Archive of SID

Iranian Accounting & Auditing Review, Spring 2007,Vol. 14 No. 47, pp 7- 20

Customer Lifetime Value (CLV) Measurement

Based on RFM Model

Babak Sohrabi1, Amir Khanlari2

1.
Assistant professor, School of Management, University of Tehran
2.
PhD student of Marketing, School of Management,
University of Tehran

(Received: 9/Apr/2007; Accepted: 10/June/2007)

Abstract
Nowadays companies increasingly derive revenue from the creation and
sustenance of long-term relationships with their customers. In such an
environment, marketing serves the purpose of maximizing customer lifetime
value (CLV) and customer equity, which is the sum of the lifetime values of
the company’s customers. A frequently-encountered difficulty for companies
wishing to measure customer profitability is that management accounting
and reporting systems have tended to reflect product profitability rather than
customer profitability. But in spite of these difficulties, Companies looking
for methods to know how calculate their customers's CLV. In this paper, we
used K-Mean clustering approach to determine customers's CLV and
segment them based on recency, frequency and monetary (RFM) measures.
We also used Discriminant analysis to approve clustering results. Data
required applying this method gathered from one branch of an Iranian
private bank which is established newly. Finally, in terms of this
segmentation, we proposed customer retention strategies for treating with the
bank customers.

Keywords: Relationship marketing, customer lifetime value, RFM model,

clustering.

Corresponding author: Email: bsohrabi @ ut.ac.ir

www.SID.ir
Archive of SID

Iranian Accounting & Auditing Review, Spring 2007,Vol. 14, No. 47 8

1. Introduction
Traditional accounting practices focus mainly on measuring tangible
assets as a statutory requirement on the balance sheet. However, nowadays it
is more usual for intangible assets such as brand, employee and customer
relationships to be the critical and often dominant determinants of
shareholder value [1]. A frequently-encountered difficulty for companies
wishing to measure customer profitability is that management accounting
and reporting systems have tended to reflect product profitability rather than
customer profitability [2]. Meanwhile most companies have accounting
systems that track costs based on functions (e.g., freight) rather than on a per
customer basis [20].
As a result, medium-volume customers tend to be the most profitable.
Unfortunately, standard accounting systems focus on periods instead of
individual customers or customer groups [16]. To avoid such twists,
customers need to be treated as a bundle of cost drivers. This is precisely the
principle of Activity-Based Accounting (ABC) [11]. It implies that
customers are the cause of activities and resources are employed to carry out
activities to serve them. Costs are thus allocated on the basis of transactions.
ABC therefore provides a fairly accurate means of measuring costs related to
customer relationships.
For years, the challenge for businesses could largely be seen as putting in
place the means of production to satisfy growing demand, and using
marketing techniques to capture customers entering the market (e.g., [7, 18,
30]). Manufacturers of goods today, however, are competing in a very
different environment, and transaction marketing (product, price, place, and
promotion, the 4Ps) alone is believed to be insufficient [13, 36]. Instead,
relationship marketing is proposed for building more unique relationships
with customers and for adding more value to goods and services than what is
possible through transaction marketing [17, 28]. Relationship marketing,
then, is not only about the 4Ps but also long- term relationships, reflecting a
transaction- relationship continuum [38]. Relationship marketing constitutes
a major shift in marketing theory and practice. Rather than focusing on
discrete transactions, it emphasizes the establishment, development and
maintenance of long-term exchanges [29]. Such relationships are thought to
be more profitable than short-term relationships as a result of exchange
efficiencies. This is especially true of customer relationships [33].
However, since not all customers are financially attractive to the firm, it
is crucial that their profitability be determined and that resources be
allocated according to the customer lifetime value (CLV). There are several
factors that account for the growing interest in this concept. First, there is an
increasing pressure in companies to make marketing accountable. Second,
financial metrics such as stock price and aggregate profit of the firm or a
business unit do not solve the problem either. Although these measures are

