Artificial Intelligence Review 14: 569–590, 2000.

Issues on the Application of Data Mining.

569
© 2001 Kluwer Academic Publishers. Printed in the Netherlands.

Customer Retention via Data Mining

KIANSING NG & HUAN LIU

School of Computing, National University of Singapore, 119260 (E-mail: {ngkians1,
liuh}@comp.nus.edu.sg)

Abstract. “Customer Retention” is an increasingly pressing issue in today’s ever-competitive

commercial arena. This is especially relevant and important for sales and services related
industries. Motivated by a real-world problem faced by a large company, we proposed a solu-
tion that integrates various techniques of data mining, such as feature selection via induction,
deviation analysis, and mining multiple concept-level association rules to form an intuitive
and novel approach to gauging customer loyalty and predicting their likelihood of defection.
Immediate action triggered by these “early-warnings” resulting from data mining is often the
key to eventual customer retention.

Keywords: customer retention, data mining, deviation analysis, feature selection, multiple
level association rules

1. Introduction

In the last decade, the increased dependency and widespread use of data-
bases in almost every business, scientific and government organization has
led to an explosive growth of data. Instead of being blessed with more
information to aid decision making, the overwhelming amounts of data have
inevitably resulted in the problem of “information overloading but knowledge
starvation”, as the human analysts are unable to keep pace to digest the
data and turn it into useful knowledge for application purposes. This situ-
ation has motivated some scientists and researchers in the fields of artificial
intelligence, machine learning, statistics and databases to put their expertise
together to form the field of knowledge discovery in databases (KDD). KDD
seeks to intelligently analyze voluminous amounts of information in data-
bases and extract previously unknown and useful knowledge (nuggets) from
them (Fayyad et al., 1996).
Active research in these fields has produced a wide range of effective
knowledge discovery techniques like ID3 (Quinlan, 1986) for classification
(used in C4.5 (Quinlan, 1993)) and Apriori (Agrawal and Srikant, 1994) for
association rule mining (used in DBMiner (Kamber et al., 1997)), to cater
to various applications in data mining. This tremendous success achieved
570 KIANSING NG AND HUAN LIU

in the research domain has also spun off a wide repertoire of high-quality,
off-the-shelf commercial data mining software/tools, like C5.0 by RuleQuest,
MineSet by Silicon Graphics and Intelligent Miner by IBM, to name a few.
Many people saw these tools as the catalyst for the success of data mining
applications. After all, many organizations are facing problems coping with
overwhelming amounts of data in their databases and are attracted by the
potential competitive advantages from data mining applications.
The availability of real-world problems and the wealth of data from the
organizations’ databases provide an excellent test-bed for us to perform
practical data mining. Our work was motivated by a real-world problem
involving a collaboration with a large company to tackle the pressing issue
of “customer retention”. Such collaborations between academia and applica-
tion domains to solve real-world problems represent a positive step towards
the success of data mining applications. The proposed approach showcases
an effective application of data mining in the sales and services related
industries, and reveals the complex and intertwined process of practical data
mining. More importantly, we demonstrate that real-world data mining is
an art of combining careful study of the domain, intelligent analysis of the
problems, and skillful use of various tools from machine learning, statistics,
and databases.
One truth about data mining applications is that even when the problem or
goal is clear and focused, the mining process still remains a complicated one,
involving multiple tasks across multiple stages. In our application, although
we are clear that we want to retain customers and the goal is to “identify
the potential defectors way before they actually defect”, it is still difficult to
know where to start. Without studying the domain, it is impossible for us
to go further. The problem and goal statement specified often gives no clue
as to “what” tasks of data mining are involved, “which” techniques are to
be applied, and “how” they are applied. Obviously, the problem of customer
retention must be further decomposed into several sub-problems such that
the knowledge derived from a task in one phase can serve as the input to the
next phase. Since the available data mining techniques and tools are designed
to be task-specific (according to the framework given in Figure 1) rather
than problem-specific, they cannot be applied directly to solve real-world
problems.
Many challenges can only be found in real-world applications. The
changing environment can cause the data to fluctuate and make the previ-
ously discovered patterns partially invalid. Such phenomenon is referred to as
“concept drift” (Widmer, 1996). The possible solutions include incremental
methods for updating the patterns and treating such drifts as an opportunity
for “interesting” discovery by using it to cue the search for patterns of
 CUSTOMER RETENTION VIA DATA MINING 571

Figure 1. A framework for classification of data mining tools (by tasks).

change (Matheus et al., 1994). In practice, we often find many organiza-

tions collecting data as a by-product of a business process. Hence, large
databases with hundreds of tables, millions of records and multi-gigabytes
in size are quite commonplace in many application domains. In our case, the
company’s databases typically logged in more than 45,000 transactions per
day. In a few months, this database alone can easily aggregate to a few giga-
bytes of data without even including the many other operational databases.
To aggravate the matter, most of the databases are of high dimensionality –
with a very large number of fields (attributes). This results in an exponential
increase in the size of the search space, and hence intolerably long compu-
tational time will be required for running any machine learning algorithms.
To resolve these problems, we need to pre-process the data set to reduce the
dimensionality of the problem. Another practical issue is that concepts can
have various levels of abstraction or taxonomies. For this reason, knowledge
or patterns have various characteristics. However, most transactional and
operational databases are usually described in terms of low-level concepts
and relations, like “freight car companies”. This is cognitively distant from
the high-level business concepts that are required for decision-making, like
“loyal-customers”. Since the generality of knowledge is needed to determine
its “interestingness” and applicability in practical problems, some forms of
generalization (and specialization) must be catered for. Our work here will
focus mainly on those challenges of the immediate need for customer reten-
tion. Obviously, there are also many other challenges for practical data mining
applications (Fayyad et al., 1996). They include challenges like “understand-
ability of patterns”, “complex relationships between fields”, “missing and
noisy data”, “data over-fitting”, so on and so forth.
Another point that cannot be stressed enough is that a data mining applic-
ation requires substantial know-how, skills and experiences from the user.
This is often not fully understood by many, thus leading to failures in data
mining applications. This work illustrates the kind of complexity involved in
572 KIANSING NG AND HUAN LIU

