is confirmed by a study in cross border banking. The researcher found that the relation
significantly developed based on the relationship itself rather than a transactional motive between foreign-service providers and the clients (Nijssen & Van Herk, 2009, p. 104).
The Scorpio Partnership Global Private Banking Benchmark (2009) indicates that worldwide a total of USD 14.4 trillion is served by the private banking industry. The assets under management dropped with 15.7% and profits dropped with 32.9% compared to 2008. This is the first decrease in this market environment in a decade. In the past the private banking industry showed a steady development with annual growth forecasts of about 10% a year (see graph 1).
Graph 1: Estimate of the size (in USD trillion) of the global private banking industry. Source: Scorpio Partnership
Despite the bad market circumstances headcount increased with 6%. The additional
headcount and the bad market circumstances also resulted in a 13.7% drop of the efficiency ratio. The market can be characterized as concentrated because the top 20 banks serve 63% of all assets under management. In the table below the largest private banks in the world are shown ranked by the assets under management.
Table 2 Top 10 2009 Private bank Global by Assets under Management ($M). Source: Scorpio Partnership
Rank Bank AuM ($ Million) Primary currency
1. Bank of America 1,501.00 USD
2. UBS 1,393.48 CHF
3. Citi 1,320.00 USD
4. Wells Fargo 1,000.00 USD
5. Credit Suisse 611.96 CHF
6. JPMorgan 522.00 USD
7. Morgan Stanley 522.00 USD
8. HSBC 352.00 USD
9. Deutsche Bank 231.19 EUR
0 5 10 15 20 25 30 2002 2003 2004 2005 2006 2007 2008 8,5 11,2 12,6 13,6 16,0 17,4 14,5 3,5 2,6 6,0 8,5 10,8 12,6 11,6
Four large-scale banks, Rabobank, ING Group, SNS Bank and ABN Amro dominate the Dutch financial retail market. In a short period a fifth player will enter the market, Deutsche Bank. Besides these large-scale banks there are eighth smaller banks active. Three of the four large-scale banks have private banking business units that are active in the Netherlands. In table three an overview is provided of the important players on the Dutch private banking market. ABN Amro Private banking is the market leader in the Netherlands.
Table 3 Private banking in the Netherlands.
ABN AMRO Private Banking Fortis MeesPierson ING Private Banking Schretlen & Co (Rabobank) Van Lanschot Bankiers
Number of offices in the Netherlands 20 12 19 5 32
Market share assets under allocation 29% 13% 6% 5% 3%
Min. wealth level of the customers 1,000,000 1,000,000 500,000 1,000,000 100,000
Number of clients per banker 130 140 160 120-140 350
The future of the private banking industry will not be the same as before the financial turmoil. PricewaterhouseCoopers identified three themes that will define the future of the private banking industry (PricewaterhouseCoopers). These identified themes were the results of the 2009 Global Private Banking and Wealth Management survey reflecting the opinion of 238 senior wealth managers across 40 countries.
The first theme is the ‘emergence of nouveau classic banking’. This concerns a higher level of transparency in the market. The trusted relationship with the client is bruised because the client thinks that some manager rated short-term revenue business higher than the client interest. As a result clients request higher standards of service and advice together with simplicity and transparency.
The second theme is adaption of business models. Business models have to be
introduced that are more profitable over the long-term. Besides that costs have to be cut, focus on the core and outsource or cease non-core activities and process. Key rationale here is that investments need to be maintained and long-term efficiency have to be prioritised.
The third and last theme relates to increasing political, fiscal and regulatory pressures. As proven in the history, financial crisis leads to new regulation. Thereby governments expect that financial institutions do not engage in unacceptable tax planning. One of the results is that off shore banking might be an ending era.
Summarized, managed assets values and trading volumes are down, clients are expecting more, the relationships are less secure and governments and regulators increase the pressure on wealth managers.
This chapter provided an overview of the financial services industry and of the credit crisis that caused a major impact on the financial markets and global economy.
6
7$2$&#'3!/$21;,!
In the previous section the background where the research is conducted has been reviewed. In this section the research design will be discussed. First, hypothesis development will be discussed, followed by the research strategy, data collection and data selection. The section will end with the description of the variable measurements.