www.SID.ir
Archive of SID

9 Customer lifetime value (CLV) Measurement based on RFM model

useful, they have limited diagnostic capability. Third, improvements in

information technology have made it easy for firms to collect enormous
amount of customer transaction data. This allows firms to use data on
revealed preferences rather than intentions [20].
Nonetheless, companies confront to CLV calculation problems yet.
Although there is many models for this purpose, but most of them are
theoretic, complex and not applicable. It is also important to point out that
most modeling approaches ignore competition because of the lack of
competitive data. Finally, how frequently we update CLV depends on the
dynamics of a particular market. For example, in markets where margins and
retention may change dramatically over a short period of time (e.g., due to
competitive activity), it may be appropriate to re-estimate CLV more
frequently. Therefore, companies need an approach to be used easily and its
input data can be gathered and updated speedy. The proposed approach in
this paper attempts to satisfy these expectations. This paper aims at
suggesting a new CLV model and customer segmentation considering RFM
model. We will also propose Customer retention strategies after segmenting
customer base. For this purpose, the plan for the article is as follows. We
first present CLV concept and its applications and what are its key drivers.
Next, we present several modeling approaches that have been adopted to
calculate CLV. This is followed by a detailed discussion of RFM as model
used in this work. Then, we present clustering outputs and strategies
suggested to treat with each customer cluster. We end the article with
concluding remarks.

2. Customer Lifetime Value (CLV)

Customer value has been studied under the name of LTV, Customer
Lifetime Value, Customer Equity, and Customer Profitability. The previous
researches contain several definitions of CLV. The differences between the
definitions are small [24]. Table 1 shows the definitions of CLV.
Considering the definitions above, we define CLV as the sum of the
revenues gained from company’s customers over the lifetime of transactions
after the deduction of the total cost of attracting, selling, and servicing
customers, taking into account the time value of money.
In other words, CLV is generally defined as the present value of all future
profits obtained from a customer over his or her life of relationship with a
firm. CLV is similar to the discounted cash flow approach used in finance.
However, there are two key differences. First, CLV is typically defined and
estimated at an individual customer or segment level. This allows us to
differentiate between customers who are more profitable than others rather
than simply examining average profitability. Second, unlike finance, CLV
explicitly incorporates the possibility that a customer may defect to
competitors in the future [20].

www.SID.ir
Archive of SID

Iranian Accounting & Auditing Review, Spring 2007,Vol. 14, No. 47 10

It is argued that customer relationships are viewed as investment

decisions and customers as generators of revenue streams. Customer
relationships also generate costs. Hence, in order to measure the customer
lifetime value, all revenues and costs pertaining to a customer relationship
must be assessed. It is then possible to calculate the current value of cash
flow streams [4]. Though, accurately estimating the revenues and costs of a
relationship remains a challenging task for a number of reasons:

Table 1: Definitions of CLV [20]

Definition Reference

The present value of all future profits generated

[19]
from a customer
The net profit or loss to the firm from a
customer over the entire life of transactions of [4]
that customer with the firm
Expected profits from customers, exclusive of
[6]
costs related to customer management
The total discounted net profit that a customer
[5]
generates during her life on the house list
The net present value of the stream of
contributions to profit that result from customer [31]
transactions and contacts with the company
The net present value of a future stream of
contributions to overheads and profit expected [25]
from the customer
The net present value of all future contributions
[34]
to overhead and profit
The net present value of all future contributions
to profit and overhead expected from the [12]
customer

• Standard accounting does not allow for allocating costs to specific

customer relationships.
• Only monetary benefits of customer relationships are taken into account.
• Revenues and costs vary over time.
• Cash flow streams are generated at different points in time and at different
levels of risk.

