This document is a thesis submitted by Vani Haridasan to SRM University in partial fulfillment of the requirements for a Doctor of Philosophy degree. The thesis examines the impact of service quality, service loyalty, and loyalty indices on the performance of service providers in the mobile telecom industry in India. It develops conceptual models to analyze the relationships between service quality, service loyalty, and customer loyalty. The findings of the study aim to provide insights to practitioners on how to improve customer loyalty and sustain competitive advantage in the mobile telecom sector.
This document is a thesis submitted by Vani Haridasan to SRM University in partial fulfillment of the requirements for a Doctor of Philosophy degree. The thesis examines the impact of service quality, service loyalty, and loyalty indices on the performance of service providers in the mobile telecom industry in India. It develops conceptual models to analyze the relationships between service quality, service loyalty, and customer loyalty. The findings of the study aim to provide insights to practitioners on how to improve customer loyalty and sustain competitive advantage in the mobile telecom sector.
This document is a thesis submitted by Vani Haridasan to SRM University in partial fulfillment of the requirements for a Doctor of Philosophy degree. The thesis examines the impact of service quality, service loyalty, and loyalty indices on the performance of service providers in the mobile telecom industry in India. It develops conceptual models to analyze the relationships between service quality, service loyalty, and customer loyalty. The findings of the study aim to provide insights to practitioners on how to improve customer loyalty and sustain competitive advantage in the mobile telecom sector.
This document is a thesis submitted by Vani Haridasan to SRM University in partial fulfillment of the requirements for a Doctor of Philosophy degree. The thesis examines the impact of service quality, service loyalty, and loyalty indices on the performance of service providers in the mobile telecom industry in India. It develops conceptual models to analyze the relationships between service quality, service loyalty, and customer loyalty. The findings of the study aim to provide insights to practitioners on how to improve customer loyalty and sustain competitive advantage in the mobile telecom sector.
QUALI TY, SERVICE LOYALTY AND LOYALTY I NDI CES ON THE PERFORMANCE OF SERVI CE PROVI DERS A THESI S Submitted by VANI HARI DASAN in Partial Fulfillment of the Requirements for the Degree of DOCTOR OF PHILOSOPHY SCHOOL OF MANAGEMENT FACULTY OF ENGI NEERI NG AND TECHNOLOGY SRM UNI VERSI TY, KATTANKULATHUR- 603 203 SEPTEMBER 2012 ii DECLARATI ON I hereby declare that the dissertation entitled CRM I N MOBI LE TELECOM SERVI CES: A STUDY ON THE I MPACT OF SERVI CE QUALI TY, SERVI CE LOYALTY AND LOYALTY I NDI CES ON THE PERFORMANCE OF SERVI CE PROVI DERS submitted for the Degree of Doctor of Philosophy is my original work and the dissertation has not formed the basis for the award of any degree, diploma, associateship or fellowship of similar other titles. It has not been submitted to any other University or Institution for the award of any degree or diploma. Place: Date: VANI HARI DASAN iii SRM UNIVERSITY, KATTANKULATHUR603 203 BONAFI DE CERTI FI CATE Certified that this thesis titled CRM I N MOBI LE TELECOM SERVI CES: A STUDY ON THE I MPACT OF SERVI CE QUALI TY, SERVI CE LOYALTY AND LOYALTY I NDI CES ON THE PERFORMANCE OF SERVICE PROVI DERS is the bonafide work of Ms. VANI HARI DASAN who carried out the research under my supervision. Certified further, that to the best of my knowledge the work reported herein does not form part of any other thesis or dissertation on the basis of which a degree or award was conferred on an earlier occasion for this or any other candidate. Dr. SHANTHI VENKATESH SUPERVISOR Assistant Professor (SG) - Marketing SRM B School, Vadapalani SRM University Chennai 600026. iv ACKNOWLEDGEMENT This thesis work has received support, sacrifice and blessings of several well-wishers. The foremost among them is my research guide, Dr. Shanthi Venkatesh (Shanthi Maam). From the day I joined to this day, Shanthi Maam has been a great, constant and reliable source of advice and guidance. In fact more than a guide she has ever been a very good mentor. It had been a great learning experience working with a person whose academic and research standards are exceedingly high to match. My overwhelming thanks are due to Shanthi Maam, for helping me in countless number of ways during the course of the research program. I am lost of words to describe her patience in correcting my thesis and her high level of commitment to perfection. I profusely thank her for believing in my abilities and educating me in the art of coherent writing. I thank Prof. J ayashree Suresh, Dean, School of Management, for her constant support throughout my research program. Her boundless enthusiasm and commitment to research is an inspiration to me. I thank her for her critical evaluation of my work, her invaluable comments on my research topic that helped me in improving the quality of the thesis. I thank her for the interest she evinced in my work and grateful to her for her timely help and assistance. I thank my Doctoral Committee members, Prof. P.T.Srinivasan, Prof. M. J Xavier, Prof. V. J Siva Kumar and Prof. U. Srinivasa Raghavan for v their advice, help and suggestions. They were very keen to provide me with constructive suggestions to enhance the quality of my thesis. I am very grateful to the Management of SSN Institutions, who have kindly sponsored me for my research at SRM University. I am grateful to Prof. B.Srinivasan, Director, SSN School of Management & Computer Applications, Prof. R. Balasubramanian, Head of Department, MBA and my fellow colleagues for providing me with a conducive environment for pursuing research in my workplace. I express thanks to the enumerators who collected the data to validate my hypothesis. I also thank the respondents for their cooperation and patience in filling the questionnaire. The greatest gratitude is due for my family that cannot be adequately expressed in words. I would not have attempted this doctoral study (with tons of commitment at home) but for their love, support and infinite patience. Above all, I thank God for his supreme grace and presence. VANI HARI DASAN vi ABSTRACT The Indian Telecom Industry has come a long way in achieving its dream of providing affordable and effective communication services to its customers. Faced with a growing market and increasing competition, companies in the telecom business are adopting to the new technological imperatives in order to out-perform their competitors. One such approach in the adoption of an information technology (IT) to move towards customers is the Customer Relationship Management (CRM). CRM efforts aim at improving service quality and service quality is found to be having a definite impact on customer loyalty. This study aims at understanding the behavioral aspects that play a greater role in understanding customer loyalty, improving service quality and thereby enhancing CRM. Data collected from the residential users of mobile services from Chennai is used for the study. In order to sustain in the competitive environment, this study provides insights to the practitioners on the paths that lead to customer loyalty. The path model uses three constructs, namely the Extended SERVQUAL Scale with seven dimensions, SERVLOYAL Scale with seven dimensions and Loyalty Indices with 3 dimensions. Using Data Envelopment Analysis (DEA), the study uses the insights from path model to provide a method to evaluate the effectiveness of CRM practices of the mobile service providers so that they can refine their strategies to improve loyal customer base. vii Although Zeithaml et al. (1996) report a strong association between overall service quality and service loyalty across multiple companies, the findings of this study clearly portray the quality-loyalty relationship with reference to the Indian Mobile segment. This underlines the importance of a multidimensional approach to service loyalty. The path models confirm the relationship between service quality, service loyalty ad loyalty indices. This study further strengthens (Bloemer et al,1999) with a clear focus on linking perceived service quality and service loyalty as a multi-dimensional perspective. The insight from this study can be used in other service sectors to measure service quality and service loyalty and develop robust CRM systems that can map and predict customer loyalty levels. viii TABLE OF CONTENTS CHAPTER NO. TI TLE PAGE NO. ABSTRACT vi LI ST OF TABLES xvi LI ST OF FI GURES xix LI ST OF SYMBOLS AND ABBREVI ATI ONS xx 1 I NTRODUCTI ON 1.1 STATEMENT OF THE PROBLEM 7 1.2 OBJECTIVES 9 1.3 HYPOTHESES 9 1.4 DELIVERABLES OF THE STUDY 10 1.5 SCOPE AND LIMITATIONS OF THE STUDY 11 1.6 CHAPTERISATION 11 1.7 SUMMARY 13 2 LI TERATURE REVI EW 2.1 INDIAN TELECOM INDUSTRY 14 2.2 CUSTOMER RELATIONSHIP MANAGEMENT 15 2.3 SERVICE QUALITY 17 2.3.1 Service Quality Definition based on Functional Quality 17 2.3.2 Positive Relationship between service quality with customer satisfaction 18 ix CHAPTER NO. TI TLE PAGE NO. 2.3.3 Positive Relationship between service quality with customer loyalty and retention 20 2.3.4 Positive Relationship between service quality with profitability 20 2.3.5 SERVQUAL in Telecom 20 2.3.6 Service Quality Measurement and Models 21 2.3.7 Service Quality Dimensions 24 2.4 SERVICE LOYALTY 25 2.4.1 Defining Service Loyalty 25 2.4.2 Dimensions of Service Loyalty 27 2.5 LOYALTY INDICES 28 2.6 LITERATURE RELATING TO LINKAGE OF SERVICE QUALITY, SATISFACTION AND LOYALTY 30 2.7 INSIGHTS AND INADEQUACIES 32 2.8 SUMMARY 34 3 RESEARCH METHODOLOGY 3.1 INTRODUCTION 35 3.2 RESEARCH DESIGN 35 3.2.1 Area of the Study 36 3.2.2 Instrument Development 36 3.2.2.1 Variables considered for the Study 36 3.2.2.2 Demographic Variables 36 3.2.2.3 Extended Service Quality Dimensions 37 x CHAPTER NO. TI TLE PAGE NO. 3.2.2.4 Service Loyalty Dimensions 39 3.2.2.5 Loyalty Indices 42 3.2.3 Sampling Method 43 3.3 PROPOSED CONCEPTUAL MODEL 44 3.4 PRE-TEST 45 3.5 DATA COLLECTION 45 3.6 SUMMARY OF RESEARCH METHODS 45 3.7 DATA EDITING, CATEGORISING AND CODING 46 3.8 DATA ANALYSIS PROCEDURES 46 3.8.1 Structural Equation Modeling 46 3.8.1.1 Measurement Model 49 3.8.1.2 Structural Model 49 3.8.1.3 Models Goodness-of-Fit Assessment 49 3.8.1.4 AMOS 52 3.8.2 Multiple Regression 53 3.8.2.1 Measurement of Variables 54 3.8.2.2 Coefficient of Multiple Determination (R 2 ) 55 3.8.2.3 Interpretation of Regression Variate 56 3.8.3 Data Envelopment Analysis 56 3.9 PROCESS OF ANALYSIS 61 3.10 SUMMARY 61 xi CHAPTER NO. TI TLE PAGE NO. 4 THE I NDI AN TELECOM I NDUSTRY: STATUS ANALYSI S 4.1 INDIAN TELECOM INDUSTRY AN OVERVIEW 62 4.1.1 Basic Service Providers 62 4.1.2 Value added Service Providers 63 4.2 THE INDIAN TELECOM INDUSTRY: HISTORICAL PERSPECTIVE 64 4.3 GROWTH OF THE INDIAN TELECOM SECTOR 66 4.3.1 Indian Telecom Industry Pre privatization 66 4.3.2 Indian Telecom Industry Post Privatization 67 4.4 INDIAN TELECOM INDUSTRY: A COMPETITIVE ANALYSIS 73 4.4.1 PEST Analysis 74 4.4.2 Michael Porters Five Forces Model 80 4.5 INDIAN MOBILE SECTOR 84 4.6 KEY TRENDS IN INDIAN TELECOM 86 4.7 DETAILS OF MOBILE SERVICE PROVIDERS CONSIDERED FOR THE STUDY 88 4.7.1 Bharthi Airtel 88 4.7.2 Aircel 89 4.7.3 Idea Cellular 90 4.7.4 Vodafone 90 xii CHAPTER NO. TI TLE PAGE NO. 4.7.5 BSNL 91 4.7.6 Reliance 91 4.8 SUMMARY 92 5 DATA ANALYSI S 5.1 INTRODUCTION 93 5.2 PROFILE OF THE RESPONDENTS 94 5.2.1 Classification based on Service Provider 95 5.2.2 Classification based on Plan 95 5.2.3 Classification based on Age 96 5.2.4 Classification based on Length of Use 97 5.2.5 Classification based on Monthly Expenditure 97 5.2.6 Classification based on Education 98 5.2.7 Classification based on Occupation 98 5.3 DESCRIPTIVE STATISTICS 99 5.3.1 Mean and Standard Deviation 99 5.3.2 Normality 104 5.4 TEST OF RELIABILITY 104 5.4.1 Service Quality Dimensions 105 5.4.2 Service Loyalty Dimensions 107 5.4.3 Loyalty Indices 108 5.5 FACTOR ANALYSIS 109 5.6 HYPOTHESES TESTING 118 5.6.1 ANOVA 118 5.6.2 Linear Regression 125 5.7 PATH MODEL 135 xiii CHAPTER NO. TI TLE PAGE NO. 5.7.1 Relationship between Service Quality and Service Loyalty 135 5.7.2 Relationship between Service Quality and Loyalty Indices 138 5.7.3 Relationship between Service Loyalty and Loyalty Indices 140 5.8 ASSESSING THE EFFECTIVENESS OF CRM PRACTICES USING DATA ENVELOPMENT ANALYSIS 141 5.8.1 Effectiveness of Service Quality on Service Loyalty 142 5.8.1.1 Behavioural Loyalty 143 5.8.1.2 Attitudinal Loyalty 149 5.8.1.3 Cognitive Loyalty 151 5.8.1.4 Conative Loyalty 152 5.8.1.5 Affective Loyalty 153 5.8.1.6 Trust Loyalty 155 5.8.1.7 Commitment Loyalty 156 5.8.2 Effectiveness of Service Quality on Loyalty Indices 157 5.8.2.1 Advocacy Loyalty Index (ALI) 157 5.8.2.2 Purchase Loyalty Index (PLI) 159 5.8.2.3 Defection Loyalty Index (DLI) 160 5.8.3 Effectiveness of Service Loyalty on Loyalty Indices 161 5.8.3.1 Advocacy Loyalty Index (ALI) 161 xiv CHAPTER NO. TI TLE PAGE NO. 5.8.3.2 Purchase Loyalty Index (PLI) 163 5.8.3.3 Defection Loyalty Index (DLI) 164 5.8.4 Summary of Data Envelopment Analysis 165 5.9 SUMMARY 165 6 FI NDI NGS 6.1 FINDINGS FROM DESCRIPTIVE STATISTICS 167 6.2 FINDINGS FROM ANOVA 167 6.2.1 Demographics on Service Quality 167 6.2.2 Demographics on Service Loyalty 168 6.2.3 Demographics on Loyalty Indices 168 6.3 FINDINGS FROM LINEAR REGRESSION 169 6.3.1 Relationship between Demographics and Service Quality 169 6.3.2 Relationship between Demographics and Service Loyalty 170 6.3.3 Relationship between Demographics and Loyalty Indices 170 6.3.4 Relationship between Service Quality and Service Loyalty 170 6.3.5 Relationship between Service Quality and Loyalty Indices 171 6.3.6 Relationship between Service Loyalty and Loyalty Indices 171 xv CHAPTER NO. TI TLE PAGE NO. 6.4 FINDINGS FROM THE PATH MODEL 172 6.5 FINDINGS FROM DATA ENVELOPMENT ANALYSIS 172 6.5.1 Effectiveness of Service Quality on Service Loyalty 172 6.5.2 Effectiveness of Service Quality on Loyalty Indices 173 6.5.3 Effectiveness of Service Loyalty on Loyalty Indices 174 6.6 SUMMARY 174 7. DI SCUSSI ON 7.1 INSIGHTS FROM LITERATURE AND GAPS 175 7.2 ATTEMPTS BY THIS STUDY TO FILL THE GAPS 176 7.3 RESEARCH IMPLICATIONS 179 7.4 MANAGERIAL IMPLICATIONS 180 7.5 CONCLUSION 181 REFERENCES 183 APPENDI X I QUESTI ONNAI RE 199 PUBLI CATI ONS BASED ON THI S RESEARCH 203 CURRI CULUM VI TAE 204 xvi LI ST OF TABLES TABLE NO. TI TLE PAGE NO. 4.1 Important dates and events in the Indian Telecom Industry 65 4.2 Status of the Indian Telecom Industry in 1993-94 66 5.1 Classification based on Service Providers 95 5.2 Classification based on Plan 96 5.3 Classification based on Age 96 5.4 Classification based on Length of Use 97 5.5 Classification based on Monthly Expenditure 97 5.6 Classification based on Education 98 5.7 Classification based on Occupation 99 5.8 Descriptive Statistics of the Interval Scaled Variables 100 5.9 Cronbachs Alpha for Service Quality dimensions 106 5.10 Cronbachs Alpha of Service Loyalty dimensions 107 5.11 Cronbachs Alpha of Loyalty Indices 108 5.12 KMO and Bartlett's Test Service Quality 109 5.13 Factor Analysis Service Quality Dimensions 110 5.14 KMO and Bartlett's Test Service Loyalty 111 5.15 Factor Analysis Service Loyalty Dimensions 111 5.16 KMO and Bartlett's Test Loyalty Indices 112 5.17 Factor Analysis Loyalty Indices 112 5.18 Confirmatory Factor Analysis Service Quality 