6.1 D:4+-3$212!/$8$*+45$,-!
Based on the research question and the literature studied a number of hypothesis were developed. Of course the specific case study context is taken into consideration when these hypothesis were developed.
The first hypothesis is directly linked to the research question stated. The research question is:
‘What is the relation between member familiarity in client service teams on the customer lifetime value of a client at a private bank?’. Research on familiarity revealed that a higher
level of familiarity resulted in better performance. There are various explanations for this phenomenon. For example higher levels of team member familiarity results in a better identification of the needed information and evaluation of groups input.
Familiarity develops positive to time. Thus when client service teams are together for a certain period servicing the same clients, the CLTV of those clients will increase. Also the less changes of client service team members a customer experiences the higher the level of team familiarity will be. Team performance in this research is measured by the CLTV. Therefore the following hypothesis is predicted.
H1: Higher levels of team member familiarity will have a positive effect on the CLTV.
In the second hypothesis will test whether there is a turning point over time on the positive effects of team familiarity on performance. Research by Katz found that member familiarity could be harmful to group performance. In his research on R&D teams this phenomenon
occurred after about three to five years (Katz, 1982, p. 93). One reason Katz mentioned is that those groups have the chance to be secluded from information.
Since the time span of this research is not as large as in the research by Katz a shorter time period will be tested in this research by splitting the period studied in half. Therefore it is hypothesized that the effect of team member familiarity decreases after eight quarters and can even be negative.
H2: The effect of team familiarity on the CLTV decreases after about eight quarters and can even be negative.
The third hypothesis relates to the global orientation of the private banking industry. The private bank studied, is serving domestic as well foreign clients. As stated earlier, and
confirmed by research from Nijssen & Van Herk (2009), it is plausible to assume that foreign clients ascribe more value to the relationship they built with the client service team. On the other hand local clients can be expected to have less affection to the client service team and to express more brand loyalty, which could be caused by customer ethnocentrism. So, the effect of team member familiarity is greater for foreign clients than for local clients. This leads to the following hypothesis:
H3: The effect of team member familiarity on the CLTV is greater for foreign clients than for local clients because foreign clients attach more value to the relationship with the client service team.
6.2 7$2$&#'3!'+,-$C-!
The general purpose of the study is to understand the link, if any, between team member familiarity and Customer Lifetime Value. This relationship is examined by using data on the individual customer level from the annex of a Dutch private bank.
6.3 7$2$&#'3!2-#&-$;:!
Until now, no research has been conducted to study the specific link between CLTV and team member familiarity. Besides that, the area on which the research is conducted, the private banking industry, is very specific. Therefore the most appropriate research method for this topic to contribute to science will be the case study method. Case study differentiates from other forms of research like experiments and surveys because one is looking for a better understanding of the context and the process that is going on (Saunders, Lewis, & Thornhill, 2007, pp. 139-140). Case study research is performed in the natural surrounding of the phenomenon. This is causing a lot of debate in the academic world because there are
questions of the generalizability of case studies (Yin, 2003, p. 10). By using multiple cases it is possible to draw conclusions that can be generalized for similar cases with similar
characteristics. Since only one case is examined, it is not to be expected that this research will generate theories. But we can expect that findings from this research could be applicable to other similar organizations with the same organization structure and culture. This case study research will be using quantitative methods. This methodology is objective, reliable and can result in general theories. The latter is the case when multiple cases are used as stated before. The quantitative methods used to test the hypotheses are standard and hierarchical multiple regression models.
6.4 E&-&!'+**$'-1+,!&,/!2&54*$!2$*$'-1+,!
There are four major reasons why this study has been conducted in the financial sector and specific the private banking industry. At first, the financial sector is experiencing a turbulent era after the emergence of the credit crisis. Confidence of the public and business dropped to very low levels and go together with a suspiciously attitude against people working in the financial sector. In other words the financial environment is a hot topic. Second, many service organizations are using state of the art CRM systems describing individual customers and are thus a great source for this type of research. Third, private banking is aimed at developing long-term, durable relationships with their client base, which is useful since this study has a
longitudinal aspect. The fourth and last reason is that the private banking segment specifically is using teams to serve their customer base.