solving real-world problems and further justifies that practical data mining is
an art that requires more than just directly applying off-the-shelf techniques
and tools. The remainder of this article is organized as follows. In Section 2,
we begin with a domain analysis and task discovery of the customer reten-
tion problem faced by the company. We then perform a top-down problem
decomposition and list various sub-problems. This is important because each
sub-problem must map to only one specific task of data mining, so that
the existing data mining tools and techniques can effectively be applied. In
Section 3, we illustrate the use of feature selection via induction to choose
the objective “indicators” (or salient features) about customer loyalty. With
this technique, “concept-drifts” – the definition of a concept change over
time (Clearwater et al., 1989) – can be captured as they take place. This
is followed by the use of deviation analysis and forecasting to monitor
these indicators for the potentially defecting customers in Section 4. Next,
in Section 5, we elaborate on the employment of multiple-level association
rule mining in predicting customers who are likely to follow the previ-
ously identified defecting customers and leave the company. These “early
warnings” of possible chain effect will enable the marketing division to
take actions or tailor special packages to retain important customers and
their potential followers before the defection takes place. Finally, Section 6
suggests some implications of this project.
Because of the confidentiality of the databases used and the sensitivity
of services provided by the company, we deliberately use “the company”
throughout the paper and describe applications through some intuitive
examples as much as possible. In explaining basic concepts, we use the
“credit” database from theUC Irvine Machine Learning Repository (Merz
and Murphy, 1996) for illustrative purpose.

2. Understanding the Domain and Problems

The business of the company is service providing. Because of the confid-

entiality and sensitivity of the company, details shall be left out. However,
in order for the readers to understand the working of the techniques used in
this data mining application, we use an imaginary example of international
telephone satellite relaying services to explain the problems and techniques.
The scenario is as follows: there are some major satellite relaying centers for
major regions around the world, and each center has its network. Companies
doing international business usually need services from three centers in order
to communicate to any business partner in the world. In other words, three
parties are involved, the sender, the company, and the receiver. The sender
and the receiver are usually determined by the need of the business. But they
 CUSTOMER RETENTION VIA DATA MINING 573

can choose the company over others or vice versa as different relayers provide
varying services and charges with contracts of various periods. Companies
(senders and receivers) can form consortiums or groups to enjoy discounts
of various sorts offered by the company. The goal of this work is to help a
relayer keep as many of its customers as possible using its relaying services.
The goal of customer retention is to retain customers before they switch to
other relayers.
Like many organizations, the dependency on information technology has
inevitably resulted in an explosive growth of data, far beyond the human
analyst’s ability to understand and make use of the data for competitive
advantages. This is also due to the fact that conventional databases and
spreadsheets used by these analysts are not designed for identifying patterns
from the databases. Neither do they possess the capability to select nor
consolidate the different sources of information from a large number of
multiple databases of heterogeneous sources. In view of these inadequa-
cies, the company involved sees “data warehousing” and “data mining” as
two intuitive solutions. Maintaining a data warehouse separately from the
transactional database allows special organization, access methods and imple-
mentation methods to support multi-dimensional views and operations typical
of OLAP. In fact, some OLAP tool can be integrated to the data ware-
house to support complex OLAP queries involving multi-dimensional data
representation, visualization and interactive viewing, while not degrading the
performance of the operational databases.

2.1. Problem identification and analysis

The first step in our analysis involves identifying opportunities for data
mining applications. This step is important because not every problem can be
solved by data mining. Some guidelines for selecting a potential data mining
application include “the potential for significant impact”, “availability of
sufficient data with low noise level”, “relevance of attributes”, and “presence
of domain knowledge”. In fact, nearly one-fifth of the whole development
time was spent on identifying the “right” problems for application, as well
as justifying the use of data mining over the conventional approaches. The
possibility that an application can be generalized to solve other similar prob-
lems in related industries is also taken into consideration. With these factors
in mind, our feasibility study has identified the problem of customer retention
as a potentially usefull data mining application.
The motivation of our work comes from the fact that the problem
of customer retention is becoming an increasingly pressing issue for
organizations in the sales (e.g., departmental stores, banking, insurance, etc.)
and services (e.g., providers of Internet and/or Telecommunications services)
574 KIANSING NG AND HUAN LIU

related industries. From an economic point of view, customer retention

directly translates to huge savings in marketing costs, as highlighted by
Coopers & Lybrand’s Vince Bowey:

“A lot of companies have not figured out what it

costs them to acquire a new customer, it’s usually
pretty shocking. We estimate that it costs three to
five times more money to acquire a new customer than
to keep the ones you have.”

