 

ain‘m
13:1; | | \.

M . .
‘v A'i “‘1‘
59%»?1

ﬂy."

.“ﬂ
'11-'15 “W
‘g'h-Whﬂ 21.2;

m.
t has ,
131‘}. 5

v

”I: ‘ ’1'?

‘fl
.41.: “ "way?!
A 53‘!

. - '17.?

.131 idi‘ﬁét'l‘ “3'!
A?q:1. '

3V

 

...
I 4* , .
' r ‘wgﬁéﬁgb
. ”I _
4:? ﬂ?

,3 1325113:

zggy'ﬁx iii-‘3

 

This is to certify that the
dissertation entitled

Examining the Inﬂuence of Loyalty Reward
Program Membership on the Behavior
Of Casino Patrons

presented by

Robert Palmer

has been accepted towards fulﬁllment
of the requirements for the

Ph.D. degree in Park, Recreation and Tourism

 

 

am

Major Prdfessor’ s Signage

December 5, 2003

 

Date

MSU is an Afﬁrmative Action/Equal Opportunity Institution

- -.--.-.-.-.--.—<~ ---- -t-.--‘-4--.--.-.-.----o-.--t-o--.-.--.-.-t-.-.. -

-.---.-.-:-.a.-‘--.-.-.-v— .—

 

 

W'
Michigan State
University

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

5119}. 51127005

yin?

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DateDue.p65.p.15

 

 

Examining the Inﬂuence of Loyalty Reward Program
Membership on the Behavior of Casino Patrons
by

Robert Palmer

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirement
for the degree of
DOCTOR OF PHILOSOPHY
Department of Parks, Recreation, and Tourism Resources

2003

ABSTRACT

Examining the Inﬂuence of Loyalty Reward Program Membership on the
Behavior of Casino Patrons

By
Robert Palmer

Numerous industries have implemented frequent buyer or loyalty rewards
programs as part of their marketing mix. As the gaming industry continues to
grow exponentially, these organizations are using player’s club programs as a
key component of their marketing strategies. This study investigates the
relationship between casino player’s club programs and its members. As the
gaming industry continues to grow exponentially, these organizations are using
player’s club programs as a key component of their marketing strategies.
Building on past loyalty studies, this study will attempt to further the knowledge of
how these programs affect the behavior of casino patrons.

Through the use of a 200,000 member casino player's club tracking
database, a detailed analysis of four years of member behavior was undertaken
to explore how behaviors have changed over the course of their membership.
The ﬁndings suggest that not all segments of a loyalty program are potentially
proﬁtable and that the costs of serving frequent (loyal) customers increase over
the course of their lifetime. A structural model was developed in order to
examine the development of attitudinal and behavioral loyalty based on data
collected from a mail survey of 3000 player’s club members. The results suggest
that perceived value is a better predictor of behavioral loyalty as compared to

attitudinal loyalty.

Overall this study had four distinct contributions: (1) It reveals that lifetime
visitation does not fully explain lifetime casino revenues and that the costs
associated with serving loyal long-term player’s club members actually increase
over their lifetime; (2) It demonstrates that commonly utilized loyalty
segmentation bases are ineffective in terms of identifying segments of a player’s
club program; (3) the development of behavioral loyalty varies across segments
of players club members, and (4) a large segment of members in the player’s
club program are potentially unproﬁtable and should not be targeted as part of
any marketing initiative. Practical and research implications will be also provided

by the author.

DEDICATION

Always remember to forget
The things that made you sad.
But never forget to remember

The things that made you glad.
Always remember to forget
The Friends that proved untrue.
But never forget to remember
Those that have stuck by you.
Always remember to forget
The troubles that passed away.
But never forget to remember
The blessings that come each day.
Irish Blessing

iv

ACKNOWLEDGEMENTS

First of all I would like to thank my parents for their years of support and love.
I would also like to thank my advisor and mentor, Ed Mahoney for all of his
support, patience, and dedication. I would also like to thank my Committee
members; Dr. Baker, Dr. Fridgen, and Dr. Spreng for your knowledge, help, and
support throughout this process. Finally, I would like to thank all of my friends
who have been so kind to deal with my hectic schedule, my traveling here and

there and everywhere and for staying with me as I achieved my goals.

TABLE OF CONTENTS

LIST OF TABLES ....................................................................................... viii
LIST OF FIGURES ....................................................................................... iv
CHAPTER 1 ................................................................................................................... 1
INTRODUCTION ............................................................................................................ 1
Product Centric to Customer Centric ............................................................................ 1
The Growth of Loyalty Rewards Programs ................................................................... 3
Growth of Casino Gaming and Casino Loyalty Programs ............................................. 5
Previous Empirical Studies of Loyalty Programs .......................................................... 9
Need and Importance of this Study ............................................................................ 10
Study Purposes and Objectives ................................................................................. 17
Organization of the Dissertation ................................................................................. 21
CHAPTER 2 ................................................................................................................. 22
LITERATURE REVIEW ................................................................................................ 22
Introduction ................................................................................................................ 22
Deﬁning Customer Loyalty ......................................................................................... 22
Satisfaction / Service Quality Research ...................................................................... 28
Customer Lifetime Value Research ............................................................................ 29
Loyalty Program Research ......................................................................................... 32
Customers Want Relationships with Companies ........................................................ 36
Loyal Customers are Proﬁtable .................................................................................. 37
Identification of Loyal Segments ................................................................................ 38
Integrated Loyalty Research ...................................................................................... 39
Deﬁnitions and Determinants of Loyalty ..................................................................... 42
Player’s Club Program Components .......................................................................... 44
Mediating and Moderating Variables .......................................................................... 49
Summary of Review of Literature ............................................................................... 51
CHAPTER 3 ................................................................................................................. 53
Methods. ..................................................................................................................... 53
Introduction to Methods .............................................................................................. 53
Study Population and Study Data ............................................................................... 53
Client Casino’s Player’s Club Database ..................................................................... 55
Mail Survey of Player’s Club Members ....................................................................... 60
Mail Survey Instrument .............................................................................................. 60
Mail Survey Sampling Method .................................................................................... 67
Mail Survey Administration ......................................................................................... 68
Preparation of Data Collected Through Mail Survey ................................................... 70
Preparation of Data Collected Through Mail Survey for Use in Structural Equation
Modeling .................................................................................................................... 71
Model Testing Using Structural Equation Modeling .................................................... 72
Limitations of data ...................................................................................................... 74

vi

CHAPTER 4 ................................................................................................................. 75

RESULTS ................................................................................................................... 75
Introduction ................................................................................................................ 75
Lifetime Visits — Lifetime Revenues Relationship ....................................................... 75
Descriptive Proﬁle of Player’s Club Members who Completed the Mail Survey .......... 82

Description of Loyal Player's Club and Disloyal Player's Club Segments Based on
Gaming Activity at the Client Casino during the Previous 12 months as Compared to

their Lifetime Gambling Behavior at the Client Casino ................................................ 86
Testing of Alternative Loyalty Signiﬁcant Segments ................................................... 91
Segmentation Based on Proportion of Total Visits to Casinos taken to Primary Casino
.................................................................................................................................. 92
Segmentation Based on Amount Budgeted for Gaming at Casinos per year .............. 96
Conﬁrmatory Factor Analysis ................................................................................... 101
' Differences of Factor Scores for Loyalty Segments .................................................. 107
Testing of Integrated Casino Loyalty Model ............................................................. 116
Model Reﬁnements .................................................................................................. 120

Using the Structural Model for Identifying Differences Among Loyalty Segments Based
on Respondents Gaming Activity at the Client Casino During the Previous 12 Months

................................................................................................................................ 127
Testing of Structural Model for Identifying Differences Among Loyalty Segments Based
on Alternative Segmentation Bases ......................................................................... 132
Testing of Rival Segments ....................................................................................... 146
Identiﬁcation of Market Segments ............................................................................ 150

CHAPTER 5 ............................................................................................................... 164
SUMMARY, DISCUSSION, CONCLUSIONS AND RECOMMENDATIONS ............... 164
Summary ................................................................................................................. 165
Discussion ............................................................................................................... 172
Conclusions ............................................................................................................. 176

Managerial Implications ........................................................................................... 178

Research Implications .............................................................................................. 181
Appendix 1 - Mail Survey Instrument ...................................................................... 183
Appendix 2: Individual Item Mean Scores for Marketing Segments ...................... 191
REFERENCES ........................................................................................................... 197

vii

Table 1:
deﬁned.
Table 2:
Table 3:

Table 4:
Table 5:
Table 6:
Table 7:
Table 8:
Table 9:

Table 10:
Table 1 1:

Table 12:

Table 13:
Table 14:
Table 15:

Table 16:

Table 17:
Table 18:

Table 19:

Table 20:
Table 21:

Table 22:
Table 23:

Table 24:

LIST OF TABLES

Selected Literature Completed on Customer Loyalty .. 25Error! Bookmark not

Summary of Research Completed on Loyalty Programs ............................. 34
Data Collected by the Client Casino’s Players Club Tracking System and

Utilized in This Study ................................................................................. .56
Membership History for the Players Club for the Client Casino ................. 58
Data Collected in Mail Survey ...................................................................... 61
Loyalty Factor Constructs Included in Mail Survey ...................................... 64
Number of Surveys Received for Each Mailing ............................................ 69
Overall Response Rate to Mail Survey ........................................................ 69

Segmentation of Client Casinos Player’s Club Members Based
On Lifetime Action Days and Lifetime Theoretical Win at the

Client Casino ................................................................................. 77
Sub-groups Comprising the High Frequency/ High Revenue Group of
Player’s Club Members ............................................................................... 81
Descriptive Proﬁle of Player's Club Members of the Client Casino that
Returned a Completed Mail Survey ................ Error! Bookmark not deﬁned.83

Survey Derived Proﬁles for Segments Based on Gaming Activity at the
Client Casino During the Previous 12 months as Compared to their Lifetime
Gambling Behavior at the Client Casino ...................................................... 87
Descriptive Proﬁle of Segments formed Based on the Proportion of Visits
(During the Previous 12 Months) to their Primary Casino ............................ 93
Proﬁle of Segments Formed Based on Total amount Budgeted perYear for
Casino Gaming ........................................................................................... 99
Results of Confirmatory Factor Analysis of Potential Loyalty Model Factors
.................................................................................................................. 103
A Comparison of Mean Loyalty Factor Scores for Segments Based on
Gaming Activity at the Client Casino During the Previous 12 months as

Compared to their Lifetime Gambling Behavior at the Client Casino .......... 109
Comparing Mean Loyalty Factor Scores Between Segments Based on
Percentage of Total Visits to Casinos taken to Primary Casino ................. 110
A Comparison of Mean Loyal Factor Scores for Groups Based on Total
Amount Budgeted for Casino Gaming per Year ......................................... 115

Summary of Path Coefﬁcients for Segments Based on Gaming Activity at the
Client Casino During the Previous 12 months as Compared to their Lifetime

Gambling Behavior at the Client Casino .................................................... 131
Summary of Path Coefficients for Segments Based on Percentage of Total
Visits to Casinos made to Their Primary Casino ........................................ 138
Summary of Path Coefﬁcients For Segments Based on Total Amount
Budgeted per Year for Trips to Casinos per Year ...................................... 145
The Six Player's Club Market Segments ................................................... 154
Behavioral and Socioeconomic Characteristics of the Six Player’s Club
Marketing Segments ................................................................................. 155

Mean Loyalty Factor Scores for the Six Player's Club Marketing
Segment ...................................................................................... 158

viii

Figure 1:
Figure 2:
Figure 3:
Figure 4:
Figure 5:
Figure 6:

Figure 7:
Figure 8:
Figure 9:

Figure 10:

Figure 11:

Figure 12:

Figure 13:

Figure 14:

Figure 15:

Figure 16:

Figure 17:

Figure 18:

Figure 19:

LIST OF FIGURES

The Main Goals of Loyalty Programs ........................................................... 4
Beneﬁts Offered to Members of Harrah's Total Rewards Tiered System ...... 8
Integrated Casino Loyalty Model ................................................................ 13
Loyalty Segments Identified in Analysis of a Mail-Order Catalog Company 39
Study Framework ....................................................................................... 54
Loyalty Program Segments Based on Lifetime Action Days and Lifetime
Revenues at the Client Casino .................................................................... 59
Elements of the Alternative Integrated Casino Loyalty Model .................... 120

Measurement Model For Alternative Integrated Casino Loyalty Model......121
Structural Model for All Members of the Client Casinos Player’s Club that
Returned a Completed Mail Survey ........................................................... 126
Results of the Testing of the Alternative Integrated Casino Loyalty Model for
those Respondents (Loyal Player’s Club segment) who had Visited the Client
Casino at Least Once Each of the Four Previous Years ............................ 129
Testing of the Alternative Integrated Casino Loyalty Model for those
Respondents (Disloyal Player’s Club segment) who have been a Member of
the Client Casinos Player's Club for Over Four Years and Either had no
Recorded Activity from January 1, 2002 to December 15, 2002 or Only Five
Percent or Less of their Total Lifetime Theoretical Win was Represented by
Gaming Activity from January 1, 2002 to December 15, 2002 at the Client
Casino ....................................................................................................... 130
Path Coefﬁcients for the Low (primary casino) Visitation Segment Whose
Percentage of Total Visits to their Primary Casino is 40% or Less of Their
Total Trips to Casinos ............................................................................... 135
Testing of the Structural Model for Persons Moderate (primary casino)
Visitation Segment Whose Percentage of Total Visits to Their Primary Casino
is 41% to 69% of Their Total Trips to Casinos ........................................... 136
Full Structural Model Results for the High (primary casino) Visitation
Segment Whose Percentage of Total Visits to their Primary Casino is 70% or
More of Their Total Trips to Casinos .......................................................... 137
Results of the Testing of the Alternative Integrated Casino Loyalty Model for
the Player’s Club Segment that Budget $20 to $700 per year (Low Casino
Spenders) for Casino Gaming ................................................................... 142
Alternative Integrated Casino Loyalty Model Results for the Player’s Club
Segment that Budget $701 to $2099 per year (Moderate Casino Spenders)
for Casino Gaming .................................................................................... 143
Structural Model Results for the Alternative Integrated Casino Loyalty Model
for the Player’s Club Segment that Budget $$2100 or More per year (High
Casino Spenders) for Casino Gaming ....................................................... 144
Results of the Testing of the Full Structural Rival Model for all
Respondents.....149
Three Stage Process for Identifying Player’s Club Marketing Segment ..... 153

CHAPTER 1
INTRODUCTION

Product Centric to Customer Centric

Marketing managers today are facing challenges that can only be
characterized as revolutionary. Many ﬁrms must now compete on a global level.
Consumers have many more product, service, and entertainment alternatives to
choose from and an enormous amount of information on which to base their
appraisals and selections. In response, marketing managers have focused on
developing strategies for increasing customer loyalty as a key part of their overall
marketing strategy. Even though loyalty building and retention strategies
Incorporate various customer relationship management tools and systems, they
continue to be based on conventional ways of acting or thinking (Berthon,
Holbrook, and Hulbert, 2000).

Although imbalance of power between customers and businesses has existed
for centuries, the power (e.g. information about alternative products and services)
now resides in the hands of customers. This power allows customers at all
stages of the supply chain to exact ever more demands from suppliers and/or
organizations (Berthon et al., 2000). Businesses have responded to this
imbalance by striving to become more “customer centric” as opposed to “product
centric” by developing new marketing approaches, including a much greater

emphasis on loyalty programs (Woods, 1998).

proﬁtability and gain a competitive advantage (Fredericks, Hurd, and Salter,
2001). Examples of these initiatives include Total Quality Management (TQM)
and the Malcolm Baldn'dge National Quality Award. Unfortunately, research has
shown that the majority of ﬁrms have not achieved the predicted economic
beneﬁts from implementing these initiatives (Anderson Anderson, Fornell, and
Lehmann, 1994; Bolton, 1998; Reichheld and Teal, 1996). However, they
spawned a great deal of emphasis among academics and practitioners on
developing new customer attentive business approaches. As a direct result of
this work, many service organizations in the 19903 focused on implementing
relationship management strategies intended to attract, maintain, and enhance
customer relationships (Bolton, 1998).

The casino gaming industry is one of the industries that have adopted a more
“customer centric” loyalty building marketing orientation. Casinos have moved
away from solely focusing on building multi-million dollar facilities and attractions
and are now placing greater emphasis on developing marketing programs
intended to forge and enhance relationships in order to retain their customer
base. This is in large part due to the increased competition brought on by the
exponential growth of casinos and other gaming opportunities (e.g. web based
gaming, lotteries) within the United States. While more persons are gaming and
revenues continue to climb, individual casinos are now directing more resources

at efforts aimed at creating longer lasting relationships with gamers.

The Growth of Loyalty Rewards Programs

One popular marketing relationship strategy is loyalty rewards programs.
Loyalty rewards programs are also referred to as afﬁnity groups, frequent buyer
programs, and customer clubs. They were ﬁrst developed by the airline industry
in the early 1980s and are now being utilized by many different tourism and
hospitality industries including casinos, hotels, and restaurants. Examples of
loyalty reward programs include Northwest Airlines WorIdPerks Plan, Harrah’s
Total Rewards Program, and Hertz Club Gold card. For example, Harrah’s Total
Rewards Program allows members access to special events, comp privileges
(e.g. free meals, hotel stays), and cash back programs. The beneﬁts provided to
the consumers vary based on their level of spending per calendar year.

A review of the history of these programs reveals that the proliferation of
loyalty programs across an industry is often the result of a progression of
competitive reaction to the introduction of a loyalty program by an industry
innovator. For example, within weeks after American Airlines introduced their
frequent ﬂyer program in the early 803, each of its major competitors had a
similar program. Interestingly, this “follow the leader” approach can be seen for
almost every industry that loyalty rewards programs are being utilized including
airlines, hotels, casinos and grocery stores. The question that must be asked is
whether the managers of these “successive” organizations analyzed the beneﬁts
of loyalty rewards programs within the context of their own marketing strategies,

or were they merely provoked by the competition.

Although the scopes of these programs vary considerably across industries,
they frequently play a critical role in the marketing strategies and account for a
signiﬁcant portion of their marketing budgets (Dowling and Uncles, 1997). While
the goals of these programs are predominantly marketing focused (depicted in
Figure 1), they can also benefit other business units and functions. For example,
tracking the services/products purchased by loyalty program members produces
information can be utilized by operation managers for their product planning
decisions. This same information can be utilized by human resources for
developing their training and hiring schedules. Additionally, loyalty club

promotions can be used to test market responsiveness.

Figure 1: The Main Goals of Loyalty Programs

 

 

 

Customer Win New
Loyalty Customers

 

 

 

 

 

 

   
   
      

Increase in
- Revenue
- Proﬁt
- Market

Share

    

 

 

Create
Communication

Build a Strong
Database

 

 

 

 

 

 

 

Support Other
Departments

 

 

 

 

 

 

Source: (Butscher 1998)

The major focus of loyalty programs is to increase revenues, proﬁt, and
market share. In order to accomplish these goals, the program must identify
current loyalty program members that are proﬁtable and attempt to increase their
usage levels and/or market share. This same information must also be utilized to
identify new markets and to build these markets using information gathered on
current loyalty program members. In addition, loyalty programs, especially
rewards programs, are a signiﬁcant source of revenues for many businesses. An
example of this is the proﬁts generated by frequent ﬂyer programs often exceed

those generated by actual ticket sales.

Growth of Casino Gaming and Casino Loyalty Programs

Although casinos and gaming have been a part of North American culture for
centuries, the passage of the 1988 Indian Gaming Regulatory Act produced a
rapid expansion of the number of casinos across North America. This act
essentially authorized commercial casino gaming on Native American lands in
approximately 31 states. In part, as a response to this act, the State of Iowa
legalized riverboat casinos on navigable waters for the purpose of stimulating
Iowa tourism. Five additional states- Illinois, Indiana, Louisiana, Mississippi, and
Missouri--had done the same by May 1994. In 1995, licensed casino table or
machine games were being operated or had been authorized in approximately 28
states and provinces (Christianson and Brinkerhoff-Jacobs, 1995). In 2000, only
Utah and Hawaii had no forms of legalized gambling (Morals, 2002). It is
estimated that the revenues generated by the US gambling industry (Casinos,

lotteries, bingo) in 2001 were $64 billion and that the United States will be the

World’s fastest-growing gambling market during the next decade (Morais, 2002).
It is estimated that legal gaming estimates globally exceeded $900 billion in
2001, with revenues accrued to casinos at $270 billion.

The rapid expansion in casinos and gaming has dramatically increased
competition for gamblers, especially heavy gamblers or what are commonly
referred to as “big-ﬁsh”. Casinos and gaming regions (Las Vegas, Reno, Atlantic
City, etc.) initially responded to this competition primarily by augmenting and
expanding their product lines (e.g. larger, more elaborate hotels and casinos)
and launching ever more costly promotions. Casinos also responded to
increasing competition by inaugurating player’s club programs and/or promotions
specially designed and targeted at loyalty club members.

Some form of a player’s club program is now offered by almost every casino.
The combination of promotions and comps aimed at loyalty club members now
account for a signiﬁcant portion of the marketing expenditures of many casinos.
Although casino loyalty rewards programs have their own unique characteristics,
they share common attributes with loyalty programs offered by other industries,
speciﬁcally frequent ﬂyer programs. For example, both casino player’s club
programs and frequent flyer programs are both service(s) oriented, consist of a
bundle of attributes that are customer speciﬁc, and there are low switching costs.
On the other hand, they differ in the sense that casino gaming is seen by some
parts of society as a vice. These similarities and differences must be taken into
account during the evaluation of these programs and is an area of interest that

has yet to been explored.

Recently, a number of casino player’s club programs have been redesigned
in an effort to increase value/beneﬁts to their members (Dowling and Uncles,
1997). As an example, Mystic Lake Casino in Minnesota (depicted in Figure 2)
provides differential and increasing beneﬁts to members based on their level of
play over the previous year. This tiered/segmented approach enables casinos to
reward and retain its most loyal customers through targeted incentives aimed at
increasing play and providing an incentive for members to achieve higher reward
levels. A review of these beneﬁts shows a clear focus on providing members with
rewards, promotions, and preferential service in order to increase both their
attitudinal and behavioral loyalty. An advantage of these tier/segmented
programs is that they take into account the inherent variability of player’s club
members and their subsequent usage behavior, thus rewarding loyal customers
or those customers that maintain a relationship with an organization over a long

period of time (O’Brien and Jones, 1995; Dowling and Uncles, 1997).

Figure 2: Beneﬁts Offered to Members of Harrah’s Total Rewards Tiered
System

Classic Gold Platinum

The BENEFITS of Membership

Rewards & Offers x x X
Restaurant Discounts* 10% 15% 20%
Gift Shop Discounts 5% 10% 15%
Participation in great Promotions x
Invitations to Parties & Events

Special Hotel Rates

Free Valet Service**

Restaurant VIP Line Passes

Dakotah! Sport & Fitness

Membership***

Personalized Host Service

Free Playworks Child Care****

Suite Upgrade

Invitation to Platinum Events

><><><><

><><><><><><><><><><

Source: (Little Six Incorporated, 2003)

While player’s club programs have continued to evolve in response to the
intense level of competition between casinos, it appears that some casinos are
making changes to their programs without a scientiﬁc basis in marketing
research, linkages with strategic priorities, or calculating the return-on-investment
from various reward incentives. Sometimes the criteria for issuing rewards are
subjective and left in the hands of casino ﬂoor managers leading to inconsistent
practices. Poor implementation practices can reduce the return-on-investment
from the loyalty rewards program. Furthermore, due to lack of research
completed on loyalty rewards programs in general, and the fact that casino

loyalty programs have only been developed over the last ﬁve to ten years, the

need to evaluate these programs is necessary for furthering the body of

knowledge on loyalty and loyalty rewards programs.

Previous Empirical Studies of Loyalty Programs

There have been two signiﬁcant empirical studies that have examined the role
and effectiveness of loyalty reward programs and how service organizations are
using these programs as a key part of their overall marketing strategies (Bolton,
Kannan, and Bramlett, 2000; Long and Schiffman, 2000). These two studies
offered new ways for deﬁning customer loyalty. Loyal customers were deﬁned as
those customers who (a) maintained a relationship with an organization over a
period of time and (b) purchased products/services at regular intervals. This lack
of attention on percentage of market share or commitment to organizations
differentiates this categorization of loyalty from that of previous studies
completed in the realms of relationship marketing and brand loyalty, and one that
should be further explored.

Although work completed on loyalty programs and their effectiveness is very
limited, there have been a number of key notable study ﬁndings. First, because
of their high purchase and visitation levels and their longer relationships with
businesses, loyalty rewards program members are comparatively more likely to
have an unsatisfactory product or service experience. The study by Bolton et al.
(2000) found that although members of loyalty rewards programs did in fact
experience a higher rate of negative service experiences, they discounted these
negative evaluations to a greater degree than nonloyalty reward program

members.

Second, since these programs are based on principles of relationship
marketing and value creation, loyalty reward program members can potentially
be segmented based on their personal values and relationships with their
respective loyalty programs. Also, there are multiple programs that a person can
belong to in a speciﬁc industry, so it can be safe to assume that persons often
belong to multiple programs within the same industry. Although the research
completed by Long and Schiffman (2000) found support for these assumptions
for members of airline frequent ﬂyer programs, they also stated that the
(revenue) performance of the loyalty reward program will vary by between and
within segments to the point that the organization may choose not to market to

certain segments since they are unproﬁtable.

Need and Importance of this Study

Although the concept of loyalty has existed for decades, there is no
consensus over what loyalty is or how it should be measured (RundIe-Thiele and
Mackay, 2001). This lack of consensus may explain in part why research on
loyalty remains so fragmented. On this same point, the impact that membership
within casino player’s club programs has on consumer behavior remains an area
that has remained relatively unexplored by marketing academics. Given the
signiﬁcance of the investments in loyalty building efforts and the amount of
research they have attracted recently, the lack of knowledge of whether
relationship marketing and loyalty programs even create “loyal” customers or
how a loyal customer is defined is somewhat of a surprise. Furthermore, there is

a very active debate and growing concerns about what constructs and measures

to use when studying loyalty. While the debate concerning loyalty measures and
constructs is interesting, some believe that the failure to agree on and utilize
consistent measures and constructs has not encouraged a ﬁeld of research that
builds substantially on previous studies. Although there are important exceptions,
there is still a shroud of mystery surrounding the concept of loyalty and how it
applies to consumer behavior

Developing a framework for validating the ﬁnancial performance of these
programs is an important step since the managerial justiﬁcation for these
programs is that they will have a positive inﬂuence on long-term ﬁnancial
performance of the organization (Anderson et al., 1994; Bolton et al., 2000;
Dowling and Uncles, 1997; Fredericks et al., 2001; Long and Schiffman, 2000;
Luxton, 2002; O’Brien and Jones, 1995; Reinartz and Kumar, 2000; Reinartz and
Kumar, 2002). Although financial evaluations of loyalty programs have
undoubtedly been undertaken by different businesses, no empirical studies have
been published that explore the relationship between player’s club programs and
ﬁnancial performance. Through understanding what segments of casino loyalty
programs are proﬁtable, it will allow for casinos to focus their limited resources on
these segments and aid in the development of de-marketing strategies for
unproﬁtable segments.

Both Bolton et al. (2000) and Long and Schiffman (2000) raised the idea of
measuring the performance of loyalty program members. They argued for the
importance of determining the affects of different components of loyalty programs

(direct mail, rewards, customer service, etc.) on creating value or loyalty

11

(attitudinal and behavioral) in customers. Evaluating the different components of
loyalty programs will create an understanding of whether and how player's club
programs build attitudinal and behavioral loyalty in a gaming setting. The
following Integrated Casino Loyalty Model (Figure 3) was formulated in order to
provide a framework with which to address the aforementioned issues and to
build our knowledge on the effects of casino player’s club programs on the
behavior of casino gamers. The components of this model are based on
previous work completed on brand loyalty (Bennett and Rundle—Thiele, 2002;
Chaudhuri and Holbrook, 2001; Rundle-Thiele and Mackay, 2001), relationship
marketing (Bolton et al., 2000; De Wulf, Oderkerken-Shroder, and lacobucci,;
2001), and switching behaviors (Ganesh, Arnold, and Reynolds; 2000). This

literature will be further discussed in Chapter 2.

Figure 3: Integrated Casino Loyalty Model

 

Direct
Mail

 

 

Preferential
Treatment

 

 

       
    
 
  
     
   
 

Interpersonal

Communications Perceived

Value

 

 

Involvement

 

Promotions &
Discounts

 

 

 

 

 

Rewards

 

 

 

 

 

 
   
   
 

 

Competitive
Factors

Afﬁnity J\

Consumer 4/
Conﬁdence

Behavioral
Loyalty

 

 

 

 

 

 

 

 

Relationship

Attitudinal ,
Commitment

Loyalty

 

 

 

 

 

 

 

13

In order to address these issues and to achieve the objectives of this study,
the major variables of interest are behavioral loyalty, attitudinal loyalty, and
perceived value. Behavior loyalty was chosen as the outcome variable for this
model because the goals of casino loyalty programs are to increase market
share and revenues. Although attitudinal loyalty and perceived value are also
important outcome variables as well, if loyalty programs fail to generate revenues
or increase market share, very few organizations would continue to utilize these
programs as part of their marketing efforts. In addition, through the inclusion of
behavioral loyalty, attitudinal loyalty, and perceived value in a single model, it will
allow for an exploration of the relationship between these constructs and to
further explore antecedents of these constructs.

No empirical work has been completed that attempts to examine the affects
that components of a loyalty program have on attitudinal and behavioral loyalty.
In order to examine the impact(s) that speciﬁc elements of casino players club
programs have on creating perceived value and attitudinal and behavioral loyalty,
the major components of player’s club programs are included in the model. From
reviewing casino player’s club programs across the United States, the major
components that were identiﬁed include: direct mail, preferential treatment,
interpersonal communications, rewards, and promotions. Through their inclusion,
we can examine to what extent each component has on creating attitudinal and
behavioral loyalty and to identify if any differences exists among segments
identiﬁed in the player’s club program. A direct result of this work will be the

ability to utilize this information in identifying development and recruitment

14

strategies that can be utilized by casino’s in improving the performance of their
player’s club programs.

The creation of attitudinal loyalty within customers is often quite complex and
although it has been found that only very few customer exhibit high levels of
attitudinal loyalty in traditional retail settings (Chaudhuri and Holbrook, 2001),
long time loyal customers within loyalty programs should experience somewhat
higher levels of attitudinal loyalty since they have been the focus of a relationship
marketing strategy for many years. As persons feel more afﬁliation towards an
organization and their conﬁdence increases that their expectations will be met on
a continually basis, this should ultimately lead to higher levels of attitudinal
loyalty. Through further examining the impact that consumer conﬁdence and
affiliation have on creating attitudinal loyalty (as opposed to perceived value), it
will allow for the deterrninination of whether player’s club programs are effective
tools for creating behavioral loyalty. In addition, it will also provide an opportunity
to further explore how segments within player’s club programs differ in terms of
their levels of trust and afﬁliation to their player’s club program and how this
impacts the creation of behavioral loyalty.

Although the ultimate purpose of these programs is to reward customers and
to encourage their continuing relationship with a ﬁrm, often these programs have
become defensive marketing strategies. For example, casinos are continually
developing new promotions, rewards, and direct mail initiatives primarily as part
of a competitive effort. They continually add beneﬁts and offer special promotions

aimed at loyalty program members in an effort to keep pace with or gain an

advantage over their competitors. Due to the fact that consumers have access to
multiple programs and there are low barriers to switching, understanding how
choice and competition affects the creation and maintenance of attitudinal and
behavioral loyalty within these programs will provide an interesting insight into
the potentially mediating role that competition plays on value and loyalty creation.

Dick and Basu (1994) contend that persons who experience higher levels of
involvement with products should experience higher levels of loyalty. A common
practice of casinos in recruiting new members for their player’s club is by
restricting access to discounts, special events or promotions to members only. A
direct result of these practices is that low involved gamers are being recruited
into the program and may represent a signiﬁcant portion of the program. This is
an important element since past studies have shown that for those persons who
are not interested in the product category, lack of personal attention and
relationship investment will not be as critical to them in their decision to continue
purchasing a speciﬁc product or service (Solomon et al., 1985). Thus,
understanding how level of involvement among casino player’s club members
moderates the development of behavioral loyalty is important and will aid in the
identiﬁcation of retention and development strategies.

The idea that all consumers want to have a relationship with ﬁrms has not
been supported in past studies (Bendapudi and Berry, 1997; Christy, Oliver, and
Penn, 1996). Furthermore, firms attempting to form relationships with customers
who do not want relationships can actually be detrimental to the customer-ﬁrm

relationship (Dowling and Uncles, 1997). Thus, persons who are more committed

16

to maintaining a relationship with an organization should experience higher levels
of behavioral loyalty since they wish to maintain a relationship with a ﬁrm and are
more receptive to loyalty based marketing initiatives. Through understanding the
level of commitment for different segments within the player’s club program, it will
allow us to identify if any differences exist and how these differences impact the
creation of behavioral loyalty.

As stated earlier, very few studies have been able to integrate both attitudinal
and behavioral loyalty. Through our access to purchase history data, it provides
an opportunity to further explore how these two constructs are related.
Furthermore, it will also allow us to deﬁne loyalty program members using data
collected from the loyalty program database and self-reported measures, and
determine if any differences exist between these two methods. Additionally, it will
also allow for the further reﬁnement of measures and antecedents of attitudinal

loyalty, behavioral loyalty, and perceived value.

Study Purposes and Objectives

The primary purpose of this study is to determine whether and how various
loyalty/players club member segments differ in their attitude and behavior toward
a casino as a result of their involvement in a loyalty program. Loyal Players Club
Members — persons who have been members for four years of the client casino’s
player’s club program and have gamed at a relatively high and consistent rate at
the client casino - will be compared to and contrasted with members of the client
casinos player’s club program and who have displayed high usage levels at the

client casino but who have signiﬁcantly decreased or stopped their visits to the

17

casino within the previous 12 months. This section segment is deﬁned as
Disloyal Players Club Members. This process will provide crucial insights for
designing a more effective player’s club including customer acquisition and
retention strategies.

A second purpose of this study is to provide an estimate of the return on
investment the casino realized from its player’s club program. Hopefully, this will
assist in creating a better understanding of the value of these programs and also
provide suggestions for enhancing this return.

Studying a speciﬁc player’s club provides a unique opportunity to empirically
examine loyalty rewards programs in a noncontractual setting and in a highly
competitive environment - the setting where the majority of these programs are
currently being utilized. It will also allow for an in-depth analysis of the return on
investments from operating loyalty rewards programs in a gaming environment.
Furthermore, it will provide data and tools to a growing industry sector that has
received relatively little academic attention.

The following objectives were developed in guide the design and conduct of

this study:

18

1. To determine the revenues of player’s club program members based on
their lifetime visits during their membership in the program

The purpose of this objective is to determine the directionality and strength of
the lifetime visits/ lifetime revenues relationship for player’s club members of the
client casino. Data (e.g. amount gamed, number of visits) collected on members
of the client casino’s player’s club program from January 1999 to December 2002
will be used to assess and compare the purchase behavior (e.g., volume of
gaming) of player’s club members.

2. To formulate, proﬁle and assess various segments within the client’s

player's club program

The purpose here is to identify substantial and exploitable segments within
the client casino’s players club. Socioeconomic characteristics (age, gender,
geographic location) and purchase behavior (frequency of visits, net revenue) of
segment members (High frequency / High revenue, High frequency / Low
revenue, Low frequency / High revenue, and Low frequency / Low revenue) will
be analyzed. Additional profiling may identify important marketing relevant sub-
segments. For example, long-term loyalty reward program members who have
not visited the casino lately or long-term loyalty reward program members whose
purchase behavior has decreased signiﬁcantly over their duration of their
membership. Data collected on members of a casino loyalty rewards program
from January 1999 to December 2002 will be used along with data collected

through the mail surveys to develop these proﬁles.

3. Compare and contrast player’s club segments identiﬁed by casino
loyalty relevant segmentations bases (e.g. annual gaming budget,
amount of total visits taken to primary casino) and determine the
effectiveness of each segmentation base in identifying relevant and
exploitable market segments

The purpose of this objective is to identify any differences that exist between

market segments in order to evaluate the effectiveness of various loyalty relevant
segmentation bases. Through the testing of multiple segmentation bases, it will
provide more information on various segmentation methods and their use within

loyalty program settings. Data collected through surveys sent to 3000 loyalty

program members will be used to complete this objective.