114 5.19 Confirmatory Factor Analysis Service Loyalty 114 xvii TABLE NO. TI TLE PAGE NO. 5.20 Confirmatory Factor Analysis Loyalty Indices 114 5.21 One-Way ANOVA Demographic Variables on Service Quality 119 5.22 One-Way ANOVA Demographic Variables on Service Loyalty 122 5.23 One-Way ANOVA Demographic Variables on Loyalty Indices 124 5.24 Relationship between Demographic Variables and SERVQUAL Dimensions 125 5.25 Relationship between Demographic Variables and SERVLOYAL Dimensions 127 5.26 Relationship between Demographic Variables on Loyalty Indices 129 5.27 Relationship between SERVQUAL and SERVLOYAL 130 5.28 Relationship between SERVQUAL and Loyalty Indices 132 5.29 Relationship between SERVLOYAL and Loyalty Indices 134 5.30 Service Provider Details 142 5.31 Input and Output Parameters Behavioural Loyalty 143 5.32 Ratios for Efficiency 144 5.33 Relative Efficiency on Behavioural Loyalty 149 5.34 Input and Output Parameters Attitudinal Loyalty 150 5.35 Relative Efficiency on Attitudinal Loyalty 150 5.36 Input and Output Parameters Cognitive Loyalty 151 5.37 Relative Efficiency on Cognitive Loyalty 152 5.38 Input and Output Parameters Conative Loyalty 153 5.39 Relative Efficiency on Conative Loyalty 154 xviii TABLE NO. TI TLE PAGE NO. 5.40 Input and Output Parameters Affective Loyalty 154 5.41 Relative Efficiency on Affective Loyalty 155 5.42 Input and Output Parameters Trust Loyalty 155 5.43 Relative Efficiency on Trust Loyalty 155 5.44 Input and Output Parameters Commitment Loyalty 156 5.45 Relative Efficiency on Commitment Loyalty 157 5.46 Input and Output Parameters Advocacy Loyalty Index 158 5.47 Relative Efficiency on Advocacy Loyalty 158 5.48 Input and Output Parameters Purchase Loyalty Index 159 5.49 Relative Efficiency on Purchase Loyalty 159 5.50 Input and Output Parameters Defection Loyalty Index 160 5.51 Relative Efficiency on Defection Loyalty 161 5.52 Input and Output Parameters Advocacy Loyalty Index 162 5.53 Relative Efficiency on Advocacy Loyalty 162 5.54 Input and Output Parameters Purchase Loyalty Index 163 5.55 Relative Efficiency on Purchase Loyalty 163 5.56 Input and Output Parameters Defection Loyalty Index 164 5.57 Relative Efficiency on Defection Loyalty 164 5.58 Details of Best Performer on various Dimensions using Data Envelopment Analysis 165 xix LI ST OF FI GURES FI GURE NO. TI TLE PAGE NO. 3.1 Conceptual Framework for the Study 44 4.1 Porters Five Forces Model 82 5.1 CFA Service Quality 115 5.2 CFA Service Loyalty 116 5.3 CFA Loyalty Indices 117 5.4 Path Diagram of Service Quality on Service Loyalty 136 5.5 Path Diagram of Service Quality on Loyalty Indices 138 5.6 Path Diagram of Service Loyalty on Loyalty Indices 140 7.1 Conceptual Framework 176 7.2 Model from Observations 178 xx LI ST OF ABBREVI ATI ONS TRAI Telecom Regulatory Authority of India DoT Department of Telecommunications MoC Ministry of Communications NTP National Telecom Policy MTNL Mahanagar Telephone Nigam Limited VSNL Videsh Sanchar Nigam Limited GSM Global System for Mobile Communications CDMA Code Division Multiple Access WLL Wireless Local Loop ARPU Average Revenue Per Unit MoU Minutes of Usage MVAS Mobile Value Added Services CRM Customer Relationship Management SERVQUAL Service Quality SERVPERF Service Performance SERVLOYAL Service Loyalty ALI Advocacy Loyalty Index PLI Purchase Loyalty Index DLI Defection Loyalty Index SEM Structural Equation Model CFA Confirmatory Factor Analysis RMSEA Root Mean Square Error of Approximation RMSR Root Mean Square Residual GFI Goodness of Fit Index xxi CFI Comparative Fit Index AMOS Analysis of Moment Structures DEA Data Envelopment Analysis DMU Decision Making Unit PEST Political, Economical, Social and Technological 1 CHAPTER 1 I NTRODUCTI ON Telecommunications is one of the prime support services needed for the rapid growth of any developing country. India has emerged as one of the youngest and fastest growing economies in the world today. One of the sectors that has shown the signs of profitability and contributed significantly to the country's economy is the telecom industry. In fact, the Indian telecom market has gained recognition as one of the most lucrative markets globally. The vast rural market holds a huge potential to drive the future growth of the telecom companies. Further, the Government's initiatives for increasing the telecom connectivity in rural areas are also likely to aid the telecom service providers to extend their services in the unconnected rural areas. Telecommunications companies recognize that becoming customer centric is very important to their long term competitive advantage, as many players offer similar features. Customer centricity depends on having a single view of the customer data that gives clear insight into the customer segments, their behavior and purchasing patterns. Telecom organizations need access to data that is accurate, reusable, and productive, so that they can create a holistic, real-time view of their customers. The history of the Indian Telecom sector goes way back to 1851, when the first operational landlines were laid by the then British Government in Calcutta. After India became an independent country in 1947, all foreign telecommunication companies were nationalized and a monopoly organization was formed by the Government of India, incorporating Post, Telephone and Telegraphs. 2 Originally, the telecom sector, like most other infrastructure sectors was owned and controlled by the Government of India. The Department of Telecommunications (DoT), reporting to the Ministry of Communications (MoC) became the key body for policy issues and regulation, apart from being the basic service provider for the entire country. By an act of Parliament, the Telecom Regulatory Authority of India (TRAI) was formed to be the regulatory agency, to monitor the activities of the telecom body. The Telecom Commission was set up with administrative and financial powers of the Government of India to deal with various aspects of Telecommunications in the year 1989. The multi-pronged strategies followed by the Telecom Commission have not only transformed the very structure of this sector but have also motivated all the partners to contribute in accelerating the growth of the sector. India is the second largest country in population and seventh largest in terms of area. Despite the government making several concentrated efforts through the Five-year Plans, to provide efficient communication system to the people of the country, due to increasing population, non-accessibility to remote and village areas and non-availability of adequate resources, the national tele-density was only 1.1 in the pre-liberalization period (till 1990), which was far low when compared to the global average of 12 for the same period. The world began to witness the changing phase of the telecom industry of India since 1994, when the Indian Government initiated the New Telecom Policy (NTP), with a broad objective to enable availability of affordable means of communication for the citizens of the country. The prime focus of the objectives of NTP was to increase the tele-density of the country by 15 in 2010. For this the country required an additional investment of Rs.5000 billion. In order to meet the investment requirements and to realize the broad objectives of the telecom policy in a phased manner, it became necessary that the government allowed private players to operate in the industry. The government also sensed the private participation imperatives for achieving the same in an efficient manner. Hence it was decided to 3 open the sector for private participation, simultaneously in both basic and mobile service segments. The other objectives of NTP 1994 include, creating a modern and efficient telecommunications infrastructure and transforming the telecommunications sector to a greater competitive environment in both urban and rural areas. Further, the NTP 1994 aimed at providing equal opportunities for all players, including the private players. Thus breaking the barriers to entry, the private sector providers forayed into the hitherto monopoly telecom sector of India, by first entering the Cellular Segment in 1994. But, the basic telecom service namely the Fixed-line Segment was at the hands of the government monopoly service provider DoT, till 1998. The very first private player to begin operations in the fixed-line segment was Bharti Telecom. The company was given license in 1998 and subsequently began its operations in the segment in 2002. The entire process of reform gained further momentum with the announcement of the New National Telecom Policy of 1999 (NTP 1999). The government started to issue licenses to multiple players to operate in all the segments of the industry. It also made provisions to discriminate the service providing functions and the regulatory functions of DoT, so that exclusive service providers are created to cater to the specific requirements of different segments of the population. The most noteworthy achievements of NTP 1999 were as follows: Creation of the Public Sector Telecom Service Providers One of the important objectives of NTP 1999, was to separate the service providing and regulatory functions of DoT. Subsequently, exclusive telecom service providers were floated to deploy the restructuring strategies in the sector, complying with the objectives laid down by NTP-1999. The policy separated the service providing function and policymaking and licensing function of the DoT. A brief note on the public sector telecom service providers is as follows: 4 The Bharat Sanchar Nigam Limited (BSNL) was floated to function as the government owned corporate entity to provide telecom services to the entire country For the metro cities of Delhi and Mumbai, the Mahanagar Telephone Nigam Limited (MTNL) became the service provider. Meanwhile, the international telecom services were entrusted with the Videsh Sanchar Nigam Limited (VSNL). As a result of the disinvestment policy adopted by the Government of India, in the late 1990s the company was sold to a private enterprise Panatone A Tata Group Company. Entry of Private Players Subsequent to the announcement made in NTP 1999, the government started issuing licenses to multiple private service providers to provide services in each telecom circle. Consequently, the private service providers kick-started their operations in fixed-line segments in the year 1994 and subsequently, the markets witnessed private players in mobile segment since 1996. With the entry of private players in the telecom segment, the sector began to witness the collapse of monopoly on the whole. Thus, the sector that was hitherto characterized by monopoly scenario gradually moved into competitive market structure, with multiple players offering a variety of services. As a result of privatization of the Indian Telecom industry, the most excited are the consumers. When there was monopoly, the users were left with no choice but to accept whatever the monopoly service provider gave to them. But in the present competitive scenario, consumers are able to exert their preferences as they are now offered a variety of options to choose from. The telecom market environment is witnessing brand new challenges both from the perspectives of service providers and service users. 5 Current Trends in Telecom I ndustry According to the Telecom Regulatory Authority of India (TRAI), the number of telephone subscriber base in the country reached 653.92 million as on May 31, 2010, an increase of 2.49 per cent from 638.05 million in April 2010. With this the overall tele-density (telephones per 100 people) has touched 55.38. According to Business Monitor International, India is currently adding 8- 10 million mobile subscribers every month. It is estimated that by mid 2012, around half the country's population will own a mobile phone. This would translate into 612 million mobile subscribers, accounting for a tele-density of around 59 per cent by 2012. Moreover, according to a study conducted by Nokia, the communications sector is expected to emerge as the single largest component of the country's GDP with 15.4 per cent by 2014. According to the latest figures made available by the Voice & Data study, Samsung posted a growth of 21.7 per cent to register revenues of Rs 5,720 crore in 2010-11 from India, from Rs. 4,700 crore in the previous fiscal. Nokia on the other hand had a flat growth, with revenues of Rs. 12,929 crore in 2010-11 from India compared to Rs. 12,900 in the previous fiscal. According to experts, home grown companies like Micromax, Karbonn, Lava, Spice and other mobile handset firms importing from China have eaten into market share of multi- national firms on the back of low-cost dual-SIM feature phones. With the availability of the 3G spectrum, about 275 million Indian subscribers will use 3G-enabled services, and the number of 3G-enabled handsets will reach close to 395 million by 2013-end, estimates the latest report by Evalueserve. According to a Frost & Sullivan industry analyst, by 2012, fixed line revenues are expected to touch US$ 12.2 billion while mobile revenues will reach US$ 39.8 billion in India. Moreover, in an attempt to boost auction of 3G spectrum, the government has allowed prospective bidders to raise short-term funds from 6 domestic market, which can be refinanced through external commercial borrowings (ECBs) within 12 months. The Indian Mobile Sector Globally in terms of mobile subscriptions, India is the worlds second largest wireless market after China. At the end of March 2011, the total wireless subscribers (GSM, CDMA & WLL (F)) base was 812 million. As a result the overall tele-density rose to 71% by end of March 2011 as against 4 % in March 2001. This growth of the sector can be clearly attributed to the favorable and improved regulatory structure, declining handset prices and innovative pre paid tariff structure. With increasing competition and the need for increasing the subscriber base in rural markets, the call rates are declining. This has led to decrease in ARPU. Therefore, in spite of unprecedented growth in the mobile subscriber base, the operator margins are declining quarter on quarter. The success of an organization depends on the degree to which they are oriented towards customers. This requires gaining customer insight in order to understand and manage the dynamics of customer expectations which forms a part of the critical success factors of any service provider. In the light of the sharp rise in the overall tele-density, the growth of this sector is attributed to the favorable and improved regulatory structure, declining handset prices and innovative pre paid tariff structure. But with increasing competition and the need for increasing the subscriber base in rural markets, the call rates are declining. This has led to decrease in ARPU. Though the Minutes of Usage (MoU) is increasing, however the same is being offset by the lowering tariffs of operations. This can be attributed to the major subscription growth that is coming from bottom of the pyramid. As ARPU declines, the challenges for operators are to increase revenues by differentiating their offerings and develop alternative streams by offering more value added services to the existing customers. The decrease in average revenue can also be attributed to the structure of the Indian Mobility Market which is largely 7 prepaid. This means that most of the subscribers added are from the bottom of pyramid with low usage resulting in low ARPU. In terms of market share, 92 % of the subscriber base in India is on pre- paid connection, with the remaining 8% on post-paid subscriptions. This has also given rise to opportunities for generating increased revenue, through exploring potential Mobile Value Added Services (MVAS) like subscription packs of news, alerts etc and more exclusive roaming services tailored to pre-paid subscribers. With the assumption that todays Value Added Services becomes tomorrows Core Services, both these types of services together are referred in this study as Mobile services. Thus in launching any of its services, the telecom service providers should focus it from the perspective of customer orientation. Todays Indian consumer is transforming rapidly. The consumer is growing richer, younger, aspiring and specific in his/her needs than ever before. Consumers now value convenience and choice at par alongside expecting better value for their hard-earned money. A range of telecom service providers are attempting to serve the needs of this emerging new-gen of Indian consumers. The last decade has witnessed a colossal transition in the paradigms with the emergence of many private players in the telecom domain, some of whom have invested in Indian operations basically looking at the promising takeaway from the changing lifestyles of Indian consumers. They have specially tailored their modus operandi to meet the life-style demands of the upwardly mobile Indian consumers of today. The bulging purses and liberalized market are making telecom being looked up as the big thing for India. 