Consequently, the following requirements have to be fulfilled in order to

measure customer profitability accurately [35]:
• An exact allocation of costs to customer relationships according to the
resources employed;

www.SID.ir
Archive of SID

11 Customer lifetime value (CLV) Measurement based on RFM model

• An estimation of all monetary and nonmonetary benefits created by the

particular customer relationship;
• A consideration of cost and revenue changes over the estimated time span
of a customer relationship;
• The discounting to present of future cash flows generated over the
estimated time span of a customer relationship; and
• An estimation of the relationship risks.
There are a large number of models which attempt to calculate CLV. At
the next section, some of most important models will be describe and then
RFM model which seem to be more simply and robust will be present in
more detail.
2.1. Models of CLV calculation
There are a lot of researches on calculating customer value. The basic
concept of these researches, however, focused on Net Present Value (NPV)
obtained from customers over the lifetime of transactions [3, 4, 19, 34].
Dwyer (1997) tried to calculate CLV through modeling the retention and
migration behavior of customers. Focused on making decision of marketing
invest, Hansotia and Rukstales (2002) suggested incremental value modeling
using tree and regression based approach. Hoekstra and Huizingh (1999)
also suggested a conceptual CLV model and categorized input data of the
model into two types, source of interaction data and time frame. Most CLV
models stem from the basic equation, although we have many other CLV
calculation models having various realistic problems. The basic model form
based upon the proposed definition is as follows
n
( Ri − C i )
CLV = ∑
i =1 (1 + d ) i −0.5

Where i is the period of cash flow from customer transactions, Ri the

revenue from the customer in period i, Ci the total cost of generating the
revenue Ri in period i, and n is total number of periods of projected life of
the customer under consideration. Therefore, the numerator is the net profit
that has been obtained at each period while the denominator transforms the
net profit value into the current value. The calculation model above is the
most basic model that ignores the fluctuation of sales and costs. Expanding
this basic model, many researchers including Berger and Nasr (1998) have
proposed CLV calculation models, which reflect the fluctuation of sales and
costs [6, 26]:
n
1
CLV = ∑ π (t ) ×
i =0 (1 + d ) i

www.SID.ir
Archive of SID

Iranian Accounting & Auditing Review, Spring 2007,Vol. 14, No. 47 12

Where π (t ) is the function of customer profits according to time t.

formulating precise π (t ) is the most important factor in calculating CLV
precisely. Colombo and Jiang (1999) developed a stochastic Recency
Frequency Monetary model to rank customers in terms of their expected
contribution. Pfeifer and Carraway (2000) proposed Markov chain models
for modeling customer relationships.
The evaluations of customer value in previous studies have treated
prediction method with regression models simply based on profits from
customers to calculate the future value of customers. That is to say,
considering the changing profit contribution obtained from customers in the
past, the existing models calculate the future worth and then define the CLV
of customers with the projected value of the future worth. Therefore, the
CLV model above is not capable of considering potential values of
customers, not available from the past profit contribution, which would be
able to be the profits of companies.
Last, the models mentioned above do not considered the defection of
customers. Although we have a customer who has very high value to our
company, this information can conclude improper marketing strategies if we
don’t pay careful attention to the possibility of the customer defection.
Hence, it is reasonable to consider the probability of individual customer’s
churn rather than to consider only the total decreasing rate of whole
customers. Verhoef and Donkers (2001) used two dimensions, current value
and potential value, to segment the customers of an insurance company.

2.2. RFM Model

To identify customer behavior, the well known method called recency,
frequency and monetary (RFM) model is used to represent customer
behavior characteristics [9, 23]. RFM models have been used in direct
marketing for more than 30 years. Given the low response rates in this
industry (typically 2% or less), these models were developed to target
marketing programs (e.g., direct mail) at specific customers with the
objective to improve response rates. Prior to these models, companies
typically used demographic profiles of customers for targeting purposes.
However, research strongly suggests that past purchases of consumers are
better predictors of their future purchase behavior than demographics [20].
The basic assumption of using the RFM model is that future patterns of
consumer trading resemble past and current patterns. The calculated RFM
values are summarized to clarify customer behavior patterns. This study
proposes using the following RFM variables [9]:
• Recency (R): the latest purchase amount.
• Frequency (F): the total number of purchases during a specific period.
• Monetary (M): monetary value spent during one specific period.