4. Develop and test an integrated casino loyalty model to examine
attitudes and behaviors of player’s club members segments in order to
determine the various effects of loyalty program memberships within a
gaming setting

The behavioral and attitudinal characteristics of player’s club members will be

identiﬁed to assess possible methods for creating and increasing attitudinal and
behavioral loyalty. This will allow for the further development of indicators and

measures of both attitudinal and behavioral loyalty. Data collected through

surveys sent to 3000 loyalty program members will be tested using a structural

equation model.

20

5. To compare, contrast and proﬁle segments derived using annual casino
gaming budgets, proportions of trips to primary casinos, and the
amount budgeted per casino trip

The purpose of this objective is to identify segments that exist within a

player’s club program and evaluate the effectiveness of loyalty programs in
serving these segments. Through the identiﬁcation of these segments, it will
allow gaming organizations to identify both high revenue and low revenue
segments. Data collected through surveys sent to 3000 loyalty program members

will be used to complete this objective. When appropriate, Independent sample t-

tests will be utilized to compare these segments.

Organization of the Dissertation

The remainder of the dissertation will be organized in the following manner.
Chapter 2 will provide an overview of relevant loyalty and loyalty program
literature. Chapter 3 obtains a discussion of the methods used in conducting this
study and the sources of data. Chapter 4 will present the results of this study.

Chapter 5 will discuss these results and provide suggestions for future studies.

21

CHAPTER 2
LITERATURE REVIEW

Introduction

This chapter examines the literature that provides a conceptual and
theoretical framework for this study. It is divided into ﬁve sections and provides a
review of the major areas of research related to customer loyalty. The ﬁrst
section focuses on customer loyalty and the various techniques that are utilized
to study loyalty. The second section will focus on the theoretical foundations that
this study has been built on and how it will add to the body of knowledge on
customer loyalty. The third section reviews past studies of loyalty rewards
programs and how this study builds on them by exploring customer loyalty
rewards programs within the gaming industry. In the fourth section the focus is
on integrated loyalty research and how its holistic approach to understanding
customer loyalty will be used in the formation of the framework for this study. The
emphasis of the ﬁfth section is model-speciﬁc literature that will be integrated and

tested in this study.

Deﬁning Customer Loyalty

The past decade has seen many ﬁrms adopting a customer focus — often
through the adoption of formal customer relationship management and/or
relationship marketing strategies. Although these initiatives were developed with
the hope of attracting new customers, often these programs attracted the ‘wrong’

type of customers (Lovelock and Wright, 2002).

22

Before understanding loyalty, the focus must ﬁrst be on the process by which
consumers and ﬁrms interact. The basis of this process is a transaction - or even
a series of transactions. Unfortunately, this process may not represent a
relationship unless there is mutual recognition and knowledge between the
parties (Lovelock and Wright, 2002). Although this process seems somewhat
straightforward, deﬁning what a relationship is and loyalty still remains quite
ambiguous.

Jacoby and Chestnut (1978) provide a review of 53 operational deﬁnitions of
loyalty that have been utilized in past research. Three main types of customer
loyalty are deﬁned: (1) loyalty as repurchase behavior, (2) loyalty as a
psychological commitment (attitude) towards a product or company and (3)
loyalty as both of the above. Previous research focusing on loyalty as repurchase
behavior uses the construct of behavioral loyalty. Frequently, behavioral loyalty is
deﬁned as the proportion of times a buyer purchases the same product or service
in a speciﬁc product or service category as compared to that of the total
purchases in that category (Neal, 2000). Attitudinal loyalty is often deﬁned as
being absolutely dedicated to a brand or product (Neal, 2000). Attitudinal loyalty
as an attitudinal commitment has often been measured using the attitudinal
loyalty construct. Although these different loyalty constructs have been studied
extensively separately, seldom have they been analyzed in a single study. One
reason for this is that studies have found that only a very few customers exhibit

attitudinal loyalty (Bolton et al., 2000; Neal, 2000).

23

The inability of ﬁrms to create attitudinal loyalty has not gone unnoticed. Neal
(2000) stated that the majority of customers exhibit convenience loyalty - or in
other words, continue to buy the same brand that they usually buy until it fails to
meet their basic needs. Although these persons exhibit high levels of behavioral
loyalty, they discontinue the use of the product/service if they either have a bad
service experience or have reason and/or incentives to reevaluate alternative
products and services. By providing “extra value” to customers, businesses are
attempting to provide incentives for customers to maintain their relationship with
them even when they experience an occasional service failure. The added value
is meant to act as a barrier to switching services. Although the work by Bolton et
al. (2000) supports the idea that loyalty program members will discount negative
service experiences to a greater degree than non-members, the lack of
knowledge about why customers who exhibited high levels of loyalty left or quit
using speciﬁc product/services is not well understood.

There are four primary service marketing-related customer loyalty paradigms:
(1) Integrated loyalty research, (2) Service quality research, (3) Customer lifetime
value research, and (4) Loyalty rewards program research. Empirical and
theoretical research related to each of the four paradigms is summarized in Table

1. Selected works from each paradigm are reviewed in subsequent sections.

24

Table 1: Selected Literature Completed on Customer Loyalty

 

 

 

 

 

 

lnman, Dyer, & Jia
(1997)

THEORETICAL
Satisfaction / Service Customer Lifetime Integrated Loyalty Loyalty Programs
Quality Value
Satisfaction is Not Customer Lifetime Connecting Do Customer Loyalty
Enough Valuation to Support Customer Loyalty Programs Really
Gale (1997) Marketing Decision with Financial Work
Making Results Dowling & Uncles
Dwyer (1997) Fredericks et al. (1997)
(2001)
Learning from E-Loyalty: Your Customer Loyalty: Behavior Learning
Customer Defections Secret Weapon on Toward an Theory: Its
Reichheld (1996) the Web Integrated Relevance to
Reichheld & Schefter Conceptual Marketing and
(2000) Framework Promotions
Dick & Basu (1994) Rothschild & Gaidis
£981)
Realize Your A Methodology for Marketing Campaign Do Rewards Really
Customers' Full Linking Customer Systems — The Create Loyalty?
Proﬁt Potential Acquisition to Secret to Life-Long O’Brien and Jones
Grant & Schlesinger Customer Retention Customer Loyalty (1995)
(1995) Thomas (2001) Luxton (2001
A Generalized Utility Customer Lifetime Customer Referral Do Your Customers
Model of Value Research: A Management: Need a Loyalty
Disappointment and Review and Future Optimal Reward Scheme?
Regret Effects on Directions Programs Uncles (1994)
Post-Choice Jain & Singh (2002) Biyalogorsky et al.
Valuation (2001)

 

A Dynamic Process
Model of Service

Customer Lifetime
Value: Market

Satisfaction is Nice:
But Value Drives

Do Loyalty Clubs
Really Make Sense?

 

Quality: From Models and Loyalty Partch (1998)
Expectations to Applications Neal (1999)

Behavioral Intentions Berger & Nasr

Boulding, Kalra, & (1998)

Staelln (1997)

Reflections on Customer Starwood, United
Gaining Competitive Proﬁtability Analysis: Woo E-Travelers
Advantage through Measurement, Beime (2000)

Customer Value
Parasuraman (1997)

Concentration, and
Research Directions
Mulhem (1999)

 

 

 

The Impact of
Technology on the
Quality - Value —
Loyalty Chain: A
Research Agenda
Parasuraman &
Grewal (2000)

 

How to Calculate the
Value of a Customer:
Measuring Customer
Satisfaction

Gurau 8. Ranchhod
(2002)

 

 

 

25

 

Table 1: Selected Literature Completed on Customer Loyalty (cont)

 

THEORETICAL (cont.‘l

 

Satisfaction / Service
Quality

Customer Lifetime
Value

Integrated Loyalty

Loyalty Programs

 

Manage Marketing
by the Customer
Equity Test
Blattberg & Deighton
(1996)

 

Modeling Customer
Relationships as
Markov Chains.
Pfeifer and Carraway
(2000)

 

EMPIRICAL

 

Understanding the
Customer Base of
Service Providers:
An Examination of

On the Proﬁtability of
Long-Life Customers
in a Non-contractual

Setting: An Empirical

Patterns of Buyer
Behavior:
Regularities, Models
and Extensions

How Hilton &
Sheraton Increased
Their Business with
Frequent Buyer

 

the Difference Investigation and Uncles, Ehrenberg Programs
between Switchers Implications for and Hammond Anonymous (1995)
and Stayers Marketing (1995)
Ganesh et al. (2000) Reinartz & Kumar

(2000)
A Dynamic Model of Customer Base Consumers say “No Customer Speciﬁc
the Duration of the Analysis: An Thanks” to Marketing
Customer’s Industrial Purchase Relationships with Raphel (1996)

Relationship with a

Process Application

Brands

 

Continuous Service Schmittlein and Anonymous (2001)

Provider: The Role Peterson (1994)

of Satisfaction

Bolton (1998)

The Loyalty Effect The Chain of Effects Implications of

Reichheld & Teal
(1996)

from Brand Trust
and Brand Affect to
Brand Performance:

Loyalty Program
Membership and
Service Experiences

 

 

 

 

 

 

The Role of Brand for Customer
Loyalty Retention and Value
Chaudhuri & Bolton et al. (2000)
Holbrook (2001)
Effects of Why it is Customer Consumption Values
Relationship Loyalty that Counts and Relationships:
Marketing on (and How to Segmenting the
Satisfaction, Measure it) Market for
Retention, and Gould (1995) Frequency Programs
Prices in the Long & Schiffman
Insurance Industry (2000)
Crosby & Stephens
(1987)

 

26

 

Table 1: Selected Literature Completed on Customer Loyalty (cont)

 

 

 

EMPIRICAL (cont.)
Satisfaction I Service Customer Lifetime Integrated Loyalty Loyalty Programs
Quality Value

 

Why Improving
Quality Doesn’t

Disloyalty Becomes
the Norm

Loyalty Programs
and their Impact on

 

 

Proﬁtability: Findings

Brand Performance:

Improve Quality (or Hiscock (2001) Repeat — Purchase
Whatever Happened Loyalty Patterns
to Marketing?) Sharp and Sharp
Kordupleski, Rust, & (1997)

Zahoric (1993)

Indirect Financial Breaking Out

Benefits from Wood (1998)

Service Quality

Danahar and Rust

(1996)

Customer The Chain of Effects

Satisfaction, Market from Brand Trust

Share, and and Brand Affect to

 

Customer Retention,
and Market Share

from Sweden The Role of Brand

Anderson et al. Loyalty

(1994) (Chaudhiru &
Holbrook (2001)

Customer Investments in

Satisfaction, Consumer

Relationships: A
Cross Country and

 

Inﬂuence Customer
Loyalty

Butt/e & Burton
(2001)

Rust and Zahoric Cross Industry
(1993) Exploration

De Wulf et al. (2001)
Does Service Failure Diagnosing

Customer Loyalty
Drivers: Partial Least
Squares vs.
Regression

Ryan et al. (1999)

 

Measuring and
Managing the
Satisfaction-Loyalty-

Assessing the
Performance of
Brand Loyalty

 

Satisfaction with
Services: Putting
Perceived Value into
the Equation
McDougall and
Levesque (2000)

 

 

 

Performance Links Measures

at Volvo Rundle-Thiele and
Gustafsson and Mackay (2001)
Johnson (2002)

Customer A Comparison of

Attitudinal Loyalty
Measurement
Approaches

Bennett and Rundle-
Thlele (2002)

 

 

27

 

Satisfaction l Service Quality Research

One of the most studied areas of customer loyalty (refer to Table 1) involves
customer satisfaction and the link between satisfaction and loyalty formation
(Bolton, 1998). While satisfaction (and perceived value) has been identified as an
important ingredient in gaining competitive advantage and an important indicator
of repurchase intentions (Parasuraman, 1997; Parasuraman and Grewal, 2000),
very little research has been able to support how satisfaction and service quality
lead to increased cost savings and/or increased revenues (Gustafsson and
Johnson, 2002). Few studies have been conducted that validate the link between
customer satisfaction and increased customer loyalty (Bolton, 1998). Also, the
conﬂicting nature of how these constructs have been operationalized makes it
unclear as to how satisfaction and service quality relate to building and retaining
customer loyalty.

A number of researchers have suggested that there is a signiﬁcant difference
between I satisfaction and loyalty. While satisfaction has been largely
operationalized as a passive condition (Fredericks, 2001), loyalty has often been

considered “199 an active or even proactive relationship with the supplier

,;

 

(Bolton, 1998). Additionally, a number of researchers contend that focusing too
much on satisfaction and customer loyalty (or duration of relationship with a
service provider) may have actually hampered a more complete understanding of
customer retention and lifetime purchases (Gale, 1997; Reichheld, 1996).

The majority of customer service and service quality studies focus on the '

measurement of purchase intentions as opposed to actual behavior. They must

28

be viewed with caution since the correlation between satisfaction and intention
measures may be a result of measurement scales, time frame, and nature of the

respondents (Monrvitz and Schmittlein, 1992; Morwitz, 1997).

Customer Lifetime Value Research

Because many companies are taking a more customer-centric approach to
marketing as opposed to a product-centric approach, there is a greater need to
understand and evaluate their relationships with customers (Jain and Singh,
2002). Customer Lifetime Value (CLV) modeling is one method for studying and
measuring the value of these relationships.

Customer Lifetime Value models have been developed for different industries
and different sized customer bases. Companies with smaller customer bases can
utilize these models to calculate the lifetime value of speciﬁc customers in order
to determine their value. At the same time, companies with a large customer
bases can utilize CLV models in order to segment their customers on the basis of
customer lifetime value (Jain and Singh, 2002; Reinartz and Kumar, 2000).

The development of CLV analytical models was stimulated by the recognition
that not all loyal customers are proﬁtable, and therefore not all customers should
be retained (Blattberg and Thomas, 1997; Blattberg and Thomas, 2000; Jain and
Singh, 2002; Reinartz and Kumar, 2000, Reinzartz and Kumar, 2002). An
example is the Dynamic Pricing Model proposed by Blattberg and Thomas
(1997) that provides a framework for managers to maximize CLV. In this model,
customer acquisition is deﬁned as a customer’s ﬁrst purchase, whereas a

retained customer is active in every time interval with no lapse periods. Because

29

of the model’s ability to forecast revenues and costs derived from both current
and future customers, it can be used to calculate revenue and costs for different
market segments.

While this model shows promise, the computational and data requirements
make its use extremely difﬁcult. Furthermore, the model is not appropriate in
situations where customers have different purchase frequencies (which often
occur in non-contractual settings) and for periods of less than four years
(Blattberg and Thomas, 1997). Also, it is quite complicated to calculate
probabilities for individual customers when large customer bases are involved
(Jain and Singh, 2002).

Customer base analysis involves the development of models to predict
whether customers will be active or inactive purchasers, and also to estimate the
length of a customer’s relationship with a ﬁrm. One of the most recognized and
frequently cited models is the Pareto/NBD Model proposed by Schmittlein,
Morrison, and Colombo (1987). Unlike other models, this research has shown
promise for the study of noncontractual service settings (Reinartz and Kumar,
2000). However, it relies on a number of assumptions that cannot be met in
regards to this study.

The ﬁrst of these assumptions is that every customer transaction must be
recorded (Reinartz and Kumar, 2000; Schmittlein, Morrison, and Colombo,
1987). Although a proportion of customers may utilize their player’s club card all
of the time or frequently, there may be times when the cards are not used, and,

as a result, some transactions may go unrecorded. A second assumption is that

30

the date of the ﬁrst transaction with the ﬁrm is known (Reinartz and Kumar,
2000). In the case of casinos, data is only recorded on loyalty rewards program
members after they join the program. Previous gaming at the casino is not
tracked, which may pose a problem since one cannot distinguish ﬁrst-time
customers and those customers who have visited previously but are only joining
the program after they have been customers of the casino for a period of time.
Finally, the model requires a minimum of 3 years of information on the revenues
and costs for each customer (Reinartz and Kumar, 2000; Schmittlein, Morrison,
and Colombo, 1987). Unfortunately, the client casino only tracks player’s club
members and not the expenses associated with servicing and marketing to
members.

The majority of CLV models suffer from the same weaknesses. The ﬁrst
weakness is that all models assume that the revenues for each person are
constant, thus failing to take into account any increases or decreases that are
common in non-contractual settings (Jain and Singh, 2002; Reinartz and Kumar,
2000). The second weakness is that only time periods of equal length can be
utilized. Since both revenues and time between purchases tend not to be
constant outside of a contractual setting (banking, cellular phone agreements),
their use in non-contractual service settings challenges the applicability of their
use in regards to this study (Reinartz and Kumar, 2000). In addition, the effect of
loyalty reward program membership on these models is unknown and requires
further exploration (Jain and Singh, 2000; Reinartz and Kumar, 2000). Although

these models may not be applicable to this study, they should be helpful in

31

evaluating the cost-beneﬁt ratio of the loyalty program and further understanding

the lifetime purchase behavior of loyalty program members.

Loyalty Program Research

Although these programs were initiated in the airline industry over 20 years
ago, the beneﬁts derived through the operation of these programs in terms of
building customer loyalty have yet to be answered. The number of studies of
loyalty programs is small compared to other types of customer loyalty studies
(Table 1). While there is a balance between theoretical and empirical studies on
loyalty programs, the majority of the empirical work has focused on credit card
programs and airline frequent ﬂyer programs.

As is the case with all loyalty based marketing strategies, the theory of
reciprocity plays a key role in understanding their use and customer subsequent
behavior. This theory postulates that people will feel obligated to exchange
objects (i.e. money, products, and services) of similar value to that of which they
received from the other transaction partner (De Wulf, et al., 2001). Both parties
will feel need to maintain the relationship in order to reduce the level of guilt
between the partners and/or to maintain the relationship (Bagozzi, 1995). Loyalty
programs attempt to provide value to the consumer by rewarding them for their
continued usage of a service provider and customers will subsequently reward
service providers through continued usage of their products/services. In addition,
these programs also provide a method of organizations of introducing switching

costs in a less evasive manner.

32

Table 2 lists past theoretical and empirical work related to loyalty programs.
Upon reviewing the theoretical work completed on loyalty programs, the titles of
the studies suggest a direction of research that questions the value of loyalty
programs. For example, Dowling and Uncles (1997) paper is titled “Do Customer
Loyalty Programs Really World. O’Brien and Jones (1995) paper is titled “Do
Rewards Really Create Loyalty’. Partch (1998) paper is titled “Do Loyalty Clubs
Really Make Sense”. The titles of these articles provide some sense about the

direction of this area of research.

33

Table 2: Summary of Research Completed on Loyalty Programs

 

 

 

 

 

 

 

 

THEORETICAL
Title I Author Year Industry Results Limitations l Future
Research
Behavior 1981 . Consumer . Appropriate long- . Need for empirical
Learning Theory: Packaged term behavior only validation
Its Relevance to Goods occurs when the o The marketing
Marketing and reinforcer meets environment is not
Promotions some need controlled, such
that not all
ROIhChi/d & behavioral notions
Gaidis can be transferred.
. Determining the
balance between
organizational goals
and the consumer.
0 Identify the role of
promotions
Do Rewards 1995 . Credit Cards 0 For the most . Need for empirical
Really Create . Telephone proﬁtable validation
Loyalty? Companies customers, higher 0 Study of Tier
. Airlines rewards offered. system based on
O’Brien and 0 Firms should utilize customer value.
Jones partnerships in
order to establish
sustainable loyalty
Do Your 1995 0 Credit Cards 0 Programs need to . Need for
Customers Need 0 Airlines be based on goals development of
a Loyalty of organization strategic orientation
Scheme? of loyalty programs
Uncles
Do Customer 1997 . Credit Cards . Best reward . Need for a focus on
Loyalty 0 Airlines programs are based the return on
Programs Really on tiers investment from
Work 0 Design programs to programs
take into account . Empirically validate
Dowling and proﬁtability return on
Uncles differences between investment from
customers different types of
loyal customers
Do Loyalty Clubs 1998 - Supermarkets - Loyalty programs . Need to determine
Really Make often backﬁre if impact of members
Sense? promotions are not only promotions on
supported non-members
Partch

 

 

 

 

 

34

 

Table 2: Summary of Research Completed on Loyalty Programs (cont)

 

 

 

 

 

 

 

EMPIRICAL
Title I Author Year Industry Results Limitations l Future
Research
How Hilton & 1995 . Hotels . Focus Loyalty 0 Need to determine
Sheraton Rewards on Top why majority of
Increased Their 15% of Program members who
Business with 0 Effective programs responded from
Frequent Buyer will use Tier promotions came
Programs Systems from only two
countries
Anonymous
Customer 1996 . Supermarkets . Focus on Top 20% 0 Use of focus
Speciﬁc of Program groups to
Marketing 0 Spend less money determine other
on advertising and elements of
Raphel more on promotions program
directed at try tiers
Implications of 2000 . Credit Card . Members of a . Need to understand
Loyalty Program loyalty program will what features of a
Membership and discount negative loyalty reward
Service evaluations program modify
Experiences for members behavior
Customer . Need to study the
Retention and ﬁnancial costs
Value versus revenues
Bolton et al.
Consumption 2000 . Airlines . Speciﬁc . Analysis of Lifetime
Values and communication Value of loyalty
Relationships: strategies must be program members
Segmenting the utilized to target is required.
Market for different segments . Identify promotions
Frequency within the club that will modify
Programs behavior of
members
Long &
Schiffman

 

 

 

 

 

35

 

The work completed by Dowling and Uncles (1997) acted as a catalyst for re-
examining the proﬁtability of loyal customers and loyalty rewards programs
(Fredericks et al., 2001; Reinartz and Kumar, 2000). According to Dowling and
Uncles (1997), the underlying assumptions that loyalty programs are built on
include: (1) Many customers want an involving relationships with the brand they
buy, (2) A proportion of buyers only purchase one brand, (3) Loyal customers are
a proﬁtable group since they represent a large part of the market and are heavy
or frequent buyers, (4) Firms have the ability to reinforce customer’s loyalty
levels, and (5) With database technologies, marketers can establish personalized
dialogue with customers. Although these are the underpinnings for many loyalty
programs, with the exception of Reichheld and Teal (1996), little research has

been completed to show that any of these assumptions hold true.

Customers Want Relationships with Companies

Very little research has verified that consumers want a relationship with the
brand they purchase, especially for lower involvement items (Dowling and
Uncles, 1997). Although a primary intent of loyalty reward programs is to build
relationships with customers by requiring that they belong to loyalty programs in
order to be eligible for certain promotions and perks, often customers who do not
want an on-going relationship are members. This incongruence could potentially
be damaging and may lead to the consumer terminating his/her relationship with
the ﬁrm (Dowling and Uncles, 1997).

Very little research has estimated to what extent (if any) loyalty programs

actually aid in the formation of attitudinal and behavioral loyalty (Bolton et al.,

36

2000). By not understanding the loyalty creation process involved through the
use of player’s club programs by casinos, they lack the information needed to
modify their respective programs and improve their marketing practices and
strategies. Due to the fact that the majority of these programs were implemented
in response to competitive offerings, as opposed to being part of an overall
marketing strategy, evaluating these programs will allow ﬁrms to determine the
value of operating these programs as opposed to alternative marketing
approaches (i.e. price cuts, increased advertising, improving facilities, etc.)

(Dowling and Uncles, 1997).

Loyal Customers are Proﬁtable

Only two empirical studies have been conducted that test the assumption that
loyal customers are proﬁtable. Their ﬁndings suggest only a weak support for this
assumed relationship (Reichheld and Teal, 1996; Reinartz and Kumar, 2000). No
published empirical research has examined the return on investment derived
from loyalty programs. Although one must assume that applied research has
been completed by businesses that have loyalty programs, the results of these
studies are not available publicly.

There is also a question concerning the assumption that loyal customers are
profitable because they are a large part of the market and are heavy/frequent
buyers. The work completed by Long and Schiffman (2000) shows that the
majority of loyalty rewards program members belong to multiple clubs. In highly
competitive environments where firms have similar offerings and switching costs

are low, one can assume that customers will belong to multiple clubs and/or have

37

limited loyalty to any particular business even though they are loyalty club

members (Bolton et al., 2000; Dowling and Uncles, 1997).

Identiﬁcation of Loyal Segments

According to Fredericks et al. (2001) and Reinartz and Kumar (2002) , firms
need to understand revenue and costs performance based on different segments
in order to develop initiatives that will better serve each targeted segment in a
cost-effective manner. In addition, they should also identify and stop servicing
unproﬁtable segments (Raphel, 1996). Although these unproﬁtable segments can
potentially be transformed into proﬁtable customers, it is important to understand
the required investment and time required to accomplish the transformation
(O’Brien and Jones, 1995; Partch, 1995). Through the identiﬁcation of these
strategies and related costs, managers will be able to develop strategies based
on data as apposed to anecdotal information or shotgun marketing approaches.
As shown in Figure 4, only two of the four segments that Reinartz and Kumar
(2002) identiﬁed are proﬁtable segments and should be the focus of any

marketing effort.

38

Figure 4: Loyalty Segments Identiﬁed in Analysis of a Mail-Order Catalog

 

 

Company
BUTTERFLIES TRUE FRIENDS
. Good ﬁt between . Good ﬁt between
company’s offerings company’s offerings
HIGH and customer’s needs and customer’s needs
PROFITABILITY . High proﬁt potential . Highest proﬁt potential
STRANGERS BARNACLES
0 Little ﬁt between . Limited ﬁt between
company’s offerings company’s offerings
LOW and customers needs and customers needs
PROFITABILITY 0 Lowest proﬁt potential . Low proﬁt potential

 

 

 

SHORT TERM CUSTOMERS LONG TERM CUSTOMERS

Source: (Reinartz and Kumar, 2002)

Integrated Loyalty Research

The literature reviewed in this section provides a better understanding of

methods by which ‘loyal’ and ‘disloyal’ customers have been deﬁned in previous

studies. The focus of four works will provide the basis for the development of this

study’s Integrated Casino Loyalty Model. The three studies include: Bennett and

Rundle-Thiele (2001), Chaudhuri and Holbrook (2001), and De Wulf, Odekerken-

Shroder, and labcobucci (2001).

39

 

The Chaudhuri and Holbrook (2001) study focused on the relationship
between the concepts of brand loyalty (both purchase loyalty and attitudinal
loyalty) and ﬁrm level outcomes (market share and relative price). More
speciﬁcally, the focus is on the chain of effects that brand trust and brand equity
have on brand performance. Examining this chain of effects allows for a detailed
analysis of how brand trust (ability of an organization to continually meets its
obligations) and brand affect (brand’s potential to elicit a positive emotional
response) affects brand performance (market share and relative price). By
testing a path model, they found a positive correlation between both brand trust
and brand equity on brand performance. They also determined that brand trust
and brand equity are two distinct constructs and that they are indicators for both
attitudinal and brand loyalty. Furthermore, Chaudhuri and Holbrook (2001) also
found that purchase and attitudinal loyalty were positively related to market share
and relative price. Although the results of this study show the importance of trust
and affect when studying loyalty, the testing of this model in a service setting
may show divergent results that require further exploration (Chaudhuri and
Holbrook, 2001).

Rundle-Thiele and MacKay (2001) examined the role of attitudinal loyalty by
testing two distinct measures of attitudinal loyalty and determining whether they
were two distinct constructs or measured the same phenomenon. The two
measures tested were an individual’s propensity to be brand loyal, and attitude
towards the acts of purchasing a speciﬁc brand. Through testing these

measures, marketing researchers will be able to utilize this information in order to

40

more accurately measure attitudinal loyalty within business service settings. Due
to the inconsistency in measuring attitudinal loyalty in previous studies, this work
provides a strong test for attitudinal measures and for their use in studying
behavioral and brand loyalty. They found no signiﬁcant relationship between the
two measures and concluded that they are distinct constructs. Although not
directly tested, their research supports the fact that multiple measures and
indicators must be utilized when studying attitudinal and behavioral loyalty.

De Wulf, Oderkerken-Shroder, and Lacobucci (2001) investigated the role of
perceived relationship investment and its role in developing behavioral loyalty in
a relationship marketing setting. They proposed that as persons perceive higher
levels of investment in customers on the part of businesses, consumers will
exhibit higher levels of behavioral loyalty. As part of their study, they tested the
impact of direct mail, preferential treatment, interpersonal communications, and
tangible rewards on relationship quality. They found that interpersonal
communications had the strongest effect on perceived relationship investment
across all three countries, while the other elements had mixed results.
Furthermore, they also found a positive relationship between perceived
relationship investment and behavioral loyalty. Although the setting for this study
was in retail environments, the elements included in this study are similar to
those within a gaming environment. The phenomenon of reciprocity has been
found to be present in a number of consumer-business relationships, but
according to Bagozzi (1995), further research is needed to determine its role and

function in every day consumer exchanges.

41

Deﬁnitions and Determinants of Loyalty
The measures and indicators that are utilized in this study were

predominantly formulated and tested in research focusing on brand loyalty.

M“.
"Won-u.

For the purpose of this study, behavioral loyalty is deﬁned as the probability
that the person intends to maintain a continuing relationship with their primary
casino. The indicators used to measure behavior loyalty include: “I will visit the
casino on my next visit” and “I intend to remain a customer of the casino.” This
deﬁnition and similar indicators have been utilized in previous studies of
behavioral loyalty in multiple settings (Chadhuri and Holbrook, 2001;
Cunningham, 1957; Jacoby and Chestnut, 1978). Although elements of player’s
club programs are aimed at developing aftitudinwaiﬂloyalty as well, the goal of
loyalty programs is the creation of behavioral loyalty. In addition, the three
studies primarily used in developing the framework of this study identiﬁed
behavioral loyalty as a major variable of interest in their studies (Bennett and
RundIe-Thiele, 2001; Chaudhuri and Holbrook, 2001; De Wulf, Odekerken-
Shroder and labcobucci ,2001).
purchase intentions and operationalized attitudinal loyalty by using both
personality trait and brand speciﬁc measures (Mellens et al., 1996). Studies by
Chaudhuri and Holbrook (2001) and Bennett and Rundle-Thiele (2002) have
found divergent results with the use of these measures, but both studies
recommend the need to further study the relationship between behavioral and

attitudinal loyalty and their antecedents. In this study, attitudinal loyalty is deﬁned

42

using the following three measures: “I am committed to this casino,” “I would be
willing to still visit this casino if it was more difﬁcult to reach,” and “I tell others
about the beneﬁts of joining the players club.” These measures have been
utilized in previous brand loyalty studies with success (Chaudhuri and Holbrook,
2001; Morgan and Hunt, 1994). These measures are modiﬁed for use in a
gaming setting.

The creation of attitudinal loyalty within customers is often quite complex and,
although it has been found that only very few customer exhibit high levels of
attitudinal loyalty in traditional retail settings (Chaudhuri and Holbrook, 2001),
long time loyal customers within loyalty programs should experience somewhat
higher levels of attitudinal loyalty since they have been the focus of a relationship
marketing strategy for many years (Bolton et al., 2000). Also, because loyalty
programs are designed to build relationships with customers, past studies of
attitudinal loyalty in similar settings have identiﬁed trust and affect as
antecedents to attitudinal loyalty with some success (Chaudhuri and Holbrook,
2001).

According to Doney and Cannon (1997), trust is based on the belief that an
organization can continue to meet the expectations and obligations of
consumers. Due to the fact that persons within the gaming environment are not
actually purchasing a product or service but rather opportunities to realize a high
return on investment, the concept of trust is important to player’s club members.
If player’s club members do not have conﬁdence that the casino will provide a

safe and fair gaming environment, no loyalty program, regardless of the perks

43

and incentives, will create attitudinal and behavioral loyalty. As the level of
consumer conﬁdence increases, this should be positively related to higher levels
of attitudinal loyalty of player’s club members. Trust is operationalized as
consumer conﬁdence and measured using the following indicators: ”I trust this
casino,” “I believe that this casino is honest in its dealings with me,” and “I rely on
this casino for my gaming.” These measures are based on previous market
studies (Chaudhuri and Holbrook, 2001; Doney and Cannon, 1997).

Past studies have shown that brand affect and consumer attachment are

Holbrook, 2001). For the purpose of this model, brand affect and consumer
attachment are modeled as afﬁliation. The indicators utilized to measure
affiliation include: “I feel good when I visit this casino,” “Being a member of this
players club makes me want to visit this casino more often,” and “Visiting this
casino gives me pleasure.” These measures are based on those utilized by
Chaudhuri and Holbrook (2001) and modiﬁed for use in a gaming environment. In
other words, as the pleasure of utilizing the services of an organization increase,
the level of attitudinal loyalty experienced by player’s club members should

ultimately increase.

Player’s Club Program Components

In order to understand the impact of the speciﬁc elements of loyalty programs
on creating both attitudinal and behavioral loyalty, these elements must be
included in the model. The idea of value creation has received signiﬁcant

attention and value creation strategies have been perceived to have the greatest

44

potential for increasing customer loyalty (Reichheld and Teal 1996). While there
are various methods for measuring perceived value, value has often been
measured using three elements — price, tangible deliverables, and intangible
attributes (Neal, 1999). Price is not a signiﬁcant issue given the nature of the
gaming product; the other two components form the backbone of loyalty
programs. Measurements of relative value for a product or service will provide a
good indicator of which service provider a customer would choose among a set
of competing providers (Neal, 1999). Integrating relative value into the casino
loyalty model being developed and tested in this study will provide a test of the
value construct and its relationship on creating attitudinal and behavioral loyalty.
The sum of the components of players club programs will be used as indicators
for perceived value and perceived value will is modeled as a latent factor.

A review of promotional literature regarding various casino loyalty programs
across the United States reveals that the major components of these programs
include direct mail, preferential treatment, interpersonal communications,
rewards, and promotions. Speciﬁc measures for these components were not
available from the literature since this is the ﬁrst empirical study to measure
speciﬁc components of loyalty programs within a gaming environment. However,
the work by De Wulf et al. (2001) provided guidance for the formation of these
measures. Due to the fact that four of the five components included in this study
were utilized in their study, building on their constructs seems appropriate due to

the lack of available measures in the literature.

45

Due to the fact that direct mail costs represent a signiﬁcant portion of casino
marketing budgets and are the dominant method utilized by casinos for
communicating with its player’s club members, its inclusion is essential to fully
understanding the impact of loyalty programs on creating attitudinal and
behavioral loyalty. Therefore, one element of the model is a measurement of
player’s club members’ (both Loyal Player’s Club members and Disloyal Players
Club members) perceptions of the extent that the client casinos and other
casinos keep them informed about the loyalty program, special events and
promotions through direct mail.

A number of previous studies provided guidance on ways to assess
perceptions of direct mail communications with customers (Anderson and Narus,
1990; De Wulf et al., 2001; Dwyer, Schurr, and Oh, 1987; Morgan and Hunt,
1994). The indicators that will measure player’s club member awareness and
perceptions of direct mail communication include: “I receive mailings about the
loyalty program on regular basis,” “I receive information about promotions
through these mailings,” and “I receive information about special events through
these mailings.” These indicators are similar to those utilized by De Wulf et al.
(2001) and were modiﬁed for use in this study.

The majority of casino loyalty programs are now utilizing tier or segmented
loyalty programs and top tier members receive preferential or special service.
Although the study by De Wulf et al. (2002) found that preferential treatment had
little effect on creating perceived value or attitudinal loyalty, this may have been

due to the setting of their study. In a retail environment, preferential treatment is

46

often not experienced by individuals due to the longer time between purchases
(De Wulf et al., 2001). Due to the fact that preferential treatment is a component
of casino loyalty programs, one would presume that it would have a greater affect
within the setting of this study. The indicators that will measure player’s club
members’ perceptions of preferential treatment are “I receive special treatment
because I am a member of the player’s club” and “Casino staff offer me more
personal service because I am a member of the players club.”

Interpersonal communications is an important component of not only casino
loyalty programs, but any relationship marketing strategy (Wood, 1998).
Interestingly, in the study by De Wulf et al. (2001), they found that interpersonal
communications had the greatest effect on predicting perceived relationship
investment. Interpersonal communication refers to the frequency of personal
exchanges that casino staff have with player’s club members as opposed to non-
player’s club members, while preferential treatment emphasizes that players club
members will receive a higher level of service as opposed to non-players club
members (De Wulf et al., 2001). The indicators used to measure interpersonal
communications are “Casino staff take the time to personally get to know me
because I am a member of the players club” and “Casino staff offer me more
personalized service because I am a member of the players club.”