This problem is especially important for organizations that have a small

customer base where each customer represents a group of a large number
of companies. The defection of any single customer means a significant
percentage loss in the revenue of the company. Naturally, the company
has a strong interest to retain each and every one of them. This is espe-
cially important when the company is facing keen competition from many
upcoming relayers in the neighboring countries. The increased competitive-
ness means that the marketing division is facing higher risk of customer
defection, which potentially escalates marketing costs due to defections.
Furthermore, the relaying activities are an entrepot in nature, i.e., an inter-
mediary center for call transfer and repackaging for conference call. Hence,
the defection of a customer to another relayer is likely to influence its asso-
ciated business alliances to also defect, in order to maintain their already
established business relationship. Thus, the possibility of a “snow-ball effect”
from a defection further gives rise to a pressing need for an effective method
to identify the less obvious associations between customers for the benefit
of marketing and customer retention. Finally, contrary to popular belief,
focusing on retaining customers is not a passive policy. This is because
existing customers can, and often do, bring in new customers through their
business associations and expansion, as well as through word of mouth.
Despite the importance of the “customer retention” problem, many organi-
zations simply cannot do anything about it, since their customers are free to
leave without warning when they are dissatisfied with the services or if better
offers come by. Fortunately, this is not the case in this relaying business.
By offering attractive discount and rebate schemes to customers, especially
the major senders (consortiums, or groups), the company can usually at least
secure them under short-term contractual agreements. This means that poten-
tial defectors cannot just pull out abruptly. Moreover, most customers are
usually “committed” by the sheer size of their regular relaying transactions
through the company, which must be pre-scheduled some time (e.g., weeks
or months) in advance. Hence, potential defectors will have to take a couple
 CUSTOMER RETENTION VIA DATA MINING 575

Figure 2. Task identification for “customer retention” in the relaying business.

of months to gradually trim-out their outgoing volumes in relaying business

before their eventual withdrawal. This gradual pulling out process offers the
company the opportunity to identify the signals of defection and to predict
the possible chain effect with each defection. “Early warnings” like these
can give the marketing department ample lead-time to investigate the causes
and take rectification actions before defections actually take place. As was
described in Matheus et al. (1994), the “interestingness” of a deviation can
be related to the estimated benefit achievable through available actions. In the
course of our work, it was observed that managers and executives have rarely
realized that the very knowledge that can help them alleviate these problems
lies no further than within the wealth of data already at their disposal. In
the context of our application, the company logs in every call, its duration,
the length of the relay, and other information into various databases daily.
This wealth of information, which has so often been underestimated and
under-utilized, can be invaluable for our data mining applications.

2.2. Task identification: Customer retention

In this application, the goal is focused and clear. We are concerned about
customer defection and the goal is to identify the potentially defecting
customers so that steps can be taken to retain them before they actually
defect. At first glimpse, it is difficult to start. The key to finding a solution
is to iteratively decompose it into some solvable sub-problems. The problem
analysis and task decomposition for our application is briefly summarized in
Figure 2.
As we can see in the figure, the main problem of customer retention is
decomposed into three sub-problems or sub-goals. In the first sub-problem,
576 KIANSING NG AND HUAN LIU

we need to identify a list of objective indicators that are representative

of customer profiles. The task is to select pertinent indicators from many
possible ones. We need to employ a reliable method that can select good
indicators from data sets sampled from various databases. The selection of
indicators is recast into a problem of feature selection for classification (Liu
and Motoda, 1998). A list of relevant attributes is identified if they are influen-
tial to the classification of customer loyalty and their likelihood for defection.
As predictive accuracy is the main concern here, a wrapper model of feature
selection is adopted (John et al., 1994). The updating of objective indicators
is accomplished by running an induction algorithm on the data sets. There-
after, we can capture concept drifts by frequently updating these objective
indicators.
The list of objective indicators is then served as the input to the next sub-
problem – identifying the list of potential defectors. This is achieved by using
deviation analysis to measure the actual performance of a customer with those
forecasted from their historical data sets for each of the objective indicators.
The task of customer retention seems completed at this stage. However,
we pursue further. For each potential defector discovered in the second phase,
we look for their associated business partners who are likely to follow suit in
order to maintain their established relaying business. Such predictions can
be obtained through the mining of multiple-level association rules from the
organization and association databases. If they can be convinced to stay with
the company, the potential defectors may have to stay.
In general, our work illustrates how the study and decomposition of
customer retention and the integration of various techniques of data mining
can give rise to an effective solution to a complicated real-world problem.
The details of our implementation follow.

3. Identifying Concept Drifts

The key to solving the problem of customer retention is to identify the list
of potential defectors and predict the consequences following each potential
defection even before they actually take place. Intuitively, this gives rise to
the need for us to first identify a set of relevant attributes or indicators that are
representative for the target concept of “customer loyalty and their likelihood
of defection”. The knowledge found can then be cross-validated against the
existing knowledge, and employed to capture concept drifts.

3.1. The conventional approach

Because of the clear importance of this task, most organizations have had to
rely on the judgments of their human experts to devise a set of “subjective
 CUSTOMER RETENTION VIA DATA MINING 577

indicators”. This has several shortcomings. To begin with, the human’s

analytical and pattern recognition abilities are extremely weak in identifying
factors that are relevant to the classification outcomes. Therefore, the set of
subjective indicators specified by the human experts will often involve uncer-
tainty or incompleteness to some extent. Nevertheless, the list of subjective
indicators is still important, because it can help us determine which of the
many databases contain important attributes and should be used. Secondly,
in the context of a dynamic environment over a period of time, the set of
indicators becomes susceptible to concept drifts. Human experts may not be
able to detect such subtle changes, especially if they have taken place gradu-
ally. Because of such dynamism, a business may be at risk of monitoring the
outdated and irrelevant indicators, or of missing important indicators.
There are other issues as well. For instance, the comprehensive logging of
daily operations in the company has caused the dimensionality of databases to
be very high (ranging from dozens to hundreds of attributes), although many
of the attributes are irrelevant or redundant to the target concept. Usually, the
raw data sets are too large for the user to monitor effectively and efficiently.
Hence, it is necessary for us to first identify a set of relevant indicators. This
is a problem of feature selection (Blum and Langley, 1997). A comprehensive
survey of various methods can be found in Dash and Liu (1997). Since our
problem requires the set of objective indicators to be highly accurate, C4.5
(Quinlan, 1993), a decision tree induction (Quinlan, 1986) method for classi-
fication, was chosen due to its proven good performance. The underlying idea
is that attributes used in a decision tree that gives high accuracy are relevant
and meaningful indicators.