1.1 STATEMENT OF THE PROBLEM The beneficiaries of the competition being consumers, the telecom players in todays environment are required to design and deploy customer-centric strategies not only to grab a share in the market but also to sustain in the market in 8 the long-run. The players have realized the importance of constant service-quality delivery to the customers for long-run sustainability. Customer relationship signifies identifying the needs of the customers and stretching out ways and means to satisfy them. To be precise, it means achieving high customer profitabilitycustomer revenues over and above customer costs, which demands matching customer expectations with customer satisfaction. The high cost of customer acquisition is making todays businesses understand the importance of retaining the customers for long-run sustainability. Customer Relationship Management (CRM) aims at narrowing the gap between the company and its customers. In Telecom Sector, CRM plays a vital role in not only bringing the customers close to the company, but also in identifying the changing behavioral pattern of the customers. In technology-dynamic markets like telecom, an efficient CRM system is essential, since the customer attrition is high due to the presence of close substitutes and near-zero switching costs. With satisfied customers becoming retained customers and the retained customers becoming loyal customers, it becomes interesting to study the effectiveness of customer loyalty through CRM implementation in the Indian Mobile sector. This study helps the mobile service providers in determining the service quality parameters that influence service loyalty parameters which in turn influence loyalty indices. Previous studies have established the influence of service quality aspects on the customer loyalty. There are some studies that reflect behavioral aspects of the customers and the relationship on loyalty. But any such contemporary management practices when adopted by certain industry sectors need to be confirmed for its effectiveness. Thus CRM which is being widely adopted by the mobile service providers needs to be measured with respect to the degree to which the customers turn advocates; the degree to which the customers return to the company for repurchase i.e. in the context of cross selling and up selling and the degree to which 9 the customers show symptoms of defection. Even though some of these aspects have been explored by previous researchers to a great extent, it is felt that not much attention has been drawn into measuring the effectiveness of loyalty, particularly for sustained presence. Hence it is felt that a closer study on this aspect would be worthwhile. The study further provides useful information and opens up new avenues for future research. 1.2 OBJ ECTIVES The study has the following objectives: 1. To identify customer perceptions of service quality of mobile services used by them. 2. To analyze the service loyalty dimensions that impact the loyalty levels of the customers 3. To assess the loyalty indices that provide directives for strengthening CRM 4. To present a path model for establishing and confirming the relationship (if any) existing among service quality, service loyalty and loyalty indices 5. To provide suggestions to the service providers to base their CRM strategies on the findings of the study. 1.3 HYPOTHESES H 01 : There is no significant relationship between demographic variables and service quality dimensions. H 02 : There is no significant relationship between demographic variables and service loyalty dimensions. H 03 : There is no significant relationship between demographic variables and loyalty indices. 10 H 04 : There is no significant relationship between service quality and service loyalty. H 05 : There is no significant relationship between service quality and loyalty indices. H 06 : There is no significant relationship between service loyalty and loyalty indices. 1.4 DELI VERABLES OF THE STUDY This study would deliver an analytical insight to the strategic policy options to the competing service providers, industry onlookers and others concerned with the industry who are involved in the CRM practices. To the academic researcher, the study would provide a knowledge base to probe further, into new horizons of research. The specific deliverables are: The study would provide insight on the impact of demographic variables on the service quality, service loyalty and customer loyalty of the mobile service provider. The confirmatory factor analysis would give a research insight as to the influence of each of the latent factors in strengthening the CRM practices, while the confirmatory path reveals the critical quality variable(s) to concentrate on in order to improve service loyalty. The study would suggest strategies for improvement to the various players to match the expectations and to sustain in the competitive environment. 11 1.5 SCOPE AND LI MI TATI ONS OF THE STUDY The following are the scope and limitations of the study. The findings of the study can be extended for further research on varied perspectives and other sectors namely, Insurance and Banking Hospitality Healthcare and other segments where service quality plays an important role in CRM. The study is confined only to the GSM and CDMA mobile user segment and as such the results derived may not be applicable to the other segments of the industry. The study limits its applicability to Chennai city and can be extended only to markets having similar demographic, geographic and socio- economic characteristics. The study applies convenience sampling method of data collection, and the limits that apply to this sampling method of data collection hold good for this study also. 1.6 CHAPTERI SATI ON Chapter 1 presents the introductory perspectives and an overview about the Indian Telecom industry. The chapter includes an introduction to the research problem, the research objectives and gives the scope and limitation of the study concepts. 12 Chapter 2 provides a detailed overview on the literature reviewed by the researcher in order to have an insight into the research studies undertaken in related areas; to identify the research gaps and also to obtain a sound knowledge about the methodological issues. Chapter 3 focuses on the research methodology for the empirical study. The research design is discussed in terms of the scale development, questionnaire design, data collection, and participant selection, and sample size design, data editing, coding and categorizing and analyzing procedures. Chapter 4 presents a comprehensive note on the path of growth trends and transitions of the Indian Telecommunication industry. It also covers a detailed competitive analysis of the Indian Telecom Industry, Chapter 5 presents the processes and results of the preliminary data analysis and hypothesis testing are discussed in detail. Profile of participants and response rate for this study are described. Preliminary data examination includes descriptive statistics of mean and standard deviation, and normality. Reliability of the measurement scales are examined, followed by the testing of hypotheses. Chapter 6 discusses on the findings arising from the research, which will throw light on the significance of the path models and the findings on the effectiveness using Data Envelopment Analysis. Chapter 7 covers the overall conclusions of the dissertation. Conclusions are drawn about the research problem and the implications for researchers and practitioners are discussed. Limitations of the study are identified and areas for further research are suggested. 13 1.7 SUMMARY This chapter provides the foundation for the thesis. It provides a context and justification for the study. The research problem and research questions addressed in the study were introduced. The main purpose of this study is to investigate ways to strengthen CRM practices by the Mobile Service Providers. Methodology underlying the purpose of the study was briefly described and justified. Definitions of the key concepts are also presented. Some delimitations of the scope are delineated and a chapter structure was provided for the whole thesis. On these foundations, the thesis proceeds with a detailed description of the research beginning with the chapter two - Literature Review. 14 CHAPTER 2 LI TERATURE REVI EW This chapter covers a detailed overview of the literature reviewed by the researcher in order to have an insight into the theoretical framework and methodological issues related to the study undertaken. For this purpose the researcher has reviewed research papers and articles published in both International and National journals and also several popular and focused books which give broad perspectives on the conceptual framework of the study undertaken. The details of the literature reviewed by the researcher are arranged chronologically in the following sections. 2.1 I NDI AN TELECOM I NDUSTRY Rekha J ain (1993) reviewed the policy changes in the Indian Telecom Sector during the initial stages of post liberalization. The author explained that in response to the business needs of faster, cheaper, and more varied modes of communication, the telecommunication sector in many countries had undergone rapid technological and structural changes. Athreya (1996) described about the significant changes in the Indian Telecommunications sector during the NTP 1993. He identified three phases of changes. First, there was a policy vacuum almost up to 1990. Second, there was a shift in telecommunication policy brought about by a paradigm shift in government economic policy. Third, difficulties were experienced in implementing the new policy. J ain (2001) explained that spectrum auctions had been used with significant success in many developed countries. He analyzed that while India was one of the early adopters of spectrum auctions, its success in service provision had 15 been low. The author examined the issues in auction design that contributed to the delay and reviewed the key elements in the design process namely a coherent regulatory framework, choice of service areas, flexibility for service area consolidation, standards and their role, convergence, managing public service regulation and managing defaults. 2.2 CUSTOMER RELATI ONSHI P MANAGEMENT Sheth and Parvatiyar (1995) focused their work on understanding the motivations of consumers to engage in relationships. They emphasized that in order to develop an effective theory of relationship marketing, it is necessary to understand what motivates consumers to reduce their available market choices and engage in a relational market behavior by patronizing the same marketer in subsequent choice situations. Baker et al (1998) noticed that it is more productive to consider Relationship Marketing to be a third dimension of the existing Transactional Model than to develop a separate paradigm. According to the authors, two-dimensional marketing model is described as reflecting the marketing of products (dimension one) to market segments (dimension two), facilitated by channels and supported by promotion and pricing. The third dimension, relationships, is introduced as a spiral diagram that elevates the customer above the two planes of products and markets, and in doing so, generates a barrier to entry of predator sellers. J oe Peppard (2000) has emphasized that CRM should not be looked from a narrow perspective. He presented a CRM framework which was based on incorporating e-business activities, channel management, relationship management and back office / front-office integration within a customer centric strategy. Veith Krner et al (2000) developed a conceptual model for CRM in Business Media. They used a case study to conceptualize the CRM practices in a financial industry. 16 Russell S. Winer et al (2001), gave a framework for the CRM concepts. They analyzed the needs for CRM as a strategy and explained in detail the intrinsic activities involved in CRM practices. Injazz and Popovitch (2003) emphasized that Customer Relationship Management (CRM) is a combination of people, processes and technology that seeks to understand a company's customers. It is an integrated approach to managing relationships by focusing on customer retention and relationship development. Payne and Frow (2005) developed a conceptual framework for customer relationship management (CRM) that helps broaden the understanding of CRM and its role in enhancing customer value and, as a result, shareholder value. The authors explored definitional aspects of CRM, and they identified three alternative perspectives of CRM. The authors emphasized the need for a cross-functional, process-oriented approach that positions CRM at a strategic level. Kumar and Reinartz (2006) explained that Customer Relationship Management (CRM) deals with applying database marketing techniques at the customer level to develop strong company-to-customer relationship. CRM involves identifying different types of customers and developing specific strategies for interacting with each customer. Examples of such strategies are developing better relationships with profitable customers, locating and enticing with new customers that will be profitable and finding appropriate strategies for unprofitable customers. Chieko Minamia and John Dawson (2008) focused on the effects of customer satisfaction with CRM, customer retention and profit management, and the effects of CRM technique on performance. CRM is regarded as the integration of relationship technology (i.e. data consolidating and data mining) with loyalty schemes. In this analysis a direct effect of CRM implementation on return on equity (ROE) was supported; however, a negative impact of customization on ROE was found. 