The use of promotions and discounts as part of casino marketing strategies is
almost as old as casinos themselves. Recently, in order to gain access to special
events or to participate in these promotions, casinos have restricted these

offerings to only persons that are members of their respective player’s club

47

program. As indicated by Dowling and Uncles (1997), this may result in
customers detecting or decreasing their rate of visitation. In addition, the rampant
use of promotions may actually create a bargain prone market that will base their
product or service purchase decisions on the availability of promotions.

To test the importance of these promotions and events, the casino loyalty
program model in this study includes measures of player’s club members’
perceptions of casino efforts to offer special events, promotions, and discounts
available to player’s club members. Building on the work of De Wulf et al. (2001),
promotions and discounts are measured by using the following indicators: “I often
receive discounts because I am a player’s club member,” “I often get invited to
special events because I am a member of the player’s club,” and “I often
participate in promotions that are only available to player’s club members.”

A consistent practice is the use of points or comp balances as a method for
tracking players gambling usage and as a method for allowing player’s club
members to utilize these points to obtain rewards. These rewards can include
show tickets, airline tickets, access to tournaments, clothing, and or cash back
(transferring points back into cash). Although not tested within the framework of
their study, De Wulf et al. (2001) suggests that rewards be included in future
studies and that it may represent a signiﬁcant tangible element of any
relationship marketing strategy.

Although no indicators for rewards were found in the literature, building on the
work De Wulf et al. (2001) and Woods (1998), rewards are deﬁned as a tangible

beneﬁt that a casino provides to player’s club members as part of the overall

48

player’s club program. The indicators used to measure rewards include: “I often
receive rewards/gifts because I am a member of the players club,” “As my level
of gaming increases, the more gifts/rewards that I receive from the casino,” “I
have used my points/comp balance to purchase gifts, tickets, or cash back,” and
“I often receive comps (free food, drinks) during my visits because I am a

member of the players club.”

Mediating and Moderating Variables

The hypothesized casino loyalty model includes one mediating variable and
two moderating variables. The mediating variable is competitive factors. Because
of the competitive nature of casino loyalty programs, and the fact that past
studies have shown that members of loyalty programs belong to multiple
programs (Long and Schiffman, 2000), the current study includes a comparison
of the perceived values of membership in a casino loyalty program compared to
their perception of the value of programs offered by other casino loyalty
programs. Interestingly, previous studies have failed to include competitive
factors as a component of their models but have recommended that they be
included in future studies (Bennett and RundIe-Thiele, 2001; Chaudhuri and
Holbrook, 2001; De Wulf, Odekerken-Shroder and labcobucci, 2001). The
indicators that will measure competitive factors include: “Compared to other
casinos, I receive good value for my money,” “I feel that I receive more value for
my money by being a member of the player’s club,” “Compared to other player’s
clubs, this program offers me more value for my money,” and “I would remain a

customer of this casino even if they stopped offering the player’s club.”

49

Involvement is one of the moderating variables. In theory, persons who
experience higher levels of involvement with products and related businesses
should experience higher levels of loyalty (Dick and Basu, 1994). Past studies
have utilized product involvement as a distinct construct (De Wulf et al., 2001) or
suggested that it be used in future studies (Chaudhuri and Holbrook, 2001). In
this study, involvement is deﬁned as the consumer’s perceptions of the
importance of the product category based on the consumer’s needs, values, and
interests. This method is identical to that utilized by De Wulf et al. (2001) and
proposed by Mittal (1995). In other words, for those persons who are not
interested in the product category, lack of personal attention and perceived
relationship investment will not be as critical to them in their decision to continue
purchasing a speciﬁc product or service (Solomon et al., 1985). In the study
completed by De Wulf et al. (2001), they found that persons who exhibited lower
levels of product involvement are less inﬂuenced by a retailer’s investment in the
relationship. This would suggest that less serious gamers are not as much
inﬂuenced by components of the loyalty program compared to more active and
serious gamblers. The indicators used to measure involvement are: “I am
someone who ﬁnds it important about where I gamble,” “I am someone who is
interested in learning about gambling and different gambling opportunities,” and
“I am someone for whom it means a lot to gamble.” These indicators were based
on those proposed by De Wulf et al. (2001) and modiﬁed for use in a gaming

environment.

50

The other moderating variable is relationship commitment. The idea that
actions taken by one party will be reciprocated in kind by another party has been
found to be a key variable in channel relationships and should also be present in
consumer-ﬁmt relationships (Bagozzi, 1995). Unfortunately, the belief that all
consumers want to have a relationship with ﬁrms has not been supported in past
studies (Bendapudi and Berry, 1997; Christy, Oliver, and Penn, 1996). Again in
theory, persons who are more committed to maintaining a relationship with a
business, in this case a casino, should experience higher levels of behavioral
loyalty since they wish to have a relationship with the business and are more
receptive to loyalty based marketing initiatives and loyalty club incentives.
Relationship commitment was measured using the following indicators: “I am
someone who likes to be a regular customer of a casino,” “I am someone who
likes building relationships with companies,” and “I am willing to go the extra mile
to remain a customer of this casino.” These indicators are based on the work of

De Wulf et al. (2001).

Summary of Review of Literature

Although a signiﬁcant amount of research has been completed on customer
loyalty, many gaps exist and important questions remained unanswered. One of
the largest gaps is how loyalty is deﬁned and measured. This proposed
Integrated Casino Loyalty Model will explore how attitudinal and behavioral
loyalty is created in a service setting. Furthermore, it will also present a strong
test for the effectiveness of loyalty reward programs and their ability to create

attitudinal and behavioral loyalty.

51

A second essentially unanswered question is whether loyal customers and/or
loyalty programs lead to increased proﬁts by ﬁrms. Although it is widely assumed
that loyal customers are more proﬁtable customers and increased loyalty leads to
greater proﬁt, very little research has been shown to support this assumption.

This study will aid in developing metrics that will allow for an evaluation of
loyalty programs in a situation where these programs are predominantly
organized, and ﬁll a signiﬁcant gap in the research on loyalty, relationship
marketing, and loyalty program research (Anderson et al., 1994; Bolton et al.,
2000; Dowling and Uncles, 1997; Fredericks et al., 2001; Jain and Singh, 2002;
Long and Schiffman, 2000; Luxton, 2002; O’Brien and Jones, 1995; Reinartz and

Kumar, 2000; Reinartz and Kumar 2002).

52

CHAPTER 3
METHODS

Introduction to Methods

This chapter provides an overview of the data collection methods and
analyses that were utilized in the completion of this study.

In terms of structure, the focus of the ﬁrst section will discuss the study
population and data that will be collected as part of this study. The framework of
this study is depicted in Figure 5. The second section will discuss the analysis of
the data included in the player’s club database. The third section will focus on the
mail survey instrument and the sampling procedures utilized to complete the mail
survey. The fourth section will focus on how the mail survey data was prepared
and analyzed. The ﬁnal section will discuss structural equation modeling and the

process by which the Integrated Casino Loyalty Model will be tested.

Study Population and Study Data

The study population for this study consists of members of a player’s club of a
large native casino in the Midwest United States. This player’s club program was
initiated in May 1998, but only data collected on members from January 1, 1999
to December 15, 2002 was utilized for the purpose of this study. To ensure
anonymity of the client casino, the casino from which the data was drawn from
will be referred to as the client casino. All members of the player’s club for the
client casino except for those players who were banned (the majority of these

bans being self-imposed) were included in various study analyses.

53

 

 

0.5.500 $me
02.0 30.6... 2.85.6.0
.0 5:85:52 -

 

 

 

 

.0002. 3.9.0. 0283....

5 0:58:00 000.00

805.95: 0.55:8
.0 5:85.52 -

 

 

 

 

momma

5:85:50» .550.

3.9.0. .0 08:02:00.5
0.: 058mm. -

 

 

 

 

.5505 5.0

0.5.6.... 08.08 20.0

0.: .0 95:53 .oEmE
00800 :5 2:05 -

 

 

1

 

 

.5603
some .0. 2:05:90.
035.6”. 3...... 0...:0... -

 

 

 

$000... 5:955? .moom .65..
0:90 00...... - .625 :22 m 9509.30
0.5ng 5.0 0.9.0...
.o. .89.. 3.93 2.90 .88 5....
5.800.... .0 9:00... - >035 :05. v 0.500.3O
0:58:00
.5220. 95:52
.0. 00mm: 5:95....30 .moom >22.
Em>o.0. 3.9.0. .0 9:00... - .625 :92 m 0%..»3O
98 .88 as: 52:0 =22
>023 .05 00395 A1 .05
5 .00me 0.580% 829mm.
.0 £0an.98 - .550... 0550 2.2.0 N 9500.30
95580 L .88.
3.0 see: 0322.8 .2 .08 I 8.: .. .5...
9... 28:35... $298
.0 €055.93 - .5505 0....me E96 — 2500.30
amen—«.5. 3.5 5500—30 3.5

3.3.3.

 

 

 

Al

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

x.o.seEe..... >335 "m 0.52“.

54

Client Casino’s Player’s Club Database

When casino patrons join the client casinos player’s club program, a record is
established for them in the casino’s computer tracking system. At the time they
join, they are required to provide a driver’s license (or other acceptable ID) in
order to ensure that they are over 21 years of age. They are also asked to
provide some additional socioeconomic information. Members are then provided
Point of Sales (POS) cards that they place into readers on slot machines or
provide to dealers at table games in order to earn points and rewards. Obviously,
this information is also used to monitor their gaming amounts and patterns over
time and to focus promotions at certain types of gamers (e.g. slot tournaments,
table game tournaments).

Data is collected on player’s club members only on those occasions where
they have inserted their card in the reader. If members fail to place their player’s
club cards into the reader or provide it to the dealer, no data will be collected for
that transaction. Clearly not all players, including persons who are player’s club
members, utilize their cards all the time. In terms of the client casino, some of the
largest gamers do not use the cards at all.

This study makes use of socioeconomic information on members and data on
their gaming behavior at the casino including variables related to “how much” and
“how frequently” they game at the casino. The player’s club information that is

utilized in various aspects of this study is described in Table 3.

55

Table 3: Data Collected by the Client Casino’s Players Club Tracking
System

 

SOCIOECONOMIC CHARACTERISTICS

Name 1

Birth Date 1

Address (Street, City, State, Zip) 1

Gender 1

Marital Status 2

Wedding Anniversary Date (If Applicable) 2
Number of Children2

Distance to Casino (Point to Point)3

 

 

PURCHASE BEHAVIOR

Number of action days 4

Total theoretical win per action day 5
Number of minutes played 6

Total number of action days per month
Total theoretical win per month
Number of minutes played per month
Total number of action days per year

. Total theoretical win per year

 

1 -— This information is mandatory and obtained at time of registration

2 — This information is not mandatory

3 - This information is calculated by Casino Marketplace and is calculated from the Zip Code
provided at time of registration

4 - Action day is deﬁned as all gambling activity in a 24 hour period

5 - Theoretical win is amount gambled x ﬂoor average (0.069) x time

6 — This is the amount of time that a player’s club member has his/her card inserted into the

machine over the course of an action day

All data collected on player’s club members for the client casino is stored in
multiple SQL databases using Casino Marketplace management software. There
are two major suppliers of casino player’s club management systems (Casino
Marketplace and Akers). In terms of use and data storage, both systems are

almost identical. The only difference between the two systems is their user

interfaces.

56

 

Information is collected on player’s club members commencing on the date
they join. Obviously, no information is available on their gaming at the client
casino prior to their membership in the club. As a result, there was no information
on how long members have gamed at the client casino previous to joining. Only
information on their gaming from the client casino was available, even though (as
the survey indicates) the vast majorities visit multiple casinos and are members
of multiple player’s club programs.

At the time the study commenced there were 196,087 player's club members
for the client casino. The growth of the client casinos player’s club since its
implementation in May 1998 is depicted in Table 4. One of the recruitment
strategies of the client casino and other casinos is to make membership within
the player’s club mandatory in order to receive/participate in promotions or
discounts, including promotions offered to bus tours. Because many of these bus
tour gamers are not nearby residents and therefore often do not become regular
visitors to the casino, over half of the members of the client casinos player’s club
had less than three action days. An action day is deﬁned as all gambling activity
in a 24-hour period at the client casino. Although persons who are inactive for a
period of at least eighteen months are removed from the player’s club, the data
for these individuals is maintained within the client casinos player’s club tracking

database for research purposes.

57

Table 4: Membership History for the Players Club for the Client Casino

 

 

 

 

 

 

 

 

 

 

 

Time Period (As of) Members Added Total Members % Change from
in the Client previous time
Casinos Player’s period
Club Program
January 1, 1999 25,789 25,789 -
July,1 1999 21,732 47,521 84 %
January 1, 2000 12,029 59,550 25%
July 1, 2000 10,858 70,408 18%
January 1, 2001 18,607 89,015 26%
July 1, 2001 32,331 121,346 36%
January 1, 2002 27,870 149,216 23%
July 1, 2002 21,407 170,623 14%
December 15, 2002 25,464 196,087 15%

 

 

 

 

Following an approach similar to that used by Reichheld and Teal (1996) and

Reinartz and Kumar (2000), player’s club members were ﬁrst segmented into

four groups based on their lifetime action days and lifetime revenues. A median

split method of segmenting was utilized because this method has been found to

be a better descriptor of the lifetime distribution as opposed to mean splits

(Bearden, Rose and Teal, 1994; Collett, 1994; Schmittlein, Cooper, and

Morrison, 1994). The four segments are shown in Figure 6.

58

 

Figure 6: Loyalty Program Segments Based on Lifetime Action Days and
Lifetime Revenues at the Client Casino

 

 

 

 

 

 

High

High Frequency/ High Frequency/

LIFETIME ACTION Low Revenue High Revenue

DAYS

Low Low Frequency / Low Frequency /

Low Revenue High Revenue

Low High
LIFETIME REVENUES

Lifetime Action Days was determined by calculating each member’s number
of (gaming) action days since they joined the client casinos player’s club
program. An action day is all gaming completed within a 24-hour period.
Segmenting members based on lifetime duration was not appropriate given the
fact that many persons included in the database visited the client casino only one
time and were later removed from the program after 18 months of inactivity.
Although this differs from the approach utilized by Reinartz and Kumar (2000),
the correlation between lifetime duration and lifetime revenues was calculated
and was not signiﬁcantly different from zero- thus providing support for the use of
a different measure.

Theoretical win is used as a measure of revenues by most casinos.
Theoretical win is the amount gambled multiplied by the floor average. In the
case of the client casino, the ﬂoor average is 6.9%. Thus, for every $100
gambled at the client casino, the casino retains $6.90. Casinos prefer theoretical

win as opposed to amount of money gamed (e.g. the amount of money a person

59

places into the slot machine or places on the table) since often when persons win
small amounts they do not cash the money out. Not removing this money from
play means that the person is actually playing on the casino’s money. Theoretical

win also removes the effects of jackpot victories and large losses.

Mail Survey of Player’s Club Members

A mail survey was utilized to obtain data to test the Integrated Casino Loyalty
Model that was not available from the client casinos players club tracking
database. Respondents were not asked for their name and address unless they
wished to be included in the incentive draw. The incentive utilized was an
opportunity to win one of ﬁve prizes of $100. Data was collected on these

individuals from April 1, 2003 to June 1, 2003.

Mail Survey Instrument

The six page survey instrument can be found in Appendix 1. All participation
in the survey was voluntary. The majority of the survey questions collected
information for segmentation purposes and tested the Integrated Casino Loyalty
Model and related constructs. General socioeconomic information was also
collected to assist in targeting marketing strategies formulated as a result of this

study. The data collected as part of this study is described in Table 5.

6O

Table 5: Data Collected in Mail Survey

 

SOCIOECONOMIC CHARACTERISTICS

Name

Address (Street, City, State, Zip)

Age

Gender

Number of adults residing in household
Number of children residing in household
Household Income

Incentive
Incentive
12
13
14
14
15

99999

 

 

GAMING BEHAVIOR

Visitation to casinos previous 12 months

Number of casinos visited in previous 12 months
Number of trips taken to casinos in previous 12 months
Amount budgeted per trip to casinos

Visitation to all casinos

Membership of players club programs for all casinos

% of total trips to casinos made to each casino
Changes in amount of visitation to each casino
Reasons forjoining (or not joining) players club program at
__Primary Casino

MODEL CONSTUCTS

Direct Mail

Preferential Treatment
Interpersonal Communications
Promotions and Discounts
Rewards

Affinity

Consumer Conﬁdence
Perceived Value
Competitive Factors
Involvement

Relationship Commitment
Attitudinal Loyalty
Behavioral Loyalty
Satisfaction 1

Switching Costs 1

Value 1

99.099.09.090
\lecno-Io-IUIJr-ooN—x

p be

_.s
_I

oppppppppopppppp

 

1 - Not tested in Model

6]

 

Table 5: Data Collected in Mail Survey (cont.)

 

 

LOYALTY PROGRAM INFORMATION

0 Use of player’s club program card Q. 9
0 Membership in other loyalty prﬂrams Q. 10

 

The ﬁrst section collects information relating to their casino visits during the
previous 12 months. Persons who had not visited any casino in the previous 12
months were asked their reasons for not visiting. They had twelve possible
reasons to select from. These twelve options were developed in conjunction with
casino management. This section also collected data on the number of casinos
visited over the previous 12 months, the number of visits to casinos over the
previous 12 months, and the amount budgeted for casino visits. A visit was
defined as a 24—hour period during which they visited a particular casino
regardless of whether they left and came back again.

The second section asked about visits to various casinos. A list of ten casinos
was selected based on the results of on-site and telephone surveys completed
for the same casino in 2001. They include nine casinos located in Michigan and
one in Ontario, Canada.

Respondents were asked (1) “whether they have ever visited that specific
casino,” (2) “whether they are a member of that casinos player’s club,” (3)
“whether they had visited that casino in the previous 12 months,” and (4) “what
percentage of their total casino trips over the previous 12 months was made to

that casino.” These questions were asked for each casino. They were then

62

 

asked about changes in their visitation over the course of their gaming lifetime to
these same casinos.

Respondents were asked whether they were visiting these casinos the same
amount, more often, or less often than in the past. Due to space constraints,
responses to Turtle Creek Casino and Leelanau Sands were combined since
both properties are operated by Traverse Bay Entertainment. In addition, the four
casinos located in the Detroit / Windsor area (Windsor Casino, MGM Grand,
Motorcity, and Greektown) were also merged into one category. Respondents
could also write in another casino they had visited but was not included in the list.
For the casinos they were visiting more frequently, they were asked about
possible reasons for this change including: “Satisﬁed with gaming experience,”
“Convenient,” “Take advantage of player’s club beneﬁts,” “Good Service,” “Loose
slots IWin Often,” and “Other.” They were asked to check all that apply. For the
casinos they were visiting less frequently, they were asked about possible

reasons for this change including: “Not satisﬁed with experience, Poor service
experience,” “Poor players club beneﬁts,” “Lack of promotions,” “Not convenient,”
“Tight slots I Did not win,” and “Other.”

The third section collects information about their primary casino. The primary
casino is the casino that they had visited most frequently during the previous 12
months, not necessarily the one where they gambled the most money. They are
ﬁrst asked whether they are a member of the player’s program and the reasons

why they joined the program (or why they did not join the program). The second

set of questions gathers data for model testing and is outlined in Table 6. Forty-

63

two separate questions were asked using a seven-point Likert scale (1 being

strongly disagree and 7 being strongly agree). The ﬁnal question asked about

how frequently they use their player’s club card while gambling.

Table 6: Loyalty Factor Constructs Included in Mail Survey

 

 

 

Construct Deﬁnition Purpose References

Direct Mai| Perceptions of the Component of - Anderson and Narus
extent that their player’s club program (1990)
primary casino keeps and an antecedent of - De Wulf et al. (2001)
them informed about perceived value - Dwyer, Schurr, and
the loyalty program, Oh ( 1987)
special events and - Morgan and Hunt
promotions through (1994)
direct mail

Preferentia| Perception that, as Component of - De Wulf et al. (2001)

Communications

players club
members, they
receive a higher level
of service as opposed
to non-players club
members

player’s club program
and an antecedent of
perceived value

 

Interpersonal
Communications

Frequency of personal
exchanges that casino
staff have with
player’s club
members as opposed
to non player’s club
members

Component of
player’s club program
and an antecedent of
perceived value

- De Wulf et al. (2001)
- Wood (1998)

 

Promotions and

Perception of the

Component of

- De Wulf et al. (2001)

 

 

Discounts frequency and player's club program - Wood (1998)
. . participation in one- and an antecedent of

time events or perceived value
promotions that are
available only to
player’s club
members

Rewards The perception of the Component of - De Wulf et al. (2001)

 

value and frequency
of gifts that are
available to player’s
club members

 

player’s club program
and an antecedent of
perceived value

 

- Wood (1998)

 

64

 

Table 6: Loyalty Factor Constructs Included in Mail Survey (cont.)

 

 

 

 

 

 

 

Construct Deﬁnition Purpose References
Afﬁnity The intensity of To identify the — Chaudhuri and
positive (or negative) intensity and Holbrook (2001)
responses that frequency of positive
player’s club affective responses
members perceive to which demonstrate
have with their higher levels of
primary casino attitudinal loyalty.
Modeled as an
antecedent of
attitudinal loyalty
Consumer The belief that their To identify the - Chaudhuri and
Confidence primary casino can perceptions that Holbrook (2001)
continue to meet their casino player’s club
expectations and programs meet the
obligations to player’s expectations of
club members members. Modeled as
an antecedent of
attitudinal loyalty
Competitive Perceived value of Due to the lack of - Bennett & Rundle-
Factors being a member of differentiation Thiele (2001)
their primary casino between programs, - Chaudhuri and
player’s club determining the Holbrook (2001)
compared to their quality of the player’s - De Wulf, Odekerken-
perception(s) of the club program of their Shroder and
value of programs primary casino labcobucci (2001)
offered by other compared to other
gaming organizations programs will
determine whether
persons are loyal to
only high quality
programs
Involvement Consumer's In previous studies, - Chaudhuri and
perceptions of the levels of involvement Holbrook (2001)
importance of the were found to - De Wulf et al. (2001)
product category moderate the - Dick & Basu (1994)
based on their needs, inﬂuence of - Mittal (1995)
values, and interests relationship marketing - Solomon et al.
strategies (1 985)
Relationship Their perceived level To determine the - Bendapudi and
- of commitment to extent to which Berry (1997)
comm'tment maintaining a persons wish to have - Christy, Oliver, and
relationship with their a relationship with Penn (1996)

 

primary casino

businesses

 

 

 

65

 

Table 6: Loyalty Factor Constructs Included in Mail Survey (cont.)

 

Construct

Definition

Purpose

References

 

Perceived Value

Frequency and
amount of tangible
and intangible
rewards they receive
by being members of
their primary casino
player’s club program

Customers feel
obligated to reward
companies that invest
in maintaining
relationships with
loyal customers

- Bagozzi (1995)
- De Wulf et al. (2001)

 

Attitudinal Loyalty

Absolute dedication or
commitment to their
primary casino

Although difﬁcult to
create by businesses
as part of relationship
marketing strategies,
persons who exhibit
higher levels of
attitudinal loyalty are
less prone to switch
service providers

- Chaudhuri &
Holbrook (2001)

- Morgan and Hunt
(1994)

 

 

Behavioral Loyalty

 

The probability that
the person intends to
maintain a continuing
relationship with their
primary casino

 

Outcome variable for
this model because
the goals of casino
loyalty programs are
to increase market
share and revenues.

 

- Chaudhuri and
Holbrook (2001)

- Cunningham (1957)

- Jacoby and
Chestnut (1978)

 

The next section asked about their

(operated by other industries)

membership in other loyalty programs

including seven questions that gathered

information about the relationship commitment construct, satisfaction, and value

constructs. The ﬁnal section collects socioeconomic information (age, gender,

number of persons residing in the household and household income).

66

 

Mail Survey Sampling Method

In order to study a consumer’s lifetime behavior, a minimum of four years of
data is required to perform any statistical analysis (Reinartz and Kumar, 2000). A
direct result of this requirement was that the High frequency/ High revenue (N=
66,081) segment was the only segment which met these requirements, as
median splits for the segments were three action days. Further, $43.31 in lifetime
theoretical win would be surpassed by any person having three years of
continuous activity at the client casino. Additionally, the client casino requested
that 218 preferred members be removed from the mail survey sample. As a
result, the ﬁnal sampling frame for the mail survey was 65,863 members of the
client casinos player’s club program.

Loyal Players Club members were deﬁned as those members who had
visited the client casino at least once each of the four previous years. A total of
5,499 members of the High frequency/ High revenue segment met this criterion.
A random sample of 1,300 persons was selected for participation in the mail
survey. Statistical analysis revealed no statistical differences between these
1,300 members and those not selected on the following variables: age, distance
to casino, lifetime theoretical win, lifetime action days, and gender.

Disloyal Player’s Club members were those members who had been a
member of the client casino’s player’s club for over four years and either: (1) had
no recorded activity from January 1, 2002 to December 15, 2002 or (2) had ﬁve
percent or less of their total lifetime theoretical win at the client casino

represented by gaming activity from January 1, 2002 to December 15, 2002. A

67

total of 3,281 members of the High frequency / High revenue segment met these
criteria. A random sample of 1,700 persons was selected for participation in the
mail survey. Disloyal Player’s Club members were over-sampled due to the fact
that this inactivity could be the result of change of address, death, and/or health
reasons. Additionally, because members were not required to submit change of
address information to the casino, the incidence of bad mailing addresses was
signiﬁcant. Again, no statistically signiﬁcant differences were found between
Disloyal Player’s Club members included in the sample and those who were not

chosen.

Mail Survey Administration

On April 1, 2003, a cover letter, six-page questionnaire, and a business reply
return envelope were mailed to 3,000 player’s club members of the client casino.
No identiﬁcation marks were placed on the returned questionnaires in order to
assure anonymity. A reminder post card was mailed to 2,640 persons on April
16, 2003. To ensure that there would be an acceptable number of cases to test
the hypothesized model using structural equation modeling, on May 2, 2003, a
follow-up (cover letter, survey questionnaire, business reply envelope) was
mailed to 600 members included in the Disloyal Player’s Club sample and 400
members included in the Loyal Player’s Club sample who had not replied. These
persons were randomly chosen from those persons who had not yet returned an
identiﬁable survey. The surveys included in the second survey were identical to
the ﬁrst mailing, but identiﬁcation marks were placed on the survey in order to

differentiate these surveys from those sent in the ﬁrst mailing.

68

The response rates for each mailing are indicated in Table 7. The number of

surveys received by both groups remained quite consistent for each mailing.

Table 7: Number of Surveys Received for Each Mailing

 

 

 

 

 

 

Number of Returned Surveys
First Mail Reminder Second Mail
Survey Postcard Survey

Loyal Player’s 258 120 126
Club

Disloyal 275 124 93
Player’s Club

TOTAL 533 244 219

 

 

 

 

 

In terms of overall response rates, a respectable response rate of 39% was
obtained (Table 8). As would be expected, the response rate was higher for the
Loyal Player’s Club segment (43%) as opposed to the Disloyal Player’s Club

segment (36%).

Table 8: Overall Response Rate to Mail Survey

 

 

 

 

 

 

 

 

 

 

Completed Original Undelivered Delivered Response
surveys sample mail sample rate 2
size'1
Loyal 504 1300 1 19 1 181 43%
Player’s
Club
Disloyal 492 1700 328 1372 36%
Player’s
Club
TOTAL 996 3000 447 2553 39 %

 

1- Delivered sample size = original sample - undelivered mail
2- Response rate = completed survey / delivered sample size

69

 

In terms of undelivered mail, a relatively high percentage of the mailings to
the Disloyal Player’s Club sample (19%) was returned as undeliverable. A large
number (185) of surveys were returned with bad addresses, 14 were returned
with a notice that the person was deceased, and 129 surveys were returned with
a notice that the person had moved and left no forwarding address, or the
forwarding order had expired.

Conversely, just 9% of the surveys sent to Loyal Player's Club members were
returned as undeliverable. Of this total, 65 surveys were returned with bad
addresses, 4 were returned with a notice that the person was deceased, and 50
surveys were returned and with a notice that the person had moved and left no
forwarding address, or the forwarding order had expired.

The high percentage of bad addresses and members who changed
addresses or had died reveals some problems in terms of database
management. It also supports the decision to oversample Disloyal Players Club

members.

Preparation of Data Collected Through Mail Survey

All surveys were coded and entered into an Excel spreadsheet. The data was
then transferred to the Statistical Package for Social Sciences, version 11.0
(SPSS). After the transfer, random checks of the data were performed in order to
ensure no data was lost during the transfer and no mistakes occurred during the

transfer.

70

Of the 996 completed surveys returned, 14 surveys were disregarded since
either no informatiOn was provided or the majority (over 95%) of the
questionnaire was unanswered.

Of the remaining 982 surveys, checks were conducted to identify any outliers.
A total of ﬁve surveys were identiﬁed as being outliers in terms of annual amount
gambled and amount of trips to casinos each year. In order to check their
responses, a check was conducted using the client casinos player’s club
database, and their responses were found to be consistent with their gambling

behavior, thus providing support for their inclusion in this study.

Preparation of Data Collected Through Mail Survey for Use in Structural
Equation Modeling

As outlined in Table 5, 48 questions gathered data to test the Integrated
Casino Loyalty Model. These questions utilized a seven-point Likert scale (1
being very strongly disagree and 7 being very strongly agree).

In order to ensure that persons did not check the same answer for all (or the
majority) of questions related to the constructs, a check was performed for each
score. Of the 982 completed surveys, a total of 25 surveys were found to have
shown a patterned response. A patterned response was deﬁned as a person who
chose the same value 95% or more of the time for the 48 questions. These
responses were not included in the model-testing phase of the analysis, nor were
they included for the conﬁrmatory factor analysis.

In order to conduct model-testing using structural equation modeling, all

construct measures were required to have a value (for tests using EQS 5. 7b). No

7]

cases could have a missing value for any construct. A total of 151 surveys were
found to have a missing value for a construct and were thus excluded from the
model-testing. As a result, 806 total surveys (32% of the total deliverable sample)

were utilized for testing the Integrated Casino Loyalty Model.

Model-Testing Using Structural Equation Modeling

The most common use of structural equation modeling is for model-
generating (Joreskog, 1993). In terms of model generation, researchers test a
hypothesized model and modify the model if necessary in order to improve its ﬁt
relative to the data being analyzed. The goal of this process is to discover a
model that is theoretically grounded and corresponds to data reasonably well.

Common statistical procedures like multiple regression, canonical correlation,
factor analysis, and ANOVA can be referred to as special cases of structural
equation modeling (Kline, 1998). The major difference between these statistical
procedures and structural equation modeling is its ability to evaluate entire
models and its reduced reliance on signiﬁcance tests. Structural equation
modeling provides researchers with the ability to prove, reject, or modify parts of
a model or the entire model.

The three most common software programs utilized for structural equation
modeling (SEM) are AMOS, LISREL, and EQS. Although other software exist
these three are most commonly used by academics and practitioners (Kline,
1998). Among these three, LISREL and EQS are most commonly used, in part,
because of the signiﬁcant amount of research and literature available in regards

to these programs and their frequent use in structural equation modeling.

72

Although AMOS was one of the ﬁrst SEM programs, published work on its use
remains quite limited as opposed to EQS and LISREL.

In this study, EQS 5. 7b was utilized to complete the analysis. EQS 5. 7b was
chosen because of its ability to handle raw data and because of its extensive
capabilities for model generation, which allow it to be used for both data
preparation and analysis. Furthermore, through its use of robust standard errors
and multiple estimation methods, it provides more ﬂexibility in use and managing
data as opposed to LISREL and AMOS. The major disadvantage of utilizing EQS
is its inability to estimate for incomplete data. In relation to the current study, this
was not a limitation since the number of cases with complete data exceeded the
minimum recommended threshold of 150 cases per group (Kline, 1998).

The ﬁrst step in model-building is to conduct a conﬁrmatory factor analysis.
Once factors are identiﬁed, the measurement model is tested. The measurement
model only models the relationships between the latent factors and their
respective indicators. After the ﬁt of the measurement model has reached an
appropriate level for the entire sample, the full structural model is then tested.
Modifications to the model are then made in order to improve the model ﬁt in
relation to the data. For testing multiple groups, the model is then tested for each
group to determine if any differences exist between groups. If additional data is
available, rival models can also be tested to determine their ﬁt in relation to the
hypothesized or revised models. The purpose of testing these rival models is to
allow for alternative explanations and/or comparisons to be made using the same

data sets.

73

Limitations of Data

The client casino player’s club database had several limitations in relation to
this study. The ﬁrst limitation was an absence of data for the period from
December 16, 2002 to April 1, 2003. Some of the persons identiﬁed as Loyal
Player’s Club members could have stopped frequenting the casino during this
period. They would therefore be incorrectly classiﬁed as Loyal Player’s Club
members.

The second limitation was that information was only obtained on player’s club
members from the client casino. As a result, this study was unable to identify
persons who may be members of multiple player’s club programs.

The ﬁnal limitation was in regards to whether the persons who joined the
client casinos player’s club were new customers to the casino or were already
existing customers. Without this information, the current study was unable to
determine how effective the client casino loyalty program was at attracting new

customers.

74

CHAPTER 4
RESULTS

Introduction

This chapter presents the results of the analyses that were conducted to
achieve the study objectives. The ﬁrst section is the results of the analyses for
estimating the lifetime visits I lifetime revenues relationship of player’s club
members. The next section presents the mail survey results. The third section
reports the results of the conﬁrmatory factor analysis. The fourth section presents
the tests of the Integrated Casino Loyalty Model for various segments. In the last
section, segments based on data gathered from the mail survey will be identiﬁed

and proﬁles provided.

Lifetime Visits - Lifetime Revenues Relationship

The ﬁrst objective of this study was to explore the relationship between
lifetime visits and lifetime revenues for members of the client casinos player’s
club. Median splits were utilized to divide the client casinos player’s club tracking
database into four groups based on lifetime action days and lifetime theoretical
win. An action day was deﬁned as all gaming completed at the client casino
within a 24-hour period. Theoretical win was deﬁned as the amount gambled at
the client casino multiplied by the ﬂoor average of the client casino. These four
groups are indicated in Table 9 and are titled High frequency/ Low revenue, High
frequency / High revenue, Low frequency / High revenue, and Low frequency /

Low revenue. Previous analyses using median splits, and as predicted by

75

 

relationship marketing, the two diagonal groups should represent the largest
percentage since a perfect relationship between any two variables would result in
both diagonals having a 50% share of the sample. If more of the player’s club
members reside in the off-diagonals, the relationship between the two variables
is less strong. From the current analysis, the largest percentage of player’s club
members (Low frequency / Low revenue (39.2%) and High frequency / High
revenue = (33.7%)) were found in the diagonals.

Of interest is that a signiﬁcant portion of the player’s club members resided in
the off-diagonals (27%), thus showing that a signiﬁcant portion of the market
(Low frequency/ High revenue (15.3%) had high revenues although their number
of lifetime action days was quite small. Furthermore, a portion of the market had
below median revenues although they visited the casino frequently (High
frequency / Low revenue (11.7%). This suggests that even though these
members visited the client casino frequently, the revenues derived from this
group was quite small. The size of the off-diagonal groups suggests that lifetime
action days by itself is not a strong predictor of lifetime revenues for members of
the client casinos player’s club.

Bivariate Pearson Correlations were calculated to test the strength of the
lifetime visits / lifetime revenues relationship. The r was 0.08 for the High
frequency / Low revenue group and 0.45 for the High frequency / High revenue
group. This indicates only a moderate linear relationship between lifetime action
days and lifetime revenues for the High frequency / High revenue group and a

very weak relationship for the High frequency / Low revenue group. The

76

correlation for the Low frequency / High revenue group was not statistically

signiﬁcant from zero and the correlation for the Low frequency / Low revenue

segment was 0.36, which suggests a moderate linear relationship. Thus, only the

two diagonal groups exhibited a moderate linear relationship between lifetime

action days and lifetime revenues, which suggests that lifetime action days is not

an adequate predictor of lifetime revenues for the off—diagonal groups.