3.2. Decision tree induction for classification

Classification is one of the most important and frequently seen tasks in data
mining: given a large set of training data of the form {A1 , A2 , . . . , An , C},
its objective is to learn an accurate model of how attribute-values (A0i s) can
determine class-labels C. “Decision trees” is one possible model (Quinlan,
1986) from which a set of disjunctive if-then classification rules can be
derived. Classification rules having high predictive accuracy (or confidence)
are employed for various tasks. First, the model can be used to perform
classification for future data having unknown class outcome – prediction.
For example, a bank manager can check a future application against the
classification model obtained from historical data to determine whether an
application should be granted a credit – a screening process. Second, since
those attributes appearing in the classification rules are influential to the
eventual outcome of the classification, the user can have a better under-
standing and insight into the characteristics for each target class. This is
578 KIANSING NG AND HUAN LIU

Table 1. Training data set for classification in credit-screening.

Id Jobless Bought Sex ... Age Savings Granted

1 No car male ... 38 $150K Yes

2 Yes jewel female ... 26 $60K Yes
...
125 Yes stereo male ... 20 $10K No

especially useful in some real-world applications where the users seek to

achieve specific classification outcomes.

3.3. Using decision tree induction to identify objective indicators

Our actual data, sampled from transactional databases residing in Oracle, has
more than 40 attributes and 60,000 periodical records. Because of its confid-
entiality, we choose the “credit” data to illustrate the idea of classification
using decision trees. The data is parially shown in Table 1. The last column
shows the class values. C4.5 is applied to the data to derive the updated
classification rules about customers of the following form:

If (Jobless = “No”) ∩ (Bought = “jewel”) ∩

(Saving > 20) → Granted = “Yes” [90%].

The attributes that appear in the classification rules are objective indicators
as they are found in the data and are considered influential to the target
concept “Granted”. For instance, from the above classification rule, the user
can conclude that attributes “Jobless”, “Bought” and “Saving” are influential
and relevant to the target class of “Granted”, while other attributes such as
“Married”, “Age”, “Sex” are not.
Periodical applications of this method allow the Marketing Department
to objectively identify the most recent set of influencing indicators in order
to capture possible concept drifts. The set of objective indicators are then
compared with the set of subjective indicators identified by the domain
experts. As a result of the cross-validation process, the eventual set of merged
indicators is more up-to-date and more reliable for gauging the loyalty of
customers and their likelihood of defecting. Monitors are then placed on
these loyalty-indicators in the data warehouse so that if any customer shows
significant deviations beyond a certain minimum deviation threshold δMin , an
exception report of defection will be triggered off. This is described in the
next section.
 CUSTOMER RETENTION VIA DATA MINING 579

4. Predicting Potential Defectors

Having identified objective indicators, we now need to identify those

customers who show signs of defecting according to these indicators and
then predict the potentially defecting customers. This is possible because the
customers’ contractual agreements with the company and their sheer size of
relaying commitments established through the company effectively prolong
a defecting process.

4.1. The conventional approach

An intuitive way of identifying the potential defectors is to monitor the

performances of customers over several periods. Similarly, a marketing
analyst in the company will prepare a periodic (for example, weekly, monthly,
or yearly) report showing the percentage change of every major customer’s
relaying volumes in the previous period to that of the previous month and
year. By doing so, the gauge of the loyalty of a customer (sender or receiver)
is tagged to the relative volume change of the customers. Sometimes, this
method does manage to identify potential defectors who pull out abruptly.
However, in most cases, such simplistic analysis can be improved.
The flaw is mainly that the comparisons made on the volume perform-
ances is relative. Even when a sender is maintaining a steady volume with
the company, it could still be defecting. For example, a sender S1 is a fast-
growing company who decides to change to another new relayer. Hence,
S1 diverts its increased volume into the new relayer, while maintaining its
existing volumes with the company. When it has eventually re-established
its relaying business over at the new relayer, it will pull out completely and
suddenly. Another situation involves an economic boom where all the senders
in the Asia Service-Route are increasing at a rate of 10%. A particular sender
in the same route having no increase in volume should become suspicious.
Besides, such comparison has overly relied on the volume performance
for indication of defection. In fact, there are many other indicators that can
also reflect the customer’s loyalty, depending on the application. The use of
multiple indicators can produce more interesting findings. In fact, the quality
and performance of the calls that are dispatched to receivers, as reflected
in some of the attributes of the Relaying Performance database, can also
reflect the loyalty of customers. There are still many other issues such as
the seasonality observed in the relaying business, incorporation of the trends
in the external environments, and the profiles of the subject under study.
Solving such complicated issues requires some predictive modeling methods
that can extrapolate the future performance from the historic performance as
referencing norms. From these defined norms, significant changes can then be
580 KIANSING NG AND HUAN LIU

derived. Our approach further compares the magnitude of every deviation to

that of the other subjects having similar profiles. Only then can we ascertain
whether the deviation actually means a defection.
We use a trend-seasonal forecasting model to predict future performance
for every customer, based on past performance in the various performance
databases. With the predicted norms to serve as the references, we can then
employ deviation analysis (Piatetsky-Shapiro and Matheus, 1994) to identify
“trimming patterns” among the customers. Without these predictions and
analysis, human analysts can barely observe any phenomenon of gradual
deviations at the initial stage of defection.