17 Robert Gee (2008), highlighted that organizations must understand what drives both value and delight for their customers. Adopting a customer centric vision enables an organization understand their customers, deliver customer delight and drive for loyalty. Different customers have different requirements and will be delighted in different ways. Database segmentation and data analysis are critical if an organization is to generate loyalty from different customer segments. A win-back strategy is recommended as previous customers are less costly to win-back compared to the costs of acquiring of new customers. Hyunchul Ahna et al (2011) points out that as the competition between mobile telecom operators become severe; it becomes critical for operators to diversify their business areas. Especially, the mobile operators are turning from traditional voice communication to mobile value-added services (VAS), which are new services to generate more average revenue per user (ARPU). That means, cross- selling is critical for mobile telecom operators to expand their revenues and profits. In this study, the authors propose a customer classification model, which may be used for facilitating cross-selling in a mobile telecom market. This model uses the cumulated data on the existing customers including their demographic data and the patterns for using old products or services to find new products and services with high sales potential. 2.3 SERVI CE QUALI TY 2.3.1 Service Quality Definition based on Functional Quality According to Berry et al (1988), Service Quality is defined as global judgement or attitude relating to the superiority of the service. Bitner, Booms and Tetreauly (1990) defines it as the customer's overall impression of the relative inferiority /superiority of the organization and its services. Carman (1990) describes the replication and testing of the SERVQUAL battery (A. Parasuraman et al; see record 1986-10681-001), which measures the perceived quality of a service situation. The scale was tested in 4 service settings different from those of the original test: a dental school patient clinic, a business 18 school placement center, a tire store, and an acute care hospital. Six basic questions of interest to the retailer were discussed: (1) the number of dimensions and how generic they are, (2) the extent to which item wording can be changed, (3) service situations that include multiple service functions, (4) the validity of analyzing differences between expectations and perception, (5) the point at which expectation information should be obtained, and (6) the relationship between expectations and importance. Asubonteng and Swan (1996) define it as the difference between customers expectations for service performance prior to the service encounter and their perceptions of the service received. These studies primarily focused on functional quality aspects (i.e., pertaining to service delivery process or how the services are delivered) and inadequately addressed technical quality aspects (i.e., issues concerning what is actually delivered). However, researchers in cellular mobile communication emphasized that technical quality attributes play an important role in forming service quality perceptions of customers. 2.3.2 Positive Relationship between Service Quality with Customer Satisfaction Danaher and Gallagher (1997) used the study of an actual hotel service delivery process and partitioned into five distinct service encounters; check-in, the room, the restaurant, the breakfast and check-out and investigated how quality factors were related to their respective encounters and how cumulative satisfaction levels impact on each other and over time. Average satisfaction levels for each of the five encounters were found to be significantly different. Moreover, there was a clear trend in the cumulative satisfaction results. Check-in resulted in high satisfaction, the room was not so satisfying and the restaurant rated the worst. Satisfaction scores rose after the breakfast experience and rose again after check-out. 19 Ko de Ruyter et al (1997) argued that there is a conceptual overlap as well as distinctions between, the role of expectations and perceptions and the questions whether service satisfaction is a super ordinate concept to quality or vice versa. They developed an empirical model which was tested in a health care setting to determine the nature of the relationship between service quality and service satisfaction. The results suggest that service quality should be treated as an antecedent of service satisfaction. It was also found that service satisfaction, in addition to service quality, is a direct function of disconfirmation and perception. Christine and Binks (1999) emphasized that within any service there is scope for considerable variation in the degrees to which both parties become involved in the relationship; beyond a certain minimum level, customers and service providers may be more or less participative. However, participative behavior should yield benefits. Customers who are more willing to share information and develop closer personal contacts might be expected to benefit in terms of a higher quality of service provision, because the provider will be more knowledgeable about their needs and expectations. Leisen and Vance (2001) emphasized that the successful standardization of service quality in the telecommunication industry across multiple nations demand, as a precondition, that the countries in question expose similar service quality dimensions and that the importance of these dimensions to overall satisfaction with the service is also similar. The authors investigated if these conditions were met in the telephone services sector of the telecommunication industry in the USA and Germany. Confirmatory factor analysis results of various alternative dimensionalities provided the best support for a five-dimensional conceptualization in both countries. However, the two countries differ in their respective importance evaluations of particular service quality dimensions, which make it difficult to provide a standardized service quality solution. 20 2.3.3 Positive Relationship between Service Quality with Customer Loyalty and Retention Ranaweera and Neely (2003) presented a holistic model of customer retention incorporating service quality perceptions, price perceptions, customer indifference and inertia. Data from a large-scale postal survey of telephone users in England showed that perceptions of service quality have a direct linear relationship with customer retention even in mass services with low customer contact. Price perceptions and customer indifference too were found to have a direct linear effect on retention. 2.3.4 Positive Relationship between Service Quality with Profitability Bloemer et al (1999) et al, argued that the relationship between perceived service quality and service loyalty is an issue which requires conceptual and empirical elaboration through replication and extension of current knowledge. The authors focus on the refinement of a scale for measuring service loyalty dimensions and the relationships between dimensions of service quality and these service loyalty dimensions. The results of an empirical study of a large sample of customers from four different service industries suggest that four dimensions of service loyalty can be identified: purchase intentions, word-of-mouth communication; price sensitivity; and complaining behaviour. 2.3.5 SERVQUAL in Telecom Wang and Lo (2002) identified that service quality, customer satisfaction and customer value are the most important factors of business success for either manufacturers or service providers. The authors have paid attention to the measurement model of service quality in Chinas mobile phone market based on the well-known SERVQUAL model, but with modification on the basis of focus group discussions and expert opinions to reflect the specific industry attributes and the special culture of China. Emphasis is then paid to the study of the dynamic relationships among service quality, customer value, customer satisfaction and their 21 influences on future behaviors after the key drivers of customer value and customer satisfaction are identified. J ohnson et al (2002) provided an overview of service quality and discussed its potential for offering a competitive advantage; to test several research propositions concerning service quality in the Thai telecommunications industry; the results indicated that perceptions and expectations of service quality level showed no significant difference. A post hoc analysis found that the telecommunication industry received excellent ratings on tangibles, particularly customer service staffs dress, and low ratings on empathy, particularly service providers interest differences. Tangibles are an aspect of service quality that is extremely important to the Thai telecommunication customer. According to Hannikainen et al, (2002), service quality is capability of a network to provide services and to fulfill user's expectations. According to Telecom Authority of India (2007), service quality is an indicator of performance of a network and of the degree to which the network conforms to the stipulated norms. 2.3.6 Service Quality Measurement and Models Berry et al (1990) conducted a study on improving customer service quality. They identified five dimensions that customers use to judge a company's service; discussed the potential causes of service role ambiguity and identified the possibilities that arise when a customer experiences a service problem. According to their study, the principal dimensions that customers use to judge a companys service are: Tangibles - The appearance of physical facilities, equipment, personnel, and communication materials. Reliability - The ability to perform the promised service dependably and accurately. 22 Responsiveness - The willingness to help customers and to provide prompt service. Assurance - The knowledge and courtesy of employees and their ability to convey trust and confidence. Empathy - The provision of caring, individualized attention to customers. Berry and Parasuraman (1992) emphasized that a well designed and implemented service quality information system raises the possibility that a company will invest service improvement money in ways that actually improve service. The five elements of the service quality information system are; (i) measure service expectations, (ii) emphasize information quality, (iii) capture customers' words, (iv) link service performance to business results, and (v) reach every employee. Carvalho and Leite (1999) extended the Parasuraman Berry - Zeithaml conjecture to assess the quality of postal services in Brazil. A qualitative stage yielded a list comprising 39 attribute items. In the quantitative stage the three- column format of a SERVQUAL questionnaire was employed to permit the computation of importance weights and tolerance widths for each attribute item. There was an inverse association between importance and tolerance of service quality attributes. Hanjoon Lee et al (2000) explained that service quality is an elusive and abstract construct to measure, and extra effort is required to establish a valid measure. The authors investigated the psychometric properties of three different measurements of health-care service quality as assessed by physicians. The multitrait-multimethod approach revealed that convergent validity was established for measures based on the single-item global rating method and multi-item rating method. Furthermore, discriminant validity for the seven health-care service quality dimensions measured by the three methods was not well established. The high levels of inter-dimensional correlations found suggested that the service quality dimensions may not be separable in a practical sense. 23 Michael K Brady et al (2002) described that service quality be measured using a performance-only index (SERVPERF) as opposed to the gap-based SERVQUAL scale. The intent of the research was to examine the ability of the performance-only measurement approach to capture the variance in consumers' overall perceptions of service quality. The results of their studies indicate that service quality is properly modeled as an antecedent of satisfaction. Parasuraman (2005) used the means-end framework as a theoretical foundation to conceptualize constructs refines, and tests a multiple-item scale (E-S-QUAL) for measuring the service quality delivered by Web sites on which customers shop online. Using two stages of empirical data collection, the study revealed that two different scales were necessary for capturing electronic service quality. The basic E-S-QUAL scale developed in the research is a 22-item scale of four dimensions: efficiency, fulfillment, system availability, and privacy. The second scale, E-RecS-QUAL, is salient only to customers who had nonroutine encounters with the sites and contains 11 items in three dimensions: responsiveness, compensation, and contact. Mahapatra and Khan (2006) provided a systematic integrated approach for modeling customer evaluation of service quality applied to technical education. The authors identified that the quality of service largely relates to human behaviour, the quality dimensions and items under each dimension of the measuring instrument widely differ depending on the application to the type of service setting. The stakeholders in an educational setting range from students to recruiters, with varying levels of interaction with the system and expectations from the system. Therefore, it is advisable to identify the minimum number of service items that suit all the stakeholders before implementing any quality improvement programme. To this end, EduQUAL, a survey-based model, has been specially developed to suit a technical education system. Anita Seth et al (2008) developed a valid and reliable instrument to measure customer perceived service quality incorporating both service delivery as well as technical quality aspects. The resulting validated instrument comprised of 24 dimensions including reliability, responsiveness, assurance, empathy, tangibles, convenience, and customer perceived network quality. 2.3.7 Service Quality Dimensions Researchers have tried to operationalize service quality from different perspectives for different service applications. Based on their conceptual / empirical studies, researchers derived and proposed different service quality dimensions for various service applications. Berry et al (1988), identified five dimensions namely, reliability, responsiveness, assurance, empathy and tangibles. The application areas are telecom companies, brokerage and insurance companies and banks. Lehtinen and Lehtinen (1991) identified five dimensions namely, physical quality, corporate quality, interactive quality, process quality and output quality. The application areas identified are restaurants, disco and pub restaurants. Rosen and Karwan (1994) identified six dimensions namely, reliability, responsiveness, tangibles, access, knowing the customer and assurance. The application areas are teaching, restaurant, bookstore and health care. J ohnson et al (1995) identified three dimensions namely, input quality, process quality and output quality. This study was based on bank customers in the UK. Siu and Cheung (2001) identified six dimensions namely, personal interaction, policy, physical appearance, promises, problem solving and convenience. The study was based on the service quality delivery of a department store chain. Alzola and Robaina (2005) identified five dimensions namely, reliability, design, guarantee, empathy and security. This study was based on Electronic Commerce B2C. Improvements in service quality can also enhance 25 competitiveness. Several dimensions of competitiveness have become relevant in India and have been researched across levels. 