Table 9: Segmentation of Client Casino’s Player’s Club Members Based on
Lifetime Action Days1 and Lifetime Theoretical Win at the Client

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Casino2
High High Low Low All
Frequency Frequency Frequency Frequency Player’s
Low High High Low Club
Revenue Revenue Revenue Revenue Members

% of player’s 11.7% 33.7% 15.3% 39.2% 100%
club members
MEAN THEORETICAL WIN PER YEAR

1999 $21.51 $869.37 $180.62 $14.80 $328.93

2000 $21.62 $1,000.26 $141.17 $14.05 $366.72

2001 $24.43 $1,206.52 $139.57 $11.36 $435.26
MEAN THEORETICAL WIN PER ACTION DAY

1999 $12.41 $104.17 $131.28 $11.90 $61.31

2000 $ 9.05 $ 50.85 $ 90.01 $11.51 $36.48

2001 $ 6.30 $ 38.11 $ 85.24 $ 9.46 $30.33

2002 $ 5.01 $ 24.10 $ 65.48 $ 7.29 $21.58
MEAN NUMBER OF ACTION DAYS PER YEAR

1999 2.5 Days 8.9 Days 1.6 Days 1.2 Days 4.0 Days

2000 3.3 Days 18.7 Days 1.7 Days 1.3 Days 7.4 Days

2001 5.8 Days 31.7 Days 1.8 Days 1.3 Days 12.1 Days
SUMMARY

% change in

Theor. Win 13.0 % 38.7 % - 22.8 % - 23.3 % 32.3 %

(1999 — 2001)

% change in

Theor. Win / - 59.6 % - 76.9 % - 50.1 % - 38.8 % - 64.8 %

Day

(1999 - 2002)

% change in

Action days 131.1 % 256.2 % 12.5 % 5.7 % 200.2%

(1999 - 2001)

 

 

 

 

 

 

1- Theoretical win was deﬁned as amount gambled x ﬂoor average (0.069) x time
2- Action day was deﬁned as all gambling activity for a 24 hour period

77

 

In terms of the overall gambling behavior of the client casino’s player’s club
members, as indicated in Table 9, the mean theoretical win per action year
increased 32% from 1999 to 2001. While the amount gambled per year
increased, the mean theoretical win per action day actually decreased 64.8%
between 1999 to 2002. There was also a 200% increase in number of action
days per year by player’s club members of the client casino from 1999 to 2001.
The combination of these analyses indicates that player’s club members of the
client casino were visiting more often but divided their annual amount of money
gamed at the client casino over more trips. So while the client casino
experienced an increase in annual gambling revenue from player’s club
members, the cost of securing this additional revenue was greater and as a
result, average gaming per visit and margins were reduced. Although direct
expenses were not obtained per gaming visit and were not available, one can
assume that there was a service cost (estimates range from $5 to $50 for an
average casino patron) for each visit to the client casino.

The next analysis focused on the four groups (High frequency / Low revenue,
High frequency /High revenue, Low frequency / High revenue, and Low
frequency / Low revenue) in order to explore their gambling behavior over the
previous four years. In terms of revenues, the group with the most potential for
proﬁtability was the High frequency /High revenue group. Between 1999 and
2001, the mean theoretical win per year increased 38.7% while their number of
action days per year increased 256%. Although these numbers showed a

signiﬁcant increase, the theoretical win per action day decreased 77% between

78

1999 and 2002 to just a meager $24.10 per action day in 2002. This further
demonstrates that the cost of securing this additional revenue was coming at an
increased cost to casinos.

Of major interest is that the Low frequency / Low revenue represented the
largest percentage of the client casino’s database (39.2%). This group has a low
rate of visitation (less than three lifetime action days) and low lifetime theoretical
win (less than $43.31). The few number of visits and the low amount gamed per
trip exempliﬁes some of the disadvantages of forcing persons to join the player’s
club programs. This forced recruitment appears to add a signiﬁcant number of
less proﬁtable members to the program, thereby increasing the cost of managing
and maintaining the player’s club program.

In relation to the off-diagonal groups, the Low frequency/ High revenue group
visited the casino less frequently (less than 3 lifetime action days) but gambled
more during each visit compared to other groups. Since they had the highest
mean theoretical win per action day as compared to the other three groups,
these persons represented a high revenue group and the goal of any player’s
club program should be to increase visitation of this group. The High frequency/
Low revenue group represented the least potentially proﬁtable group since they
gambled small amounts but visited the casino often. Although their theoretical
win per year increased 13% between 1999 and 2002, their theoretical win per
year in 2002 was only $24.43, which suggests that the costs of marketing to

these individuals is less likely to be recouped.

79

The High frequency / High revenue group of the client casino was further
divided into four different sub-groups using a median split of 21 lifetime action
days and $145.45 lifetime theoretical win. This segmentation process was
undertaken in order to further examine the highest revenue group of the client
player’s club program and to identify potential sub-groups. The results are shown
in Table 10.

Sub-group 2 comprised over half (57.2%) of the High frequency / High
revenue group. These player’s club members had over 21 lifetime action days
and more than $145.45 in lifetime theoretical win. This sub-group of player’s club
members is obviously important to the casino since their theoretical win per
action year ($2,000.88 for 2001) was six times greater than the other three sub-
groups and their mean number action days per year (47.2 days for 2001) was
twice that of the other sub-groups.

Player’s club members of the client casino comprising sub-group 3 had the
highest theoretical win per action day ($42.68) as compared to the other three
segments. This shows that an important group of player’s club members exist in
the “off-diagonal” and that lifetime visits by itself is not effective at fully explaining

lifetime revenues.

80

Table 10: Sub-groups Comprising the High Frequency I High Revenue
Group of Player’s Club Members

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sub- Sub- Sub- Sub- All
group 1 group 2 group 3 group 4 High
frequency
High
Revenue
Group
% of High
High revenue
roup
MEAN THEORETICAL WIN PER YEAR
1999 $64.89 $1,418.21 $455.39 $60.57 $869.37
2000 $83.85 $1,656.33 $358.57 $58.48 $1,000.26
2001 $124.31 $2,000.88 $369.45 $69.78 $1,206.52
MEAN THEORETICAL WIN PER ACTION DAY
1999 $19.06 $134.52 $212.96 $37.03 $104.17
2000 $10.72 $ 65.59 $ 95.09 $20.81 $ 50.85
2001 $ 7.71 $ 47.40 $ 79.55 $17.57 $ 38.11
2002 $ 5.48 $ 30.73 $ 42.68 $11.56 $ 24.10
MEAN NUMBER OF ACTION DAYS PER YEAR
1999 5.7 Days 12.9 Days 2.7 Days 2.1 Days 8.9 Days
2000 12.2 Days 27.9 Days 3.8 Days 3.0 Days 18.7 Days
2001 22.6 Days 47.2 Days 5.3 Days 4.6 Days 31.7 Days
SUMMARY
% change in
Theor. Win 91.6 % 41.1 % - 23.3 % 15.2 % 38.7 %
(1999 - 2001)
% change in
Theor. Win / Day - 71.2 % - 77.2 % - 80.0 % - 68.8 % - 76.9 %
(1999 — 2002)
% change in #
of Action days 296.5 % 266.6 % 96.3 % 119.6 % 256.2 %
(1999 — 2001)
Median Splits

Lifetime Action Days

= 21 Action Days

Lifetime Theoretical Win

= $145.45

81

 

With the exception of sub-group 3, all the other sub-groups exhibited
increases in theoretical win per year between 1999 and 2001. However, it is
important to recognize that all sub-groups exhibited a signiﬁcant decrease in
theoretical win per action day ranging from 68.8% to 80% from 1999 to 2001.
This, coupled with the signiﬁcant increases (96% to 296%) in number of action
days per year during the same time period, indicates that these members were

also spreading their annual gaming over more trips.

Descriptive Proﬁle of Player’s Club Members who Completed the Mail
Survey

As previously discussed, a mail survey was utilized to gather information
about player’s club members’ overall gambling behavior and to gather data
required to test the Integrated Casino Loyalty Model. A descriptive proﬁle of the
respondents is shown in Table 11.

Only 2.9% of the persons who returned a completed survey stated that they
had not visited a casino during the previous 12 months. The three most cited
reasons for making no casino visits were: (1) they were not winning (50%), (2)
they had lost too much money (32.1%), or (3) they lost interest in gambling
(25%).

The 97.1% who had visited a casino in the previous 12 months visited, on
average 6.9 casinos and made 22.5 trips to casinos during the previous 12

months. They planned to budget an average of $129.81 per visit to a casino.

82

Table 11: Descriptive Proﬁle of Player’s Club Members of the Client Casino
that Returned a Completed Mail Survey

 

 

Frequency I Mean

 

GAMBLING BEHAVIOR

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Visited any casino the previous 12 months
Yes 97.1% -
No 2.9% _
Reasons for not visiting a casino in the previous 12
months
Lost interest in gambling 25.0% 1 -
Lost too much money 32.1% -
Lack of time 17.9% -
Not winning 50.0% -
Health declined 7.1% -
Don’t like atmosphere at casinos 17.9% ..
Number of casinos visited the previous 12 months - 6.9 2
Number of visits to casinos the previous 12 - 22.5 trips 7
months
Amount budgeted for each casino visit - $129.81 2
PLAYERS CLUB PROGRAM
Member of the player’s club of the casino visited
most often the previous 12 months
Yes 98.1% -
No 1.9% ..
Reasons for joining their primary casino’s player's
club
Joined player's club to receive promotions 75.2% 3 -
Joined player’s club to receive comps or gifts 79.4% -
Joined player’s club to get more value for my money 48.1% _
Joined players club to receive specialized service 23.7% -
Joined player’s club card for social reasons 13.2% -
Reasons for NOT joining their primary casino’s
player’s club
Do not want to be tracked 11.1% 4 -
Don’t want comps or gifts 5.6% -
Don’t go to any one casino enough 38.9% -
The benefits of joining are not worth it 33.3% -

 

 

 

1 - Percentage based on those respondents who did not visit a casino in previous 12 months

2- Percentage only based on those persons who visited a casino in previous 12 months

3 - Percentage based on those respondents who did join the players club of its primary casino

4- Percentage based on those respondents who did not join the players club of its primary
casrno

83

Table 11: Descriptive Proﬁle of Player’s Club Members of the Client Casino
that Returned a Completed Mail Survey (cont.)

 

I Frequency 1 Mean

 

PLAYER’S CLUB PROGRAM (cont.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Overall use of player’s club cards at all casinos - -
All the time 56.8% -
Most of the time 33.3% -
Sometimes 7.6% -
Rarely 1.5% -
Never 05% -

MEMBERSHIP IN OTHER TYPES OF LOYALTY PROGRAMS
Frequent Flyer Program(s) 29.7% -
Grocery Store Prggams) 50.0% -
Car Rental Program(s) 5.0% -
Hotel Prgqram(s) 15.3% -
Retailer Frequent Buyer Program(s) 5.6% -
Music and Book Club(s) 10.1% -
Restaurant Prggram(s) 25.9% -

SOCIOECONOMIC CHARACTERISTICS

A93 - 63.3 years
21 — 30 0.6% .
31 — 40 3.1% .
41 - 50 10.3% -
51 - 60 24.4% -
61 - 70 33.8% -
Over 70 27.7% -

Gender
Male 34.1% -
Female 65.9% -

Number of Adults residing in Household - 1-91

Number of Homes with Children 75% -

Household Income
Under $40,000 47.8% -
$40,001 - $59,999 24.8% -
$60,000 - $99,999 19.3% -
$100,000 - $149,999 6.8% -
$150,000 - $199,999 0.9% -
Over $200,000 0.3% -

 

 

 

84

A number of survey questions focused on respondents’ relationship with their
primary casino, which is deﬁned as the casino that they visited most frequently
over the previous 12 months. Almost all (98.1%) respondents stated that they
were a member of the player’s club at their primary casino. The dominant
reasons that they joined the player’s club program at their primary casino were to
receive comps and gifts and to have access to promotions, or in other words,
forced recruitment. The very small percentage who were not members of the
player’s club program at their primary casino did not join because they did not
believe they would make enough visits to any one casino to make joining
worthwhile. In other words, they recognized that they would not be signiﬁcant or
loyal enough customers to beneﬁt from being a player’s club member. Also, they
either visited the casino when there was not a player’s club related promotion in
progress, did not come on a bus tour, or did not consider the usual joining
incentives (promotions) sufficient.

In terms of use of their player’s club card, only 56.8% stated that they used it
all the time. Since a number of customer lifetime value models require each
transaction to be noted, their subsequent use in studying loyalty programs within
a gaming setting is quite limited. This, in addition to some ﬁndings of this study
that will be discussed later, indicates that although player’s club tracking data is
very useful for some analyses, additional information is needed to have a

complete picture of their casino gaming behavior, including behavior at other

casinos.

85

In terms of their propensity to participate in loyalty programs operated by
other types of businesses (e.g. travel, retail, grocery stores), 50% were members
of grocery store frequent buyer programs, 29.7% were members of frequent ﬂyer
programs, 25.9% were members of restaurant programs, and 15.3% were
members of hotel programs. This suggests that there were some player’s club
members who were more prone to participate in loyalty building programs.

The mean age was 63.3 years and 65.9% of respondents were female. A
signiﬁcant portion of respondents (47.8%) had a household income of less than
$40,000 per year, which was likely an artifact of the sample being retired persons

or senior citizens on ﬁxed incomes.

Description of Loyal Player’s Club and Disloyal Player’s Club Segments
Based on Gaming Activity at the Client Casino During the Previous 12
months as Compared to their Lifetime Gambling Behavior at the Client
Casino

Loyal Player’s Club segment members must have visited the client casino at
least once during each of the last four years. Disloyal Player’s Club members
were identiﬁed as members who had participated in the client casino’s program
for over four years, but (1) had no recorded player’s club card activity from
January 1, 2002 to December 15, 2002 or (2) only ﬁve percent or less of their
total lifetime theoretical win was attributable to gaming activity from January 1,
2002 to December 15, 2002. Table 12 provides a summary of the descriptive

statistics for both the Loyal Player’s Club and Disloyal Playefs Club segments.

86

Table 12: Survey Derived Proﬁles for Segments Based on Gaming Activity
at the Client Casino During the Previous 12 months as Compared
to their Lifetime Gambling Behavior at the Client Casino

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I Loyal I Disloyal I ALL
GAMBLING BEHAVIOR
Visited any casino the previous 12
months
Yes 98.2% 96.1% 97.1%
No 1.8% 3.9% 2.9%
Reasons for not visiting a casino in the
previous 12 months
Loss interest in ﬂbllﬂ 0% 1 77.7% 25.0% 1
Loss too much money 60% 66.6% 32.1%
Lack of time 20% 44.4% 17.9%
Not winning 100% 100% 50.0%
Change in family status - 22.2% 7.1%
No casino’s close by 20% 11.1% 7.1%
Health declined - 22.2% 7.1%
Don’t like atmosphere at casinos 60% 22.2% 17.9%
Number of casinos visited the previous 7.46 casinos 6.31 casinos 6.9 casinos
12 months "-
Number of visits to casinos the 2385 trips 20‘“ trips 222 trips
previous 12 months 2
Amount budgeted for each casino visit2 $120.32 $139.39 $129.81
Total amount budgeted for all casino $2.869.63 $2,844.95 $2,881.78
visits the previous 12 months 2
PLAYERS CLUB PROGRAM
Membership in the player’s club
program of their primary casino (the
casino they visited most often in the
previous 12 months)
Yes 98.3% 97.9% 98.1
Reasons for joining their primary
casinos players club
Joined players club to receive promotions3 723% 70.2% 751%
Joined players club to receive comps or 74.3% 76.0% 79.4%
_gifts
Joined players club to get more value for 46.3% 44.9% 48.1%
my money
Joined players club to receive specialized 20.4% 24.8% 23.7%
serVIce
Joined players club card for social reasons 11.6% 13.5% 13.2%

 

1 - Percentage based of those respondents who did not visit a casino in previous 12 months
2- Percentage only based of those persons who visited a casino in previous 12 months
3 - Percentage based of those respondents who did join the players club of its primary casino

87

 

Table 12: Survey Derived Proﬁles for Segments Based on Gaming Activity
at the Client Casino During the Previous 12 months as Compared
to their Lifetime Gambling Behavior at the Client Casino (cont.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Loyal Disloyal ALL
OVERALL USE OF PLAYERS CLUB CARD AT ALL CASINOS
All the time 58.0% 55.6% 56.8%
Most of the time 33.0% 33.7% 33.3%
Sometimes 7.6% 7.2% 7.6%
Rarely 0.6% 2.6% 1.6%
Never 0.4% 0.9% 0.6%
MEMBERSHIP IN OTHER TYPES OF LOYALTY PROGRAMS
Frequent Flyer Program(s) 32.5% 26.9% 29.7%
Grocery Store Program(s) 48.3% 51.8% 50.0%
Car Rental Program(s) 6.0% 3.9% 5.0%
Hotel Program(s) 15.2% 15.3% 15.3%
Retailer Frequent Buyer Program(s) 5.6% 5.6% 5.6%
Music and Book Club(s) 9.2% 11.0% 10.1%
Restaurant PLogram(s) 26.3% 25.5% 25.9%
SOCIOECONOMIC CHARACTERISTICS
£199 64.1 years 62.4 years 63.3 years
21 - 30 0.2% 1.1% 0.6%
31 - 40 1.3% 5.0% 3.1%
41 - 50 10.7% 9.9% 10.3%
51 - 60 23.6% 25.3% 24.4%
61 - 70 36.6% 30.9% 33.8%
Over 70 27.6% 27.9% 27.7%
Gender
Male 31.1% 33.4% 34.1%
Female 68.9% 66.6% 65.9%
Number of Adults in Household 195 adults 1-89 adults 1-91
Number of Homes with Children 7.4% 7.6% 7.6%
Household Income
Under $40,000 45.1% 50.8% 47.8%
$40,001 - $59,999 26.6% 23.0% 24.8%
$60,000 - $99,999 20.6% 18.0% 19.3%
$100,000 - $149,999 7.0% 6.7% 6.8%
$150,000 - $199,999 0.5% 1.2% 0.9%
Over $200,000 0.2% 0.2% 0.3%

 

88

 

Both Loyal Player’s Club and Disloyal Player’s Club segments were active
visitors to casinos. Just a small percentage more of the Loyal Player’s Club
members (98.2%) had visited a casino during the previous 12 months compared
to disloyal segment members (96.1%). For those persons who had not visited a
casino in the previous 12 months, the dominant reasons for both segments were
“they were not winning” and “they had lost too much money.” For the Disloyal
Players Club segment, 77% stated that they became inactive as a result of
losing interest in gambling, compared to 0% of inactive Loyal Player’s Club
segment members. Although there were slight differences between the
responses of these two segments, due to the small size of the inactive Loyal
Player’s Club and Disloyal Player’s Club segments, the validity of any statistical
tests would be questionable.

The only signiﬁcant difference (t = -2.54, p = 0.01) in gaming behavior was
that Loyal Player’s Club members had visited more casinos over the previous 12
months. No statistically signiﬁcant (p < 0.05) differences existed in relation to the
number of trips to casinos or the amount budgeted per trip. Although on average
Loyal Players Club members visited more casinos and made more trips, the
Disloyal Player’s Club segment actually budgeted (set aside more money) per
trip. 80 again, even those classiﬁed as Loyal Players Club members were
distributing the money allocated for gaming across more trips to casinos.

The vast majority of both Loyal Players Club (98.3%) and Disloyal Player’s
Club (97.9%) segments were members of the player’s club for their primary

casino. There were no differences in the reasons why Loyal Player’s Club and

89

Disloyal Player’s Club members joined the player’s club program for their primary
casino since they both primarily joined in order to receive gifts and comps and to
be included in various promotions.

In terms of their socioeconomic characteristics, two-thirds of both segments
were female. The mean age of 64.1 for the Loyal Player’s Club segment and
62.4 for the Disloyal Player’s Club segment suggests that player’s club members
were predominantly seniors. In terms of household income, the largest
percentage for both Loyal Player’s Club and Disloyal Player’s Club segments had
household incomes of under $40,000 per year, which is likely an artifact of
having a large majority of the sample being senior citizens on pensions or ﬁxed
incomes.

It was interesting that there were no statistically significant differences (p <
0.05) between Loyal Player’s Club and Disloyal Player’s Club members, which
poses a particular problem when it comes to designing and targeting of
marketing strategies including communications and promotions.

It is important to iterate that the client casino was frequently not the primary
casino. The client casino was only the primary casino (casino visited most often
in the previous 12 months) for 27.1% of the survey respondents. This is an issue
when it comes to customer value modeling because even though a casino
player’s club member may have a relationship with a casino for many years and
may be characterized as a ‘onal’ member, they may actually do more of their
gaming at another casino. Utilizing information from only one source (e.g.

player’s club tracking data) has the potential for misidentifying loyal and disloyal

90

members since it may, for some members, only track a small portion of their
overall gaming. In other words, player’s club members can be identiﬁed as loyal
customers based on their relative (to other members of the same program)
number of trips they make to a casino, but not be loyal in terms of their allocation

of their total gaming dollars.

Testing of Alternative Loyalty Signiﬁcant Segments

The failure to identify signiﬁcant differences between Loyal Player’s Club and
Disloyal Player’s Club segments raises questions about its use as the basis for
segmentation. Although it has been effectively used in studies of other product
category loyalty programs, its validity in gaming environments had not been
documented in previous studies. Further analysis was therefore directed at
identifying a potentially more distinguishing and marketing exploitable base for
segmenting casino player’s club programs.

The ﬁrst alternative segmentation base tested was the proportion of their total
casino visits made to their primary casino. As indicated in Chapter 2, the most
common method of deﬁning behavioral loyalty is proportion of purchases. In
other words, a person who purchases 100% of a certain product from one
supplier is deemed to exhibit high levels of behavioral loyalty as opposed to a
person who purchases a certain product from multiple suppliers.

Another segmentation base tested is the amount of money budgeted per year
for visits to casinos. As stated in Chapter 1, a common strategy that a number of
casinos utilize in designing their player’s club programs is the use of a tier

system. These tiers are based on the amount gamed each year and members

91

are placed in a tier based on their level of play and are evaluated on an annual
basis in order to determine whether they are promoted to the next tier, remain in

the same tier, or demoted.

Segmentation Based on Proportion of Total Visits to Casinos Taken to
Primary Casino

The descriptive proﬁle of segments based on proportion of visits to their
primary casino is reported in Table 13.

Three segments were formed based on their survey reported percentage of
their casino visits over the previous 12 months to their primary casino. To ensure
there were adequate cases to run structural models for each segment and
accurate comparisons could be made, the cut-off points were selected to ensure
that all three segments were of roughly equal size. Obviously, only persons who
had visited a casino in the previous 12 months could be included in this analysis.
Persons who made 40% or less of their total visits to casinos over the previous
12 months to their primary (most frequently visited) casino were deﬁned as Low
(primary casino) Visitation segment. Persons who directed 70% or more of their
total visits to casinos over the previous twelve month visits to their primary casino
were labeled as High (primary casino) Visitation segment. Moderate (pn'mary
casino) Visitation segment directed 41% to 69% of their total twelve month visits

to casinos to their primary casino.

92

Table 13: Descriptive Proﬁle of Segments Based on the Proportion of Visits
(During the Previous 12 Months) to their Primary Casino

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

40% or less 41 % to 69% 70% or ALL
of visits to of visits to more of
their their visits to
primary primary thelr
casino casino primary
casino
LOW MODERATE HIGH
VISITATION VISITATION VISIT ATION
GAMBLING BEHAVIOR
Visited any casino the previous
12 months
Yes 100% 100% 100% 97.1%
No - - - 2.9%
Number of casinos visited the 8.3 7.7 5.1 6.9
previous 12 months 1 casinos casinos casinos casinos
Number of visits to casinos the 16.0 22-1 28.7 222 trips
previous 12 months 1 “195 trips trips
Amount budgeted for each $138.02 $138.46 $112.41 $129.81
casino visit1
Total amount budgeted for all $2,208.32 $3,059.97 $3,226.17 $2,881.78
casino visits the previous 12
months 1
PLAYERS CLUB PROGRAM FOR PRIMARY CASINO
Membership in the player’s club
program of their primary casino
(the casino they visited most
often in the previous 12 months)
Yes 98.7% 98.8% 98.0% 98.1
No 1.3% 1.2% 2.0% 1 .9%
Reasons for joining their primary
casino’s player’s club
Joined player’s club to receive 73.7% 76.1% 70.8% 75.2%
promos 2
Joined player’s club to receive 78.4% 80.7% 73.8% 79.4%
COITIpS
Joined player’s club to get more 51.3% 49.7% 41.0% 48.1 %
value for my money
Joined player’s club to receive 22.5% 22.7% 24.3% 23.7%
spgialized service
Joined player’s club card for social 11.9% 15.5% 11.5% 13.2%

reasons

 

 

 

 

 

1 - Mean based only of those respondents who had visited a casino in previous 12 months
2 - Percentage based of those respondents who were a member of players club for primary

casino

93

 

Table 13: Descriptive Proﬁle of Segments Based on the Proportion of Visits
(During the Previous 12 Months) to their Primary Casino (cont.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

40% or less 41 °/o to 69% 70% or ALL
of visits to of visits to more of
their their visits to
primary primary their
casino casino primary
casino
LOW MODERATE HIGH
VISITATION VISITATION VISITATION
OVERALL USE OF PLAYERS CLUB CARD AT ALL CASINOS
All the time 57.6% 54.5% 55.7% 56.8%
Most of the time 34.7% 37.6% 30.7% 33.3%
Sometimes 5.2% 6.1% 11.5% 7.6%
Rarely 1.3% 1.6% 1.4% 1.6%
Never - 0.3% 0.7% 0.6%
MEMBERSHIP IN OTHER TYPES OF LOYALTY PROGRAMS
Frequent Flyer Program(s) 29.7% 30.7% 30.8% 29.7%
Grocery Store Program(s) 66.5% 51.9% 39.6% 50.0%
Car Rental Program(s) 7.6% 4.3% 3.9% 5.0%
Hotel Prggram(s) 17.4% 17.4% 14.1% 15.3%
Retailer Frequent Buyer Program(s) 6.4% 5.0% 7.2% 5.6%
Music and Book Club(s) 9.7% 13.4% 10.2% 10.1%
Restaurant Program(s) 31.4% 26.7% 22.3% 25.9%
SOCIOECONOMIC CHARACTERISTICS
#96 62.6 years 62.5 years 63.1 years 63.3
21 - 30 0.9% 1.3% - 0.6%
31 - 40 1.7% 2.2% 4.4% 3.1%
41 -50 11.3% 11.1% 10.5% 10.3%
51 - 60 25.2% 25.7% 26.2% 24.4%
61 - 70 37.4% 36.5% 28.9% 33.8%
Over 70 23.5% 23.2% 29.9% 27.7%
Gender
Male 32.5% 38.7% 33.4% 34.1%
Female 66.1% 61.3% 66.6% 65.9%
Number of AduIts in Household 1.93 adults 1.97 adults 2.01 adults 1.91
Number of Homes with Children 7.9% 71% 65% 75%
Household Income
Under $40,000 49.5% 39.3% 51.6% 47.8%
$40,001 - $59,999 24.0% 26.0% 26.0% 24.8%
$60,000 — $99,999 18.6% 24.2% 15.5% 19.3%
$100,000 - $149,999 5.4% 9.8% 6.2% 6.8%
$150,000 - $199,999 2.0% 0.7% - 0.9%
Over $200,000 0.5% - 0.4% 0.3%

 

94

 

In terms of casino gaming behavior, the Low (primary casino) Visitation
segment visited more casinos on average during the previous 12 months (8.3
casinos) compared to the High (primary casino) Visitation segment (5.1 casinos).
Members of the High (primary casino) Visitation segment completed almost twice
as many visits to casinos during the previous 12 months (28.7 trips) in contrast to
the Low (primary casino) Visitation segment (16.0 trips). Clearly, the High
(primary casino) Visitation segment concentrated more trips to fewer casinos
including their primary casino. This marketing attractiveness was offset partially
because the Low (primary casino) Visitation segment budget $26 more per
casino trip (an average of $138.02 per visit) than the High (primary casino)
Visitation segment (an average of $112.41 per visit). Because the Low (primary
casino) Visitation segment made half as many trips, their annual casino gaming
budget was over $1,000 less than the Moderate (primary casino) Visitation
segment and High (primary casino) Visitation segments. These results provide
support for the ﬁndings of the database analysis completed as part of the ﬁrst
objective of this study which suggest that on a per-visit basis, loyal customers
spend less as opposed to disloyal or infrequent visitors. Although more revenues
were derived from “loyal” player’s club members over the course of a year, the
cost of obtaining this revenue was higher for loyal members as opposed to
disloyal members.

A very high percentage of all three of these segments - Low (primary casino)

Visitation, Moderate (primary casino) Visitation, and High (primary casino)

95

Visitation — were members of their primary casinos player’s club and they joined
primarily to receive comps and gifts and access to promotions.

It was somewhat disappointing, but not unexpected given the results of the
ﬁrst segmentation analysis that there were few differences in terms of overall
player’s club card use, membership in loyalty programs operated in other
industries and socioeconomic characteristics across all three segments. All three
segments had moderate participation rates in other types of loyalty programs and
only 0.6 years separated the mean age for all three segments. All three
segments were roughly two-thirds female and the majority had a household
income of less than $40,000 per year. As a result of this lack of differences, it is
more difﬁcult to target speciﬁc recruiting strategies at members of the three
segments.

So all in all, while it is relatively easy to segment based on proportion of
casino visits made to their primary casino, the information produced is only
moderately useful in terms of the design, targeting, recruitment and retention of

player’s club members.

Segmentation Based on Amount Budgeted for Gaming at Casinos per Year

Members of the client casino who responded to the mail survey were
segmented into three roughly equal segments based on what they reported they
had budgeted for gaming at casinos over the previous 12 months. Again, only
respondents (a) who had visited a casino in the previous 12 months, and (b) that
provided both the amount budgeted per trip and the number of trips taken to

casinos during the previous 12 months were included in this analysis. Of the total

96

982 cases, 813 respondents (83%) met these criteria and were included in the
analysis. The descriptive statistics for these segments are presented in Table
14.

Player’s club members who responded to the survey and budgeted $20 to
$700 per year for casino gaming were identiﬁed as the Low Casino Spending
segment. Conversely, those persons who budgeted over $2100 per year for
gaming at casinos were deﬁned as the High Casino Spending segment. Persons
who budgeted $701 to $2099 per year for visits to casinos were identiﬁed as the
Moderate Casino Spending segment.

Unlike segments that resulted from the other two segmentation bases, there
were statistically signiﬁcant (p < 0.05) differences across the segments including
number of annual visits to casinos, amount budgeted for gaming per casino visit,
and total amount budgeted for casino gaming per year. In terms of value (the
amount budgeted per year for casino visits), one member of the High Casino
Spending segment was equal to seventeen persons located in the Low Casino
Spending segment. The High Casino Spending segment also visited twice as
many casinos in a 12-month period than the Low Casino Spending segment.

Although all three segments were similar with regard to their high participation
in the player’s club program of their primary casino, almost all (99.3%) of the
High Casino Spending segment are members of their primary casinos player’s
club and 87.3% joined in order to receive comps. Comps were a less important
reason for joining a player’s club for the Low Casino Spending and Moderate

Casino Spending segments. In addition, maybe in order to receive more points

97

that can be utilized for comps, almost two-thirds (65.2%) of the High Casino
Spending segment utilized their player’s card all the time compared to 51.7% for
the Moderate Casino Spending segment and 46.0% for the Low Casino
Spending segment.

The High Casino Spending and Moderate Casino Spending segments
generally had higher participation rates in other loyalty programs operated by
other industries including frequent ﬂyer programs, hotel programs, and restaurant
programs. All three segments had a similarly high participation rate in grocery
store programs.

Interestingly, the High Casino Spending segment had a more even proportion
of men and women gamers in contrast to the other two segments. The most
prominent and expected socioeconomic difference was that on average, the High
Casino Spending segment had a higher household income as compared to the
other two segments. The High Casino Spending segment also had the youngest
mean age (61 years) and had fewer children residing in the household as
compared to the other two segments. This suggests that they have the income
and free-time to support their gaming expenditures and frequency of trips to
casinos.

While there were only a few important behavioral and socioeconomic
differences identified between the segments, differences in the mean loyalty
factor scores provided further insight into the effectiveness of these segmentation

bases for their use in studying gaming loyalty programs.

98

Table 14: Proﬁle of Segments Formed Based on Total amount Budgeted
per Year for Casino Gaming

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$20 to $701 to $2100 or ALL
5700 $2099 per more per
per year year year
budgeted budgeted budgeted
for casino for casino for casino
gaming gaming gaming
LOW MODERATE HIGH
CASINO CASINO CASINO
SPENDING SPENDING SPENDING
GAMBLING BEHAVIOR
Visited any casino in previous 12
months
Yes 100% 100% 100% 97.1%
Number of casinos visited the 4.26 6.69 9.78 6.9
previous 12 months 1 casinos casinos casinos casinos
Number of visits to casinos the 8:23 18_-33 39.84 222 trips
previous 12 months 1 "'95 "'95 "'95
Amount budgeted for each casino $65.21 $108.73 $217.06 $129.81
visit 1
Total amount budgeted for all $536.38 $1,993.02 $8,647.67 $2,881.78
casino visits the previous 12
months 1
PLAYERS CLUB PROGRAM FOR PRIMARY CASINO
Membership in the player’s club
program of their primary casino
(the casino they visited most often
in the previous 12 months)
Yes 96.6% 98.5% 99.3% 98.1
No 3.4% 1.5% 0.7% 1 .9%
Reasons for joining their primary
casino’s player’s club
Joined player’s club to receive 67.6% 72.6% 76.4% 75.2%
promos 2
Joined player’s club to receive 62.5% 78.6% 87.3% 79.4%
comps
Joined player’s club to get more 36.4% 46.2% 54.9% 48.1%
value for my money
Joined player’s club to receive 15.4% 19.2% 33.1% 23.7%
specialized service
Joined player's club card for social 16.5% 13.2% 9.1% 13.2%

reasons

 

 

 

 

 

1 - Mean based only on those respondents who had visited a casino in previous 12 months
2 - Percentage based of those respondents who were a member of primary casinos players club

99

 

Table 14: Descriptive Proﬁle of Segments Formed Based on Total amount
Budgeted per Year for Casino Gaming (cont.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$20 to $701 to $2100 or ALL
5700 $2099 per more per
per year year year
budgeted budgeted budgeted
for casino for casino for casino
gaming gaming gaming
LOW MODERATE HIGH
CASINO CASINO CASINO
SPENDING SPENDING SPENDING
OVERALL USE OF PLAYERS CLUB CARD AT ALL CASINOS
All the time 46.0% 51.7% 65.2% 56.8%
Most of the time 41.8% 35.9% 29.3% 33.3%
Sometimes 9.1% 10.0% 4.8% 7.6%
Rarely 1.9% 1.9% 0.7% 1.6%
Never 1.1% 0.4% - 0.6%
MEMBERSHIP IN OTHER TYPES OF LOYALTY PROGRAMS
Frequent Flyer Program(s) 23.2% 31 .2% 34.5% 29.7%
Grocery Store Program(s) 47.1% 57.9% 50.9% 50.0%
Car Rental Program(s) 2.6% 6.8% 6.2% 5.0%
Hotel Progﬂnﬁ) 9.9% 19.2% 18.9% 15.3%
Retailer Frequent Buyer Program(s) 4.8% 7.1% 4.4% 5.6%
Music and Book Club(s) 6.6% 12.4% 13.8% 101%
Restaurant Prgggms) 25.7% 27.8% 26.5% 25.9%
SOCIOECONOMIC CHARACTERISTICS
Age 63.2 years 63.3 years 61.0 years 63.3
21 - 30 0.8% 0.4% 1.1% 0.6%
31 - 40 5.7% 0.8% 3.0% 3.1%
41 -— 50 12.5% 10.5% 11.9% 10.3%
51 — 60 17.9% 26.5% 31.9% 24.4%
61 — 70 31.2% 38.5% 33.0% 33.8%
Over 70 31 .9% 23.3% 19.3% 27.7%
Gender
Male 32.2% 35.3% 40.4% 34.1%
Female 67.8% 64.7% 59.6% 65.9%
Number of Adults in HousehoId 1.83 adults 1.95 adults 2.01 adults 1.91
Number of Homes with Children 119% 60% 65% 75%
Household Income
Under $40,000 56.2% 43.2% 38.1% 47.8%
$40,001 - $59,999 23.0% 28.8% 24.2% 24.8%
$60,000 - $99,999 16.2% 23.1% 20.8% 19.3%
$100,000 - $149,999 4.7% 3.9% 14.0% 6.8%
$150,000 - $199,999 - 0.9% 2.1% 09%
Over $200,000 - - 0.4% 0.3%

 

 

 

 

 

IOO

 

Conﬁrmatory Factor Analysis

Before the Integrated Casino Loyalty Model was tested and before using the
loyalty model constructs to distinguish between segments, the validity and
reliability of these constructs ﬁrst needed to be tested. Although validity and
reliability measures for the majority of the constructs were available in literature,
the measures had to be adapted for use in a gaming environment. Measures for
the construct of rewards were not available in literature so they were based on
previous studies completed by the client casino. Because there is support in the
literature for the majority of the loyalty constructs, a conﬁrmatory factor analysis
was utilized to test the constructs. Conﬁrmatory factor analysis has proven to be
quite effective in the validation of constructs in previous studies utilizing structural
equation modeling. The constructs that were tested as part of the conﬁrmatory
factor analysis included the components of the loyalty program (direct mail,
preferential treatment, interpersonal communications, promotions and discounts,
rewards), antecedents of attitudinal loyalty (afﬁnity, consumer conﬁdence),
moderating and mediating factors (competitive factors, involvement, relationship
commitment). Although modeled as latent constructs, perceived value, attitudinal
loyalty and behavioral loyalty were also included in the analysis in order to
identify how segments differed in terms of their mean summed scores.