4.2. Deviation analysis

Deviation analysis is the discovery of significant changes or deviations of

some pre-defined measures from its normative value over a time period in
a data set (Piatetsky-Shapiro and Matheus, 1994). In most applications, the
measured normative value is expressed as the expected value (expectation Et )
of some time-series, or as a forecasted value calculated from applying some
mathematical models like the “seasonal model” that describes the series. In
our work, a deviation δt for time t is given by:

(At − Et )
δt =
Et

where At is the Actual value for the indicator, and Et is the Expected value
for the indicator over a time period of the time-series.
If the analysis detects any deviation δt exceeding a certain user-specified
“minimum deviation threshold δMin ”, i.e., δt > δMin in some pre-defined
measures in the temporal database, it suggests that a significant deviation has
occurred. Some exception reports are generated. Since significant deviations
from the norms are unexpected, they should be “interesting” to the user.
Such statistical analysis method is widely employed in data mining to
discover a few really important and relevant deviations among a multitude
of potentially interesting changes in the temporal databases. Without such a
method, most of the changes are normally “drowned out” by the mass of data
(Matheus et al., 1994) and will remain unnoticed. Even if human analysts
were able to detect the more abrupt pattern changes in the time-series, it
would be extremely difficult to monitor such a large number of deviations
over a long period of time. Nevertheless, finding these patterns is interesting
in discovering higher-level relationships.
 CUSTOMER RETENTION VIA DATA MINING 581

4.3. Forecasting

The measured normative value Et of a time-series is often calculated by

applying some mathematical models. This section will briefly introduce
the concept of time-series forecasting. Forecasting is an area of predictive
modelling that involves the building of an appropriate mathematical model
from historical data, followed by applying the model to forecasting future
measures. Most management decisions today depend on information from
forecasting. This is especially important for big organizations. Given the
large number of time-series to be forecasted periodically, computer-based
quantitative modelling is the only realistic alternative.

4.3.1. Developing a trend-seasonal forecasting model

In general, there are two common classes of forecasting models used in
Management Science. They are time-series models and causal models. In the
former, a series of future “performances” are predicted based on a period
of historical behaviours, whereas in the latter, it is predicted based on other
known and quantifiable factors that will affect “performances”. In our work,
we adopt the former because of the wealth of available historical data in
databases, which can serve as the basis for the training and adjustment of
the model. The causal model has many complications and difficulties that are
still under research.
The particular forecasting model adopted here is a type of multiplic-
ative trend-seasonal model. A multiplicative seasonal model means that the
expected measure in any season/month t within a year is given by A ∗ St ,
where A is the base value of current estimate and St is the seasonal index
for period t. A seasonal index defines the ratio of the actual value of the time
series (week, month, or quarter) to the average for the year. Hence, a value for
St above 1 means that the expected measure in that period exceeds the base
value A, and vice versa. A seasonal model is adopted because the perfor-
mance of a relaying business (e.g., the import and export volumes) follows
a definite pattern that repeats itself cyclically over years. For example, peak
volumes are always expected before the seasons of Christmas and New Year.
In addition, since the relaying business also exhibits a constant but steady
growth of business volumes and customers over years, a trend component is
incorporated into the multiplicative seasonal model. Combining these compo-
nents into a single model, we have a multiplicative trend-seasonal model,
which is used to predict the various measures. The model can generally be
stated by the expected deseasonalized forecast for the period (t − 1) + B with
seasonal adjustment:

Ft = [(A + B ∗ (t − 1)) + B] ∗ St
582 KIANSING NG AND HUAN LIU

where Ft : Forecast for month t in [1 . . . 12], A: Base value of the current

estimate, B: Slope (gradient) of the trend line, and St : Seasonality Indices for
month t.
To forecast, the user should specify three types of data sets: the warm-
up data, training data and forecast data. First, the warm-up data set, which
comprises a selected range of historical data, is used to compute the initial
estimates of the base value A, the slope of the trend line B and the seasonality
indices St for each month t. This provides the initial unadjusted forecasting
model. Next, the training data set is selected from the next period of data
that is not used in the warm-up data set. This step uses the current estimates
of A, slope of the trend line B, and seasonality indices St to extrapolate
the forecasted measures Ft . The difference between the computed forecast
Ft and the actual measures is then used to adjust the estimates of A, B
and the seasonality constants of St based on “exponential smoothing” (to be
discussed next). This step is crucial as it adjusts the above three factors in
the forecasting model proportionally according to the fluctuations observed
in the actual measures from the predicted forecast measures. Last, a range
of future periods after the training data are selected to form the forecast data
set. The predicted measures in these periods are extrapolated using the trained
base-value A, slope-value B, and seasonal indices St .

4.3.2. Exponential smoothing

The method of exponential smoothing works like an auto-pilot designed to
continuously use the forecast error in one period to correct and improve the
forecast of the next period. On the basis of comparing the forecast Ft with
the actual measure in that period, the adjustment method can compute new
estimates for base At , the slope of the trend line Bt and the seasonality indices
St by adjusting three smoothing constants – α, β, and γ respectively.
In the following example, we will only make use of the base value A
and its smoothing constant α for illustration. From the historical data, we
obtain an estimate of A = 10, 000 which is then used as the forecast for
the actual outcome in 1995. Due to an economic boom in 1995, the actual
performance is 11,000. Thus, we had a forecast error of 11,000 – 10,000 =
1,000. Considering any random fluctuation, we adjust the estimate of A for
the 1996 forecast by the fraction of the forecast error that we attribute to the
actual shift in the value of A. We can specify this fraction (in the range 0 to
1) in the form of a smoothing constant. For instance, if α is set to 0.1, we
are actually attributing 10% of the current forecast error to an actual shift in
the value of A and 90% to randomness. In general, the closer the smoothing
constant is to 1, the larger the fraction of the forecast error we are attributing
to an actual shift.
 CUSTOMER RETENTION VIA DATA MINING 583

Usually, these smoothing constants are left to the control of the end-user
in a dynamic environment, although empirical experiments have shown that
a value between 0.10 and 0.30 for all the smoothing constants often results
in reliable forecasts. However, if the user expects the level of the estimate to
change permanently in the immediate future because of some special circum-
stances, then a larger value of a smoothing constant (like 0.7) should be used
for a short period of time. Once the computed level of the forecasting model
has changed in accordance with these special circumstances, the user should
then switch back to a smaller value of the smoothing constant.