www.SID.ir
Archive of SID

13 Customer lifetime value (CLV) Measurement based on RFM model

As mentioned, this approach models three dimensions of customer

transactional data to classify customer behavior [39]. The first dimension is
recency, which indicates the length of time since the start of a transaction.
Meanwhile, the second dimension is Frequency, which indicates how
frequently a customer purchases products during a particular period. Finally,
monetary value measures the amount of money that customer spending
during a period [27].
A large number of studies specifically in loyalty programs areas
considered RFM. For instance, Jonker et al. (2004) demonstrated the use of
RFM value in direct-mailing; they proposed an optimization strategy for
customer segmentation and marketing employing stochastic dynamic
programming. Also, Buckinx and Van den Poel (2005) and Fader et al.
(2005) proved that RFM variables can predict accurately the CLV. They
showed how RFM variables can be used to build a CLV model that
overcomes many of its limitations. They also showed that RFM are
sufficient statistics for their CLV model.

3. Research Methodology
We selected 214 RFM data from customer base of the given case. For
data analysis, K- Means Cluster Analysis is used to cluster the Bank's
customers. This procedure attempts to identify relatively homogeneous
groups of cases based on selected characteristics, using an algorithm that can
handle large numbers of cases. However, the algorithm requires you to
specify the number of clusters. The parameter was set to 8, since eight
( 2 × 2 × 2 ) possible combinations of inputs (RFM) can be obtained by
assigning high or low, according to the average R (F, M) value of a cluster
being less than or greater than the overall average R (F, M). The RFM values
of customers were normalized as follows. The profit
form, x ′ = ( x − x ) /( x − x ) , was used to normalize the F (frequency)
S L S

and M(monetary) values, since F and M positively influenced CLV or

loyalty. The cost form, x ′ = ( x L − x) /( x L − x S ) , was used to normalize the
R value, since it negatively impacted CLV. x ′ and x represented the
normalized and original R (F, M) values, while x L and x S represented the
largest and smallest R (F, M) value of all customers. After clustering
customers's data to segment them based on RFM model, we also propose
customer retention strategies after segmenting customer base (Fig. 1).

www.SID.ir
Archive of SID

Iranian Accounting & Auditing Review, Spring 2007,Vol. 14, No. 47 14

RFM Model
Transaction Recency
Data of
Services
Frequency

Monetary

Customer
Retention Customer
Strategies Segmentation

Fig. 1 Conceptual framework of research

4. Empirical Results
Table 2 presents the clustering results, listing eight clusters, each with the
corresponding number of customers and their average R, F and M values.
The last column also shows the overall average for all customers. These, for
each cluster, were compared with the overall averages. If the average R (F,
M) value of a cluster exceeded the overall average R (F, M), then H (High)
was included and otherwise L (Low) was assigned. Two last row of the table
show the RFM pattern for each cluster and the number of cases in each
cluster respectively. Table 3 shows distances between final cluster centers,
these results can be used only for descriptive purposes. The numbers show
distance between a cluster and other ones that computed based on distance
between cluster centers.

Table 2: Final Cluster Analysis

Cluster
average
1 2 3 4 5 6 7 8
Recency 23 81.48 41 58.68 72.86 75 5.33 54.39 51.47
Frequency 12 22.6 26.67 10.14 57.49 12 10.33 78.61 28.73
Monetary 15000 905.96 173.66 1658.93 543.71 15000 1956 351.44 4448.71
Cluster L, L, H, L, L, H, H, H,
H, L, L L, L, L H, L, L H, H, L
Type H H L L
Number of
3 50 6 28 84 22 3 18
Cases

www.SID.ir
Archive of SID

15 Customer lifetime value (CLV) Measurement based on RFM model

Table 3: Distances between Final Cluster Centers

Cluster 1 2 3 4 5 6 7 8
1 1.158 1.024 .978 1.237 .591 .893 1.297
2 1.158 .465 .301 .418 .951 .880 .721
3 1.024 .465 .295 .510 1.075 .463 .623
4 .978 .301 .295 .578 .909 .607 .802
5 1.237 .418 .510 .578 1.100 .948 .323
6 .591 .951 1.075 .909 1.100 1.176 1.269
7 .893 .880 .463 .607 .948 1.176 .976
8 1.297 .721 .623 .802 .323 1.269 .976