Factors were derived through principal axis factoring and were rotated using
the varimax method. The number of factors was determined using a 1.00 eigen-
value cut-off. Based on the results of the conﬁrmatory factor analysis, none of the

individual measures were removed from the analysis. In terms of Cronbach’s

lOl

alpha scores, twelve of the thirteen constructs had scores above 0.70 and the
one remaining construct (Relationship Commitment) had an acceptable score of
0.64. The results of the conﬁrmatory factor analysis are included in Table 15.

Factors representing the components of casino loyalty programs (Direct Mail,
Preferential Treatment, Interpersonal Communications, Promotions and
discounts, Rewards) all had Cronbach’s alpha scores greater than 0.80,
suggesting a high degree of reliability. With only one exception (direct mail), the
means of the summed loyalty factor scores were less than 4.00 for all constructs.
This suggests that the player’s club members viewed the components of player’s
club programs as having only a moderate effect, thus questioning the impact of
loyalty programs on player’s club members who participated in this study.

The two antecedents to perceived value (Afﬁnity and Consumer Conﬁdence)
had high Cronbach’s alpha scores. The score for afﬁnity was 0.80 and 0.85 for
consumer conﬁdence. The mean scores for both of these loyalty factors were just
above 4, which again suggests that respondents only show moderate levels of

trust and conﬁdence in their primary casino’s player’s club programs.

i02

Table 15: Results of Conﬁrmatory Factor Analysis of Potential Loyalty

Model Factors

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Construct Mean Standard Factor
Deviation Loadings
v1 — Direct Mail (0t=.96)
I receive mailings about the loyalty program on a 4.68 2.27 .953
regular basis
I receive information about promotions through these 4.85 2.21 .981
marlin s
I receive information about special events through 4.88 2.19 .976
these mailing_s
V2 — Preferential Treatment (or=.83)
I receive special treatment because I am a member 304 2.12 .924
of the players club
Casino staff offers me more personalized service 2.40 1.90 .924
because I am a member of the plgyer's club
V3 — Interpersonal Communications (a=.94)
Casino staff takes the time to personally get to know 2.40 1.87 .975
me because I am a member of the player's club
Casmo staff often hold personal conversations with 2.28 1.83 .975
me because I am a member of the player’s club
V4 — Promotions & Discounts (or=.86)
I often receive discounts because I am a member of 3.85 2.35 .888
the player's club
I often get invited to special events because I am a 3.57 2.28 .897
member of the plgyer's club
I often participate in promotions that are only 3.36 2.26 .856
available to player's club members
V5 — Rewards (or=.82)
I often receive rewards/gifts (clothing, hotel rooms, 3.31 2.40 .862
tickets) because I am a member of the player’s club
As my level of gaming increases, the more 3.52 2.32 .865
gifts/rewards I receive from the casino
I have used my points or comp balance to purchase 4.61 2.38 .710
gifts, tickets, or cashback
I often receive comps (free food, drinks) during my 4.11 2.49 .788
visits because I am a player’s club member
vs — Afﬁnity (0t=.80)
I feel good when I visit this casino 4.54 1.80 .890
Being part of the player’s club makes me want to visit 3.62 2.08 .749
this casino more often
Visiting this casino gives me pleasure 4.65 1.78 .890

 

Scale (1 = Do not agree to 7 = Strongly agree)

103

 

Table 15: Results of Conﬁrmatory Factor Analysis of Potential Loyalty

Model Factors (cont.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Construct Mean Standard Factor
Deviation Loadings
V7 — Consumer Conﬁdence (ct=.85)
I trust this casino 4.05 1.92 .925
I believe that the casino is honest in its business with 4.27 1.90 .926
me
I rely on this casino for my gaming 4.10 1.89 .776
V8 — Perceived Value (or=.93)
The casino makes great efforts to retain its players 3.55 2.09 .899
club members
The casino makes great efforts to improve its ties 3.32 2.00 .935
with player’s club members
The casino really cares about keeping its player’s 3.31 2.00 .935
club members happy
The casino asks for input and ideas from player’s 2.66 1.90 .787
club members
The casino continually enhances the beneﬁts of the 3.09 1.99 .876
player's club program
V9 - Competitive Factors (0L=.77)
Compared to other casinos, I receive good value for 4.17 1.94 .879
m mone
I feel that I receive more value for my money by 4.05 2.13 .877
being a member of the player’s club
Compared to other player’s club, this program offers 3.85 2.08 .913
me more value for my money
V10 — Involvement (or=.70)
I am someone who finds it important about where I 3.89 2.07 .788
gamble
I am someone who is interested In learning about 3.38 2.08 .839
gambling and different gambling opportunities
I am someone for whom it means a lot to gamble 2.80 1.81 .756
V11 — Relationship Commitment (or=.64)
I am someone who likes to be a regular customer for 3.13 1.96 .846
a casino
I am someone who likes building relationships with 3.26 2.00 .775
companies
I am willing to go “the extra mile” to remain a 3.55 2.01 .659
customer of this casino

 

Scale (1 = Do not agree to 7 = Strongly agree)

104

 

Table 15: Results of Conﬁrmatory Factor Analysis of Potential Loyalty
Model Factors (cont.)

 

 

 

 

 

 

 

 

 

 

 

Construct Mean Standard Factor
Deviation Loadings

V12- Attitudinal Loyalty (a=.74)

I am committed to this casino 2.99 2.00 .844

I would be willing to still visit this casino if it was more 3.12 1.93 .816

difficult to reach

I tell others about the benefits ofjolnlng the player‘s 3.68 2.30 .772

club
V13 — Behavioral Loyalty (or=.81)

I will visit this casino on my next trip 4.67 2.05 .916

I intend to continue to be a customer of this casino 5.09 1.83 .916

 

 

 

 

Scale (1 = Do not agree to 7 = Strongly agree)

In terms of the moderating and mediating factors (Competitive factors,
Involvement, Relationship commitment), competitive factors and involvement had
Cronbach’s alpha scores greater than 0.7 and relationship commitment had a
score of 0.64. The mean scores for all three of the constructs was below 4,
suggesting that the player’s club members (at least the ones that completed a
mail survey) perceived that the player’s club program of their primary casino was
only moderately competitive as compared to alternative programs. The low mean
scores for involvement and relationship commitment were likely due to the forced
recruitment practices of gaming organizations and the inclusion of less “serious"
gamers within casino player’s club programs.

Three factors (Perceived Value, Attitudinal Loyalty, and Behavioral Loyalty)
were modeled as latent factors. Although not directly measured in terms of the

model, having information of the mean scores for these factors was useful for

105

 

evaluating the segmentation bases and how the segments differed in terms of
these major factors of interest. All three had Cronbach’s alpha scores greater
than 0.70 (0.93, 0.74, and 0.81 respectively), which suggest a high level of
validity. Using a seven-point Likert scale (1 being strongly disagree to 7 being
strongly agree), the mean scores for respondents included in this study were
similar to those found in previous loyalty studies. Both perceived value and
attitudinal loyalty were quite low (less than 3.5) which suggests that very few
player’s club members exhibited attitudinal loyalty and they viewed the value
derived through their membership in the loyalty program of their primary casino
as being quite limited. These ﬁndings are not surprising given the low loyalty
factor scores for loyalty program components and their subsequent use as
antecedents of attitudinal loyalty. Of interest is that mean scores for the
individual measures of behavioral loyalty were just below 5.0, which suggests
that player’s club members exhibited moderate levels of behavioral loyalty
towards their primary casino.

In summation, the loyalty factors included in the Integrated Casino Loyalty
Model exhibited high levels of validity. Of interest is that with the exception of
behavioral loyalty, the mean loyalty factor scores for all of the indicators were
centered on 4.00. Since a seven-point Likert scale was used, this suggests that
the respondents perceived the factors included in this study only moderately.
These low mean scores and the subsequent higher mean score for behavioral

loyalty may suggest that additional factors (e.g. satisfaction, convenience loyalty)

106

may be involved in the development of behavioral loyalty but are not being

directly tested within the hypothesized model.

Differences of Factor Scores for Loyalty Segments

As discussed earlier in this chapter, there were few distinguishing behaviors
or socioeconomic characteristic differences based on the proposed segmentation
base and the ﬁrst alternative segmentation base (proportion of total trips to
casinos made to primary casino). While there were some important behavioral
and socioeconomic differences identified between segments based on amount
budgeted per year for casino gaming, more robust proﬁles would assist in
designing more effective casino player’s club programs. Furthermore, in order to
further test the segmentation base proposed in this study, the mean loyalty factor
scores for the two segments (Loyal Player’s Club members and Disloyal Players
Club members) are presented in Table 16.

Table 16 shows that there were a few important statistically signiﬁcant
differences between player’s club segments formed based on the data collected
as part of client casinos players club tracking system. The Loyal Players Club
segment (those members who had visited the client casino at least once each of
the four previous years) and Disloyal Player’s Club segment (those members
who had been a member of the client casinos player’s club for over four years
and either had no recorded activity from January 1, 2002 to December 15, 2002
or only five percent or less of their total lifetime theoretical win was represented
by gaming activity from January 1, 2002 to December 15, 2002 at the client

casino) had statistically signiﬁcant differences (p < 0.05) for promotions and

107

discounts available to player’s club members at their primary casino (promotions
and discounts), perceived value of being a member of the player’s club for their
primary casino (perceived value), and their comparison of the player’s club of
their primary casino compared to other casino’s players clubs (competitive
factors). Of interest is that the mean loyalty factor scores for the Disloyal Players
Club segment were actually higher than those for the Loyal Player's Club
segment. Since one would assume the opposite to occur, this further provides
support that this segmentation method actually leads to the misidentiﬁcation of
loyal customers and is not an effective segmentation method for use by gaming
organizations.

Although the proposed segmentation base was able to delineate a number of
signiﬁcant differences between segments, the need to further test alternative
segmentation bases could potentially highlight further differences between
segments within casino player’s club programs and aid casinos in their marketing

efforts.

108

Table 16: A Comparison of Mean Loyalty Factor Scores for Segments
Based on Gaming Activity at the Client Casino During the
Previous 12 months as Compared to their Lifetime Gambling
Behavior at the Client Casino

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Loyal Player’s Disloyal
Club Members Player’s Club
Members
Mean SD. Mean 8.0. t p
Direct Mail 4.68 2.18 4.96 2.12 1.93 0.054
Preferential 2.67 1.81 2.85 1.96 1.36 0.174
Treatment
Interpersonal 2.29 1.74 2.40 1.87 0.85 0.398
Communications
Promotions & 3.48 2.01 3.76 2.05 2.04 0041*
Discounts
Rewards 3.84 1.95 3.94 1.92 ' 0.71 0.478
Afﬁnity 4.18 1.56 4.38 1.63 1.81 0.070
Consumer ' 4.09 1.66 4.21 1.69 1.07 0.284
Conﬁdence
Perceived Value 3.07 1.76 3.34 1.79 2.15 0032*
Competitive 3.89 1.56 4.13 1.57 2.17 0030"
Factors
Involvement 3.28 1.57 3.47 1.58 1.89 0.071
Relationship V 3.26 1.52 3.31 1.53 0.55 0.581
Commitment
Attitudinal 3.19 1.67 3.32 1.71 1.15 0.250
Loyalty
Behavioral 4.79 1.83 4.95 1.75 1.30 0.193
Loyalty
*p<005

Scale (1 = Do not agree to 7 = Strongly agree)

In regards to the ﬁrst alternative segmentation base, the segments formed
based on the proportion of total visits to their primary casino (most often visited)
again identiﬁed very few differences based on the mean scores for the loyalty
factors. The means scores for each loyalty factor for the three segments are

provided in Table 17.

109

Table 17: Comparing Mean Loyalty Factor Scores Between Segments
Based on Percentage of Total Visits to Casinos taken to Primary

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Casino
Visit primary Visit primary Visit primary
casino 40% or casino 41% to casino 70% or
less 69% more
LOW MODERATE HIGH
Mean SD. Mean SD. Mean 8.0.

Direct Mail 4.98 1.93 4.96 2.12 4.58 2.31
Preferential 2.68 1.83 2.85 1.91 2.72 1.87
Treatment
Interpersonal 2.25 1.76 2.35 1.79 2.39 1.83
Communications
Promotions 8 3.57 2.05 3.72 2.00 3.55 2.06
Discounts
Rewards 3.95 1.83 3.99 1.90 3.71 2.05
Afﬁnity 4.22 1.61 4.44 1.54 4.20 1.61
Consumer 3.97 1.64 4.25 1.64 4.20 1.74
Conﬁdence
Perceived Value 3.24 1.71 3.29 1.75 3.07 1.80
Competitive 3.97 1.51 4.07 1.54 4.01 1.63
Factors
Involvement 3.37 1.43 3.50 1.53 3.22 1.63
Relationship 3.28 1.47 3.37 1.52 3.20 1.56
Commitment
Attitudinal 3.17 1.54 3.37 1.67 3.22 1.74
Loyaﬂy
Behavioral 4.52 1.82 4.97 1.71 5.05 1.83
Loyalty

 

 

 

 

 

 

 

 

Scale (1 = Do not agree to 7 = Strongly agree)

All three segments (Low (primary casino) Visitation, Moderate (primary
casino) Visitation, and High (primary casino) Visitation) perceived the
components of the loyalty program (Direct mail, Preferential Treatment,

Interpersonal Communications, Promotions and Discounts, and Rewards) only

110

moderately since only direct mail had a mean score higher than 4.0 (based on a
seven-point Likert scale with 1 being strongly disagree and 7 being strongly
agree). Maybe as a direct result of these low scores, all three segments exhibited
low levels of attitudinal loyalty (Low (primary casino) Visitation = 3.17, Moderate
(primary casino) Visitation = 3.37, and High (primary casino) Visitation = 3.22)
and perceived value (Low (primary casino) Visitation = 3.24, Moderate (primary
casino) Visitation = 3.29, and High (primary casino) Visitation = 3.07).
Surprisingly, the Moderate (primary casino) Visitation segment actually had the
highest mean scores for the loyalty factors of perceived value and attitudinal
loyalty.

Independent sample t-tests were used to identify statistically signiﬁcant
differences between the segments. In comparing Low (primary casino) Visitation
and High (primary casino) Visitation segments, signiﬁcant differences were found
only for the perceived assessment of the amount of direct mail sent to player’s
club members (t = 2.15, p < 0.03) and their level of behavior loyalty (t= -3.14, p <
0.002). Of signiﬁcant interest is that although the Low (primary casino) Visitation
segment had higher mean scores for seven loyalty factors (including three of the
ﬁve components of the loyalty program) as compared to the High (primary
casino) Visitation segment, the High (primary casino) Visitation segment had
higher perceived levels of attitudinal loyalty (3.22) and behavioral loyalty (5.05)
as compared to the Low (primary casino) Visitation segment scores of 3.17 for

attitudinal loyalty and 4.52 for behavioral loyalty. These ﬁndings suggest that

111

items not included in the loyalty model may have a role in the development of
behavioral loyalty.

In comparing the Low (primary casino) Visitation and Moderate (primary
casino) Visitation segments, the only signiﬁcant difference was for behavioral
loyalty (t= -2.81, p < 0.001). In comparing the Moderate (primary casino)
Visitation and High (primary casino) Visitation segments, the only signiﬁcant
difference was the assessment of direct mail sent to player’s club members by
their primary casino (t= 2.10, p < 0.04). As a result of the lack of differences
found between these segments, this further exempliﬁes that the traditional
behavioral loyalty segmentation base (proportion of purchases) is ineffective at
identifying segments of a casino player’s club program.

The second alternative segmentation base tested as part of this study was
based on the total amount budgeted for gaming at casinos per year. As indicated
earlier, this segmentation base proved quite effective at identifying both
behavioral and socioeconomic differences between segments. Through the use
of independent sample t-tests, signiﬁcant differences were found between the
three segments. The mean scores for the Low Casino Spending, Moderate
Casino Spending, and High Casino Spending segments are presented in Table
18.

In comparing the Low Casino Spending and High Casino Spending
segments, signiﬁcant differences were found for all factors (p < 0.05) with the
exception of the perceived level of interpersonal communications that staff of the

primary casinos have with player’s club members as opposed to non-members.

112

Furthermore, as one would expect, mean scores for the High Casino Spending
segment were higher as compared to the mean scores for the Low Casino
Spending segment.

In comparing the Low Casino Spending segment to the Moderate Casino
Spending segment, signiﬁcant differences were found for the perceived amount
of direct mail (t = - 2.76, p < 0. 001), the availability of rewards (t = -3.38, p <
0.001), the level of involvement that player’s club members had in regards to
casino gaming (t = -2.49, p < 0.01), and their perceived level of behavior loyalty (t
= -2.76, p < 0.006). The signiﬁcant difference for direct mail and rewards is most
likely due to the fact that Low Casino Spending segment do not often have
access to rewards and special events due to their low level of gaming.

In comparing the Moderate Casino Spending and High Casino Spending
segments, signiﬁcant differences were found for four of the ﬁve components of
the loyalty program. These four components included the assessment of direct
mail (t = -3.23, p < 0.001), level of interpersonal communications (t = -2.00, p <
0.04), availability and frequency of promotions and discounts (t = -3.45, p <
0.001), and the perception of the quality of the rewards made available to
player’s club members (t = —4.10, p < 0.001). These ﬁndings show support that
the current practice of casinos focusing on the High Casino Spending segment
actually has an effect on the gaming behavior of these player’s club members —
albeit a moderate one. Statistically signiﬁcant differences were also identiﬁed for
these two segments for their perceived level of relationship commitment (t = -

3.60, p < 0.001) and their levels of attitudinal loyalty (t = -2.88, p < 0.004).

113

Of signiﬁcant interest is that all three segments again rated the components
of the loyalty program quite low, especially interpersonal communications. This
suggests that although gaming organizations are using these programs as a key
component of their marketing strategy, player’s club members included in this
study perceived the impact of these programs only moderately. As a result of
these low scores, alternative marketing strategies should be identiﬁed and tested

to determine their potential in marketing to these segments.

114

Table 18: A Comparison of Mean Loyal Factor Scores for Groups Based on

Total Amount Budgeted for Casino Gaming per Year

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$20 to $700 $701 to $2099 $2100 or more per
per year per year year
LOW MODERATE HIGH

Mean SD. Mean SD. Mean 8.0.
Direct Mail 4.19 2.15 4.47 2.14 5.33 1.99
Preferential 2.44 1.76 2.71 1.79 3.00 2.00
Treatment
Interpersonal 2.26 1.86 2.25 1.67 2.53 1.92
Communications
Promotions 8 3.15 1.91 3.51 2.03 4.09 2.00
Discounts
Rewards 3.21 1.78 3.80 1.84 4.46 1.94
Afﬁnity 4.01 1.50 4.28 1.64 ' 4.48 1.58
Consumer 3.89 1.60 4.15 1.71 4.28 1.65
Conﬁdence
Perceived Value 3.06 1.79 3.15 1.74 3.38 1.78
Competitive 3.71 1.52 3.99 1.57 4.30 1.55
Factors
Involvement 2.95 1.54 3.31 1.46 3.80 1.54
Relationship 2.92 1.44 3.20 1.51 3.66 1.51
Commitment
Attitudinal 2.89 1.69 3.18 1.64 3.56 1.65
Loyalty -5
Behavioral 4.50 1.82 4.95 1.75 5.14 1.68
Loyalty

 

 

 

 

 

 

 

 

Scale (1 = Do not agree to 7 = Strongly agree)

In summation, the most effective segmentation method (in terms of
identiﬁcation of market segments) is the method of segmentation that is based on
the amount budgeted for casino visits per year. Although this may be a by-
product of the current practices of gaming organizations and their use of tier
systems, it could also be a result of the ineffectiveness of player’s club programs

in serving lower spending customers. Although the rate varies across casinos, it

115

is generally assumed that one should never provide more than 10% of players’
theoretical win as part of a player’s club program. In other words, for every $100
gambled, the casino will put aside $0.60 for that person in terms of player’s club
rewards or cash back. This is not a signiﬁcant amount when compared to the

amount of money needed to earn this comp balance.

Testing of Integrated Casino Loyalty Model

Although very few differences (behavioral, socioeconomically, and loyalty
factor scores) were identiﬁed between segments formed with the ﬁrst two
segmentation bases, there may be differences with regard to how behavioral and
attitudinal loyalty is actually developed within these segments. Testing the
Integrated Casino Loyalty Model allowed for a strong test of these segmentation
bases as well as provided insight into how attitudinal and behavioral loyalty is
developed in regards to these segments.

As stated previously, the proposed Integrated Casino Loyalty Model was used
as the framework for this test. Building on past loyalty studies, the model was
developed to take into account special aspects of casino gaming and player’s
club programs. The following components of player's club programs were
included: Direct mail, preferential treatment, interpersonal communications,
promotions and discounts, and rewards. Since loyalty programs have been
identiﬁed as value-added programs, the components of the player’s club were
modeled as antecedents of perceived value. Through the inclusion of afﬁnity and

consumer conﬁdence as part of this model, a detailed analysis of how attitudinal

116

loyalty is created and its subsequent relationship with perceived value and
behavioral loyalty was possible. Finally, the use of behavioral loyalty as the
outcome variable was supported by the fact that the majority of organizations
utilize loyalty programs as a method for increasing usage and/or market share.

When testing latent structural models, the ﬁrst step is to the test the
measurement model (no paths between latent variables). The purpose of the
measurement model is to determine whether each indicator variable is loading
onto their speciﬁed latent variables. Once the measurement model has reached
an acceptable level of ﬁt in regards to the data (structural equation modeling
uses covariance matrixes), the full hypothesized structural model (paths between
latent variables and the introduction of mediating and moderating variables) is
tested to determine its ﬁt relative to the data being tested. As is the case in most
modeling efforts, the hypothesized model is often modiﬁed using the results of
the Lagrange Multiplier test. The Lagrange Multiplier tests provide suggestions
for model modiﬁcations to improve the chi-square/degrees of freedom ratio, thus
attempting to improve the ﬁt of the model relative to the data (mail survey
respondents for whom mean scores could be calculated for all loyalty factors).
When utilizing the results of the Lagrange Multiplier test, any modiﬁcations
should be not only based on statistical results, but should also make sense
theoretically.

Previous studies have raised questions related to the reliability of using chi-
square tests as the sole indicator of model ﬁt (Fan, Thompson and Wang, 1999).

Recognizing the need for other measures of ﬁt, a number of ﬁt indices have been

117

developed and are now included in all structural equation modeling statistical
programs. The most commonly accepted ﬁt indices include the Comparative Fit
Index (CFI; Bentler, 1990), Incremental Fix Index (IFI; Bollen, 1989), and
Standardized Root Mean Squared Residual (SRMR). Cut-off values for CF I and
IFI of these indices are set at 0.90 and SRMR is below 0.05 (Bentler, 1990;
Hoyle and Panter, 1995).

The most cited alternative ﬁt index is the Root Mean Square Error of
Approximation (RMSEA) which focuses on the degree of ﬁt between the data
being tested and the proposed model. According to MacCallum et al. (1996), the
RMSEA value should be small (preferably in the area of 0.06 or lower), but a
score of less then 0.08 has been generally accepted in literature (MacCallum et
aL,1996)

As stated previously, the ﬁrst step was the testing of the hypothesized
Integrated Casino Loyalty Model. Testing showed that the ﬁt of the hypothesized
measurement model was acceptable based on the results of the ﬁt indices
(Comparative Fit Index = 0.97, Incremental Fit Index = 0.97, Standardized Root
Mean Squared Residual = 0.03). The test also indicated that the x2 (111.03)
divided by the degrees of freedom (9) resulted in a score of 12.3. This score
suggests that the likelihood of ﬁnding these results in a general population is
quite unlikely (p = 0.0001), thus rejecting the model. In addition, the test of the full
structural model resulted in a signiﬁcant number of errors (error terms variances
set to zero, linear dependence between variables, and disturbance terms set to

correlate). These errors suggest that indicator variables were loading on multiple

118

factors, resulting in problems of multicollinearity. The culmination of these errors
made the results un-interpretable.

As a result of these issues, an alternative model was developed based on the
results of the Lagrange Multiplier test and from reviewing past literature. The
results of the Lagrange Multiplier test suggested the removal of direct mail,
preferential treatment, and competitive factors since these indicators loaded on
multiple factors. Although of major interest, competitive factor loaded on multiple
factors and was linearly dependant on multiple factors, thus reducing the
reliability of the results of the model testing effort. The results also proposed
placing involvement and relationship commitment as indicator variables for
behavioral loyalty and removing the paths between these two variables and that
of the two latent constructs of perceived value and attitudinal loyalty. Since these
variables were modeled as moderator variables, these changes were completed
by removing the paths between involvement and relationship commitment and
both perceived value and attitudinal loyalty. As a result of these modiﬁcations, a

proposed alternative model was developed and is presented in Figure 7.

119

Figure 7: Elements of the Alternative Integrated Casino Loyalty Model

 

 

    
 

 

 

 

 

 

   

 

 

 
   
 

 

  

 

 
   
 

 

  
  

 

 

 

 

 

 

  
     
 

 

 

 

 

 

 

 

V1
INTERPERSONAL V6
COMMUNICATION
F1 INVOLVEMENT
PERCEIVED
V2 VALUE
REWARDS
F3
BEHAVIORAL
v3 LOYALTY
PROMOTIONS 8:
DISCOUNTS
V4 F2 V7
AFFINITY ATTITUDINAL RELATIONSHIP
LOYALTY COMMITMENT
V5
CONSUMER
CONFIDENCE

 

 

Model Reﬁnements

The results of the measurement model of the Alternative Integrated Casino
Loyalty Model are presented in Figure 8. The test of the measurement model
includes the estimation of the reliability coefﬁcients of the constructs, as well as

the determination of the convergent and discriminant validity of the instrument.

120

Figure 8: Measurement Model For Alternative Integrated Casino Loyalty
Model ( N = 806 Cases)

 

 

   
 

 

 

 

   

 

 

   
   

 

 

 

   

 

 

 

 

 

 
    
   

  

 

 

 

 

 

 

 

 

 

    
     

 

 

    

 

   
 

 

 

 

V1
079 _. INTERPERSONAL 0.61 (FIXED
COMMUNICATION
F1
PERCEIVED
V2 VALUE
0-64 -> REWARDS
V3
0.82 -> PROMOTIONS& .56 (HXED)
DISCOUNTS
V4 0 82 F2
- ATTITUDINAL
0.37 —> AFFINITY ._ LOYALTY
V5
0.61 —9 CONSUMER
CONFIDENCE
vs 0.69 (FIXED)
073 _. INVOLVEMENT
F3
BEHAVIORAL
LOYALTY
V7
022 RELATIONSHIP L40
' COMMITMENT -
MODEL FIT INDICES
x2/df 4.6
RMSEA 0.06
CFI 0.98
IFI 0.98
SRMR 0.03

 

 

 

Standardized coefﬁcients, covariance, and error terms shown
All paths are significant at p < 0.05

Bagozzi and Yi (1991) suggested two methods for assessing convergent
validity. Convergent validity is achieved if each indicator (e.g. Direct Mail,
Rewards and discounts, Afﬁnity, etc.) loads signiﬁcantly on its respective latent
factor. From reviewing the results Of the measurement model test, all indicators
loaded signiﬁcantly on their respective latent variables, thus suggesting that
convergent validity was achieved. A second and more stringent test Of
convergent validity should be utilized with sample sizes of 1,500 or more cases.
With this test, convergent validity is achieved when the squared factor loading is
greater than 0.5. Although the number of cases included in the testing of the
measurement model was only 806 cases, of the seven variables included in the
measurement model, ﬁve of the seven variables had squared factor loadings
greater than 0.5 which suggests that convergent validity was achieved (Bagozzi
and Yi, 1991).

Discriminant validity is the degree to which items differentiate among
constructs or measure distinct concepts. TO assess discriminant validity, it
requires examining variables included in the models (both measurement and
structural) to ensure that they are not perfectly correlated (correlation equal to 1).
For the Alternative Integrated Casino Loyalty Model, all correlations between
variables are signiﬁcantly different (p <05) and no correlations are higher than
the recommended 0.85 level. This suggests that discriminant validity was

achieved and that all Of the variables and factors being modeled are in fact

distinct.

122

In order to assess the validity of the results of the Conﬁrmatory Factor
Analysis (CFA), multiple goodness-Of-ﬁt measures were utilized to determine
how well the proposed alternative measurement model ﬁt in relation to the data
(Mail survey respondentS’ for whom mean scores could be calculated for all
loyalty factors). Obviously, if the results of the ﬁt indices are quite strong, this
suggests that the model being tested is effective at explaining the data.

As indicated in Figure 8, the ratio of chi-square to degree of freedom is 4.6.
As suggested by Bollen (1989), the ratio of chi-square to degrees of freedom
should exceed 3. Since our ratio exceeds this threshold, the alternative
measurement model was statistically signiﬁcant from zero (p < 0.05). Any ratio
below 3 suggests that the model may not be statistically different from zero and
that the model may be under-identiﬁed or not measuring all of the factors
included in model (Kline, 1998). In relation to the ﬁt indices, the measurement
model showed that the model exceeds all of the cutoff values (Conﬁrmatory
Factor Index= 0.98, Incremental Fit Index= 0.98, and Standardized Root Mean
Squared Residual= 0.003). In terms of the alternative fit index, the Root Mean
Square Error of Approximation (RMSEA) was .06, which is below the acceptable
cutoff. The results of these scores suggest that the measurement model Of the
Alternative Integrated Casino Loyalty Model ﬁt the data quite well.

The results of the test Of the full structural model for all 806 player’s club
members who responded to the survey and that were included in the model
testing portion of this study are indicated in Figure 9. All model coefﬁcients were

significant at p < 0.05. Although both effects were moderate, perceived value had

a greater effect (0.38) than attitudinal loyalty (0.29) in predicting behavioral
loyalty. The correlation between perceived value and attitudinal loyalty is quite
strong with a correlation coefﬁcient of 0.73, which suggest that although these
two latent constructs are distinct, they are highly correlated and may (in future
cases) be non-distinct (if the correlation exceeds 0.90).

Each of the exogenous variables loaded signiﬁcantly on its related latent
variable. Both interpersonal communications and rewards had moderate effects
on perceived value with path coefﬁcients Of 0.61 and 0.77 respectively. In relation
to attitudinal loyalty and as found in previous studies, both afﬁnity and consumer
confidence were signiﬁcant predictors Of attitudinal loyalty. Although to a lesser
degree, promotions and discounts had a moderate effect on attitudinal loyalty. As
one would expect in a loyalty program setting, relationship commitment was a
strong predictor of behavioral loyalty. In summation, with path coefficients
ranging from 0.52 to 0.90, this suggests that the indicator variables are good
predictors of the latent constructs. A supplementary W-test was performed to test
for co-linearity because of the strength Of the loading Of relationship commitment
on behavior loyalty and afﬁnity on attitudinal loyalty. The w-test showed that all
variables are in fact distinct and that discriminant validity was achieved.

In terms Of model ﬁt, the structural model had a relatively good ﬁt with the
data for the 806 player’s club members who responded to the survey based and
for whom scores for all of the loyalty factors could be calculated. This is clearly
indicated by the results of the fit indices (Comparative Fit Index= 0.98,

Incremental Fit Index= 0.98, Standardized Root Mean Squared Residual= 0.03,

124

and Root Mean Square Error of Approximation: 0.06). In terms Of variance
explained, the overall r2 for the full structural model for the entire sample was
0.39. This suggests that although the indicator variables are loading quite
strongly onto their appropriate latent constructs, a signiﬁcant amount of the
variance is not being explained by the Altemative Integrated Casino Loyalty
Model. This suggests that other factors may be involved in the development Of

behavioral loyalty and are not being directly tested as part of this model.

125

5505 2:9 5:0 new .oocm:m>oo .mEoOEooo nouEmucﬁm

 

  

 

 

 

 

       

 

  
      

 

     

 

 

 

 

 

 

 

 

 

 

    

 

   
  

 

   

 

   

  

 

 

 

  
   
 

mozmoazoo
and mm23w200 Al
48. mm j<mm>o m>
mod 1.2mm
mod _n: >.S<>o._
mad _.._0 ._<z_o::t< . E2E< I
86 $921 Nu. a o ¢>
3 Si
men—z. tn. ANDOE $230020
w szEOEOma I
n>
6
n
ezmstzzoo
6.:mzoEﬁmm mom<>>mm I
n> >536; w3<> ~>
._<mO_><Imm om>_mommn_
nu. F...—
. 20.20.232.28
ezm2m>so>z_ 8o D<zowmmammez_ I
m> _.>

 

 

 

 

 

38 u 2V >625
=32 Edi—zoo n 3533. «2: 2:0 MERE moEmuo 52.0 05 no Eon-=02 =< Lo“— _ooos. _anabm ”a 2:9“.

Using the Structural Model for Identifying Differences Among Loyalty
Segments Based on Respondents Gaming Activity at the Client Casino
During the Previous 12 Months

The Altemative Integrated Casino Loyalty Model was tested for both the Loyal
Player's Club and Disloyal Player’s Club segments (segments based on
respondents gaming activity at the client casino during the previous 12 months)
in an effort to identify distinguishing Characteristics. Recall that the Loyal Players
Club segment had visited the client casino at least once each of the four previous
years. The Disloyal Player’s Club segment members have been members of the
client casinos player’s club for over four years and either had no recorded activity
from January 1, 2002 to December 15, 2002 or only ﬁve percent or less of their
total lifetime theoretical win was represented by gaming activity from January 1,
2002 to December 15, 2002 at the client casino. The structural model for 406
players that comprised the Loyal Player’s Club segment is presented in Figure
10. The structural model run for 376 Disloyal Player's Club segment is presented
in Figure 11. Table 19 shows the path coefﬁcients for these two segments
compared to all 806 player’s club members who completed a mail survey.

In comparing the models for both segments, the only statistically signiﬁcant
difference (p < 0.05) was for the path coefﬁcients between involvement and
behavioral loyalty. Of interest is that the Disloyal Player's Club segment actually
had a higher mean score for involvement (3.47) as compared to the Loyal
Player’s Club segment (3.24), but this higher level of involvement did not
translate into an increase in behavioral loyalty. Although there are no other

statistically signiﬁcantly differences between the segments, for both segments,

127

rewards had a signiﬁcantly large effect on perceived value. The effect between
afﬁnity and attitudinal loyalty and also between relationship commitment and
behavioral loyalty is also quite large. For both segments, perceived value is a
stronger predictor Of behavior loyalty than attitudinal loyalty. However, the effect
sizes for attitudinal and perceived value on behavior loyalty are larger for the
Disloyal Player’s Club segment. This adds additional support for the conclusion
that this segmentation base may not effectively identify loyal customers since
loyalty literature insists that loyal customers should have higher levels Of
attitudinal loyalty which should ultimately lead to higher levels of behavioral
loyalty. The ﬁndings Of this model test are contrary to that proposed by loyalty
literature.