4.4. Using deviation analysis and forecasting to discover potentially

defecting customers

In our implementation, the normative values Et for each of the five to seven
“indicators” are first developed through a trend-seasonal forecasting model
(Levin et al., 1992), based on the customer’s historical performance in the
temporal databases. With an annual size of more than 50,000 records avail-
able for each indicator to train the forecasting model, the accuracy of the
predicted normative values can be increased. This should also be credited to
the additional factors taken into consideration in our approach:
− the use of seasonal indeces adjusts the forecast according to the annual
seasonal pattern in the time-series, and
− the use of exponential smoothing allows more weight to be assigned to
the recent data, thus taking into account the current circumstances, like
a recent economic downturn in South-East Asia.
With the normative values of every indicator forecasted for every
customer, deviation analysis is performed to detect those customers who
show significant deviations, or δt > δMin . These customers are deemed to
be potentially defecting and this warrants a further “interestingness valid-
ation” (Matheus et al., 1994). In other words, their deviations are further
compared with those δSR of the (aggregated) customers operating in the
same service-route. By doing so, we take into account the trends in the
external environment and the profiles of the subject under consideration. This
will ensure that a “real” deviation is exclusive only to the specific subject
and not some general phenomenon experienced by other subjects too in the
same service-route. For example, the Asian boom in the mid of last decade
had generally boosted the volumes of those in the Asia service-route. Simi-
larly, the recent Asian economic crisis also causes an overall reduction in
the volumes of those senders and receivers in the Asia service-route. An
illustration of one such analysis is shown in Figure 3.
In the analysis, senders S1 , S4 and consortium C2 showed significant
deviations δ1 which satisfied the δMin of –10%. These are further compared
584 KIANSING NG AND HUAN LIU

Result of deviation analysis

Query for deviation
Select distinct Sender Sender c.f. ServiceRoute Interest’g
From Indicator I1 S1 (-11%) c.f. SR1 (+01%) Yes
Where δMin ≤ -10% S4 (-11%) c.f. SR5 (-10%) No
And Date = “Jan” C2 (-15%) c.f. SR3 (-01%) Yes

Figure 3. An illustration of deviation analysis: (left) – A SQL-like query interface for

“monitor” setting. (right) – An “interestingness” evaluation of the deviation analysis.

with the average deviation δSR of the aggregated customers in their respective
service-routes. In S1 ’s case, the general population in SR1 performs reas-
onably well (a positive deviation +1%), suggesting that S1 ’s deviation is
unexpected and thus interesting. In S4 ’s case, the general population in SR5
performs equally badly (a deviation of –10%), suggesting that S4 ’s deviation
should be expected and thus not interesting. As mentioned above, we can
usually relate S4 ’s kind of deviation to some regional event like the current
economy turmoil in Asia that affects all the relaying operations in the Asia’s
service-routes. If no such explanation can be found, then it would mean that
all the customers in the service-route are declining.
If consistent deviations are also observed across the set of indicators for
“deviating” customers like S1 and C2 , then a periodic exception report is
produced to alert the domain experts on these possible “defectors”. Domain
knowledge and insights are then applied to verify the findings for each of
these cases and the suspected potential defectors will be monitored closely
for the subsequent periods. Persistent deviations are strong signs of likely
defection. Besides performing deviation analysis on the Customer concept,
similar analysis can also be applied to investigate and identify upcoming
or weakening Markets (continents and countries) and service-routes for the
purpose of marketing.

5. Avoiding the Chain Effect

Many organizations would be content if they can predict the potentially

defecting customers. Nevertheless, this work goes one step further. We ask
who else will likely follow suit for each of the potential defectors. Such
association knowledge is especially important to a relaying business. This
is because the choice of a particular relayer linking different senders and
receivers via different service-routes and Markets is usually dictated by a few
 CUSTOMER RETENTION VIA DATA MINING 585
Table 2. A database of Message-Transactions for association rule mining.

Date-Time Msg-Id In-Msg Sender Out-Msg Receiver ...

970501-1210 AB0012 M1 Hong Kong M4 Amsterdam ...

970501-1230 RX3766 M3 Shanghai M7 Jakarta ...
...
970601-0115 AO9912 M7 Frankfurt M2 Kuala Lumpur ...

major players. Since they carry very large relaying volumes, they have great
influence over the smaller companies. Hence, the defection of a major sender
will encourage similar behavior in their associated business partners who
will attempt to preserve established relationships. This can inflict a severe
dent to the financial health of a relayer. Hence, there is much incentive for
the Marketing Department to have a full picture of the consequences from
an identified potential defection. If we wait until a “chain effect” becomes
observable to the human analysts, it would be too late. In short, preventive
measures should also be taken to take care of the followers when a potential
defector is detected as they can also influence the major players to change
their stands.