2.4 SERVICE LOYALTY 2.4.1 Defining Service Loyalty Service loyalty is the degree to which a customer exhibits repeat purchasing behavior from a service provider, possesses a positive attitudinal disposition toward the provider, and considers using only this provider when a need for this service arises. This definition suggests service loyalty is a matter of degree, ranging from the completely loyal customer to one who will never consider using a provider in the future. According to this definition, an extremely loyal customer is one who (a) regularly uses a service provider, (b) really likes the organization and thinks very highly of it, and (c) does not ever consider using another service provider for this service. Conversely, an extremely non-loyal person (a) will never use the provider again, (b) has negative feelings toward the organization, and (c) welcomes suggestions about other providers and is willing to try any other provider. This three-dimensional definition is consistent with Zeithaml, Berry, and Parasuramans (1996) operationalization of the loyalty to company factor in their behavioral- intentions battery. The five items they use to measure loyalty include (1) saying positive things about the company, (2) recommending the company to someone who seeks advice, (3) encouraging friends and relatives to do business with the company, (4) considering the company the first choice to buy services, and (5) doing more business with the company in the next few years. Thus, in effect, their measure includes items from all three dimensions of the proposed service loyalty definition listed above. Jacoby et al (1978) have explored the psychological meaning of loyalty in an effort to distinguish it from behavioral (i.e., repeat purchase) definitions. Their analysis concludes that consistent purchasing as an indicator of loyalty could be invalid because of happenstance buying or a preference for convenience and that 26 inconsistent purchasing could mask loyalty if consumers were multi-brand loyal. Because of these possibilities, the authors conclude that it would be unwise to infer loyalty or disloyalty solely from repetitive purchase patterns without further analysis. Czepiel and Gilmore (1987) emphasized that essentially social nature of service encounters, a short-run phenomenon, provides the occasions in which buyer and seller negotiate the terms of their exchange relationship, a long-run phenomenon. Defined as the mutual recognition of special status between exchange partners, exchange relationships insure efficacy for the buyer, as they mitigate market volatility for the seller. Understanding how economic exchange is played out against a background of social exchange can yield actionable insights. Murray (1991) explored the information needs of service consumers. In the purchase decision process, search behavior is motivated in part by perceived risk and the consumer's ability to acquire relevant information with which purchase uncertainty can be addressed. Marketing theory suggests that consumers use information sources in a distinctive way to reduce the uncertainty associated with services. Dick et al (1994) explained that Customer loyalty is viewed as the strength of the relationship between an individual's relative attitude and repeat patronage. The relationship is seen as mediated by social norms and situational factors. Cognitive, affective, and conative antecedents of relative attitude are identified as contributing to loyalty, along with motivational, perceptual, and behavioral consequences. Kandampully (1998) explained the premise of `quality of service as the competitive edge in gaining market leadership which has been well recognized both in academic research and by leading service organizations. However, it has become increasingly important for organizations to find ways, not only to reach the top, but to maintain that leadership in an ever increasing competitive market-place. In order to protect their long-term interest, service organizations are seeking ways to forge 27 and to maintain an on-going relationship with their customers. It is argued that long- term superiority of a service firm is dictated by the organizations ability to maintain their relationship with the customer by offering `service loyalty: a demonstration of the organizations commitment to maintain the service promise. The author argues that service loyalty precedes customer loyalty. The author emphasizes how a firms service employees develop the emotional connection with customers which leads to exceptional service and the ability to exceed customer expectations. Gustafsson et al (2005) in their study of telecommunications services, examine the effects of customer satisfaction, affective commitment, and calculative commitment on retention. The study further examines the potential for situational and reactional trigger conditions to moderate the satisfaction-retention relationship. The results support consistent effects of customer satisfaction, calculative commitment, and prior churn on retention. Prior churn also moderates the satisfaction-retention relationship. Recently to evaluate Service Loyalty, a Servloyal construct based on seven dimensions has been introduced by Sudhahar et al (2006). The authors emphasize that besides the behavioural and attitudinal dimensions there figure cognitive, conative, affective, trust and commitment dimensions. 2.4.2 Dimensions of Service Loyalty A review of the literature suggests the service loyalty construct consists of three separate dimensions: behavioral loyalty, attitudinal loyalty, and cognitive loyalty. Behavioral Loyalty: Early definitions of loyalty focused almost exclusively on its behavioral dimension. In particular, loyalty was interpreted as a form of customer behavior (such as repeat purchasing) directed toward a particular brand over time. Although current thought infers that loyalty includes more than just a behavioral dimension, some researchers continue to measure loyalty exclusively on the behavioral dimension. 28 Attitudinal Loyalty: Scholars have questioned the adequacy of using behavior as the sole indicator of loyalty. Day (1969), in particular, criticized behavioral conceptualizations of loyalty and argued brand loyalty develops as a result of a conscious effort to evaluate competing brands. Others have suggested this attitudinal dimension includes consumers preferences or intentions (e.g., Pritchard 1991). After Days criticism, attitude gained increasing attention as an important dimension of loyalty. Over time, scholars began to consider customer loyalty as having two dimensions: behavioral and attitudinal. Cognitive Loyalty: In additional to the behavioral and attitudinal dimensions, a few scholars include what has been termed a cognitive form of loyalty (Lee and Zeiss 1980). Some studies suggest loyalty to a brand or store means it comes up first in a consumers mind when the need for making a decision as to what to buy or where to go arises, while others operationalize loyalty as a customers first choice among alternatives. Similarly, Dwyer, Schurr, and Oh (1987, p. 19) argue that being committed to a relational exchange virtually precludes considering other exchange partners)) such customers have not ceased attending to alternatives, but maintain their awareness of alternatives without constant and frenetic testing. This suggests alternative organizations are not seriously considered by truly loyal customers when subsequent purchases are made) a viewpoint strongly supported by others. That is, a customer who is considered extremely loyal does not actively seek out or consider other firms from which to purchase. 2.5 LOYALTY INDI CES Bob E. Hayes (2007) felt while many objective measures of customer loyalty exist (e.g., defection rate, number of referrals), customer surveys remain a frequently used way to assess customer loyalty. There are a few reasons for the popularity of customer survey use in customer experience management. First, customer surveys allow companies to quickly and easily gauge levels of customer loyalty. Companies may not have easy access to objective customer loyalty data or may simply not even gather such data. Second, results from customer surveys can be 29 more easily used to change organizational business process. Customer surveys commonly include questions about customer loyalty as well as the customer experience (e.g., product, service, support). Used jointly, both business attribute items and loyalty indices can be used (e.g., driver analysis, segmentation analysis) to identify reasons why customers are loyal or disloyal. Finally, objective measures of customer loyalty provide a backwards look into customer loyalty levels (e.g., defection rates, repurchase rates). Customer surveys, however, allow companies to examine customer loyalty in real-time. Surveys ask about expected levels of loyalty- related behavior and lets companies look into the future regarding customer loyalty. Customers' ratings of a set of loyalty questions suggest that there are three, very general, loyalty constructs, Advocacy, Purchasing and Defection; Advocacy Loyalty : reflects the degree to which customers will advocate of the company Purchasing Loyalty : reflects the degree to which customers will increase their purchasing behavior Defection Loyalty : reflects the degree to which customers will show symptoms of defecting to competitors The evidence from previous studies (Bob E. Hayes, 2007) shows that the Advocacy Loyalty Index (ALI) ,the Purchasing Loyalty Index (PLI) and Defection Loyalty index (DLI) measure three different types of loyalty. Even though these types of loyalty are correlated (advocates tend to be purchasers), the relationship between the ALI and PLI is not perfect, suggesting that these loyalty indices measure unique constructs. Customer loyalty is not a unidimensional construct, but rather a multidimensional construct that can help reliability measured. When we say a customer group has high vs. low loyalty, we need to clarify to which loyalty we are referring. It is possible that a given customer group can have different levels of loyalty (e.g., high advocacy, low purchasing). It is clear that a blanket statement about levels of "customer loyalty" can be ambiguous. 30 2.6 LI TERATURE RELATI NG TO LI NKAGE OF SERVI CE QUALI TY, SATI SFACTI ON AND LOYALTY Torsten J. Gerpott et al (2001) have done a study on the relationship between customer satisfaction (CS), customer retention (CR) and customer loyalty (CL) in the German Mobile Telecom Market. They developed a two-staged model in which overall CS has a significant impact on CL which in turn influences a customer's intention to terminate/extend the contractual relationship with his mobile cellular network operator (CR). They identified mobile service price and personal service benefit perceptions as well as (lack of) number portability between various cellular operators as supply-related variables which has strong effects on CR. Mobile network operators' perceived customer care performance had no significant impact on CR. Chih-Ping-Wei and I-Tang-Chiu (2002) studied on churn prediction in telecom. The authors emphasized that a mobile service provider wishing to retain its subscribers needs to be able to predict which of them may be at-risk of changing services and will make those subscribers the focus of customer retention efforts. In response to the limitations of existing churn-prediction systems and the unavailability of customer demographics in the mobile telecommunications provider investigated, they proposed a churn-prediction technique that predicts churning from subscriber contractual information and call pattern changes extracted from call details. The technique is capable of identifying potential churners at the contract level for a specific prediction time-period. Moon-Koo Kim et al (2004) had studied the effects of customer satisfaction and switching barrier on customer loyalty in Korean mobile telecommunication services. The study was undertaken when the Korean telecom industry was shifting its strategic focus away from attracting new customers, towards retaining existing customers through the promotion of customer loyalty. J ohn Hadden et al (2007) analyzed that a business incurs much higher charges when attempting to win new customers than to retain existing ones. As a 31 result, much research has been invested into new ways of identifying those customers who have a high risk of churning. However customer retention efforts have also been costing organisations large amounts of resource. In response to these issues, the authors suggested that the next generation of churn management should focus on accuracy. Sung Ho (2007) used the customer relationship management perspective to investigate customer behavior. The authors differentiate between customers through customer segmentation; tracks customer shifts from segment to segment over time, discovers customer segment knowledge to build an individual transition path and a dominant transition path, and then predicts customer segment behavior patterns. Koustuv Dasgupta et al (2008) examined the communication patterns of millions of mobile phone users, to study the underlying social network in a large- scale communication network. The key purpose is to address the role of social ties in the formation and growth of groups, or communities, in a mobile network. In particular, the study focused on the evolution of churners in an operators network spanning over a period of four months. Based on their findings, they proposed a spreading activation-based technique that predicts potential churners by examining the current set of churners and their underlying social network. The efficiency of the prediction is expressed as a lift curve, which indicates the fraction of all churners that can be caught when a certain fraction of subscribers were contacted. Ying-Feng Kuo et al (2009), studied the relationships among service quality, perceived value, customer satisfaction, and post-purchase intention in mobile value-added services. The main findings are as follows: (1) service quality positively influences both perceived value and customer satisfaction; (2) perceived value positively influences on both customer satisfaction and post-purchase intention; (3) customer satisfaction positively influences post-purchase intention; (4) service quality has an indirect positive influence on post-purchase intention through customer satisfaction or perceived value; (5) among the dimensions of service quality, customer service and system reliability is most influential on perceived 32 value and customer satisfaction, and the influence of content quality ranks second; (6) the proposed model is proven with the effectiveness in explaining the relationships among service quality, perceived value, customer satisfaction, and post-purchase intention in mobile added-value services. Fujun et al (2009) proposed and tested an integrative model to examine the relationship among service quality, value, image, satisfaction, and loyalty in China. The authors reveal that service quality directly influences both perceived value and image perceptions, that value and image influence satisfaction, that corporate image influences value, and that both customer satisfaction and value are significant determinants of loyalty. Thus, value has both a direct and indirect (through satisfaction) impact on customer loyalty. Other variables mediate the impact of both service quality and corporate image on customer loyalty. 2.7 I NSI GHTS AND I NADEQUACI ES The review of literature has provided insights on the following areas: 1. The impact of privatization of telecom sector on the mobile service providers and the subsequent penetration into the market has given rise to cut-throat competition among the mobile service providers. 