Then, discriminant analysis is used to determine whether clusters could

be used to distinguish the sample customers (whether statistically
significant). The analysis rejected the null hypothesis H 0 because the P-
values were significant (P<0.05). The result confirmed that these clusters can
significantly distinguish sample customers (Table 4). Table 5 also shows
Eigenvalues, as first column depicts, three canonical discriminant functions
were used in the analysis and Eigenvalues are significant.
Table 4: Wilks' Lambda test
Exact F Approximate F
Number of
Variables

Lambda

Statistic

Statistic
Step

Sig.

Sig.
df1

df2

df3

df1

df2

df1

1 df2
1 .034 1 7 206 830.060 7 206 .000
2 2 .003 2 7 206 507.037 14 410 .000
3 586.3
3 .000 3 7 206 400.53 21 .000
3

Table 5: Eigenvalues
Function Eigenvalue % of Variance Cumulative % Canonical Correlation

1 40.181 72.9 72.9 .988

2 10.619 19.3 92.2 .956
3 4.317 7.8 100.0 .901

Figure 2 shows group centroids graphically; at this figure the location of

each cluster center has been depicted. As shown, X and Y axis are function 1
and 2 amounts respectively.

www.SID.ir
Archive of SID

Iranian Accounting & Auditing Review, Spring 2007,Vol. 14, No. 47 16

10
1
8
8
6
7

2 5 3 6

-2 4

2
-4
Function2

-6

-8
-10 0 10 20

Function 1

Fig. 2 Cluster number of cases

Finally, table 6 shows classification results; at each row, predicted group
membership for each cluster has been showed, as seen, all cases but case 5
predicted to assign to only one cluster (100% fit), so according to
calculations, 99.5% of original grouped cases correctly classified.
As mentioned earlier, each cluster represents a market segmentation.
Customers in clusters with the pattern R (L) F (H) M (L) are considered to
be loyal but not high valuable, purchased recently, purchase frequently, but
spend not regularly with the bank. They are silver (not gold) customers.
Bank should satisfy their needs perfect and provide various services to match
with their needs to promote transaction values. Clusters with the pattern R
(L) F (L) M (L) may include new customers who have only recently visited
the bank. Customers in such clusters may be trying to develop closer
relationships with the bank. These customers may become gold customers.
Clusters with the pattern R (L) F (L) M (H) and also the pattern R (H) F (L)
M (H) may include high valuable customers but not loyal to this bank, they
should be treated specially to transact more. Finally, clusters with the pattern
R (H) F (L) M (L) or the pattern R (H) F (H) M (L) include those who very
rarely visited the site and made very few or low value transactions. They are
valueless customers and bank should provide extra services to attract such
customers.

www.SID.ir
Archive of SID

17 Customer lifetime value (CLV) Measurement based on RFM model

Table 6: Classification Results

Cluster Predicted Group Membership
Number of Total %
Case 1 2 3 4 5 6 7 8
1 3 0 0 0 0 0 0 0 3 100
2 0 50 0 0 0 0 0 0 50 100
3 0 0 6 0 0 0 0 0 6 100
4 0 0 0 28 0 0 0 0 28 100
5 0 1 0 0 83 0 0 0 84 98.8
6 0 0 0 0 0 22 0 0 22 100
7 0 0 0 0 0 0 3 0 3 100
8 0 0 0 0 0 0 0 18 18 100