5.1. The conventional approach

The conventional approaches include information exchanges between

relayers, senders, and receivers. They offer only subjective and often unreli-
able prediction of the association relationships among the customers. From
the earlier data analysis, we have identified that the Transactions data-
base, having some 120,000 records per month, contains attributes incoming
messages and outgoing messages for every relaying transaction. Table II
depicts the design of the Transactions database.
From such transactional records, we can mine for association rules
(Agrawal et al., 1993) which represent the transactional relationships between
messages, senders and receivers in the notation of an association rule:

M1 → ? [s% support, c% confidence]

Although the mining of association rules at the message level will give
a good idea of the association relationships between senders and receivers,
knowledge of this level does not provide much business value for our applic-
ation. This is because knowledge at too low a level (over-specific) will end
up looking like the raw data and having little general meaning. Since most
586 KIANSING NG AND HUAN LIU

Figure 4. A taxonomy of the “Customer” concept in a relaying business.

concepts in a real-world context involve multiple levels of abstractions, it

causes many problems for an application. Hierarchical concepts are also
present in our application domain. For example, our analysis of the concepts
Customer and Market reveals some taxonomies as follows:
− {groups  consortiums  senders} ∈ Customers
− {continents  countries  relayers} ∈ Market
As an example, we will elaborate on the concept of Customer. The taxonomy
is shown in Figure 4. A group is an association of consortiums that forms
for the purpose of negotiating a better volume rebate/discount or qualifying
for a better charging scheme. Most consortiums will therefore form groups
with their associated business partners. Hence, a group consists of many
consortiums, and each consortium owns several individual senders. With this
hierarchy, we can deduce a trivial association rule, which is “a consortium
is associated with some other consortiums in the same group.” Therefore,
the defection of one of them may affect their partners in the group. This is
because the remaining may fail to reach the minimum volume quota in order
to enjoy the rebate or discount. Since the association rules at a low level, say at
the sender level, are too specific to have much application value, we mine for
generalized association rules (Srikant and Agrawal, 1996), or generate rules
that are as general as possible by taking the existing taxonomies into account.
A similar approach to finding multiple-level association rules (Han and Fu,
1996) is employed here.

5.2. Multiple-level association rules

Many databases in the real-world are transaction-oriented and do not contain

class labels. The most popular example is the supermarket’s bar-coded
transactional data. Generally, the task of mining association rules over a trans-
 CUSTOMER RETENTION VIA DATA MINING 587

actional database can be formally stated as follows: Let I = {i1 , i2 , . . . , im }

be the set of items in the database. Each transaction T in the database D has
a unique identifier and contains a set of items called an itemset. An itemset
with k items is called a k-itemset. The support of an itemset is the percentage
of transactions in D that contain the itemset. An association rule is a condi-
tional implication among itemsets, A → B, where itemsets A, B ⊂ I . The
support of the association rule is given as the percentage of the transactions
that contain both A and B, and the confidence is given as the conditional
probability that a transaction contains B, given that it contains A.

5.2.1. Predicting the followers of a potential defector

The Transactions database contains the transactional associations between
senders and receivers with respect to messages. We can derive association
rules between customers having the following form for a message:

Sender 1.3.3 → Receiver 2.1.3 [s% support, c% confidence]

Associations at this level may be too specific to be of any value to

a business. We therefore generalize the associations to higher taxonomy
levels, in accordance with the hierarchical taxonomy of Customer (like the
one in Figure 4). For example, group G1 includes consortiums C1.1 , C1.2 ,
and C1.3 . Group G2 contains consortiums C2.1 , C2.2 , and C2.3 . Employing
the algorithms presented in Han and Fu (1996), association rules at the
consortium and group level are generated:
− Consortium: C1.3 → C2.1 [30% support, 50% confidence]
− Group: G1 → G2 [50% support, 70% confidence]
The first association rule is interpreted as “if consortium C1.3 defects, so
may C2.1 (50% confidence)”. The second association rule says “if group G1
defects, so may G2 (70% confidence)”. The following example illustrates the
chain effect of a defection based on the taxonomy of Customer. If consortium
C1.3 defects, consortiums C1.1 and C1.2 within the same group G1 are likely
to defect too because of their established business transactions. So will C2.1 ,
due to its association with C1.3 found in the above association rule. This will
further affect the consortiums C2.2 and C2.3 within the same group of G2 .
Similarly, should group G1 defect, G2 is likely to follow suit. Eventually, all
the consortiums in these two groups will defect together to another relayer. It
is imperative to detect this kind of chain reaction before it is too late.
The mining of multiple-level association rules enables the company to
regularly predict possible chain reactions of a defection, thus giving it a good
chance to take pertinent actions before a major customer starts leaving for
a new relayer and thereby influencing its associated business partners not to
leave either. The knowledge of associations also allows the marketing depart-
588 KIANSING NG AND HUAN LIU

ment to customize attractive schemes for the identified customers’ alliances

in order to attract more volumes from them. It should be highlighted here that
association rules may sound relatively easy to be detected by human analysts.
However, in a database containing several millions of transactional records,
these associations are simply “drowned out”.