2. This rising level of competition in turn has forced the mobile service providers to focus on various customer centric strategies wherein the service providers concentrate on sustained competitive advantage. 3. Customers on the other hand are well educated, demanding and willing to spend for better value for money. This and the above aspects have turned the attention of the service providers to depend on a technology-driven strategies where there is a need to collect, maintain, manage and analyse customer details and their purchasing behaviour which would help them to bundle up offers which would satisfy a specific need of a specific segment of customer in a specific 33 period of time. This is the Customer Relationship Management, a contemporary management practice. 4. Though CRM evolved in a very short duration, its application is widespread in services which involve huge set of customer base. 5. The impact of CRM practices in various countries and on various industry verticals gives a positive insight as to the level of maturity of its practices. 6. When deciding to choose a telecommunication service provider, customers often take into account the service quality, service price and customer service. While the first two factors can be controlled by the telecom company, the last factor seems to be the most influential and the hardest to get right as well. One outcome of good customer service, which depends on the service quality of the telecom providers positively results in a certain degree of behavioral attitude within the customers which in turn grows into loyalty. Thus a loyal customer may turn advocate to the company; purchase more frequently, purchase more quantity or purchase more variety of services or if disloyal switch to competitors. The literature reviewed above indicates certain inadequacies in the following areas: 1. While there are individual studies on CRM implementations (Payne & Frow, Finnegan & Willcocks, Robert Gee), Service Quality (Danaher and Mattsson,Wang and Lo, Johnson and Sirikit,Gale, Rudie and Wansley,Zeithaml & Bitner), Relationship between CRM and Customer Loyalty (Mosad Zineldin, Majumdar, Terblanche and Boshoff, Gustafsson et al) and Loyalty Indices (Bob E. Hayes), it is noticed that a study on the linkage especially with the changing demographic patterns of the customers, if any, arising out of these concepts seems to be missing. Hence it is felt that such an existing 34 gap can be further explored and studied and the present attempt is towards such a linkage. It is also a fact that the CRM practices in Indian Telecom is in the matured stage. Hence such a linkage study particularly in the telecom industry would be timely to provide managerial insights in order to further strengthen the CRM practice in this highly competitive industry. 2. Identifying the influence of service quality dimensions, service loyalty dimensions on each of the loyalty indices, namely advocacy loyalty index, purchase loyalty index and defection loyalty index. 3. Even though some of these aspects have been explored by previous researchers to a great extent, it is felt that not much attention has been drawn into measuring the effectiveness of loyalty, particularly for sustained presence. 2.8 SUMMARY This chapter provides an overview of the literature reviewed by the researcher. The crux of each material reviewed by the researcher is presented chronologically. It also includes a section on the insights obtained and inadequacies present in the literature review, which the present study is aiming to bridge. 35 CHAPTER 3 RESEARCH METHODOLOGY 3.1 I NTRODUCTION Methodology is the key to finding answers to the question that initiates the research and therefore comprises a very important part of any study. The research methodology that is chosen has to enhance the reliability and generalization of the results. It should furthermore be cost effective, efficient and versatile. This balance can only be obtained by using the appropriate methodology for gathering data. As previous research on strengthening of CRM practices by the Indian mobile service providers through customer loyalty are limited, the current study can be defined as an early attempt with the aim to develop the body of knowledge regarding the existing phenomenon. 3.2 RESEARCH DESIGN The research design adopted for this study was descriptive in nature. Descriptive research is a type of conclusive research that has its major objective as the description of something- usually market characteristics or functions (Malhotra, 2006). The current research used a quantitative approach. Quantitative research methodology seeks to quantify the data and typically applies some form of statistical analysis (Malhotra, 2006). The benefits of quantitative research are that researchers rely on large samples to show statistical effect and can generalize the findings from the sample to the population (Churchill and Iacobucci, 2002). 36 3.2.1 Area of the Study The data for testing the proposed models were collected among the resident mobile subscribers in Chennai. Chennai, capital of Tamil Nadu is the fourth most populous metropolitan area and the fifth most populous city in India. The population of Chennai is 4.6 million (2011 census). 3.2.2 I nstrument Development The instrument was designed using scales from previous related research. Since this study attempts to verify the relationships between the constructs, quantitative research methodology is employed. Instrument is in Appendix I. Scales for the independent and the dependent variables were adopted from the extant literature due to their relevance to the study's context and past measurement reliability. 3.2.2.1 Variables Considered for the Study In order to comply with the objectives laid down for the study, certain independent variables were considered for analyzing behavioral pattern of the mobile subscribers, which are as follows: Demographic Variables Extended Service Quality Parameters Service Loyalty Parameters Loyalty Indices 3.2.2.2 Demographic Variables The Demographic Variables considered for the study are: Age Length of Use 37 Monthly Expenditure Education Occupation 3.2.2.3 Extended Service Quality Dimensions The service quality aspects of the mobile service providers were analyzed based on the following dimensions. These dimensions are identified based on the SERVQUAL Construct. This extended SERVQUAL (Seth et al, 2008) instrument determines service quality structure by combining both functional as well as technical quality (i.e., network quality in cellular mobile context) attributes. Thus the extended SERVQUAL construct has the following seven dimensions: Reliability - refers to the ability to perform the promised service dependably and accurately. Contact employees perform the service right the first time The service provider provides the service at the promised time The customer is kept well-informed about the progress of their complaints Billing system is accurate and error free Bills are received on time. Responsiveness - refers to willingness to help customers and provide prompt service Contact employees give you prompt service The customers complaints / queries are taken seriously The customers complaints are resolved quickly The service provider is always willing to help you. 38 Assurance - refers to knowledge and courtesy of employees and their ability to inspire trust and confidence Contact employees are friendly and polite while handling your complaints / queries They have the adequate knowledge of tariffs and plans of service providers The behaviour of contact employees instils confidence in you You feel safe in your transactions with your service provider Empathy - refers to caring, individualized attention the firm provides its customers For lodging the complaints service provider is easily accessible They have your best interests in heart Contact employees give you individual attention The employees understand your specific needs Tangibles - refers to appearance of physical facilities, equipment, personal and communication materials Service provider's physical facilities are visually appealing Contact employees appear neat Materials associated with the service(such as pamplets etc) are visually appealing Convenience- implies flexible and comfortable facilities to suit the customer needs. They have convenient business hours Ease of lodging the complaints / queries Your service provider provides flexibility in the payment of bills Application formalities are simple 39 Customer Perceived Network Quality - an indicator of network performance in terms of voice quality, call drop rate, network coverage and network congestion. Your service provider provides sufficient geographical coverage (on highways, inside the buildings and basement) You experience minimum premature termination of calls during conversation (ie call drops) You get clear and undisturbed voice Your call gets connected to the called person during the first attempt most of the time You are able to make calls at peak hours 3.2.2.4 Service Loyalty Dimensions According to (Chaudhuri and Holbrook, 2001; Bandhopadhyay and Martell, 2007), Service Loyalty has two dimensions, namely, Behavioural Loyalty and Attitudinal Loyalty. Recently, to evaluate Service Loyalty, a Servloyal construct (Sudhakar et al, 2006) based on seven dimensions has been introduced. Besides the behavioural and attitudinal dimensions, there figures cognitive, conative, affective, trust and commitment dimensions. The seven dimensions are explained with the items. Behavioural Loyalty - refers to the customers willingness to continue the relationship with the provider in the short period. It does not imply a commitment to the provider, but simply expresses the degree of loyalty in the immediate future. I will transact with this mobile service provider again for future needs I will try new services that are provided by the service provider I will recommend other people to patronize this service provider 40 I will say positive things to other people about the services provided by the mobile service operator Attitudinal Loyalty - refers to the predisposition towards the provider deriving from a psychological process. I will continue to patronize this mobile service provider / service even if the service charges are increased moderately I have strong preference to this service provider I will continue to use this service provider regardless of changes in the service I am likely to pay little more for the services when situation arises Cognitive Loyalty - is the psychological state of mind which is based on the information available to the consumer about the service performance which indicates that one service provider is preferable to its alternatives. Cognition can be based on prior or vicarious knowledge or on recent experience-based information. Loyalty at this state of mind is directed toward the brand because of this "information" (attribute performance levels). To me, this service provider would rank first among others I would continue this service for a long period of time I will deal exclusively with the service provider The service of the provider reflect a lot about who I am Conative Loyalty - is the phase of loyalty development which is influenced by repeated episodes of positive affect toward the service provider. Conation, by definition, implies a brand-specific commitment to repurchase. Conative loyalty, then, is a loyalty state that contains what, at first, appears to be the deeply held commitment to buy. In effect, the consumer desires to repurchase, but similar to any "good intention," this desire may be an anticipated but unrealized action. 41 I have found this provider better than others Repeatedly, the performance of this provider is superior to that of its competitors Affective Loyalty refers to the state when the customer feels involved with the service provider. Affective loyalty is the state of mind that is developed as a result of a liking or attitude toward the service provider on the basis of cumulatively satisfying usage occasions. This reflects the pleasure dimension of the satisfaction definition and pleasurable fulfillment. The loyalty exhibited is directed at the degree of affect (liking) for the service provider. I like the terms of the service provider I like the performance and services of the provider I have a positive attitude toward this provider I am satisfied with my decision to stay with this provider Trust loyalty - refers to the state which is developed as a result of the belief on the competence of the service provider. If the service provider has the required expertise, to perform his/her activities, carry out his/her obligations or accomplishes his/her promises, the consumer gets a feeling of security about the service provider. The employees of the service provider give individual attention The employees understand my specific needs and go out of their way to help me The employees respond caringly when I share my problems The personnel at the service provider are filled with professionalism and dedication Commitment Loyalty - regards the deepest involvement of the customer in the relationship with the provider. The customers show high degree of continued association with the service provider. 42 I am committed to the service provider Even when I get to hear any negative information about the provider, I would still continue with this service provider I like switching from one service provider to another My continued association with the service provider is important to me. 3.2.2.5 Loyalty I ndices Bob E. Hayes, 2007 identifies key drivers of loyalty for service providers and introduces new customer loyalty metrics designed to help companies increase revenue through new and existing customers. The instrument to measure loyalty indices were used to understand the respondents perception about their level of agreement with respect to questions, each on a five point Likert Scale (1 Strongly Disagree and 5 Strongly Agree). He has identified three loyalty indices namely, Advocacy Loyalty I ndex (ALI ) - reflects the degree to which customers will be advocates of the company. Likelihood to choose the service again Likelihood to recommend Likelihood to pass on positive feeling about the services Likelihood to continue purchasing the same products / services Purchasing Loyalty I ndex (PLI ) - reflects the degree to which customers will increase their purchasing behavior. Likelihood to increase frequency of purchasing Likelihood to purchase different products / services 43 Defection Loyalty I ndex (DLI ) - reflects the degree to which customers will show symptoms of defecting to competitors. Likelihood to continue this only as a supplementary connection Likelihood to switch to a different provider 3.2.3 Sampling Method Convenience sampling was used because the sampling units were accessible, easy to measure and required the least time and cost expenditure (Ko, 1995).. It also takes into account the operational difficulties associated in the process of data capturing. However, by going through this method of sampling it was ensured that all units of population were adequately represented. In determining the sample size the following factors were taken into consideration: Dispersion of the population Time taken by the respondents to complete the questionnaire Resources required completing the survey Respondents willingness to part with information. Burns and Bush (1999) believe that sample size affects the accuracy of results. Sample size also has a direct impact on the appropriateness of the statistical techniques chosen (Hair et al., 1998). The size of the sample for this research was designed in accordance with the criterion for applying the analytical technique chosen. Since this study used Structural Equation Modeling and regression a large sample was required for this multivariate technique. The sample size was fixed for around 550, as a small sample size might provide less accurate estimates of the degree of relationship among the variables due to unstable correlations (Bordens and Abbott, 1996). All variables in the study were identified through a review of the relevant research and literature. External validity refers to the researchers ability to generalize the results of the study from the sample to the population from which the 44 sample was drawn (Gay, 1987). One of the threats to external validity occurs if the sample is not representative of the population. The best protection for equal representation occurs when a sample has been randomly selected (Fink, 1995). Sampling error refers to chance variation among the means (Gay, 1987) due to influences not under the control of the researcher. Sampling errors are normally distributed and can be controlled if a sufficiently large number of subjects are selected from the population. Sampling error for this study was controlled through the selection of a sample large enough to be representative of the population. Measurement error refers to inconsistencies in measurements. The measurement error was controlled by employing instruments which have been shown to be reliable and valid. In order to address possible measurement error reliability tests were employed on the scale items. 