5. Conclusions and further research

As marketing strives to become more accountable, we need metrics and
models that help us assess the return on marketing investment. Many CRM
researches pertain to develop a comprehensive model of customer
profitability since the question ‘Who are profitable customers?’ is a starting
point of CRM. CLV is one such metric. Many models have been researched
to calculate CLV of a customer. The easy availability of transaction data and
increasing sophistication in modeling has made CLV an increasingly
important concept in both academia and practice. In this paper, we suggested
a CLV model considering the recency, frequency and monetary at the same
time. It clusters customers into segments according to their lifetime value
expressed in terms of RFM. Moreover, clustering customers into different
groups helps decision-makers identify market segments more clearly and
thus develop more effective strategies. The authors recommend researchers
to work on more measures and consider it comprehensively or rather
compare various CLV models in a specific industry.
Acknowledgement
The authors express their appreciation to the editor and anonymous
reviewers. The authors would also like to thank the participating bank
employees for their cooperation.

References

1. Amir, E. and Lev, B. (1996). "Value- relevance of nonfinancial

information: The Wireless Communications Industry", Journal of
Accounting and Economics, 22, pp.3- 30.

2. Balachandran, S. (2005). "How to build a culture of cost management",

www.SID.ir
Archive of SID

Iranian Accounting & Auditing Review, Spring 2007,Vol. 14, No. 47 18

Financial Times, p.16.

3. Bayo´n, T., Gutsche, J. and Bauer, H. (2002). "Customer equity

marketing: touching the intangible", European Management Journal,
20(3), pp.213- 222.

4. Berger, P. D. and Nasr, N. I. (1998). "Customer lifetime value:

marketing models and applications", Journal of Interactive Marketing,
12(1), pp.17- 30.

5. Bitran, G. R. and Mondschein, S. (1996). "Mailing decisions in the

catalog sales industry", Management Science, 42(9), pp.1364- 1381.

6. Blattberg, R.C. and Deighton, J. (1996). "Manage marketing by the

customer equity test", Harvard Business Review, pp.136- 144.

7. Brookes, R. and Palmer, R. A. (2004). The new global marketing reality,

Basingstoke’ Palgrave.

8. Buckinx, W. and Van den Poel, D. (2005). "Customer base analysis:

Partial defection of behaviorally- loyal clients in a non- contractual fmcg
retail setting", European Journal of Operational Research, 164 (1),
pp.252- 268.

9. Chan, C. C. H. (2005). "Online auction customer segmentation using a

neural network model", International Journal of Applied Science and
Engineering, 3(2), pp. 101- 109.

10. Colombo, R. and Jiang, W. (1999). "A stochastic RFM model", Journal
of Interactive Marketing, 13(3), pp. 2- 12.

11. Cooper R. and Kaplan R. S. (1988). "Measure costs right: make the right
decisions", Harvard Business Review, 66, pp. 96- 103.

12. Courtheoux, R. (1995). Customer retention: how much to invest.

Research and the Customer Lifecycle, New York, NY: DMA.

13. Denison, T. and McDonald, M. (1995). "The role of marketing: Past,

present and future", Journal of Marketing Practice, 1(1), pp. 54- 76.

14. Dwyer, F. R. (1997). "Customer lifetime valuation to support marketing

decision making", Journal of Interactive Marketing, 11(4), pp.6- 13.

15. Fader, P. S.; Hardie, B. G. S. and Lee, K. L. (2005). "Counting your

customers the easy way: An alternative to the Pareto/ NBD model"
Marketing Science, 24 (2), pp. 275- 284.

www.SID.ir
Archive of SID

19 Customer lifetime value (CLV) Measurement based on RFM model

16. Goebel, D. J.; Marshall, G. W. and Locander, W. B. (1998). "Activity-

based costing: accounting for a market orientation", Industrial Marketing
Management, 27, pp. 397- 510.

17. Gro¨nroos, C. (2000). Service management and marketing, Chichester’

John Wiley and Sons.

18. Gummesson, E. (1999). Total relationship marketing, Oxford’

Butterworth- Heinemann.