Data mining will be used to study the subject. Specifically four sources of input from this financial institution will be used. The first source represents monthly-generated standard reports that are composed by the financial department. These reports consist of data for each client about for example turnover, margin, assets, segments, investment ratio’s and investment portfolio distribution. The reports over a period of four years are used and modified in such a way that a database is formed that can be used for the analysis and the calculation of the CLTV. The second source of data is derived from CRM systems that can deliver information concerning for example contact moments and team composition. Especially this last element is essential for this study. The third source of date consists of monthly-generated financial statements of the annex. In these statements costs of for example staff, marketing costs, travel expenses, housing and ICT are specified. Besides the revenues from the two major product categories, interest margin and equity commissions are specified. The fourth and last input came from the human resource department, which provided details such as position, gender and age of the employees who worked at the annex in the period studied.
These sources will be merged to one database that can be used for analysis. The final database will consist of data as from the third quarter of 2005 until the third quarter of 2009. This results in data from 17 quarters. This period is chosen because of two reasons. The first reason implies a limitation caused by the case study subject. The annex is able to deliver comparable, consistent data since the third quarter of 2005. Before this period the necessary reports were generated in a different way and on an irregular basis. The second reason is that research by Katz revealed that the positive effects of team member familiarity disappear over time (Katz, 1982). Thus the longest possible research period has been used for this study.
The final database is composed for one annex of the private bank, which is geographically located in the centre of the Netherlands. This annex is characterized by a relative large and diverse database. The diversity is expressed by the fact that it serves various segments such as international clients, institutions & charity banking and professionals & executives. In the first reporting period there were 1,023 persons registered as a client of the studied annex. This total client database however has been reduced because a number of parameters were applied. These parameter are stated below:
- The first parameter. The client was active and in the possession of an investment portfolio.
- The second parameter. A different banker and investment advisor serve the client in the first and the last period.
- The third parameter. The client remains an active client from the third quarter of 2005 to the third quarter of 2009.
- The fourth parameter. The outliers resulting from the variable assets under
administration were eliminated to increase validity. The selected cases have assets under administration between the range of EUR 100,000 and EUR 25,000.000. Applied these parameters reduce the total database to a sample of 178. This sample will form as the basis of the study.
6.5 F&)*$!5$&20#$5$,-!
In this case study eight variables are computed. The dependent variable in the regression models is the CLTV. The independent variables are Average team tenure, Number of changes
over time, Total age difference, Total gender changes and Same team for more than eight quarters. The control variables are the Assets under Allocation Q3 2009 and the Portfolio model.
6.5.1 >02-+5$#!*1%$-15$!8&*0$!'&*'0*&-1+,!
There is an extensive amount of research on how to calculate CLTV. Research by Donkers, et al. revealed in an extensive comparison of various CLTV models that the less complicated models were quite ass well able to predict CLTV as the more complicated ones (2007, p. 182). Despite the fact that the more complicated models comprehends more relational variables. Based on this knowledge it is not necessary to use a very complicated model. Further, it is not necessary to achieve the most accurate CLTV calculation since only the development of this figure related to the level of familiarity in a customer service team will be
analyzed. Therefore the basic model is modified in such a way that it fits with data available from the systems. This model is stated below.
(2)
Where:
t = quarter
n = length of customer relationship
D = customer retention rate per quarter (see description 2)
REt = equity revenues earned from customer in quarter t (see description 3)
RTt = treasury revenues earned from customer in quarter t (see description 3)
Ct = cost of servicing customer in quarter t (see description 4)
r = yearly discount rate (see description 1)
The different elements of the formula are discussed below:
1. The discount ratio is used to discount future cash flows to the present value. In this way the value for each cash flow is reduced for each time period by which it is removed from the period that has been specified as the present.
The discount ratio used in this formula is the average development of the net earnings per share over a 10-year period. The net earnings per share can be easily calculated by dividing the net income with the average number of outstanding shares. In the
company studied, the net earnings per share served as a benchmark to indicate the growth target for the next year. In this way the discount rate derived expresses the expected value development of the company. The average development of the net earnings per share and in this case the discount rate is 12.74%.
2. The customer retention (D) rate is calculated for every year and for each segment resulting in twelve different retention rates to be used in the CLTV formula. To calculate the retention rate, the number of customers in a specific segment served in
t=1 is compared with the number of customers in the same segment that are still
served in the fourth quarter. The number of defected clients is then divided by the total number of customers served by totalling the number of customer served in each