3.3 PROPOSED CONCEPTUAL MODEL In developing the proposed research model, we examine and confirm the relationship between service quality, service loyalty and loyalty indices which are multidimensional variables. Also the influence of the demographic variables on each of service quality, service loyalty and loyalty indices are examined. These multidimensional variables become the latent factors in strengthening the CRM implementation by the Indian Mobile Service Providers. Figure 3.1 Conceptual Frame Work for the Study 45 3.4 PRE-TEST The purpose of questionnaire pre-testing is to ensure that item wording, flow of questions, suitability of measurement scales, instructions, and other aspects of the questionnaire are understandable. Churchill and Iacobucci (2002) stress that; data collection should never begin without an adequate pre-test of the instrument. Therefore, although the items used in this study were taken from the extant studies which had established the measurement, this step was still necessary. It was because those wordings have different meanings and connotations in different cultural contexts (Sekaran, 2003). Questionnaire pre-testing can help to rectify any inadequacies beforehand (Sekaran, 2003). After the questionnaire had been developed, the same was pre-tested using a pilot study with 50 subscribers in Chennai. Interviews with respondents indicated that the item wording was clear and easy to understand. In addition, analyses of descriptive statistics indicated no skewness or kurtosis. Scale reliability was performed on each measure. 3.5 DATA COLLECTION The questionnaires including covering letter, were personally distributed to customers of mobile services, during April-J une, 2011. Convenience sampling method was used to collect the data from customers. Further, the data was specifically collected from residential mobile customers, who had been using the services for at least six months. Finally, of the 550 surveys individually administered, 523 questionnaires were received at a response rate of 95 per cent. On further filtering, 490 responses were found to be completely filled. 3.6 SUMMARY OF RESEARCH METHODS The research design utilized is a quantitative research approach in cross- sectional time frame and the participants being residential mobile subscribers. Measurement scales used in this study were taken from prior studies. Questionnaires were pre-tested to confirm the clarity of the contents involved. Data were collected from residential mobile subscribers in Chennai. The sample size was 550. After the 46 data was collected from the survey, data was processed through several steps to fine- tune their quality and these steps are discussed in the following sections. 3.7 DATA EDITI NG, CATEGORI SI NG AND CODI NG This section describes briefly three steps to deal with the data before they were used for analysis. These steps included data editing, coding and categorizing. Their purpose is to ensure that the data are reasonably good and of assured quality for the later analysis (Sekaran, 2003). Data editing was firstly conducted to check the degree of questionnaire completion. This step was initially done at the same time when the respondents returned the questionnaires. The interviewers checked the completion of all questions on the spot and asked for further information if any question was unanswered. A second editing was conducted after the data were collected. Then coding was conducted to facilitate the entry of responses and avoid later confusion (Sekaran, 2003). Finally, data were keyed into the Microsoft Excel sheet by using actual scores marked by the respondents. There were no missing data. After the data were entered in the worksheet, analysis of the data would be conducted by importing the data into the SPSS data editor and into AMOS 19.0. The procedures for examining the data are described in the following section. 3.8 DATA ANALYSI S PROCEDURES This section firstly introduces the main multivariate technique adopted to test the hypothesis involving the constructs in this study. It then outlines the procedures of analysing the data and statistical techniques employed for the analysis. 3.8.1 Structural Equation Modeling Structural Equation Modeling (SEM) techniques are considered today to be a major component of applied multivariate statistical analyses and are used by biologists, economists, educational researchers, marketing researcher, medical researchers, and a variety of other social and behavioural scientists. SEM offers 47 researchers a comprehensive method for quantification and testing of theoretical models. Although the statistical theory that underlies the techniques appeared decades ago, a considerable number of years passed before SEM received the widespread attention it holds today. In its broadest sense, SEM models represent translations of a series of hypothesized cause-effect relationships between variables into a composite hypothesis concerning patterns of statistical dependencies (Shipley, 2000). The relationships are described by parameters that indicate the magnitude of the effect (direct or indirect) that independent variables (either observed or latent) have on dependent variables (either observed or latent), by enabling the translation of hypothesized relationships into testable mathematical models. Once a theory has been proposed, it can then be tested against empirical data. The process of testing a proposed theoretical model is commonly referred to as the confirmatory aspect of SEM (Raykov and Marcoulides, 2000). Another aspect of SEM is the so-called exploratory mode. This aspect allows for theory development and often involves repeated applications of the same data in order to explore potential relationships between variables of interest (either observed or latent). SEM can also be used to test the plausibility of hypothetical assertions about potential inter-relationships between constructs and their observed measures or indicators. Latent variables are hypothesized to be responsible for the outcome of observed measures. In other words, the score on the explicit questionnaire would be an indicator of the construct or latent variable. Researchers often use a number of indicators or observed variables to examine the influences of a theoretical factor or latent variable. It is generally recommended that researchers use multiple indicators (preferably more than two) for each latent variable considered in order to obtain a more complete and reliable picture than that provided by a single indicator (Raykov and Marcoulides, 2000) because both observed and latent variables can be independent or dependent in a proposed model. The main focus of the current research study is the validation of the hypothesized causal models on the influences of demographic variables, service 48 quality and service loyalty on customer loyalty, using the Structural Equation Modeling (SEM) technique. SEM is the most appropriate method for the current research problem due to the following characteristics: 1. Ability to represent unobserved concepts in the causal relationships as well as accounting for measurement error in the estimation process. The latent variables in this research were service quality dimensions namely, reliability, responsiveness, assurance, empathy, tangibles, convenience and customer perceived network quality; service loyalty dimensions namely, behavioural, attitudinal, cognitive, conative, affective, trust and commitment ; loyalty indices namely, advocacy loyalty index, purchase loyalty index and defection loyalty index. 2. Estimation of multiple and interrelated cause and effect dependence relationships. Of the two major measurement processes in SEM, the first measurement process is known as measurement model analysis. A measurement model is a sub model of SEM that: (1) specifies the indicators (observable or measurable variables) for each construct (unobserved concept or dimension) and (2) assesses the reliability of each construct for estimating the causal relationships (cause and effect relationships). The reliability is the degree to which the observed or measured indicators are consistent in their measurements. The second sub model of SEM is the structural model, which is basically the set of one or more dependence relationships linking the hypothesized models constructs. The hypothesized relationships are translated into a series of structural equations (similar to regression equations) for each dependent latent construct. In the next stage, SEM estimates series of such separate, but interdependent structural equations simultaneously. Anderson and Gerbing (1988) suggest a two-stage approach to causal modeling using SEM, in which the measurement model is first confirmed and then the structural model is built. If the measurement model provides an acceptable fit to 49 the data, the structural model then provides an assessment of the extent of relationships between the hypothesized constructs (Byrne, 1994). We have followed this two stage approach using the SEM software AMOS 19.0 with maximum likelihood estimation. 3.8.1.1 Measurement Model In the measurement model, we can assess the contribution of each scale item, as well as incorporate how well the scale measures the concept (its reliability) into the estimation of the relationships between dependent and independent constructs. This procedure is similar to performing a factor analysis of the scale items and using the factor scores in the regression. 3.8.1.2 Structural Model The structural model converts the hypothesized causal paths (or, proposed relationship between constructs) into a series of structural equations for each dependent construct. In this process, some dependent constructs become independent constructs in subsequent relationships, giving rise to an interdependent nature of the structural model. The extent of the causal effect of an independent construct on a dependent construct is known from the path coefficient values. The structural model result provides the estimated path coefficient values and calculated t value for each path coefficient. The estimated coefficients are checked for significance by looking at the significance of the statistic. 3.8.1.3 Models Goodness-of-Fit Assessment The purpose of assessing a models fit is to determine the degree to which the hypothesized model as a whole is consistent with the empirical data of the study. The literature offers a wide range of goodness of fit indices. Unfortunately, the literature is unequivocal as to which represent a good estimation of fit (Diamantopoulos and Siguaw, 2000). In part this might be due to the fact that some indices of fit are affected by the sample size, estimation method, model complexity, violation of the underlying assumptions of multivariate normality and 100-variable 50 independence (Byrne, 1998). As a result, since there is no agreement in the literature as to which are the best, multiple descriptive indices of fit were used to further assess the goodness-of-fit of the model. It has been suggested that the assessment of model adequacy must be based on multiple criteria that take into account theoretical, statistical, and practical considerations (Byrne, 1998). Two considerations in the selection of best fit indices were kept in mind. These involved a reference of most cited indices in the methodological literature on Confirmatory Factor Analysis CFA and the extent to which these indices are less affected by sample size, estimation procedures, violations of assumptions or some combinations of both (Chaudhuri, 1995, Diamantopoulos and Siguaw, 2000, Byrne, 1998). The first measure of fit included in the AMOS output is the chisquare value ( 2 ). The chi-square statistic is the traditional measure for evaluating overall model fits in covariance structure models and provides a test of perfect fit in which the null hypothesis is that the model fits the population data perfectly (Diamantopoulos and Siguaw, 2000). A statistically significant 2 causes rejection of the null hypothesis, implying imperfect model fit and possible rejection of the model. Thus the aim in confirmatory analysis and Structural Equation Modeling, contrary to conventional hypothesis testing procedures, is not to reject but to accept the null hypothesis. By consensus, a large value of 2 indicates that the model fit poorly with the population and subsequently small values of 2 correspond to good fit (J oreskog and Sorbom, 1996). Despite its importance in providing a formal significance test of the covariance structure hypothesis, 2 had led to problems of fit. The Chi-square statistic has been found to be very sensitive to both normality (particularly excessive kurtosis), and the size of the sample, and it also assumes that the model fits perfectly well in the population. This last point has an unrealistic impact in the empirical analysis since the model might fit well the sample but it is very unlikely to happen in the population. For the above reasons, instead of regarding 2 as a test statistic, the literature considered it as a goodness (or badness)-of-fit measure (Diamantopoulos and Siguaw, 2000). Among the common indicators used to judge a good fit of the model to the population were the Root Mean Square Residual (RMR) and the Root Mean 51 Square Error of Approximation (RMSEA). These indices are based on their discrepancy function and have been considered as important to supplement an investigators judgment. RMSEA is recognised as one of the most informative criteria in covariance structure modeling and considered the first Goodness of Fit - GFI to be reported in the study (Byrne, 1998). In addition, whenever using maximum likelihood as a unique method of estimation, RMSEA needs to be reported because it has been found to yield consistent results across estimation procedures when the model is well defined (Sugawara and MacCallum, 1993) and also is not biased by sample size (Widaman and Thomson, 2003). Furthermore, RMSEA is particularly important when determining the number of constructs because it decreases when an additional construct reduces the F (minimal population discrepancy function) substantially, but increases if including the additional construct only reduces F slightly (McNight, Choudhury and Kacmar, 2003). Values less than 0.05 are considered of as good fit and values as high as 0.08 represent reasonable errors of approximation in the population (MacCallum and Austin, 2000). It is also recommended to report the 90 percent interval around the RMSEA values to assess the precision of RMSEA estimates as such information helps the assessment of model fit (MacCallum, Browne and Sugawara, 1996). A good precision of the RMSEA value is indicated by a narrow confidence interval reflecting how well the model fits the population covariance matrix (Byrne, 1998). The Root Mean Square Residual (RMR) represents the average residual value derived from the fitting of the variance-covariance matrix for the hypothesised model to the variance-covariance of the sample data and a value of 0.05 or less is considered of a good fit (Chaudhuri, 1995). Although, RMR values are highly regarded, they are difficult to interpret due to their residual susceptibility to the sizes of the observed variances and covariances (Byrne, 1998). Additional information regarding the fit of the model was obtained from the values of the following indices of fit which were first introduced by J oreskog and Sorbom (1982). 