The model ﬁt for both the Loyal Player’s Club segment (Comparative Fit Index
= 0.98, Incremental Fit Index = 0.98, Standardized Root Mean Squared Residual
= 0.03, Root Mean Square Error of Approximation = 0.07) and Disloyal Player’s
Club segment (Comparative Fit Index = 0.98, Incremental Fit Index = 0.98,
Standardized Root Mean Squared Residual = 0.03, Root Mean Square Error of
Approximation = 0.07) were acceptable. In terms of the overall variance
explained, the Alternative Integrated Casino Loyalty Model explained more of the
variance for the Disloyal Player’s Club segment (0.40) as Opposed to the Loyal
Player’s Club segment (0.31), which suggests that although very few signiﬁcant
differences were identiﬁed, the model is much better at explaining the behavior Of
the Disloyal Playefs Club segment as compared to the Loyal Player’s Club

segment.

128

:26...» DEC. Cote new 62.9.98 .mEoOEooo 852%:ma

 

 

own.
36
wad
mad
No.0

QM

wmoaz. ._._u_ .500:

am j<mm>o
mzmm

_n:

_LO

<mw2m
sax

 

 

«

 

 

._.Zw_2._._2_200
n=IwZO_.r<4mm

h>

 

 

 

 

kzm§w>40>z_
o>

 

 

a

     

mod

>._.._<>O.d
..<m0_><Im_m

Mn.

 

>._.._<>O._
...<Z_DD.:.C.<

Nu.

 

w3._<>
Ow>_w0mma

v".

    
                    

     

 

woszEZOO
KNEDwZOO

m>

 

 

 

 

>._._Z_n_n_<
v>

 

 

 

wkZDOOmE
w mZOCCEOmn.

m>

 

 

 

mom—(gum
N>

 

129

 

 

 

vmd

 

ZO_._.<O_ZD_2$_OO
._<ZOmmwn_mw.rZ_

_.>

I

 

 

894 u z 23> 0:355
Son. 05 we comm coco ammo.— um 0530 28:0 9: sex»; on; on? aceEmom n20 9.96:. .93.:
mazes—Looms”. omega L8 .235. 3.93.. 0530 “Xanadu... 035522 65 Co 953... 65 Co 3:53. 5.. 959".

 

wOZmDEZOO

5.6% 959 Cote ocm .Oocmtgoo 95.0508 nouEmucSm
mus—DwZOO ll

  

 

 

 

 

 

 

     
     

 

 

    

 

 

 

 

 

 

 

 

 

     

       

    

 

 

 

 

 

 

 

 

 

84. «m .:<mm>O m>
no.0 mimm
wad _n: .559
mad E0 2253:? . . LE2E< .I
So «moss. «u. 8 o e>
«a 5%.
mmoaz. FE Jun—OE wkzaoowa
w szEOZOma I
m>
ezmztzzoo
szszﬁSmm maxim”. I
5 >559 w3<> ~>
D<mO_><:mm Dm>momma
nu. v".
2929222200
ezm2m>so>z_ Nod 2284.34.52 I
o> _.>

 

 

 

 

 

 

as u z. 2.80 26.6 65 .6 «com .3 .3888 2 23

.v basses E9: 3333. 9...an .5 3230501 mm; 5.5 30:202.... 0:533 .20... :9: be $0..

.8 accede oz". 2:0 .0 «com .2. LonEooon 3 NooN ._. boa—=2. E9: >=>zo< 32003. e: to... 355

new 93> Son. :30 .8 2:0 9.8.6:. moEmmo 20:0 05 no .3822 a coon 022 9.3 35:53 CEO
mLoAmE .9333 mace—scones”. owe—t .6“— EUOE 5.93... 2..an 68952:. 033523 on. Co mazes... ”E. 95m."—

Table 19: Summary of Path Coefficients for Segments Based on Gaming
Activity at the Client Casino During the Previous 12 months as Compared
to their Lifetime Gambling Behavior at the Client Casino

 

 

 

 

 

 

 

 

 

 

 

 

PATH COEFFICIENTS
Loyal Player’s Disloyal ALL
Club Player’s Club

Interpersonal Communications
-) Perceived Value 0.54 0.62 0.61
Rewards
-) Perceived Value 0.73 0.78 0.77
Promotions
-) Attitudinal Loyalty 0.53 0.52 0.56
Afﬁnity
-) Attitudinal Loyalty 0.84 0.89 0.82
Consumer Confidence
-) Attitudinal Loyalty 0.75 0.79 0.79
Involvement
9 Behavioral Loyalty 0.67 0.48 0.68
Relationship Commitment
-) Behavioral Loyalty 0.92 0.88 0.90
Perceived Value
6“) Attitudinal Loyalty 0.68 0.70 0.73
Perceived Value
-> Behavioral Loyalty 0.35 0.36 0.38
Attitudinal Loyalty
-> Behavioral Loyalty 0.25 0.32 0.29

 

 

 

 

 

All paths are statistically signiﬁcant at p < 0.05

131

 

Testing of Structural Model for Identifying Differences Among Loyalty
Segments Based on Alternative Segmentation Bases

The next segments that the Alternative Integrated Casino Loyalty Model was
tested on were formed using the first alternative segmentation base Of
percentage Of total visits to casinos that were made to their primary casino (the
casino they visited most frequently over the previous 12 months). The results of
the test of the structural model for the Low (primary casino) Visitation segment
(those who made 40% Of less of their annual visits to their primary casino) are
presented in Figure 12. The results of the structural model for the Moderate
(primary casino) Visitation segment (those who visited their primary casino 41%
to 69% Of their total visits) and the High (primary casino) Visitation segment
(those who visited their primary casino 70% or more Of their total visits) are
presented in Figure 13 and Figure 14. A summary of the path coefficients for all
three segments (Low (primary casino) Visitation, Moderate (primary casino)
Visitation, and High (primary casino) Visitation) compared to that of the 806 total
respondents included in the testing of the Alternative Integrated Casino Loyalty
Model are included in Table 20.

In terms of the relationship between perceived value and attitudinal loyalty on
behavior loyalty, no pattern could be discerned from the results. From past
studies (Chaudhuri and Holbrook, 2001; De Wulf et al., 2001, Reichheld and
Teal, 1997), one would assume that attitudinal loyalty and perceived value would
be positively related to behavioral loyalty (or in this case, the proportion of total
trips made to their primary casino). As indicated by the results, the Low (primary

casino) Visitation segment actually had a larger attitudinal loyalty - behavior

132

loyalty path coefﬁcient than either of the other two segments. The effects
between promotions - attitudinal loyalty and also relationship commitment -
behavioral loyalty are the only paths that show a consistent pattern between
segments. Although no pattern could be discerned, the correlation between
perceived value and attitudinal loyalty was quite signiﬁcant with correlations of
0.60, 0.72, and 0.65 respectively for the Low (primary casino) Visitation,
Moderate (primary casino) Visitation, and High (primary casino) Visitation
segments, which suggests that for all three segments, the correlation between
these two factors is constant.

In terms of perceived value, the Moderate (primary casino) Visitation segment
had a larger effect (0.45) as compared to the Low (primary casino) Visitation
segment (0.32) and High (primary casino) Visitation segment (0.33). For the Low
(primary casino) Visitation segment, relationship commitment had a more
moderate effect (0.57) as compared to the effects for the entire sample (0.90)
and the Moderate (primary casino) Visitation (0.83) and High (primary casino)
Visitation (0.86) segments.

In terms of model ﬁt, with the exception Of the Root Mean Square Error of
Approximation for both the Low (primary casino) Visitation and Moderate
(primary casino) Visitation segments and the Standardized Root Mean Squared
Residual for the High (primary casino) Visitation segment, all of the remaining ﬁt
indices were above acceptable limits. Since multiple indices are utilized to
ascertain the ﬁt of models to their respective data sets, having one or two indices

fail to achieve acceptable limits does not equate to the models being rejected

133

(Kline, 1998). AS a result, the structural models for all three segments are
accepted. In terms of overall variation explained, the overall variance explained
for the Low (primary casino) Visitation, Moderate (primary casino) Visitation, and
High (primary casino) Visitation segments are 0.37, 0.37, and 0.39 respectively.
This suggests that although the model ﬁts the data (the mean loyalty factor
scores for all three segments), a Sizeable portion Of the variance remains
unexplained. This suggests that alternative factors may have an inﬂuence on the
development Of behavioral loyalty and that are not included as part Of the
Alternative Integrated Loyalty Model being tested as part of this phase of the

study.

134

$505 060. 8:0 new .0ocmtm>oo 005.0508 “30520205

 

 

mum.
mod
3.0
3.0
«Pd

wé

mm j<mm>o

1.210
E.
E0

(mm—2m.

mmoaz. PE Jun—OE

CENX

 

 

«

 

 

 

  

 

 

 

 

  

 

 

a

88qu 8580 3 8.: .53 :2: Co 86.. 3 as. m. 6580 .52.: E05
8 3.05 .83. Co 0002020.". 00023 E0Em0m cos—3.05 35000 30:..an 30.. 05 Low 55.05000 50m "NV 059".

ezwztzzoo
aimzoﬁﬁmm

5 >559

D<mo_><zmm
m...—
ezmzm>so>z_ .
8 o
o>

 

t4<>04
4<Z.DD._._.Fr<

Nu—

 
             

 

wD.._<>
Dm>.mommn_

E

    

 

 

 

NOZMOEZOO
mus—DwZOO

m>

 

 

 

     
     

 

>._._Z_n_n_<
¢>

 

 

 

wHZDOOwE
w mZO_._.O_ZOmn_

m>

 

 

 

  

 

mam—(ZEN.
N>

 

 

 

 

29.540.235.200
4<ZOmmwamwkZ_

—.>

 

 

135

 

 

 

 

 

 

   

 

     
     

 

 

    

 

 

 

 

 

 

 

 

 

 

     

  

 

 
  

 

 

 

 

   
   

 

 

 

   

 

 

 

 

 

.9505 £50. .95 new .00cmtm>oo 6.5.0508 00523.55
mozmozzoo
$28200 AI
Km. um. ....<m.m.>0 m>
mod mzmw
mod .u.. >556.
wad Io ._<z_o:.:t< KEEN? AI
mod <wm.>.m. N". and ¢>
3 3%.
mmoEz. tn. Juno—z 2.2 3 o 0 m. o
w mzoﬁozomn. AI
m>
pzmztzzoo
aimzoﬁﬁmm Suﬁsm”. Al
5 >559 m3<> ~>
._<mo_><_._mm 8.23%...
n“. I
2929232200 AI
hzm2m>.o>z_ . . . ._<zo.mmwn.mm»z_
o> 2 o S o F>
.Emuz.

002.000 3 mat... .80... 5.9.... *0 $00 3 o\. 3 m. 2.30 ENEE. 0.2.... 2 3.05 .30... 00 09350.3.
002.3 .cwEmcm cos—3.05 32.000 >383: 28000.2 0:029... .8 .30.... .8303“ 05 *0 9500.... 6.. 0.50.“.

c2610 080. 5:0 0:0 02.0.5500 0.5.0508 00520290

 

 

gm.
mod
had
had
0.20

in

mmoaz. tn. Jun—OE

Na ._._<mm>o
mzmm

_n__

:0

$02....
.63...

 

 

.

 

 

szitﬁiou
n._ImZO_...<n.mm

h>

 

 

 

 

 

...2w_2w>..O>Z_
o>

 

 

 

~wvmuz. 00:.000 2 0&0... .80... :03... “—0 0.6.2 .6 $2 0. 0:.000 ESE..."— ._.05 2 0:05 .80...
00 0002020; 009.; «:0Em0w 20.03.05 35000 Fig.5. :2... 0..“ .3 0:300”. .0202 .953-Em =3". "3. 0.59”.

>....._<>OA
..<mO.><Iwm

nu.

  

    

>...._<>O.._
..<Z.DD..._._F<

Nu.

 

         

 

m34<>
Qm>_momma

Pu.

 

moszEZOO
mmZDwZOO

m>

 

 

 

     
     

 

 

>.._Z_n.n.<
v>

 

 

 

wHZDOOmE
w $20....020mm

n>

t

 

 

 

 

 

mom<>>mm
N>

 

 

 

   

 

ZO...<O.ZDS=200
4<ZOmmmmmmkz.

v>

 

 

137

Table 20: Summary of Path Coefﬁcients for Segments Based on

Percentage of Total Visits to Casinos made to Their Primary
Casino

 

PATH COEFFICIENTS

 

Low (primary
casino)
Visitation

Moderate
(primary
casino)

Visitation

High
(primary
casino)

Visitation

ALL

 

Interpersonal
Communications
_)

Perceived Value

0.66

0.57

0.58

0.61

 

Rewards
.9
Perceived Value

0.77

0.76

0.75

0.77

 

Promotions
9
Attitudinal
Loyaﬂy

0.44

0.50

0.58

0.56

 

Afﬁnity

9
Attitudinal
Loyaky

0.84

0.85

0.84

0.82

 

Consumer
Conﬁdence
_>
Attitudinal

Loyalty

0.78

0.71

0.81

0.79

 

Involvement
-) Behavioral
Loyany

0.69

0.73

0.68

0.68

 

Relationship
Commitment
9

Behavioral
Loyany

0.57

0.83

0.86

0.90

 

Perceived Value
<--)

Attitudinal
Loyalty

0.60

0.72

0.65

0.73

 

Perceived Value
.9

Behavioral
Loyaﬂy

0.32

0.45

0.33

0.38

 

 

Attitudinal
Loyalty
9

Behavioral
Loyalty

 

0.37

 

0.20

 

0.36

 

0.29

 

All paths statistically signiﬁcant at p < 0.05

138

 

The second alternative segmentation base tested used the annual amount
budgeted for gaming at casinos as its criteria. As in previous analyses, three
segments were formed to ensure that each segment had adequate cases to test
the Alternative Integrated Casino Loyalty Model. The three segments analyzed
were the Low Casino Spending segment (those who budgeted $20 to $700 per
year), Moderate Casino Spending segment (budgeted $701 to $2099 per year),
and High Casino Spending segment (budgeted $2100 of more per year). The
results for the testing of the structural models are presented in Figures 15, 16,
and 17 for the Low Casino Spending, Moderate Casino Spending, and High
Casino Spending segments respectively. A summary of the path coefﬁcients for
all three segments based onannual amount budgeted for gaming compared to
that of the 806 player’s club members who returned a completed survey and
included in the model testing component of this study are presented in Table 21.
All paths for the models for all three segments were signiﬁcant at p < 0.05 level.

While the effect size between interpersonal communications and perceived
value was much greater for the Low Casino Spending segment (0.71) as
compared to the Moderate Casino Spending segment (0.59) and High Casino
Spending segment (0.62), the inverse was true in regards to the effect size
between rewards and perceived value. This suggests that as player’s club
members increase their rate of casino gaming, rewards appear to play a stronger
role in predicting the development of perceived value. This is likely due to the fact
that persons in the Low Casino Spending segment often do not have access to

rewards as a result of their low level of play. In relation to attitudinal loyalty,

139

afﬁnity was the most signiﬁcant predictor with loadings of 0.89, 0.90, and 0.90 for
the Low Casino Spending, Moderate Casino Spending, and High Casino
Spending segments respectively.

In relation to the latent factors, attitudinal loyalty had a much smaller effect
(0.15) on behavioral loyalty for the Low Casino Spending segment and is lower
as compared to the path coefﬁcient between attitudinal loyalty and behavioral
loyalty for the Moderate Casino Spending segment (0.32) and the High Casino
Spending segment (0.40). These ﬁndings suggests that attitudinal loyalty is
almost non-existent among members in the Low Casino Spending segment and
as the amount of gaming increases, the effect of attitudinal loyalty increases, but
only moderately so.

In terms of perceived value, although no pattern could be discerned, the
strength of the path coefﬁcient between perceived value and behavioral loyalty
for the Low Casino Spending segment suggest that at the lower gaming levels,
perceived value has a greater impact on creating behavioral loyalty. This is likely
due to the fact that the Low Casino Spending segment exhibits relatively low
levels of attitudinal loyalty.

Of interest is that in regards to the High Casino Spending segment, both
perceived value and attitudinal loyalty moderately effects behavior loyalty with
respective path coefﬁcients of 0.40 and 0.39. This suggests that behavioral
loyalty is affected equally by both perceived value and attitudinal loyalty and that
the loyalty programs (as hypothesized) is effective in developing perceived value,

attitudinal loyalty, and behavioral loyalty, albeit for this one segment.

140

In terms of model ﬁt, with a few exceptions (Root Mean Square Error of
Approximation for both the Low Casino Spenders and High Casino Spenders
segments and the Standardized Root Mean Squared Residual for the low
spending segment), all of the remaining ﬁt indices were above acceptable limits.

In terms of variance explained, the Alternative integrated Casino Loyalty
Model explained more of the variance for the Low Casino Spenders segment
(0.43) and High Casino Spenders segment (0.53) as compared to that of the
Moderate Casino Spenders segment (0.27). Although the differences in variance
explained by the Alternative Integrated Casino Loyalty Model are quite
significant, the low rate of variance explained for the Moderate Casino Spenders
segment is potentially due to members of this segment exhibiting patterns closer

to that of the Low Casino Spenders and High Casino Spenders segments.

141

c.5020 00:0. .90 0:0 .00:0..0>00 900.0508 80.0.0055

 

 

#00.
no.0
00.0
00.0
N...0

v.0

mmoaz. .5“. JmDOE

Na j<mm>0
0200

E.

.00

$0.)...

.me

 

 

..

 

 

...2w_2.:S=200
Q.IwZO.F<..mm

h>

 

 

 

 

sz§m>40>z_
0>

 

 

.

SE u 20 02.5.00 00.000 0010000000 00.000 30.: 00> .00 03.» 2 am» 00030 .05 E08000
020 0.0.0... 05 .00 .0005. 3.0.3.. 0:.000 0000.005. 02005003 0.... 00 00.000... 05 50 02.00: "0.. 0.50.".

      

mud

>...._<>O..
l.<w.O.><Imm

mu.

    

>0...<>04
4<Z.QD.E......<

Nu.

 

w34<>
Om>_w0mwa

_.n_

      
                 

        

 

moszEZOO
mum—230200

m>

 

 

 

00.0

 

>._..Z.u.n.<
v>

 

 

 

2.230005
.0 wZO...O.20m.n.

n>

 

 

 

m0m<>>wm
N>

 

 

{.0

 

ZO.H<O.ZDS=>.OO
..<ZmeMn.N.M...Z.

_.>

 

 

142

02,000 0.0.0. .000 000 .0000..0>00 0.00.0508 000000005

 

 

 

 

 

 

     

 

 
  

   
   

 

 

    

 

 

 

 

 

 

 

 

 

     

   

 

 

 

  

 

 

   

  
  

 

 

 

    
  

002000.200
00.220200 AI
0.0. Na .3505 m>
v0.0 1.2mm
00.0 _n... >5<>O._
00.0 .00 42.2.0000: >020"? Al
00.0 (mm—2m. Nu— v>
0N .ENX
mmoaz. ._._n_ 15002 00230005
0 0200.05.00“. AI
0>
m
50.20.2200
2:020:50”. 0002.300 Al
..> >556; 03<> ~>
..<m.0.><...0m 00200000
0“— wu—
20:<0_2:_2.200 AI
02020392. . . 22000000002.
0> E 0 mm o ..>

 

 

 

 

 

.000 u 2. 00.800 00.000 510000000 00.000 0.0.0005: .00.. .00 0003 0.
.05 000000 .00. “000.000 020 0.0.0... 05 .0. 00:00”. .0005. 5.05.. 00.000 000.0000. 02.00.32 ”0.. 0.00.“.

0.5000 0000. .000 000 .0000..0>8 0.00.0508 005000000

 

 

000.
00.0
00.0
00.0
.....0

_..v

0m0.02. .5". 40002

N0 .:<00>0
0200

.0.

.00

.0020
..me

 

 

.

 

 

0202:2200
a.1020....<3wm

b>

 

 

 

 

 

hzwﬁm>30>z.
0>

 

 

 

5

>.....<>03
._<w.0.><...mm

0n.

    

>.....<>0...
._<Z.n.3._......_.<

Nu.

 

         

 

      

..ZmE...0w>Z_
Dm>_w0mwn.

wu—

 

002090200
0.0230200

0>

 

 

 

     
     

 

  

>..._Z_n.u.<
¢>

 

 

 

00.230005
.0 020....02010

0>

t

 

 

 

 

0Dm<>>wm
N>

144

 

 

 

 

20...<0.Z3.2.200
3(200mwammhz.

—.>

Al

 

 

.000 u z. 00.0.00 00.000 510000000 00.000 00...: .00» .00 0.05. .0 00 Na» «000:0 .05 «000.000
00.0 0..0>0.n. 05 .0. .0005. 3.020.. 00.000 0000.000. 0200.004. 05 .0. 00:00”. .0005. 0.0.02.0 "t. 0.00....

Table 21: Summary of Path Coefﬁcients For Segments Based on Total

Amount Budgeted per year for Trips to Casinos per Year

 

PATH COEFFICIENTS

 

Low Casino
Spenders

Moderate
Casino
Spenders

High Casino
Spenders

ALL

 

Interpersonal
Communications
9

Perceived Value

0.71

0.59

0.62

0.61

 

Rewards
9
Perceived Value

0.66

0.75

0.80

0.77

 

Promotions
9
Attitudinal
Loyahy

0.48

0.46

0.54

0.56

 

Afﬁnity

..>
Attitudinal
Loyauy

0.89

0.90

0.90

0.92

 

Consumer
Conﬁdence
.9
Attitudinal
Loyaky

0.77

0.77

0.76

0.79

 

Involvement
-) Behavioral
Loyauy

0.75

0.74

0.67

0.68

 

Relationship
Commitment
_)
Behavioral
Loyany

0.89

0.83

0.84

0.90

 

Perceived Value
(n)
Attitudinal

LOLEI'H

0.62

0.61

0.70

0.38

 

Perceived Value
.9

Behavioral
Loyany

0.56

0.26

0.40

0.73

 

 

Attitudinal
Loyany
_)

Behavioral
Loyany

 

0.15

 

0.32

 

0.39

 

0.29

 

All paths are statistically signiﬁcant at p < 0.05

145

 

Testing of Rival Model

As stated in Chapters 1 and 2, a signiﬁcant amount of effort has focused on
exploring satisfaction within a marketing context. Although both loyalty and
satisfaction have individually received much attention from academics, very little
research has empirically examined the relationship between these two
constructs. Although a few studies have explored the link between customer
satisfaction and increased customer loyalty, more work is needed to investigate
the relationship between two of the most widely studied constructs in consumer
behavior (Bolton, 1998). Given the conﬂicting nature of how these constructs
have been operationalized, it is important to ﬁrst discuss how satisfaction and

service quality are related to building and retaining customer loyalty.

A number of researchers have suggested that there is a signiﬁcant difference
between satisfaction and loyalty. While satisfaction has been largely
operationalized as a passive condition (Fredericks, 2001), loyalty has often been
considered to be an active or even proactive relationship with the supplier
(Bolton, 1998). In addition, studies suggest that 65-70% of encounters that meet
or exceed the expected threshold of customer satisfaction usually only involve
hygiene factors, thus having no impact on customer loyalty (Buttle & Burton,
2001; Romano, 1995).

A number of key ﬁndings were revealed from the tests of the full structural
model (Figure 18) for the entire sample. First, all of the indicators were signiﬁcant
with loadings ranging from 0.71 to 0.95. In terms of path coefﬁcients, the results

suggest that satisfaction is a better predictor of behavioral loyalty (0.62) than

146

attitudinal loyalty (0.12). These findings provide support for previous satisfaction
literature that suggests that satisfaction is a strong predictor of behavioral loyalty.
Furthermore, the strength of the relationship between attitudinal loyalty and
satisfaction (.50) suggests that these two constructs are related to a moderate
degree, thus providing support that it is also a moderate predictor of attitudinal
loyalty. As depicted in the model, there is a moderately signiﬁcant relationship
between perceived value and both attitudinal loyalty (0.61) and satisfaction
(0.47); while having a very weak negative effect on behavioral loyalty. These
ﬁndings which are consistent with literature that was reviewed suggest that
perceived value is an important construct and is signiﬁcantly related to both
attitudinal loyalty and satisfaction. The low negative path coefﬁcient between
perceived value and behavioral loyalty is likely due to the fact that the effects of
perceived value are mediated by both attitudinal loyalty and satisfaction.

In terms of fit indices for thealternative structural model, all of the ﬁt indices
were above acceptable levels (Comparative Fit Index = 0.99, Incremental Fit
Index = 0.99, Standardized Root Mean Squared Residual = 0.03, Root Mean
Square Error of Approximation = 0.06). The strength of the paths coefficients
between the indicator variables and the latent constructs, coupled with the high
level of model ﬁt suggests that the rival model is quite efﬁcient at predicting
behavioral loyalty of casino players club members.

These ﬁndings indicate that behavioral loyalty is only created in persons who
are highly satisfied. In addition, those persons who exhibit attitudinal loyalty are

most likely highly satisﬁed with their service experiences.

147

In terms of comparing the alternative casino loyalty program model and the
proposed hypothesized model, a number of similarities and differences were
uncovered. The ﬁrst similarity is the strength of the relationship between
perceived value and attitudinal loyalty for both models. This suggests that
although satisfaction was included in the rival model, its impact on the strength of
the relationship between these two variables suggest that the relationship is quite
robust. In terms of key differences, the strength of satisfaction in predicting
behavioral loyalty suggest that satisfaction is a strong predictor of behavioral
loyalty. This provides support that with the absence of satisfaction, attitudinal
loyalty and behavioral loyalty will not likely be developed in customers of casino

loyalty programs.

148

 

 

 

 

~83. seam

ad \ cad

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   
  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m~m>uom
cozomwmzmm 2.6
No.0 5.6
«>338 makes
3.0
end
3.30.. n . - o2m> _ ,
68328 O o oozoocod mod 93:5
V
> Ed 2d .
v 0.20m Gd 3 o :m>aom
>=m>04
350332
mhd end
Nae—3.x 3234‘

 

 

 

 

 

 

a? u z 3528.8”. __a .2 .255. .95. 3.32:5 .3". 2: so 95mm» 2: so 333m "8 2:9“.

149

Identiﬁcation of Market Segments

The final section will focus on the identiﬁcation of actionable player’s club
market segments. A three stage sequential segmentation approach shown in
Figure 18 was utilized in an effort to identify exploitable market segments. The
first step was to segment client casino player’s club members that responded to
the mail survey based on the amount they budget annually for casino gaming.
The decision to use this segmentation base was based on its performance
compared to the two other segmentation bases tested in identifying
distinguishable market segments and its current use by a number of casinos for
marketing purposes. Three segments were identified: (1) $20 to $700, (2) $701
to $2099, and (3) $2100 or more per year. As stated previously, these cut-offs
were chosen to ensure that adequate cases were available to run structural
models for each segment and that all three segments are roughly equal size.

The first stage segments were then further segmented based on their
proportion of visits to their primary casino. Again, the primary casino may or not
be the study’s client casino. Since an objective of casino's players club programs
are to increase market share, understanding how player’s club members allocate
their gaming visits/purchases is important. Only the $20 to $701 and $2100+
segments were segmented based on their proportion of visits to casinos since
these groups had significantly different results in regards to their gambling
behavior, socioeconomic characteristics, and loyalty factor scores. If the third
segment was included ($701 to $2099 segment), the segments identiﬁed by

these process may overlap to a greater degree and impair the ability of

150

identifying specific market segments. Four segments were identified: (1) player’s
club members who budget $20 to $70 annually for casino gaming and visit their
primary casino (the casino they visited most frequently over the previous 12
months) 40% or less of their total visits to casinos over the previous 12 months
(3.7%), (2) player’s club members who budget $20 to $70 annually for casino
gaming and visit their primary casino 70% or more of their total visits to casinos
over the previous 12 months (8.8%), (3) player's club members who budget over
$2100 annually for casino gaming and visit their primary casino 40% or less of
their total visits to casinos over the previous 12 months (4.0%), and (4) player’s
club members who budget over $2100 annually for casino gaming and visit their
primary casino 70% or more of their total visits to casinos over the previous 12
months (9.0%).

The last stage involved segmenting the four second stage segments based
on the amount they budget for gaming per casino visit. They were divided into
segments that budget $1 to $60 per visit and more than $120 per visit. Again, the
middle segment ($61 to $120 per visit) was not selected since their behavior and
socioeconomic characteristics were not signiﬁcantly different from both
segments. This segmentation base was selected in order to identify the most
proﬁtable player’s club segments (in terms of amount budgeted per casino visit).

The end result of this process was the identification of eight market segments.
Two segments only contained 0.4% and 0.2% of total respondents who returned
a mail survey and were eliminated from further analysis based on the minute size

of the segment does not justify a targeted marketing strategy. Although the

151

Infrequent Big Spenders segment is quite small (0.6%), it has signiﬁcant revenue
potential and was retained in the analysis. The ﬁve additional segments that were
identiﬁed and profiled in the next section include: (1) Disloyal low spenders
(2.3%), (2) Loyal low spenders (6.8%), (3) Transient big spenders (3.3%), (4)
Frequent loyal low spenders (1 .0%), and (5) Loyal big spenders (5.7%).

The socioeconomic characteristics, behavioral characteristics, and loyalty
factor mean scores of the six market segments identiﬁed through this process
are presented in Table 22, Table 23, and Table 24. The mean scores for the
individual survey items are included in Appendix 4. Testing of the Alternative
Integrated Casino Loyalty Model for these segments was not undertaken do to

the lack of sufﬁcient cases for each segment.

152

 

 

$5 .m

 

 

 

$06

 

 

 

 

 

 

20:. .0 $2.

 

 

 

 

 

 

 

 

 

 

 

 

$9.?

 

 

 

 

 

 

{swam

 

 

 

So» ..on

 

 

29: Lo 83»

 

 

 

”mg ..0 $9?

 

 

 

 

 

 

 

 

 

Qewﬁ

 

 

 

 

 

.3th

 

 

Law» ..on

Nmmuz

 

 

amen» 3 was»

 

 

 

 

20E .5 $2

 

 

 

 

 

 

 

 

 

 

 

{and

 

 

 

 

 

ﬂmucoaw
mm. as. x .............. €233.26
8%:on <13
26.. .93.. A .............. a... .
.5209“. . t 8» 3
ﬁnd
Emucmaw
m_m ‘ ............. nub-F \ ON?“ hm>o
E225...
.VoNd
at: c3 . F»
2%:on ..xomd
O_m A .............. Q
.5382. t... . an; ..o>0
«and
935% A ..............
So. .98.. 3...: co» . w»
iced
nth . cur» ..o>0
28:on .xkd
30l— ‘ .............. a: I
_m>o_m_o . t cm» 3

 

$mdm

 

 

 

Lag Lon

 

w4n2<m
._<._.O._.

 

 

 

 

 

can» 3 on»

 

 

 

mm¢_ ..0 £5?

 

 

 

 

 

 

 

 

macoEmom 939.55. 33.0 9.96... 933.252 Lo. 3305 003w 09...... "mp 2:9“.

153

 

 

0528 9.0

 

 

 

 

 

 

om; .o>O 0. mg... .90. n6 SoE .0 $9. 905. .o 09% 2%:on 9m .93..
05mm”. 9.0 9%:on
oodow 0. 8. E 0. mg... .99 .0 90E .0 Axon m5.... .o 09% 26.. F50. 2362".
0:58 .Em
om; .m>O 0. mg... .99 .0 mm... .0 $3 20$. Lo 09% 9.2:.on 9m 2.22.2...
9.68 oco
om; .96 o. a... .99 .o 99: 5 so. 8% 0. om... 22.5% mm. 233:...
0598 9.0
8.08 o. 8. a 0. mo... .99 .o 22.. a so. 85 0. ow... 92.5% 26.. 3.6..
9..me .Em
00.8w 0. oo. 5 0. mg... .90. .0 mmb. .0 $3 comm 0. oma 2%:on .50.. _w>o.m.o
moEmao .mo>

2.... 3m coyomusm EsoE<

 

o... 3...... ho omen—Leech.

 

Lon. touomgm EsoE<

 

 

 

£558 .922... 9.6 $28.“. 5w 9: "um 29¢

154

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

:0co
$3. 9... E as... 5 9.0.0.. 9.0.3 9.0. 5 08.. 05.08 005.... .05
:0:0
..>... K $0.8 9.0.8 9.0.0.. 9&8 $0.0m mac... 9.7.8 39...... .65
E3080
#8. $0.8 $0.8 90.8 $0. .N $0. 5 ms.<m 9.. 0:38 39...... ..m.>
8:.80
metn. .0. :0..m..m.> ..0 0.0m.
...... .oz..m<0 >523...
.00 En. 00:05
8:600 Em 8580 0.9 8:.mmo ad 8:.000 N9 8:.08 Nm 8:.08 Om 8:.me .0 00:52
50> .0n.
9.... 3m 9.... 0.8 9.... m. R mg... m. 9.... ..m. .3... w... 8.; 00903..
0...... .0n.
9.8% 8.03 8.9.3 23% Nwdma «00$ 0 05:50 00903..
50> .0n.
8.023 8de2 3.8.3 85% 8.05» 8.38 m 0......00 00902...
- ,mo..><:mm.oz..z<0.
oi. 3 $5. $0..” $3... $0.. .x. E 8.; .93 .0 00980.0...
9...? «on... 9...... some so: as; 9.2.80 .98 .0 89:00.0...
$00 $0.. eon...” $0.0 $0... $2 0.0590 .98 .0 89:00.0:
. , . , . _ ,. , . . _ . - _ ._ . 8:0..mmho<m<..o 04:520.
0:00:80 0.00:000 0.00:000 0.0.0500
0.00:000 .50.. .93.. 9m Em 0.00:000 .50..
9m .93.. «c0300.... $0.80.... 20300.2. 30.. .83.. .0>o.m_n.

 

 

 

 

 

 

 

 

3:0Emom 8:00:05. 03.0 9.0.3.... 50 05 .0 0039.23.20 0.Eo:oo0o.oow 0:0 .0..o_>0..0m "MN 0.00...

155

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$8 $8 $8 $8 $8 $2 .9852. .8598”.
$2 $8 $. - $8 $.. .0320 .80 n8 0.2.2
A0.E0.mo.n.
$0 $2 - $2 $0 $v .930 .889. 0.98.
$8 $8 $8 - $0 $.V .9859. .90:
$0 $2 $. - $8 - 3.8.8.. .998. .8
$8 $2 $.. $88 $8 $8 .9598... 0.90 88.0
$8 $8 $8 $8 $8 $8 3598.. a... .889.
.. __. H _. 055.00”...805040050....z__.=..._0mmm=m_....
0:080. .9000
8 $2 $8 - $2 $0 5. v.8 8.0 0.0.6.. 852.
00.200 005.98%
$8 $8 $8 $2 $2 - 2.82 2 ago 0.93... 85%
50:0... x... .0. 03.m>
$8 $8 $8 $8 $8 $8 22.. .8 o. 8.0 0.96... 852,
05m .0 0mEoo
$8 $02 $8 $8 $8 $8 9.82 o. 8.0 208.. 852.
0:0..0Eo.m
$8 $8 $8 $8 _ . $8 $8. 9.80. o. 86 90.6... 852,
.. , , ., . . . . . r .. _. 0330 00.0.2...“ 02.2.0300.._0200<um
0.00:030 0.00:030 0.00:000 0.00:0m0
0.09.000 .50.. 8.3.. gm Em 0.00:000 .50..
9m .03.. «c0300.. 30.80.; E03092. .50.. .23.. 8.3.0.0

 

 

 

 

 

 

 

 

3:00. 0808000 0:50me 03.0 01050.... x.0 05 .0 00.080.00.20 3808002000 0:0 .a.0.>0:0m "mm 0.00...