19. Gupta, S. and Lehmann, D. R. (2003). "Customers as assets", Journal of

Interactive Marketing, 17(1), pp.9- 24.

20. Gupta, S.; Hanssens, D.; Hardie, B.; Kahn, W.; Kumar, V.; Lin, N.;
Ravishanker, N. and Sriram, S. (2006). "Modeling Customer Lifetime
Value", Journal of Service Research, 9(2), pp.139- 155.

21. Hansotia, B. and Rukstales, B. (2002). "Incremental value modeling",

Journal of Interactive Marketing, 16(3), pp.35–46.

22. Hoekstra, J. C. and Huizingh, E. K. R. E. (1999). "The lifetime value

concept in customer- based marketing", Journal of Market Focused
Management, 3(3- 4), pp. 257- 274.

23. Hsieh, N. C. (2004). "An integrated data mining and behavioral scoring
model for analyzing bank customers", Expert Systems with Applications,
27, pp. 623- 633.

24. Hwang, H.; Jung, T. and Suh, E. (2004). "An LTV model and customer
segmentation based on customer value: A case study on the wireless
telecommunication industry", Expert Systems with Applications, 26, pp.
181- 188.

25. Jackson, D. R. (1994). "Strategic application of customer lifetime value

in the direct marketing environment", Journal of Targeting Measurement
and Analysis for Marketing, 3(1), pp. 9- 17.

26. Jain, D. and Singh, S. S. (2002). "Customer lifetime value research in

marketing: a review and future directions", Journal of Interactive
Marketing, 16(2), pp. 34- 45.

27. Jonker, J. J.; Piersma, N. and Poel, Dirk V. (2004). "Joint optimization of
customer segmentation and marketing policy to maximize long- term
profitability", Expert Systems with Applications, 27, pp. 59- 168.

28. Lindgreen, A. and Wynstra, F. (2005) "Value in business markets: What

www.SID.ir
Archive of SID

Iranian Accounting & Auditing Review, Spring 2007,Vol. 14, No. 47 20

do we know? Where are we going?" Industrial Marketing Management,

34(7), pp.732- 748.

29. Morgan, R. M. and Hunt S. D. (1994). "The commitment–trust theory of

relationship marketing", Journal of Marketing, 58, pp. 20- 38.

30. Parvatiyar, A. and Sheth, J. N. (2000). "The domain and conceptual

foundations of relationship marketing" In Sheth, J.N. and Parvatiyar A.
(Eds.), Handbook of relationship marketing, Thousand Oaks, California’
Sage Publications, pp. 3- 38.

31. Pearson, S. (1996). Building brands directly: creating business value

from customer relationships, London: MacMillan Business.

32. Pfeifer, P. E. and Carraway, R. L. (2000). "Modeling customer

relationships as Markov Chains", Journal of Interactive Marketing,
14(2), pp.43- 55.

33. Reichheld, F. F. and Sasser, E. W. (1990), "Zero defections: quality

comes to services", Harvard Business Review, 68, pp. 105- 111.

34. Roberts, M. L. and Berger, P. D. (1989). Direct marketing management,

Prentice- Hall: Englewood Cliffs, NJ.

35. Stahl, H. K.; Matzler, K. and Hinterhuber, H. H. (2003). "Linking

customer lifetime value with shareholder value", Industrial Marketing
Management, 32(4), pp.267- 279.

36. Tapscott, R. and Caston, A. (1993). Paradigm shift, New York’

McGraw- Hill.

37. Verhoef, P. C. and Donkers, B. (2001). "Predicting customer potential

value an application in the insurance industry", Decision Support
Systems, 32, pp.189- 199.

38. Webster, J. F. E. (1992). "The changing role of marketing in the

corporation", Journal of Marketing, 56(4), pp. 1- 17.

39. Yao, J.; Li, Y. and Chew, L. T. (2000). "Option price forecasting using
neural networks", The International Journal of Management Science, 28,
pp. 455- 466.

www.SID.ir