6. Conclusions

In the course of our work, we have identified some interesting objective indi-
cators among a large number of attributes. The finding has verified our earlier
conjecture on the limitations of human capabilities. In addition, preliminary
experiments on the historical data sets have successfully identified some
already defected customers long before they showed prominent signs of
defection. This work is significant because our approach can be generalized
into solving similar problems in the sales and services related industries, like
Telecommunications, Internet Service Providers, Insurance, Cargo Trans-
shipment, etc. For instance, a popular strategy used by many companies in the
services industry is using attractive promotions and discounts to “lure” new
customers into short-term services under them. Even department stores in the
sales industry come up with their own VIP smart cards in a bid to retain their
customers. We would like to highlight that the information from the logs and
databases can potentially be turned into valuable knowledge for competitive
advantage. For example, the customers’ particulars and their profiles (like
mobile-phone or Internet usage patterns) could be mined for predicting a list
of potential defectors among them. Since most customers are bound to the
services of a company for at least a period of time (usually around a year),
special offers can be made to those who show signs of dissatisfaction.
Many ideas presented here can in fact be modified to suit various applica-
tions of similar needs. For instance, although the third sub-task in our work
is made possible by the availability of the transactional associations, unique
to a business of hierarchical structure, there are many other kinds of asso-
ciations in different problem domains. Spatial associations can be identified
and applied in some property-related problems while sequential associations
can be found in a sales transactions database and applied to predicting future
purchases in E-business. These different associations in different problem
domains can help infer valuable knowledge.
One of the goals of this work is to show that the maturity of data mining
has reached a point where large-scale applications to practical problems are
desirable and feasible. This work will hopefully create some sort of chain
effect in motivating the strategic use of data mining in business applications
where conventional approaches fall short. The success of practical appli-
 CUSTOMER RETENTION VIA DATA MINING 589

cations serves to remind the executives and managers that understanding

the underlying concepts behind data mining methods remains the key to a
successful data mining application. There is no such thing as a data mining
tool that can fit into every kind of problem. Neither is there any short-cut
solution to a complicated problem. A detailed analysis, a good design and a
systematic development are necessary for a successful application. This work
further demonstrates that in the context of practical applications data mining
is very much an art.

Acknowledgements

We would like to thank Farhad Hussain and Manoranjan Dash for helping us
finalize this version of the paper, and the company involved in the project to
make this application possible although it is unfortunate that the identity of
the company cannot be mentioned. We are also indebted to the anonymous
reviewers and the editor for their detailed constructive suggestions and
comments.

References

Agrawal, R., Imielinski, T & Swami, A. (1993). Database Mining: A Performance Perspective.
IEEE Trans. on Knowledge and Data Engineering 5(6): 914–925.
Agrawal, R. & Srikant, R. (1994). Fast Algorithms for Mining Association Rules in Large
Databases. In: Proceedings of the 20th VLDB Int’l Conference, Santiago, Chile, 487–499.
Blum, A. & Langley, P. (1997). Selection of Relevant Features and Examples in Machine
Learning. Artificial Intelligence 97: 245–271.
Clearwater, S., Cheng, T., Hirsh, H. & Buchanan, B. (1989). Incremental Batch Learning. In
Segre, A. (ed.) Proceedings of The Sixth International Workshop on Machine Learning,
366–370. Morgan Kaufmann Publishers, Inc.
Dash, M. & Liu, H. (1997). Feature Selection Methods for Classifications. Intelligent Data
Analysis: An International Journal 1(3). http://www-east.elsevier.com/ida/free.
htm.
Fayyad, U., Piatetsky-Shapiro, G. & Smyth, P. (1996). From Data Mining to Knowledge
Discovery: An Overview. In Fayyad, U., Piatetsky-Shapiro, G., Smyth, P. & Uthurusamy,
R. (eds.) Advances in Knowledge Discovery and Data Mining, 495–515. AAAI Press /
The MIT Press.
Han, J. & Fu, Y. (1996). Attribute-Oriented Induction in Data Mining. In Fayyad, U.,
Piatetsky-Shapiro, G., Smyth, P. & Uthurusamy, R. (eds.) Advances in Knowledge
Discovery and Data Mining, 399–421. AAAI Press / The MIT Press.
John, G., Kohavi, R. & Pfleger, K. (1994). Irrelevant Feature and the Subset Selection Problem.
In Cohen, W. A. H. H. (ed.) Machine Learning: Proceedings of the Eleventh International
Conference, 121–129. New Brunswick, N.J.
590 KIANSING NG AND HUAN LIU

Kamber, M., Han, J. & Chiang, J. (1997). Metarule-Guided Mining of Multi-Dimensional

Association Rules Using Data Cubes. In Proceedings of the 3rd Int’l Conference on
Knowledge Discovery and Data Mining (KDD-97), 207.
Levin, R., Rubin, D., Stinson, J. & Gardener, E. (1992). Quantitative Approach to Manage-
ment (8th Edition), Ch. 3, Forecasting, 103–134. McGraw-Hill.
Liu, H.& Motoda, H. (1998). Feature Selection for Knowledge Discovery Data Mining.
Boston: Kluwer Academic Publishers.
Matheus, C., Piatetsky-Shapiro, G. & McNeil, D. (1994). An Application of KEFIR to the
Analysis of Healthcare Information. In Proceedings of the 11th Int’l Conference on
Artificial Intelligence AAAI-94, Workshop on Knowledge Discovery in Databases, 25–36.
Merz, C. & Murphy, P. (1996). UCI Repository of Machine Learning Databases. http://
www.ics.uci.edu/∼mlearn/MLRepository.html. Irvine, CA: University of Cali-
fornia, Department of Information and Computer Science.
Piatetsky-Shapiro, G. & Matheus, C. (1994). ‘The Interestingness of Deviations. In Proceed-
ings of the 11th Int’l Conference on Artificial Intelligence AAAI-94, Workshop on
Knowledge Discovery in Databases, 25–36.
Quinlan, J. (1986). Induction of Decision Trees. Machine Learning 1(1): 81–106.
Quinlan, J. (1993). C4.5: Programs for Machine Learning. Morgan Kaufmann.
Srikant, R. & Agrawal, R. (1996). Mining Quantitative Association Rules in Large Relational
Tables. In Proceedings of ACM SIGMOD Conference on Management of Data. Montreal,
Canada.
Widmer, G. (1996). Recognition and Exploitation of Contextual Clues via Incremental
Meta-Learning. In Saitta, L. (ed.) Machine Learning: Proceedings of the Thirteenth
International Conference, 525–533. Bari, Italy.