156

m .Ammmdm I 80.9.3 N .68de 5.23 F 05 0885 20:339. .8 mmtomoﬁo .EoEmmm comm .2 28m _8tom2mo .59: 25353. -F

809.com» B>Ov 0 new .Ammmdmrm - 25de m .30..de I 000605 v .Ammmdm I 25.00%

 

 

 

 

 

 

 

 

 

 

 

 

3. tm 2 ma 5 F 3 9:85 2288:
29.330:
2.38 ow 938 5 958 tw 9.8m ow $58 3 £38 me 5 use. 932 ho 53:52

29:8 $5 298“. $8 235.1. $3 0688 $5 mason. $mo mason. mm
was. $8 222 $8 was. $3. was. $8 was. $5 222 $t $960
98> mm 299» mm 28> ow 28> mv 28> mo 2mm) mo 0m<

mo_hm_mmho<m<:o QEOzOomQUOm
ﬂotsam ﬂotsam mhwucwam ﬂoucomm
2%:on Bo.— _w>o._ 9m 9m mboucoaw 304
Pm _m>o._ 20:09.". E228... E3622. 26.. .83.. _w>o_m_n_

 

 

 

 

 

 

 

 

3:03 ﬂawEmwm maze—.35. n20 9.822.“. 5m 9: ho mom—£38220 o_Eo:ooco_oom can _Eo_>m:om "nu 2cm...

7
5

l

32mm 29.25 H n 3 35¢ .0: on u S gnaw

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

H H H 3 H a H 8 H m H N... H u H 3 H v H 3. H o H E. 213 _sosmﬁm

H H H 3 H N H S H n H S H m H 2 H o H .3 H v H ad 213 .2525

H H H 3, H N H S H v H 8 H N H 2 H o H m.“ H m H 3 205258 958%?

H H H 3 H u H mm H n H S H m H 3 H o H 3 H v H 3 2029.25

H F H E. H N H m... H n H 3 H v H 3 Ho H mm H m H .3 220$ 9,2888

H H H S H N H 3, H n H mm H v H 5 H o H 2 H m H 2. gas 828.8
_N H v... _UN H v... —v H OH. H” —. H 9v Hm H Em mm H Em mocwuccoo .mEamcoo
H23. HNHB. H15. H23 H33 HmHS. be?
H :3. H13 HNHE. H33 H33 H23 €533.
«45085

H H H 3. H m H S H N H mm H v H on H o H 3 H n H 3 w 22555
weesmoEzEEoo

H H H .3 H N H on H m H 3 H m H mm H m H 3 H v H 3. 5828.95
Emcbmwt.

H H H 3 H N H 2 H m H 3 H v H 2 H o H 2 H v H 3 @5535

H F H E H m H mm H N H E, H H. H 3 H m H on H m H 3 :22 .85

mmmoom «055. >550. 5E
232.on 2%:on mhmucmam €06:on
92.525 26.. .83.. 9m 9m whmucmnm 26..
2m .83.. 25:59.”. $222... E3692. 30.. .30.. .8335

 

 

 

 

95895 959.55. 33.0 9.93.; x5 9: .8 mmzoom .88". 53.3.. 53.2 Gm 228.

158

The following segment proﬁles were based on the results of the mail survey. It
is estimated that the 196,087 members of the client casinos player’s club made a
total of 4,41 1,957 trips to casinos, budgeted a total of $572,716,203.08 for

gaming, and visited on average 6.9 casinos over the previous 12 months.

Disloyal Low Spenders Segment

The Disloyal Low Spender segment make on average 8.8
trips to 8.6 casinos, an average of just over one trip to each
casino they visit per year. This indicated that they are not
loyal to any one casino. Another sign of their lack of loyalty is
that only 34.7% of this segment stated that the client casino
was their primary casino. This segment has relatively low
market value since they have the second lowest average
amount budgeted for casino gaming per trip ($36.52) and
total annual casino gaming ($314.38). They are one of the
oldest segments (65 years) and have the lowest average
household income.

Although they represent 2.3% of the client casinos player’s
club members (or 4510 players), they only account for 0.7%
of the total annual casino gaming (or $4,009,013.42). It
would appear that many members of this segment were
compelled to join the player’s club program in order to
receive promotions (83%), suggesting that they are not very
interested in being a member of any speciﬁc player’s club
program. Although they rank different components of their
primary casinos player’s club quite highly, they exhibited
very low levels of perceived value, attitudinal loyalty, and
behavioral loyalty. The beneﬁts of directing more marketing
attention at this segment is questionable given the low
number of visits they make to casinos and the relatively low
amounts they budget annually for casino gaming.

159

Loyal Low Spenders Segment

Although the Loyal Low Casino Gaming segment comprise
6.8% (or an estimated 13,333 members of the client casino’s
players club), they only represent 4.7% (or $26,917,661.54)
of the total annual casino gaming by player’s club members.
Although a significant portion joined the players club
program of their primary casino in order to
receive/participate in promotions, they rate the components
of their primary casinos player’s club lower than the other
ﬁve segments. This relatively low content may be in part due
to the fact this segment had the lowest average budgeted
amount for gaming on a per trip basis ($32.22), the second
lowest amount budgeted for gaming on an annual basis
($318.39), and the fact that many casinos have stopped
marketing to and offering player's club perks to these lower
spending segments.

The client casino is the primary casino of more than half
(58.2%) of this segment, thus demonstrating a moderate
degree of loyalty. This segment makes an average of 13.3
visits to casino’s annually and assigns a relatively low
amount of importance to casino gaming. This suggests that
members of this segment are not “serious” casino gamers
and likely view gambling as more of a social activity.
Because of their relatively low amount that they spend in
casinos, coupled with the relatively few visits they make to
casinos suggests that marketing directed at this segment
should focus on nongaming related activities (free trips,
food festivals, etc).

Infrequent Big Spenders

Although Infrequent Big Spenders represent only 0.6% of the
client casino’s player’s club (or an estimated 1176
members), they spend on average $224.17 per trip and
represent 1.7% (or an estimated $9,736,175.45 per year in
casino gaming revenues). In comparison, although this
segment is 1/16 the size of the Disloyal Low Spending
segment, they game almost twice as much at casinos each
year than the Disloyal Low Spending segment. This segment
makes on average 1.3 trips to casinos annually which
suggest that they are loyal to a speciﬁc casino, although only
25.1% stated that the client casino was their primary casino.

160

The dominant reason why this segment joined the player's
club program at their primary casino is to receive comps or
gifts (66%), which suggest that this segment is interested in
receiving comps (Free drinks, foods, etc.) during their visits.
Since this segment has an average age of 46 years (almost
20 years younger as compared to the other segments) and
an above average household income, there is potential to
develop this segment into Loyal Big Spenders. Although this
segment are less involved gamers, targeted marketing
strategies might be focused on this segment in order to build
loyalty and to increase their awareness of gaming
opportunities at casinos.

Transient Big Spenders

Transient Big Spenders represent 3.3% (or an estimated
6470 members) and 1.2% of the total gaming (or
$6,872,594.44 per year in annual casino gaming) of the
client casinos player's club program. Of interest is that
although Transient Big Spenders perceive the components
of the player's club program for their primary casinos quite
highly, this level of content has not led this segment to
exhibit high levels of behavioral loyalty (ranked ﬁfth out of six
segments). This segment has the second highest mean (4.3)
for the importance of where they gamble and interest for
learning about new gambling activities, which suggests that
this segment enjoys visiting different gaming establishments.

These results coupled with the fact that this segment had
one of the lowest mean scores (3.8) in terms of whether they
would remain a customer of their primary casino if they
stopped offering the player’s club program suggests that
membership (or the loss of beneﬁts) derived through their
membership in the player’s club program would not result in
a change of their gambling behavior. Since this segment
completes an average of 21.5 trips to casinos per year, any
marketing strategy should focus on the excitement of
gambling as opposed to the value derived through the
membership in the player’s club. Possible promotions
include free trips to exotic gaming locations and gaming
tournaments.

161

Frequent Loyal Low Spenders

Frequent Loyal Low Spenders made on average 88.5 trips to
10.6 different casinos over the previous 12 months which is
much greater than the other ﬁve segments. Although this
segment only represents 1.0% of the client casino’s player’s
club, they account for almost 4.7% of total casino trips by
player’s club members. While they make many visits to
casinos annually, they budget $45.00 per trip which
suggests that this segment spreads their annual gaming
expenditures across more trips. Although this segment takes
four times as many trips compared to the Transient Big
Spenders segment, they budget almost $1300 less per year
for casino gaming.

Although 100% of this segment stated that they joined the
casino to receive comps and 70% to receive promotions,
they ranked the components of the player’s club program
quite low. Interestingly, this segment exhibited higher levels
of perceived value, attitudinal loyalty, and behavioral loyalty
as compared to the other ﬁve segments. This indicates that
the player’s club program may have little effect on building
behavioral loyalty within this segment. Since a major reason
why they joined the player’s club program at their primary
casino is to receive comps, less structured promotions and
rewards should be utilized in marketing to this segment.
Examples of these promotions can include frequent buyer
meal cards and or offering bonus points during off-peak
hours to increase visitation during slow periods. As a result
of the .average age of this segment being the oldest of any
segment (66 years) and their low household income, the
potential of increasing their level of casino gaming is
questionable.

Loyal Big Spenders

Although the Loyal Big Spenders segment only represents
5.8% (or 11,373 members of the client casinos player’s
club), they represent 13.1% of the total annual casino
gaming ($75,025,822.60) and 11.7% of annual visits to
casinos (or an estimated 516,198 visits annually). These
results, coupled with their average per trip budget of $258.16
makes this segment an obvious candidate for targeted
marketing initiatives since they have a high household
income (significantly higher as compared to the other five

162

segments) which suggests that they likely can maintain this
level of gaming.

This segment had the highest mean scores for all the loyalty
factors included in this study with the exception of consumer
conﬁdence. Obviously, this segment is a desirable market
but the client casino has not been very successful in winning
the majority of their gaming as indicated by the fact that only
14.1% of this segment stated that the client casino was their
primary casino. Although this segment ranked the
components of their primary casinos player’s club program
quite high and exhibited high levels of both attitudinal and
behavioral loyalty, they also reported that they would remain
a customer of their primary casino if the player’s club
program was not offered. This suggests that this segment
would be open to new customer incentive plans.

in summation, although each of these six segments represent a signiﬁcant
share of the player’s club program of the client casino, the need to determine to
what extent the behavior of each segment can be modiﬁed through the use of
targeted marketing strategies must be better understood. Although the Disloyal
Low Spenders segment should probably not be the focus of any additional
marketing effort, the findings suggest that with certain targeted enhancements,
player’s club programs could further solidify the casino’s relationships with
Infrequent Big Spenders and Transient Big Spenders segments. This would
require additional marketing strategies (e.g. full service vacation packages with
the primary focus not being casino gaming, concerts, etc.) that are often offered

in practice outside the scopes of the loyalty programs.

163

CHAPTER 5
SUMMARY, DISCUSSION, CONCLUSIONS AND RECOMMENDATIONS

The ﬁrst chapter of this dissertation provided an overview of the setting of this
study, the need for this study, and the proposed objectives that framed the study.
The second chapter focused on studies with an emphasis on loyalty programs,
and also other research that contributed to the development of the theoretical
framework for this study. This same literature was also used in the development
of an Alternative Integrated Casino Loyalty Model that was used to examine the
development of attitudinal and behavioral loyalty by player’s club members. The
third chapter focused on the methods utilized to conduct the mail survey of
members of the client casino’s player's club and how these members were
chosen from the client casino’s player’s club tracking database. The fourth
chapter presented the findings of this study and how they related to each of the
five research objectives.

This final chapter will brieﬂy summarize important ﬁndings of this study and
discuss how they relate to the study objectives. The ﬁndings of this study will also
be compared to previous loyalty research in order to identify any similarities
and/or differences. The practical implications of this study are especially how the
client casino and other casinos might use the findings to enhance their player's
club programs and marketing practices. The last section provides

recommendations for future loyalty and loyalty programs research.

164

Summary

As the use of loyalty programs continues to grow in a variety of industries,
including gaming and tourism, it is essential to better understand their role,
importance, and effectiveness in creating and building loyalty. This research has
veriﬁed that understanding customer expectations and response to casino loyalty
programs and the importance they place on various features (e.g. incentives,
specialized promotions) will better enable casinos to maximize their return on
their player’s club. Understanding degrees of loyalty and what is effective in
building loyalty among different player’s club segments is also important in the
design of casino player's clubs.

Although more research is necessary to more fully understanding loyalty
development and its antecedents in the context of casino gaming, this research
sheds important light on loyalty creation, the impact(s) of casino’s player’s club
programs, and it contributes to building literature on the topic. No previous
published research is available that empirically examines revenues of casino
loyalty programs, but the ﬁndings of this study support research completed in
other industries by Bolton et al. (2000), Dowling and Uncles (1997), and Reinartz
and Kumar (2000) that suggest that not all members/segments of a loyalty
program are profitable. It is also the ﬁrst empirical study that documents that
costs associated with serving casino player’s club members actually increase
over the course of their membership. Including individual components (Direct
mail, interpersonal communications, preferential treatment, promotions and

discounts, and rewards) of a casino player’s club program as part of the

165

structural model provided information on how each component contributes to the
development of perceived value, attitudinal loyalty, and behavioral loyalty.

This study also provides additional evidence of the robustness and relevance
of a number of loyalty constructs and measures identified and assessed in
previous loyalty studies. The results suggest that loyalty constructs can be
modified to make them more applicable across different products/industries and
that loyalty studies completed for use in different settings and studies completed
across multiple settings can be compared to a certain degree. Although much
more work is needed to expand and validate loyalty constructs and measures for
different products, industries, and loyalty building programs, the results
demonstrate that they can be used to study loyalty programs in a noncontractual
service setting. Also, because this study included perceived value, attitudinal
loyalty, and behavioral loyalty in a single model, it allowed for an empirical
examination of the relationship between different loyalty constructs.

The ﬁrst study objective was to examine the revenues of the client’s player’s
club members in relation to number of lifetime visits they had made since they
ﬁrst joined. The ﬁndings reveal that there is only a moderate relationship between
number of lifetime visits and lifetime revenues. This suggests that total lifetime
visits are not a strong predictor of lifetime revenues, at least in relation to the
client casino’s current player’s club members. Although the approach utilized in
this study differs from that utilized by Reichheld and Teal (1996) and Reinartz
and Kumar (2000), their approach resulted in ﬁnding a correlation near zero

between lifetime duration and lifetime revenues. A moderate correlation between

166

lifetime visits and lifetime revenues suggests that revenues do not increase
signiﬁcantly over their tenure in the player’s club.

The focus of the second objective was to identify and proﬁle segments within
the client casino's player’s club tracking system based on their lifetime action
days and lifetime revenues. The analysis showed that a signiﬁcant portion of the
client casino’s player’s club program members including the Low revenue/Low
frequency and Low revenue/High frequency segments are potentially less
proﬁtable due to their low amount gamed per visit. Also, the player’s club
segments identiﬁed as being potentially proﬁtable including the High
revenue/High frequency and High revenue/low frequency, the revenues derived
through these segments were decreasing on a per trip basis. While they were
taking more trips to casinos a year, they distributed their annual casino gaming
budgets over more trips, thus reducing the proﬁtability per trip. These ﬁndings
need to be verified for other casino player’s clubs.

The third research objective was to compare and contrast segments formed
by casino loyalty relevant segmentation bases in order to determine their
effectiveness in identifying relevant and exploitable market segments. Segments
formed based on their gaming activity at the client casino during the previous 12
months as compared to their lifetime gambling behavior at the client casino
showed no signiﬁcant differences in regards to gaming behavior (number of trips
to casinos over previous 12 months, amount budgeted per casino visit, and
annual amount budgeted for casino gaming), socioeconomic characteristics (age,

household income), or loyalty factor scores (e.g. perceived value, attitudinal

167

loyalty, and behavioral loyalty). This segmentation base was ineffective at
identifying market segments of the client casino’s player’s club program.

Because no differences were found between Loyal Player’s Club and Disloyal
Player’s Club segments, two alternative segmentation bases were tested. One of
the most commonly used behavioral loyalty measures - proportion of total visits
to the most frequented casino (primary casino) - proved not to be an effective
way to segment loyalty program members. Very few marketing relevant and
exploitable differences were identiﬁed between the Low (primary casino)
Visitation, Moderate (primary casino) Visitation, and High (primary casino)
Visitation segments. The second alternative segmentation base was the total
amount budgeted annually for casino gaming that was obtained from the mail
survey. Significant differences, including casino gaming behavior (number of
visits to casinos over previous 12 months, amount budgeted per visit to casinos),
socioeconomic characteristics (household income), and loyalty factor scores
(with the exception of interpersonal communications) were found between the
Low Casino Spenders, Moderate Casino Spenders, and High Casino Spenders.

In order to determine if and how attitudinal and behavioral loyalty is created in
a casino player’s club program setting, the fourth objective utilized the Alternative
Integrated Casino Loyalty Model to further examine this process. The alternative
model was developed because the hypothesized model developed using the
ﬁndings of previous loyalty studies was rejected due to a high chi-square /
degrees of freedom ratio. The major differences between the hypothesized and

alternative models were the exclusion of direct mail, preferential treatment, and

168

competitive factors, as well as removing the paths between involvement and
perceived value, involvement and attitudinal loyalty, relationship commitment and
perceived value, and relationship commitment and attitudinal loyalty.
Subsequent tests of the proposed Alternative Integrate Casino Loyalty Model
determined that overall, perceived value has a greater inﬂuence on predicting
casino behavioral loyalty, but only moderately so.

In order to further test the segmentation bases applied in this study,
differences across segments in terms of how they developed attitudinal and
behavioral loyalty were examined using the Alternative Integrated Casino Loyalty
Model. As a result of testing of the segments formed using the proposed
segmentation base (segments based on player’s club gambling behavior at the
client casino during the previous 12 months), no signiﬁcant differences were
found between the Loyal Player’s Club segment and the Disloyal Player’s Club
segment in terms of their path coefficients for the Alternative Integrated Casino
Loyalty Model. Of interest is that attitudinal loyalty was a stronger predictor of
behavioral loyalty for the Disloyal Player’s Club segment compared to the path
coefficient for the Loyal Player’s Club segment. One would also assume that
disloyal customers would exhibit lower levels of attitudinal loyalty since previous
studies have assumed that higher levels of attitudinal loyalty lead to an increase
in behavioral loyalty (Bolton et al., 2000; Long and Schiffman, 2000). For both
segments, perceived value was a stronger predictor of behavioral loyalty as

opposed to attitudinal loyalty.

I69

The three segments (Low (primary casino) Visitation, Moderate (primary
casino) Visitation, and High (primary casino) visitation) identiﬁed by the ﬁrst
alternative segmentation base (segments based on proportion of total visits to
casinos taken to the primary casino over the previous 12 months) again proved
ineffective at identifying significant differences among segments. Although this
segmentation base has been used extensively in previous loyalty studies, no
pattern could be discerned from the subsequent path coefﬁcients for each
segment for the alternative casino loyalty model. Of interest is that the Low
(primary casino) Visitation segment had the largest path coefﬁcient between
attitudinal loyalty and behavioral loyalty. This ﬁnding is contrary to arguments put
forth in the literature that suggest that as a customer becomes more dependent
on a service provider, their level of attitudinal loyalty should increase (Chaudhuri
and Holbrook, 2001).

Segments derived through the second alternative segmentation base
(segments based on annual budget for casino gaming) showed signiﬁcant
differences in terms of their path coefﬁcients derived through the testing of the
Alternative Integrated Casino Loyalty Model. These differences are similar in
ways to the findings of some previous loyalty studies in terms of direction and
strength of differences between segments. This segmentation base was effective
at identifying segments of a casino player’s club program based on how
behavioral loyalty was developed. Furthermore, this segmentation base was also
successful in identifying segments that differed in terms of their gaming behavior,

socioeconomic characteristics, and loyalty factor scores.

170

The ﬁnal research objective was explored by the identiﬁcation of eight
exploitable market segments using data collected from the mail survey. Although
two of these segments were dropped due to having insufﬁcient sample sizes, the
other six segments identiﬁed were further analyzed in order to identify differences
for their gaming behavior, socioeconomic characteristics, and loyalty factor
scores. The Alternative Integrated Casino Loyalty Model was not tested for these
segments due to insufficient sample sizes. Two of the six remaining segments
(representing 10% of the client casino’s player's club program) were determined
to be potentially unproﬁtable, as the Disloyal Low Spenders segment budgeted
very little for gaming per casino visit and rarely visited the same casino more
than once a year and the likelihood of recouping marketing costs is quite unlikely.
Although Loyal Low Spenders tended to frequent their primary casino quite often,
they budgeted relatively little for each casino visit and as a result, the costs of
marketing to these individuals (as well as servicing them during their visits) would
likely not be realized. Furthermore, two of the remaining four segments
(Infrequent Big Spenders, Frequent Loyal Low Spenders) did not derive much
benefit from their membership in the loyalty program due to their low rate of play
and their low rate of visitation to their primary casino. This low rate of
contentment with the player’s club program was based on the mean scores
calculated for each segment using data collected as part of the mail survey that

focused on the components of their primary casino’s player’s club program.

171

Discussion

A major purpose of this study was the development and testing of a
hypothesized Integrated Casino Loyalty structural model. As mentioned above,
this model was ultimately rejected but the process yielded a number of ﬁndings
being uncovered. As hypothesized, the model had relatively few errors and ﬁt the
data relatively well, at least until the mediating variable (competitive factors) and
the moderating variables of involvement and relationship commitment were
added to the model. It is likely that the problem of including these variables
related to the measurement indicators used for competitive factors. As suggested
by Chaudhuri and Holbrook (2001) and De Wulf et al. (2000), current measures
for competitive factors overlap somewhat with current measures of perceived
value and attitudinal loyalty, thus causing discriminant and convergent validity
issues. Although competitive factors are important elements that should be taken
into account by loyalty studies, issues still remain in terms of their application and
use in modeling customer loyalty.

Testing the Altemative Integrated Casino Loyalty Model produced a number
of key results. As predicted by loyalty literature (Chaudhuri and Holbrook, 2001;
De Wulf et al., 2000; Neal, 1998), market segments differed in terms of their
levels of perceived value and attitudinal loyalty and the inﬂuence of these two
factors on the development of behavioral loyalty. In relation to past value
research (Reichheld and Teal, 1996, Reinartz and Kumar, 2000), the strength of
the correlation between perceived value and attitudinal loyalty suggest that these

two constructs, although highly correlated, are in fact distinct. By including

172

perceived value, attitudinal loyalty, and behavioral loyalty as part of the
Alternative Integrated Casino Loyalty Model, it was ascertained in this context
that perceived value had a greater effect on behavioral loyalty as compared to
attitudinal loyalty. Although these findings could be due to the fact that player’s
club members included in this study exhibited low levels of attitudinal loyalty, it
may also be a result of these members viewing player’s club programs more as a
value-added service rather than a relationship-building process.

Finally, a large amount of the variance remained unexplained by the
Alternative Integrated Casino Loyalty Model, which suggests that additional
factors not included in this model may play a signiﬁcant role in the development
of behavioral loyalty within a casino player’s club setting. These factors may
potentially include perceived service quality and switching costs. Based on the
findings of the rival model, satisfaction has shown to me an extremely important
mediating variable and one that needs to be further studied in relation to other
loyalty constructs.

In terms of the performance of the segmentation bases, the segmentation
base proposed by Bolton et al. (2000) and Long and Schiffman (2000), which is
based on their gaming activity at the client casino during the previous 12 months
as compared to their lifetime gambling behavior at the client casino, did not
effectively identify signiﬁcant market segments within the client casino’s player’s
club program. Although this base proved somewhat effective at identifying
segments within frequent ﬂyer programs and credit card programs, the poor

performance of this segmentation base in identifying statistically different

173

segments may be due to the source of secondary data (client casino) for the
current investigation. Only 27.1% of player’s club members who returned a
completed mail survey stated that the client casino was their primary casino.
Player’s club members whose primary casino was also the client casino ranked
the client casino lowest in terms of mean loyalty factor scores as compared to the
other nine casinos included in the study. These ﬁndings suggest that the client
casino is not a market leader and is not maintaining its player's club members
and/or their market share. The poor performance of the client casino and erosion
of their market share may have contributed to the misidentiﬁcation of Loyal
Player’s Club and Disloyal Player’s Club members since members may have
transferred their gaming from the client casino as a result of poor service
experiences, lack of promotions and discounts, lack of personal attention, and
the prospect of better reward programs offered by competitive casinos.

As mentioned previously, although the ﬁrst alternative segmentation base
(proportion of total visits to casinos taken to primary casino over the previous 12
months) has been used for decades in studying loyalty, its application in studying
casino player’s club programs is quite limited due to the sparse differences
identified between market segments in terms of their gaming behavior,
socioeconomic characteristics, and loyalty factors. Potential reasons for its failure
to identify significant market segments may be due to the extensive level of
competition between casinos that is being currently experienced in Michigan. As
a result of new casinos being built and renovations and expansions being

completed on current casinos, numerous casinos are using extensive player’s

174

club promotions to boost awareness of these new initiatives. Although quite
costly, these promotions do increase awareness of new gaming opportunities
and could potentially draw current (and new) casino gamers into their
establishments. As a result, although persons may frequent their primary casino
less often, the abundance of promotions may induce player’s club members to
roam from casino to casino in order to take advantage of promotions.

Total annual amount budgeted for casino gaming proved to be the most
effective of the three segmentation bases in identifying marketing relevant and
exploitable player’s club segments. Further research should be directed at
determining if similar results are produced for other player's clubs and also its
potential application for other industries/products. The results of applying this
segmentation base offers support for casinos, including the client casino, which
are exploring the potential benefits of tiered player’s club programs. Tiered
player’s club programs provide incentives for player’s club members not only to
increase their gaming (e.g., comp balances are earned faster, access to VIP
areas, etc.), but also offer incentives for them to maintain their level of gaming in
order get promoted to the next tier or to maintain their current level since they are
re-evaluated on an annual basis. In addition, it allows casinos to focus their
marketing and staff resources on servicing higher revenue customers as
opposed to spreading their resources across the entire program. However,
additional studies are needed to determine effectiveness and costs associated

with tiered player’s club programs.

175

Conclusions

This study resulted in a number of findings that shed light on loyalty and how
it relates to casino garners, the design of player’s club programs, and even
casino management. Although there has been considerable published research
on loyalty in other product industries, there is not much scientiﬁc literature related
to loyalty and loyalty creation with respect to casino gaming. It appears that many
of the loyalty building programs implemented by casinos are based more on a
‘follow the leader’ design rather than analytical studies of the effectiveness and
return-on-investment from different elements of casino player’s club programs.
Although the player’s club tracking mechanisms (e.g. player’s club cards provide
a means of assessing levels of play) are available to the client casino and other
casinos, very little effort has been taken to design experiments aimed at
determining the effectiveness of various player’s club promotions and incentives.

Casino garners usually have a variety of product choices available to them
and they appear to be taking advantage of this situation. In line with the ﬁndings
of Long and Schiffman (2000), due to the low cost of switching, the growing
number of casinos to choose among, and marketing efforts aimed at attracting
new customers, casino gamers have opportunities and reasons to visit multiple
casinos and to be members of multiple players’ club programs. In part, this is
because memberships are often mandatory to participate in promotions,
including casino bus tour incentives. As indicated by our ﬁndings, the majority of
the client casino’s members are members of multiple casino player’s club

programs and many only visit the casino that one time. Clearly, casinos are not

176

trying or willing to take the time to distinguish between members that exhibit
potential for loyalty development when recruiting player’s club members.

Very little correlation was found between perceived quality of the player’s club
programs and their behavioral loyalty. Some segments, including the Loyal Low
Spending segment and Infrequent Big Spenders segment actually exhibited
higher than expected levels of both attitudinal and behavioral loyalty than one
would expect based on their ratings of the components of their primary casino’s
player’s club programs. Similar to the ﬁndings of Bolton et al. (2000), player's
club members appear to maintain relationships with casinos even if they are
dissatisﬁed with the service features (e.g. player’s club incentives and beneﬁts).
The extent to which a player’s club program can mediate poor service
experiences was not measured in this study, but presents an interesting area for
future research.

In line with previous loyalty modeling efforts (Bolton et al., 2000; Chaudhuri
and Holbrook, 2001; De Wulf et al., 2000), a signiﬁcant portion of the variance
was not explained by the Altemative Integrated Casino Loyalty Model. This low
amount of variance explained may be in part attributable to the complex nature of
loyalty and the exclusion of factors that may have a greater effect on creating
attitudinal and behavioral loyalty.

In line with conclusions made by Bolton et al. (2000), Dowling and Uncles
(1997), Reichheld and Teal (1996), Reinartz and Kumar (2000), loyalty club
programs offer a unique setting for loyalty studies and the results of these studies

may differ greatly from that of other loyalty research. As proposed in other loyalty

177

studies (Chaudhuri and Holbrook, 2001; Wood, 1998), very few player's club
members exhibited high levels of attitudinal loyalty. This suggests that either
casino loyalty programs fail to create attitudinal loyalty or the creation of

attitudinal loyalty is quite complex and very rarely occurs in practice.

Managerial Implications

While the primary purposes of this study were to develop and test a casino
loyalty model and the effectiveness of alternative loyalty related segmentation
bases, it provides some practical information for casinos and suggests that they
need to develop more effective marketing and loyalty building strategies, and
programs. This study raises some major questions about how the client casino
(and presumably other casinos) is designing and managing their loyalty building
programs including forced recruitment, infrequent contacts, and the quality of
service delivery. Casinos should re-assess as an industry, the purpose and
utilization of player’s club programs as part of their overall integrated marketing
strategies. In addition to evaluating and enhancing loyalty building effectiveness
of the components of their player’s club, the client casino also needs to identify
alternative solutions (e.g. improve customer service, restrict access to player’s
club to high end members, integrated marketing campaigns with loyalty programs
in other industries) in order to build customer loyalty. Although the results of this
study show that “loyal” (or frequent visitors) members may have lower revenues
per trip compared to “disloyal” (or infrequent visitors) members, further work is

needed to extend these ﬁndings to other gaming organizations.

I78

It is evident that casinos need much work to improve interpersonal
communications between player’s club members and the casino and its
staff/management. Interpersonal communications was determined to be a
signiﬁcant predictor of perceived value. The cost of improved interpersonal
communications is less than other components including direct mail, discounts
and promotions, and rewards. In addition, improved communication between
staff and player’s club members will allow feedback and expectations to be
obtained from player’s club members on a regular basis.

While the customer tracking feature of the players club is potentially very
useful, the maintenance of the player’s club tracking system lists needs to be
improved. Due to forced recruitment, not dropping persons who are inactive for
several years, or checking for bad addresses and deceased members, the size
of these databases is very large. Because casinos rely heavily on direct mail for
their primary communication with their player's club members, having a
signiﬁcant number of bad addresses for members signiﬁcantly increases the
costs associated with these mailings and reduces the return on investment from
direct mail campaigns. In relation to the client casino, having so many persons
included who are not living in Michigan, who have stopped gaming, or who have
decided not to gamble at the client casino any longer, complicates the task of
analyzing the data to guide marketing and player’s club decisions.

This study also raises questions as to whether casinos should continue to rely
on loyalty programs as their primary marketing tool. Low terms of perceived

value, attitudinal loyalty, and behavioral loyalty raise concerns about the

179

effectiveness of the players club relative to the investment. Gaming organizations
should study the use of alternative marketing and promotion strategies (e.g. price
discounts, increased advertising, improving service, improving facilities) in order
to determine their acceptance by player’s club members and/or speciﬁc
segments within the player’s club program. Additional work is also needed on
identifying strategies and the associated costs for the conversion of unproﬁtable
segments to proﬁtable.

The focus of any casino loyalty program should focus on retaining and further
developing persons comprising the top spending tiers or those having a minimum
theoretical win of at least $300 per year. As indicated by management of the
client casino, a casino should not re-invest more than 10% of a player’s club
member’s theoretical win to market/reward/service that customer. So, for a
person who has a theoretical win of $300, the maximum amount that the casino
should spend to reward this individual is $30.00, which is not a signiﬁcant amount
to reward a loyal customer. Although loyalty programs work well for airlines, the .
cost of rewarding customers is minimal since reward redemption seats on each
ﬂight are limited and partner businesses actually pay the airline for each point
they reward as part of joint marketing agreements. It must also be noted that
loyalty programs were initially designed for frequent customers and not as a
means for rewarding all customers (Wood, 1998).

Finally, there is an obvious need to develop a structured research program
that collects both purchase (database) and behavioral (surveys, focus groups)

data from and about player’s club members. Although there is great potential

180

value from analyzing player’s club tracking data and numerous advances have
been made in database analyses and software, this information does not include
customer expectations and perceptions, visits to other casinos, or how they

would react to new player’s club features and beneﬁts.

Research Implications

There are very few empirical investigations of loyalty programs, which are a
major element of marketing in the casino gaming industry. The results from this
study suggest a need for additional research.

First, additional research should be undertaken to extend this model for the
study of loyalty programs operated by other industries to further test these
indicators. Although similarities exist between loyalty programs operated by
casinos compared to those operated in other industries, especially service
sectors, further empirical work is needed in the study of loyalty programs in other
noncontractual service settings.

Further work is also needed to improve on the Alternative Integrated Casino
Loyalty Model that developed and tested in this study. Since a signiﬁcant portion
of the overall variance was left unexplained as modeled, additional factors may
need to be examined to enhance understanding of how attitudinal and behavioral
loyalty are created. Further work is also needed on developing measures for
competitive factors.

It would also be beneficial to test the model and the segmentation bases
using data obtained from a gaming organizations that operate multiple casinos

(e.g. Harrahs, MGM) to extend the use of this model for larger gaming

181

organizations. Also, obtaining data for a longer time period (i.e., over six years)
could potentially identify true casino dropouts and determine whether they
stopped gambling or switched service providers. The work by Ganesh et al.
(2000) demonstrates the value of studying the differences that exist between
switchers and stayers.

Additional statistical analyses are needed in order to examine the relationship
between lifetime duration and lifetime revenues. Through the use of hierarchical
linear modeling, the revenues for each member (or segment) can be potentially
calculated for each time period (e.g. six months, twelve months) in order to
identify whether any patterns exist. This is an important step since the majority of
customer lifetime value models assume that a linear relationship exists between
these two variables, which may not be the case.

Finally, an area of inquiry that would be of interest is to test how various
loyalty program initiatives (promotions, rewards) affect members of a loyalty
program. Although this type of experimental research would be complex, it would
put the concept of relationship marketing to the test and truly determine the
extent to which casino player’s club programs can modify casino gamers’

behavior.

182

Appendix 1 - Mail Survey Instrument

Researchers at Michigan State University are conducting a study of casino loyalty program
members with a special focus on the impacts of these programs on your gaming behavior. You
have been randomly selected from a list of casino patrons. Your response is very important
because we are only sending 3000 questionnaires for all of Michigan. The information that you
provide to us will be used as part of a larger research project on how frequent buyer programs in
multiple industries affect consumers purchasing behavior. We are M completing this survey for
any casino or gaming organization.

To do this, we need your help. Your participation in this survey is completely voluntary. You can
refuse to participate at anytime and choose not to answer any of the questions. By completing
and returning this survey. you indicate your consent to participate in this voluntary study. Your
responses will remain conﬁdential. The data and results will be reported in summary fashion in
order for no person to be identiﬁable. We will not have a record of your name on the completed
survey unless you wish to participate in our appreciation drawing for one of ﬁve prizes of $100.00.
All of your data will be held in the strictest conﬁdence and will be protected to the maximum
extent allowed by law. Your name and any information that you provide will not be released to
any person, casino, or gaming organization.

If you have any questions regarding the importance of this survey, contact Dr. Edward Mahoney
by phone at 517.353.5190 x113 or e-mail mahoneve@msu.edu. If you have any questions
regarding your rights as a study participant, or are dissatisﬁed at any time with any aspects of this
study, you may contact — anonymously - if you wish, Dr. Ashir Kumar, MD, Chair of the University
Committee on Research Involving Human Subjects at 517.355.2180, 202 Olds Hall Michigan
State University, East Lansing MI 48823 or e-mail ucrihs@msu.edu.

Please take ﬁve to ten minutes to complete and return this important survey. Thank you.

Dr. Edward Mahoney

Associate Professor — MSU

Robert Palmer
PhD. Candidate - MSU

 

A visit is deﬁned as all visits to one casino within a 24 hour period.

 

1. In the past 12 months, have you visited a casino?

CI Yes

 

1a. Why did you not visit a casino in the last 12 months? (Check all that apply)

 

Cl Loss interest in
gambling

Cl Lost too much money

Cl Lack of time

 

III I became unemployed

0 Change in family
status

Cl No casino(s) close by

 

El Bad service
experience at casino(s)

Cl Switched to other
forms of gambling

Cl Don’t like atmosphere
at casinos

 

 

Cl Not winning

 

Cl My health declined

 

Cl Other:

 

183

 

2. How many casinos have you visited in the last 12 months?

3. How many visits did you take to casinos in the last 12 months?

4. Approximately, how much do you budget to spend on each visit to a

casino? $

5. Please complete the following table. (Check all that apply)

 

Indicate which

Indicate which

Indicate which

For the last 12

 

 

 

 

 

 

 

 

 

 

 

 

 

casinos you casinos that you casinos you months, indicate
have ever are member of have visited in what percentage
visited? its players the last 12 of your total
club/slots months? visits were to
program? each casino?
Turtle Creek —
Traverse City Cl Yes Cl Cl %
Leelanau Sands — Cl Yes
Suttons Bay Cl Cl %
Kewadin — CI Yes
Upper Peninsula Cl Cl %
Little River - Cl Yes
Manistee Cl Cl %
Victories — Cl Yes
Petoskey Cl Cl %
Windsor Casino - Cl Yes
Windsor Cl Cl %
Greektown - Cl Yes
Detroit El Cl %
MGM Grand — El Yes
Detroit Cl Cl %
Motor City - Cl Yes
Detroit Cl Cl %
Soaring Eagle — Cl Yes
Mt. Pleasant E! El %
Other: Cl Yes
El E1 %

 

 

 

 

 

184

 

6. For the casino(s) that you have visited, please state whether you are
visiting the casino the same amount, more often, or less often and the
reasons why you changed your amount of visits? (Check all that apply)

 

Turtle Keewadin Little Victories Detroitl Soaring Other:
Creek/ Casinos River Casino Windsor Eagle
Leelanau
Sands

 

 

SAM E E] Cl E] El Cl C) C]

 

MORE

 

Satisfied
with Cl C] C] CI Cl C] CI

gaming
Expeﬁence

 

Convenient

 

Take
advantage 0 0 Cl 0 D D D
of
promotions
/special
events

 

Take
advantage Cl Cl 0 D D D D
of players
club
beneﬁts

 

Good
service

 

Loose slots
/Win often 0 D

 

Other:

 

LESS

 

Not
satisﬁed 0 D D C] D D U

with
expenence

 

Poor service
expenence

 

Poor
players club
beneﬁts

U
D
D
D
D

 

Lack of
promotions

 

Not
convenient

 

Tight slots /
Did not win

 

 

Other:

DUDE!
DUDE]
0000
USED
DUDE
DDUD
DDDD

 

 

 

 

 

 

 

 

185

 

For the next set of questions, please use the casino that you indicated
in QUESTION 5 as being the casino that you completed the greatest
percentage of your visits

7. Are you currently a member of its players club/slots club?

—' Go to Question 7a
Cl Yes

—> Go to Question 7b

DNo

 

7a. Why did you join its players club/slots program? (Check all that apply)

Needed to join in order to receive El Wanted to receive comps
promotions / special invitations or gifts
0 Get more value for my money Cl Wanted to receive

specialized service
0 Social (Friends/Family are members) Cl Other:

 

 

 

 

7b. Why are you not a member of its players club/slots program? (Check all
that apply)
Do not want to be tracked Cl Do not want to receive
comps or gifts
0 Do not visit one casino enough to CI The value of the rewards
make joining the program worthwhile offered does not make it
worth joining
0 Had a poor service experience Cl Other:

 

 

 

186

8. For the casino that you indicated in QUESTION 5 as being the casino that
you completed the greatest percentage of your visits, for each statement,
please indicate how strongly you agree or disagree.

(1 = Very Strongly Disagree, 7 = Very Strongly Agree)
Note: If you are not a member of its players club, please still answer

all of the relevant statements in question 8.

WRITE THE NAME OF THE CASINO THAT YOU COMPLETED THE
GREATEST PERCENTAGE OF YOUR VISITS:

 

 

COMMUNICATIONS

 

lreceive mailings about the loyalty program ona 1 2 3 4 5 6 7
regular basis

 

 

 

 

lreceive information about promotions through 1 2 3 4 5 6 7
these mailings

lreceive information about special events 1 2 3 4 5 6 7
through these mailings

SERVICE

lreceive special treatment becauselama 1 2 3 4 5 6 7

member of the players club

 

Casino staff takes the time to personally get to 1 2 3 4 5 6 7
know me because I am a member of the players
club

 

Casino staff often hold personal conversations 1 2 3 4 5 6 7
with me because I am a member of the players
club

 

Casino staff offers me more personalized service 1 2 3 4 5 6 7
because I am a member of the players club

 

PROMOTIONS

 

loften receive discounts becauselamamember 1 2 3 4 5 6 7
of the players club

 

loften get invited to special events becauselam 1 2 3 4 5 6 7
a member of the players club

 

loften participate in promotions that are only 1 2 3 4 5 6 7
available to players club members

 

REWARDS

 

loften receive rewards/gifts (clothing, hotel 1 2 3 4 5 6 7
rooms, tickets) because I am a member of the
players club

 

As my level of gaming increases, the more 1 2 3 4 5 6 7
gifts/rewards I receive from the casino

 

 

l have used my points or comp balance to 1 2 3 4 5 6 7
purchase gifts, tickets, or cashback

 

 

 

 

 

 

 

 

187

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I often receive comps (free food, drinks) during 1 2 3 4 5 6 7
my visits because I am a players club member

RELATIONSHIP WITH PLAYERS CLUB MEMBERS

The casino makes great efforts to retain its 1 2 3 4 5 6 7
players club members

The casino makes great efforts to improve its ties 1 2 3 4 5 6 7
with players club members

The casino really cares about keeping its players 1 2 3 4 5 6 7
club members happy

The casino asks for input and ideas from players 1 2 3 4 5 6 7
club members

The casino continually enhances the benefits of 1 2 3 4 5 6 7
the players club program

EXPERIENCE

I trust this casino 1 2 3 4 5 6 7
I believe that the casino is honest in its business 1 2 3 4 5 6 7
with me

I rely on this casino for my gaming 1 2 3 4 5 6 7
I feel good when I visit this casino 1 2 3 4 5 6 7
Being part of the players club makes me want to 1 2 3 4 5 6 7
visit this casino more often

Visiting this casino gives me pleasure 1 2 3 4 5 6 7
I am usually satisfied with my visits to this casino 1 2 3 4 5 6 7
Overall and overtime, I am satisﬁed with this 1 2 3 4 5 6 7

casino

 

RELATIVE VALUE

 

Compared to other casinos, I receive good value
for my money

 

I feel that I receive more value for my money by
being a member of the players club

 

Compared to other players club, this program
offers me more value for my money

 

I would remain a customer of this casino even if
they stopped offering the players club

 

SWITCHING COSTS

 

Visiting this casino more often accelerates the
amount of points that l accumulate (e.g. bonus
points)

 

| visit this casino more often in order to
concentrate my points I comp balance at one
casino

 

Because of the points and benefits of the players
club, | find it difficult to visit other casinos

 

 

 

 

 

 

 

 

LOYALTY

 

 

I am committed to this casino

IIIZI3I4I5I6I7

 

188

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I would be willing to still visit this casino if it was 1 2 3 4 5 6 7
more difﬁcult to reach

I will visit this casino on my next trip 1 2 3 4 5 6 7
I intend to continue to be a customer of this 1 2 3 4 5 6 7
casino

I am willing to go “the extra mile” to remain a 1 2 3 4 5 6 7
customer of this casino

I tell other’s about the benefits of joining the 1 2 3 4 5 6 7
players club

I am loyal to this casino 1 2 3 4 5 6 7

 

9. How often do you use your players club card when gambling at a casino?

 

C] All the
time

 

Cl Most of the
time

Cl Sometimes Cl Rarely Cl Never

 

 

The purpose of the next set of questions is to obtain some general
information on your purchase behavior. Any information you provide
will NOT be released to any individuals or organizations.

10.Which of the following loyalty or frequent buyer programs are you a
member? (Check all that apply)

 

Cl Frequent Flyer Programs
(Worldperks, etc)

0 Grocery Store Programs
(Kroger’s, Rite Aid, etc)

 

Cl Retailer Frequent Buyer
Programs
(LL. Bean, REI, etc.)

[3 Music and Book Clubs
(Barnes and Noble, BMG,
etc.)

 

Cl Car Rental Programs
(Hertz Gold, etc.)

[3 Hotel Programs
(Hilton Honors, etc.)

 

 

 

Cl Restaurant Programs
(Subway, etc;

El Other:

 

 

 

 

 

11.The next set of questions is to obtain some general information on your
purchase behavior. For each statement, please indicate how strongly you
agree or disagree. (1 = Very Strongly Disagree, 7 = Very Strongly

Agree)

189

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I am someone who finds it important about where 1 2 3 4 5 6 7

l gamble

I am someone who is interested in learning about 1 2 3 4 5 6 7

gambling and different gambling opportunities

I often use my players club card when l gamble 1 2 3 4 5 6 7

I am someone for whom it means a lot to gamble 1 2 3 4 5 6 7

I am someone who likes to be a regular customer 1 2 3 4 5 6 7

for a casino

I am someone who likes building relationships 1 2 3 4 5 6 7

with companies

I always try to get the best value for my money 1 2 3 4 5 6 7
12. Your Age: _ years

13. Your Gender: Cl Male El Female
14. How many people, including yourself, reside in your household?

_ Adults Children (under 18)

15. What was your annual gross household income in 2002?

 

Cl Under $40,000 CI $40,000 - CI $60,000 -
$59,999 $99,999
[:1 $100,000 - CI $150,000 - CI Over
$149,999 $199,999 $200,000

 

 

 

THANK YOU FOR PARTICIPATING IN THIS IMPORTANT STUDY
Please return your completed survey as soon as possible in the postage
paid envelope provided

As a thank you for completing this survey, please ﬁll out the following information
to have your name included in the drawing for 5 prizes of $100.00 each. Please
remember that all information is strictly conﬁdential and that your name or
information provided will not be released to any person or organization.

 

Name
Address
E-mail Address

 

 

 

 

 

 

190

 

Aooemm 29.96 H s 3 wean no: on u 5 23w

 

93.0
292a 05 Co LonEoE m Em .

omsmoon oE 5.3 20385280

 

 

 

 

 

 

 

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

0m ON 54 NN 0N mé .9523 v.9. coco team 3.80
53.0 mLoxma
05 Co LonEoE m Em . 8:809
oE >65. 9 «om 3.9.8th
Ym ON NF NN we mm 9 oE: 05 moxm. tom 2.30
m:o_umo_:=EEoU $530825... .
n20 90 m. ho LonEoE w Em
. 3:88 00.23 um~=mco£wa
in EN we «N we oN 20E oE meoto tmﬁ. 9..me
g
9.. Lo LonEoE m Em _ 8389.
F...» o.m Qm van NN 9N “5.5mm: ..ﬂeoam. 9.8.2 .
ucoEumou... Esp—280......
m c...mE .. :95 mEo>o .33 m
Em o.v Nm oi Em me Sonm coszLoE. 9.89 .
m c.._mE c :95 mcozoEoL
Om Qm On 3. 3” me 50% 8.65.8... 389 _
mama
5.33.. o co Emamoa 3.98.
Qm ad ad wd Nm mé 9,: Scam 35.8,: 9.82 .
. , x , . ., =m§ .625
2%:on 9%:on 920:on 9%:on
9%:on 30.. .83.. 9m 9m 9%:on 264
2m .93]. «coauoi 23.95:. «co—Got... 25.. F504 _m>o.m_o

 

 

 

 

 

 

 

 

mEoEmom margins. Low moLoom cue—2 E3. .3229: ”N 59.3.:

191

Aooeum 29.25 H h 8 02mm «on on u S 23w

 

 

QN

v4.

NM

05

F6

 
 

LonEmE 93.0 9o m. m Em _
omamoon w._m_> >E mezzo Amxctu
.38 35 mano 0282 auto .

 

mum

o6

Om

Ym

 

xomnzmmo .o .996:
.35 $9.22. 9. oocm.mn
ano ..o 858 >E vow: o>ms .

 

NV

Nd

QM

YN

MN

md

0:68

05 E9: 0282 _ mEmBothm
20E 05 $3920....

mEEmm h_o .o>o. >E m<

 

9v

 

YN

Nd

9N

VN

mm

93.0 wama
9: Co LonEoE m Em . 8283
363.0: .mEooL .92. .mc_£o.ov

mEQmEmSoL ozoom: cote .

mo. miom

 

m4»

0d

5m.

oN

0N

m6.

maonEmE ago
whim... 9 o.om..m>m >.:o 2m 55
28.6an c. $3.0.th coco .

 

Qm

mN

Em

QM

QN

QN

92o Qoxmﬁ 05 Co
LonEoE m Em . $388 $55
.m.ooaw 9 page. .om coco .

 

m4.

 

mN

 

o.v

md

 

 

QN

 

m6

 

nae ﬂoxma
05 Lo LonEoE m Em . $388
$5586 9.89 coco .

wE:oow.Q w mcozoEoLn.

 

 

9%:on
9m .90..

 

who—Ewan
30.. .83...
Emacs...—

 

92:..on
9m
28.2.9...

92:.on
9m
ucosaotc.

 

 

2%:on
Bo.— .m>o._

 

9%:on
26..
_goaﬁ

 

 

 

Scoov mEoEmow 959.55. .8 woLoow :35. E2. 3323?... ”N 5.2.3.3

192

.3me 29.95 H n 0. 35m «0: on n 5 0.3m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mEEmm
wé o4. m.m 0v o6. 0m >E 6. 0580 £5 :0 3m: .
oE 5.2, mmoc.m3n 9. :. «mono;
3. m4 3. 0... mm o... m. ocamo 9: 65 99.8 _
Z. 3. E. 3. 5m Nm 0:68 mi. 32:
oocouwcoo 6E3.mcoo,.
oSmmo.
on 2. we 3 m... 3. 9: 82m 9.38 25 9.2.0.5
cal
£5 :m_> 9 “cm; OE mmme
mm. mm on mm mm mm 32o 29% 05 Lo :8 9.6m
oswmo
om 3. m... Nm 0... 0... £5 E; _ 8;; noon _8:
. a _. . . _ 22E...
wLoocomw mLoucoqm 963:on whoocoaw
9%:on 26.. _m>o.. 9m 9m 9.2:.on 26..
9m _m>o._ E6362". $6.29.... «5339:... 26.. .m>o.. .m>o.m_o

 

 

 

 

 

 

 

3.63 mEoEmom 329.65 .6. moLoom :moE E3. 63223:. ”N x5593

193

.0050 29.0.5 u s on 00.50 «09 on u 5 0.00m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 

 

 

 

 

 

 

 

 

 

 

93.0 0.0 m. 09. 99090
00000.0 >05 .. 90>0 09.000 0.9.
.1. 50 0.0 0.3 0.0 .20 .0 .0E09030 0 9.0E0. 0.303.
>090E NE .0. 03_0>
0.0E 0E E090 E050... 0.9.
Ev 0.0 m0 v.0 «.0 9.0 93.0 0.0>0.0 .0990 0. .00.00E00
93.0 0.0 0. 09. .0 .09E0E
0 09.09 >9 >090E >E .0.
0.0. >0 0.0 N0 «.0 0.0 03.0> 0.0E 0>.000. _ .09. .00. .
090E E .0. 03_0> 000 .0
09 0.0 0... 0... >0 N0 . .009.000 .0990 0. 00.00E00
9,...000999095oo
93.0 .0 0990909 09. 00090990
v.0 v.0 0.0 00 man ...0 3.0399900 09.000 09 ..
0.09E0E 93.0 E09 000.0.
0.0 0.0 v.0 «.0 v0. N0 090 .309. .0. 00.00 09.000 09...
09 209E0E
93.0 0.0>m.0 0.. 09.000.
0.0 0.0 N0 00 00. 0.0 .3090 00.00 >..00. 09.000 09 k
0.09E0E 93.0
0.0>0_0 5.2, 00.. 09. 0>0.0E. 0.
S. 00 0.0 «.0 EN «.0 09090 900.0 00me 09.000 09..
0.09E0E 93.0 0.. 9.0.0. 0.
we 0 0.0 m0 0.0 F0 09090 .00..w $me 09.000 09...
. . H . ,. 3.0200285.
060925 0.00900w 0.0.0900w 0.00900m
0.00925 269 .0>o.. 9m 9m 0.00925 26..
9m .06.. 990300.". 990.090.... 9903.00.09. 26.. .0>09 _0>o.0.n.

 

 

 

 

 

 

 

 

3900. 390Em0m 59:09.05. .0. 00.00m 9005. E09. .0323“...— .N 0909002

194

.0050 2099.0 u A. 0. 00.00 .09 on u 5 0.000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

93.0 0.0 m. 09. 9.9.0.
N... 0.0 0.0 0.0 0.0 0.0 .0 09.0909 09. .3090 0.09.0 =0. _
9004
o. 930...... 0.0E 002, .. .. 09.000
.0 4.0 0.0 0.0 0.0 0.0 0.9. .0; =90 2 05:? 09 28; _
0.0 0.0 0.0 0.0 0.0 v.0 09.000 0.9. 0. 00...EE00 Em .
9.0..o..._09.03_9<
09.000
0.9. .0 .0E0.030 0 9.0E0. 0.
0.0 0.0 0.0 0.0 0.0 0.0 ..0..E 09x0 09... om 0. 09....2. Em .
00.900E00 5.2. 0 .9090..m.0.
0.0 0.0 0.0 0..V 0.0 0.0 09.0.59 00v... 092. 0900E00 E0 .
09.000 0 .0. .0E0.030 .0.: 0.
0.0 0.0 0.0 0.0 0.0 0.0 0 09 0. 00.... 092. 0900E00 Em .
.90E..EE00 0.9090006”.
0.9Emm 0. .0. 0 0900E
0.0 0.0 ..0 0.0 0.. 0.0 .. E092. .0. 0900E00 Em .
009.9390 0 9..9E0 90.0....
090 09..9Emm .3090 09.9.00. 9.
0.0 0.0 0.0 0.0 0.0 0.0 00.00.09. 0. 092. 0900E00 E0 .
0.9EM0 . 0.092. .3090 .90900E.
0... v.0 0..q 0.0 0.0 0.0 .. 009.. 092. 0900E00 E0 .
, . . ..w..90E0_>.0.>9.
0.009000 0.009000 0.009000 0.009000
0.009000 26.. _0>o.. Em 0.0 0.009000 26..
0.0 _0>o.. E0300... 90.090... 9030099. 26.. _0>o.. _0>o.0.o

 

 

3900. 0.90Em0m 09009.0: .0. 00600 900.2 E0: .030.>.09. ”0 090900.?

195

00.00 >.090..w u x. 0. 00.00 .o: on n 5 0.000

 

 

 

 

 

 

 

 

 

 

09.000 0.9. .0 .0Eo.030
0.0 v... 0... 0.0 0... 0... m 09 0. 0399900 0. 090.9. .
m...
0.0 .1. 0.0 0.0 0... 0... 08: >9. :0 09.000 0.9. .0; =95 _
>..0>o.. .0.o.>090m
0.009000 0.009000 0.009000 0.009000
0.009000 26.. _0>o.. 0.m 05 0.009000 26..
0.0 _0>o.. .9030... 90.090; 90302.9. 26.. _0>o.. .0>o.0.o

 

 

 

 

 

 

 

3900. 0.90E000 0900.005 6. 00.000 900.... E0: .030.>.09. .0 0909002

196

 

REFERENCES

Achrol. RS. (1991 ). Evolution of the Marketing Organization: New Forms for
Turbulent Environments. Journal of Marketing, 55 (October 1991), 77-93.

Anderson, E.W., Fornell, C., & Lehmann, DR. (1994). Customer Satisfaction,
Market Share and Proﬁtability: Findings from Sweden, Journal of
Marketing, 58 (July), 53-66.

Anderson, J.C., Narus, J.A. (1990). A Model of Distributor Firm and Manufacturer
Firm Working Relationships. Journal of Marketing, 54(1), 42-58.

Anonymous. (2001). Consumers say “No Thanks” to Relationships with Brands.
Direct Marketing, 64(1), 20-25.

Anonymous. (1996). How Hilton & Sheraton Increased Their Business with
Frequent Buyer Programs. Direct Marketing, 59(5), 43.

Au, N., & J.S.P. Hobson. (1997). Gambling on the internet: A threat to tourism.
Journal of Travel Research, 35, 77-81.

Bagozzi, RP. (1995). Reﬂections on Relationship Marketing in Consumer
Markets. Journal of Academy of Marketing Science, 23(4), 272-277.

Bearden, William, Rose, R.L., & Teel, J.E. (1994). Correlates of conformity in the
Consumption of Illicit Drugs and Alcohol. Joumal of Business Research,
30(1), 25-31.

Beatty, S., Kahle, L.R., & Homer, P. (1988). The Involvement-Commitment
Model: Theory and Implications. Journal of Business Research, 16(2),
149-167.

Bendapudi, N., & Berry, LL. (1997). Customers Motivations for Maintaining
Relationships with Service Providers. Journal of Retailing, 73(1), 15-37.

Bennett, R., & Rundle-Thiele, S. (2002). A Comparison of Attitudinal Loyalty
Measurement Approaches. Brand Management, 9(3), 193-209.

Berger, P.D., & Nasr, NJ. (1998). Customer Lifetime Value: Marketing Models
and Applications. Joumal of Interactive Marketing, 12 (Winter), 17-30.

Berthon, P., Holbrook, M.B., & Hulbert, J.M. (2000). Beyond Market Orientation:

A Conceptualization of Market Evolution. Journal of Interactive Marketing,
14(1), 50-66.

197

Beirne, M. (2000). Starwood, United Woo E-Travelers. Brandweek, (October 16),
6.

Biyalogorsky, E., & Gerstner, E., & Libai, B. (2001). Customer Referral
Management: Optimal Reward Programs. Marketing Science, 20(1), 82-
95.

Blattberg, R.C., & Deighton, J. (1996). Manage Marketing by the Customer
Equity Test. Harvard Business Review, (July-August), 136-144.

Blattberg, R.C., & Deighton, J. (1991). Interactive Marketing: Exploiting the age
of Addressability. Sloan Management Review, 33(1), 5-14.

Blattberg, R.C., Glazer, R. & Little, JD. (1994). The Marketing Information
Revolution. Boston, MA: Harvard Business School Press.

Blattberg, R.C., & Thomas, JS. (1997). Dynamic Pricing Strategies to Maximize
Customer Equity. Unpublished Manuscript, Northwestern University,
Evanston, IL.

Blattberg, R.C., & Thomas, J.S. (2000). Valueing, Analyzing and Managing the
Marketing Function Using Customer Equity Principles. Unpublished
Manuscript, Northwestern University, Evanston, IL.

Bloemer, J.M.M., & Kasper, HD. (2000). The Complex Relationship between
Customer Satisfaction and Customer Loyalty. Journal of Economics
Psychology. 16(JUIY). 31 1-329.

Bollen, K.A. (1989). Structural equations with latent variables. New York: John
Wiley.

Bolton, R.N., Kannan, P.K., & Bramlett, MD. (2000). Implications of loyalty
program membership and service experiences for customer retention and
value. Academy of Marketing Science Journal, 28 (Winter 2000), 95-108.

Bolton, RN. (1998). A Dynamic Model of the Duration of the Customer’s
Relationship with a Continuous Service Provider: The Role of Satisfaction.
Marketing Science, 17(1), 45-65.

Boulding, W., Staelin, Kalra, A., & Ziethaml, V. (1993). A Dynamic Process Model
of Service Quality: From Expectations to Behavioral Intentions. Journal of
Marketing Research, 30(February), 7-27.

Bowman, D, & Lee-Pingle, S. (1991). Buyer Behavior in Business to Business
Services: The case of Foreign Exchange. International Journal of
Research in Marketing, 14 (5), 499-508.

198

Butscher, SA. (1998). Customer Clubs and Loyalty Programmes. Hampshire,
England: Gower.

Buttle, F., & Burton, J. (2002). Does Service Failure Inﬂuence Customer Loyalty.
Journal of Consumer Behavior, 1(3), 217-227.

Chaudhuri, A., & Holbrook, MB. (2001). The Chain of Effects from Brand Trust
and Brand Affect to Brand Performance: The Role of Brand Loyalty.
Journal of Marketing, 65 (April 2001), 81 -93.

Christiansen, E.M., Brinkerhoff-Jacobs, J. (1995). Gaming and Entertainment: An
Imperfect Union. Cornell Hotel and Restaurant Administration Quarterly,
36(2), 79-84.

Christy, R., Oliver, 6., & Penn, J. (1996). Relationship Marketing in Consumer
Markets. Journal of Marketing Management, 12(1-3), 175-187.

Collett, D. (1994). Modeling Survival Data in Medical Research. London:
Chapman and Hall.

Crosby, L.A., & Stephens, N. (1987). Effects of Relationship Marketing on
Satisfaction, Retention, and Prices in the Life Insurance Industry. Journal
of Marketing Research, 24 (November), 404-411.

Cunningham, SM. (1956). Brand Loyalty — What, Where, How Much? Harvard
Business Review, 34(January-February), 116-128.

Danahar, P.J., & Rust, R. (1996). Indirect Financial Beneﬁts from Service Quality.
Quality Management Journal, 3(2), 63-75.

De Wulf, K., Oderkerken-Shroder, G., and lacobucci, D. (2001). Investments in
Consumer Relationships: A cross country and Cross-Industry Exploration.
Journal of Marketing, 65( October 2001), 33-50.

Dick, A.S., & Basu, K. (1994). Customer Loyalty: Toward and Integrated
Conceptual Framework. Journal of the Academy of Marketing Science,
22(2), 99-113.

Diller, H. (1996). Customer Loyalty as a Marketing Objective. Marketing ZFP,
18(2), 81-94.

Doney, RM, and Cannon, JP. (1997). An Examination of the Nature of Trust in
Buyer-Seller Relationships. Journal of Marketing, 61 (April), 99-113.

199

Dowling, G.R., & Uncles, M. (1997). Do customer loyalty programs really work?
Sloan Management Review, 38(Summer), 71-82.

Dwyer, RF. (1997). Customer Lifetime Valuation to Support Marketing Decision
Making. Journal of Direct Marketing, 11(Fall), 6-13.

Dwyer, F.R., Schurr, RH, and Oh, S. (1987). Developing Buyer-Seller
Relationships. Journal of Marketing, 51(2), 11-27.

Fredericks, J.O., Hurd, R.R., & Salter, J.M. (2001). Connecting customer loyalty
to ﬁnancial results. Marketing Management 10(1), 26-32.

Gale, B. (1997). Satisfaction is not enough. Marketing News, October 27, 18.

Ganesh, J., Arnold, M.J., & Reynolds, K.E. (2000). Understanding the Customer
Base of Service Providers: An Examination of the Differences between
Switchers and Stayers. Joumal of Marketing, 64(July 2000), 65-87.

Garu, C., & Ranchhod, A. (2002). How to Calculate the Value of a Customer:
Measuring Customer Satisfaction: A platform for calculating, predicting,
and increasing customer proﬁtability. Journal of Targeting, Measurement
and Analysis for Marketing, 10(3), 203-219.

Gould, G. (1995). Why is it Customer Loyalty that counts (and how to measure
it). Managing Service Quality, 5(1), 15-22.

Grant, A.W.H, & Schlesinger, A. (1995). Realize Your Customers’ Full Proﬁt
Potential. Harvard Business Review, September-October, 59-72.

Gustafsson, A., & Johnson, MD. (2002). Measuring and Managing the
satisfaction-loyalty-performance links at Volvo. Journal of Targeting,
Measurement and Analysis for Marketing, 10(3), 248-258.

Hiscock, J. (2001). Disloyalty Becomes the Norm. Marketing, June(2001), 24-25.

lnman, J.J., Dyer, J.S., & Jia, J. (1997). A Generalized Utility Model of
Disappointment and Regret Effects on Post-Choice Valuation. Marketing
Science, 16(2), 97-111.

Jacoby, J., & Chestnut, RM. (1978). Brand Loyalty Measurement and
Management, New York: Wiley.

Jain, D., & Singh, SS. (2002). Customer Lifetime Value Research in Marketing:

A Review and Future Directions. Journal of Interactive Marketing, 16(2),
34-46.

200

Joreskog, KG. (1993). Testing Structural Equation Models. In k>a> Bollen & j.S.
Land (Eds), Testing Structural Equation Models (294-216). Newbury Park,
CA: Sage.

Kahan, R. (1998). Using Database Marketing Techniques to Enhance your One-
to-One Marketing Initiatives. Journal of Consumer Marketing 15(5), 491 -
493.

Kahn, B.E., Kalwani, M.U., & Morrison, D.G. (1986). Measuring Variety Seeking
and Reinforcement Behaviors Using Panel Data. Journal of Marketing
Research, 23(May). 89-100.

Keaveney, S. (1995). Customer Switching Behavior in Service Industries: An
Exploratory Study. Journal of Marketing, 59(April), 71 -82.

Kelly, S., Hoffman, K.D., & Davis, MA. 1993). A Typology of Retail Failures and
Recoveries. Journal of Retailing, 69( Winter), 429-452.

Kline, RB. (1998). Principles and Practice of Structural Equation Modeling. New
York: Guildford Publications.

Kordupleski, R., Rust, R.T, & Zahoric, A.J. (1993). Why Improving Quality
Doesn’t Improve Quality (or Whatever Happened to Marketing). California
Management Review, Spring, 82-95.

Little Six Incorporated (2003).
http://www.mysticlake.com/clubmystic/beneﬁts.htmI

Long, M.M., & Schiffman, LG. (2000). Consumption Values and Relationships:
Segmenting the Market for Frequency Programs. Journal of Consumer
Marketing, 1 7(3), 214-232.

Luxton, R. (2002). Marketing Campaign Systems — The Secret to Life-Long
Customer Loyalty. Journal of Database Marketing, 9(3), 248-258.

Massey, W.F., Montgomery, DB, and Morrison, D.G. (1970). Stochastic Models
of Buyer Behavior. Cambridge: MIT Press.

Mellens, M., Dekimpe, MG, and Steenkampe, J.B.E.M. (1996). A Review of
Brand Loyalty Measures in Marketing. Tijdschrift voor Economie an
Management, 41(4), 507-533.

Mohs, J. (1999). Frequency Marketing. Retail Report, 12(4), 3.

Morais, RC. (2002). Casino Junkies. Forbes, 169 (10), 66-70.

201

Morgan, R.F., & Hunt, SD. (1994). The Commitment-Trust Theory of
Relationship Marketing. Journal of Marketing, 58(July), 20-38.

Moritz, V (1997). Why consumers don’t always accurately predict their own future
behavior. Marketing Letters, 8(1), 57-70.

Moritz, V., 8. Schmittlein, D. (1992). Using Segmentation to Improve Sales
Forecasts based on Purchase Intent: Which intenders actually buy.
Journal of Marketing Research, 29(November), 391-405.

Mulhern, F.J. (1999). Customer Proﬁtability Analysis: Measurement,
Concentration, and Research Directions. Journal of Interactive Marketing,
13 (Winter), 25-40.

Neal, W.D. (2000). For Most Customers, Loyalty isn’t an Attitude. Marketing
News, April 10, 7.

Neal, W.D. (1999). Satisfaction is Nice, But Value Drives Loyalty. Marketing
Research, Spring 1999, 21-23.

O’Brien L., & Jones, C. (1995). Do Rewards Really Create Loyalty. Harvard
Business Review, 73, 75-82.

Oliva, T., Oliver, R.L, & McMillan, I. (1992). A Catastrophe Model for Developing
Service Satisfaction Strategies. Journal of Marketing Research,
1 7(November), 460-469.

Oliver, R. (1980). A Cognitive Model of the Antecedents and Consequences of
Satisfaction Decisions. Journal of Marketing Research, 17(November),
460-469.

Oliver, R. (1997). Satisfaction. A Behavioral Perspective on the Consumer.
Boston: Richard D. lrwin/McGraw-Hill.

Parasuraman, A. (1997). Reﬂections of Gaining Competitive Advantage through
Customer Value. Journal of the Academy of Marketing Science, 25(2),
154-161.

Parasuraman, A., & Grewal, D. (2000). The Impact of Technology on the Quality
— Value — Loyalty Chain: A Research Agenda. Journal of the Academy of
Marketing Science, 28(1), 168-174.

Partch, K. (1998). Do Loyalty Clubs Really Make Sense. Supermarket Business,
(Jan 1998), 98.

202

Pfeifer, P.E., & Carraway, R.L. (2000). Modeling Customer Relationships as
Markov Chains. Journal of Interactive Marketing, 14(2), 43-55.

Raphel, M. (1996). Customer Speciﬁc Marketing. Direct Marketing (June 1996),
22-27.

Reichheld, F., & Sasser, W.E. (1990). Zero defections: Quality comes to service.
Harvard Business Review, 14(March), 495-507.

Reichheld, F. & Schefter, P. (2000). E-Loyalty: Your Secret Weapon on the Web.
Harvard Business Review, July-August, 78, 105-113.

Reichheld, F ., & Teal, T. (1996). The Loyalty Effect. Boston, MA: Harvard
Business School Press.

Reichheld, F. (1996). Learning from Business Defections. Harvard Business
Review, March-April, 56-69.

Reinartz, W.J., & Kumar, V. (2002). The Mismanagement of Customer Loyalty.
Harvard Business Review, July 2002, 86-103.

Reinartz, W.J., & Kumar, V. (2000). On the proﬁtability of long-life customers in a
non-contractual setting: An empirical investigation and implications for
marketing. Journal of Marketing, 64 (October 2000), 17-35.

Romano, C. (1995). The Morphing of Customer Service. Management Review,
84 (12), December, 13-19.

RundIe-Thiele, S., 8 MacKay, MM. (2001). Assessing the Performance of Brand
Loyalty Measures. Journal of Services Marketing, 15(7), 529-546.

Rust, R.T., & Zahoric, A.J. (1993). Customer Satisfaction, Customer Retention,
and Market Share. Journal of Retailing, 69 (Summer), 193-215.

Sheth, J.N., & Parvatiyar, A. (1995). Relationship in Consumer Markets:
Antecedents and Consequences. Journal of the Academy of Marketing
Science, 23(4), 255-271.

Sheth, J.N., & Sisodia, R. (1995). Improving the Marketing Productivity in
Marketing Encyclopedia: Issues and Trends Shaping the Future, Jeffrey
Heilbrunn, ed. Chicago: American Marketing Association/NTC Publishing,
217-237.

Schmittlein, David, Morrison, L.G., & Colombo, R. (1987). Counting your

Customers: Who are They and What will They do Next. Management
Science, 33(January). 1-24.

203

Solomon, M.R., Surprenant, C., Czepiel, J.A., and Guttman, EC. (1985). A Role
Theory Perspective on Dyadic Interactions: The Service Encounter.
Journal of Marketing, 49(1), 99-1 1 1.

Thibault, J.W., & Kelley, H.H. (1959). The Social Psychology of Groups. New
York: John Wiley and Sons.

Thomas, J.S. (2001). A Methodology for Linking Customer Acquisition to
Customer Retention. Journal of Marketing Research, 38, 262-268.

Uncles, M. (1994). Do your Customers Need a Loyalty Scheme. Journal of
Targeting, Measurement, and Analysis for Marketing, 2(4), 335-350.

Uncles, M., Ehrenberg, A.S.C., & Hammond, K. (1995). Patterns of Buyer
Behavior: Regularities, Models and Extensions. Marketing Science, 14(3),
G71-G78.

Uncles, M., & Laurent, G. (1997). Editorial, Special Issue on loyalty. International
Journal of Research in Marketing 14, 399-404.

Wood, N. (1998). Breaking Out. Incentive Mag, (May 1998), 59-64.

Woodruff, R.B., Cadotte, E.R., & Jenkins, R.L. (1983). Modeling Consumer
Satisfaction Processes Using Experience-Based Norms. Journal of
Marketing Research, 20(August), 296-304.

Zeithaml, V.A., Berry, L.L., & Parasuraman, A. (1996). The Behavioral
Consequences of Service Quality. Journal of Marketing, 60 (April), 31-46.

204

   

uliliilillllliljll(111il