52 Goodness-of-fit index (GFI) is an indicator of the relevant amount of variances and covariances accounted for by the model and thus shows how closely the model comes to perfectly reproducing the observed co-variance matrix (Diamantopoulos and Siguaw, 2000). Acceptable fit is obtained when values of the GFI are greater than 0.90 (Jreskog and Srbom, 1993). The GFI can be classified as an absolute index of fit because it basically compares the hypothesised model with no model at all (Chaudhuri, 1995). Information was collected from incremental or comparative indices that show the fitness of the default model compared to the independence model (null model). The incremental indices are considered better indicators of fit than the absolute indices because they exhibit relative independence from sample size (Widaman and Thomson, 2003). The study considered two indices under this category as suggested in the SEM literature (Bentler, 1990, Bentler and Bonnett, 1980). One is the Bentler-Bonnetts Normed Fit Index (NFI), regarded as practical choice. The second index, Comparative Fit Index (CFI) is a revised version of NFI that takes into consideration the size of the sample. The values for both indices range from 0 to 1 with values close to 1 representing a good fit (Bentler, 1990). A third Index recommended to be reported among the incremental fit indices is the Non-normed Fit Index (NNFI) also called Tucker-Lewis Index (TLI) (Byrne, 1998). Normed and Non-normed fit indices are recognised as very popular adjuncts to more traditional statistics in Structural Equation Modeling to help assess the quality of a model (Bentler, 1990). The final measure of fit is Hoelters (1983) critical N (CN) that is very different to the previously mentioned indices as it relates to the appropriate size a sample must have in order to accept the fit of a given model on a statistical basis. (Hoelter, 1983). A CN value >200 has been deemed an appropriate sample size (Hoelter, 1983). However, this cut-off value has been challenged as some considered higher CN values to yield an adequate fit of the model (Chaudhuri, 1995). 3.8.1.4 AMOS AMOS is an acronym for analysis of moment structures or, in other words, the analysis of mean and covariance structures. An interesting aspect of 53 AMOS is that, although developed within the Microsoft Windows interface, the program allows the researcher to choose from three different modes of model specification. Using the one approach, AMOS Graphics, one can work directly from a path diagram; using the others, AMOS VB.NET and AMOS C#, and one can work directly from equation statements. The choice of which AMOS method to use is purely arbitrary and bears solely on how comfortable the researcher feels in working within either a graphical interface or a more traditional programming interface. AMOS Graphics allows for the estimates to be displayed graphically in a path diagram. 3.8.2 Multiple Regression Multiple regression is a dependence technique which is most widely used to analyze the relationships between a single dependent variable and a set of independent variables (Hair et al., 1998). This method succeeds in measuring the joint influence of the explanatory variables on the dependent variable, and, for each of these explanatory variables, assesses the effect on the outcome variable that is attributable to that explanatory variable alone (Harraway,1995). It is a statistical tool that is usually used when both dependent and independent variables are metric (Hair et al., 1998) according to Pallant (2001), multiple regression is able to provide the information about the model as a whole (all subscales), and the relative contribution of each of the variables that make up the model (individual subscales). The application of multiple regression analysis falls into two broad categories of research problems: prediction and explanation (Hair et al. 1998). One basic purpose of regression analysis is to predict the dependent variable with a set of independent variables. In doing so, multiple regression fulfils one of the two objectives (Hair et al. 1998). The first objective is to maximize the overall predictive power of the independent variables as represented in the variate. It is explicitly designed to make errors of prediction as small as possible using the least squares criterion for overall smallness. The second objective is to compare two or more sets of independent variables to assess the predictive power of each variate. The second purpose of multiple regression analysis is to provide an assessment of the degree and 54 direction (positive or negative) of the linear relationship between independent and dependent variables by forming the variate of independent variables (Hair et al. 1998). Multiple linear regression analysis can be used to examine the effects of some independent variables on the dependent variable while controlling (i.e. held constant) for other independent variables. In order to calculate statistical predictions, the regression technique seeks to establish a rectilinear relationship between the variables concerned. Subsequently, the equation of a straight line is of important value, and is denoted by Y = bX + a where Y =predicted score; b =slope; X =X intercept and a =Y intercept. The following equation is therefore used when a number of predictor variables are employed to predict a criterion variable in multiple linear regression. Y =a +b1X1 +b2X2 + +bkXk +e where Y is the predicted value on the dependent variable, a is the Y intercept (the value of Y when all the X values are 0), X represents the various independent variables and b is the coefficients assigned to each of the independent variables during regression (Tabachnick and Fidell 2001). In this study, multiple regression analysis is the technique used to examine the relationship between Service Quality Dimensions and the Loyalty indices; Service Loyalty Dimensions and Loyalty Indices and Service Quality dimensions and Service Loyalty Dimensions in order to derive a path diagram. 3.8.2.1 Measurement of Variables An important issue in multiple regression analysis is whether the measurement of the dependent and independent variables is appropriate for this type of analysis. Multiple regression analysis accommodates only quantitative explanatory variables, measured on an interval or continuous scale. 55 3.8.2.2 Coefficient of Multiple Determination (R 2 ) The coefficient of determination (R 2 ) was used as an estimate of the predictive power of the regression model. R 2 values measure the percentage of the total variance of the dependent variable about its mean that is explained or accounted for by the independent variable (Lewis-Beck 1993). It is the squared product-moment correlation coefficient and its value can vary between 0 and 1. Thus, the closer this is to 1 the better the fit of the model (i.e. the better the independent variables are at accounting for or explaining the variation in the response variable) because if R 2 is 1 then the regression model is accounting for all the variation in the outcome variable. According to Hair et al. (1998), if the regression model is properly applied and estimated, it can be assumed that the higher the value of R 2 , the greater the explanatory power of the regression equation, and therefore the better the prediction of the dependent variable. However, there is no hard-and-fast statistical argument for deciding what level of R 2 is high enough. The R 2 value can only improve by adding more variables to the model, even when their contribution is very small or accidental. A development of the standard R 2 is adjusted-R 2 ; this is a measure of fit which take into account the number of independent variables and the sample size. While the addition of predictor variables will cause the R 2 to rise, the adjusted R 2 may fall if the added predictor variables have little explanatory power and are statistically insignificant (Hair et al. 1998; Newton and Rudestam 1999). Hence, adjusted R 2 is a less biased measure for the variance explained by the model; therefore, it was used in this study for the interpretations of explanatory power. It is to be noted that the predictive power of the regression is directly influenced by the sample size included in the regression analysis. The higher the sample size, the lower the R for a given number of independent variables at a given significance and power levels. This has been taken into consideration in assessing the overall model fit for the regression equations in this study. A further discussion on R 2 values with special reference to the results of regression analysis obtained in this study will be presented. 56 3.8.2.3 I nterpretation of Regression Variate After the final model has been derived and the predictive power of the regression model has been estimated, the final task is to interpret the regression parameter by evaluating the estimated regression coefficients for their explanation of the dependent variable. The regression coefficient captures the effect of one variable while controlling for (i.e. holding constant) the other variables in the model. Two coefficient values are of interest for interpretation. These are the unstandardised and standardised regression coefficients, or b and beta, respectively. The unstandardised regression coefficients (b-coefficients) represent the amount of change in the dependent variable associated with a one-unit change in that independent variable, with all other independent variables held constant (Newton and Rudestam 1999). 3.8.3 Data Envelopment Analysis Data Envelopment Analysis (DEA) is a linear programming based technique for measuring the relative performance of organisational units where the presence of multiple inputs and outputs makes comparisons difficult. There is an increasing concern with measuring and comparing the efficiency of organisational units such as local authority departments, schools, hospitals, shops, bank branches and similar instances where there is a relatively homogeneous set of units. The usual measure of efficiency, i.e.: Data Envelopment Analysis (DEA) is an increasingly popular management tool. DEA is commonly used to evaluate the efficiency of a number of producers. A typical statistical approach is characterized as a central tendency approach and it evaluates producers relative to an average producer In contrast, DEA 57 compares each producer with only the "best" producers. By the way, in the DEA literature, a producer is usually referred to as a decision making unit or DMU. In DEA, there are a number of producers. The production process for each producer is to take a set of inputs and produce a set of outputs. Each producer has a varying level of inputs and gives a varying level of outputs. For instance, consider a set of banks. Each bank has a certain number of tellers, a certain square footage of space, and a certain number of managers (the inputs). There are a number of measures of the output of a bank, including number of checks cashed, number of loan applications processed, and so on (the outputs). DEA attempts to determine which of the banks are most efficient, and to point out specific inefficiencies of the other banks. A fundamental assumption behind this method is that if a given producer, A, is capable of producing Y(A) units of output with X(A) inputs, then other producers should also be able to do the same if they were to operate efficiently. Similarly, if producer B is capable of producing Y(B) units of output with X(B) inputs, then other producers should also be capable of the same production schedule. Producers A, B, and others can then be combined to form a composite producer with composite inputs and composite outputs. Since this composite producer does not necessarily exist, it is typically called a virtual producer. The heart of the analysis lies in finding the "best" virtual producer for each real producer. If the virtual producer is better than the original producer by either making more output with the same input or making the same output with less input then the original producer is inefficient. The subtleties of DEA are introduced in the various ways that producers A and B can be scaled up or down and combined. To illustrate how DEA works, consider an example of three banks. Each bank has exactly 10 tellers (the only input), and we measure a bank based on two outputs: Checks cashed and Loan applications. The data for these banks is as follows: 58 #Bank A : 10 tellers, 1000 checks, 20 loan applications #Bank B : 10 tellers, 400 checks, 50 loan applications #Bank C : 10 tellers, 200 checks, 150 loan applications Now, the key to DEA is to determine whether we can create a virtual bank that is better than one or more of the real banks. Any such dominated bank will be an inefficient bank. Consider trying to create a virtual bank that is better than Bank A. Such a bank would use no more inputs than A (10 tellers), and produce at least as much output (1000 checks and 20 loans).Clearly, no combination of banks B and C can possibly do that. Bank A is therefore deemed to be efficient. Bank C is in the same situation. However, consider bank B. If we take half of Bank A and combine it with half of Bank C, then we create a bank that processes 600 checks and 85 loan applications with just 10 tellers. This dominates B (we would much rather have the virtual bank we created than bank B). Bank B is therefore inefficient. Another way to see this is that we can scale down the inputs to B (the tellers) and still have at least as much output. If we assume (and we do) that inputs are linearly scalable, then we estimate that we can get by with 6.3 tellers. We do that by taking .34 times bank A plus .29 times bank B. The result uses 6.3 tellers and produces at least as much as bank B does. We say that bank B's efficiency rating is .63. Banks A and C have an efficiency rating of 1. Using Linear Programming Data Envelopment Analysis is a linear programming procedure for a frontier analysis of inputs and outputs. DEA assigns a score of 1 to a unit only when comparisons with other relevant units do not provide evidence of in efficiency in the use of any input or output. DEA assigns an efficiency score less than one to (relatively) inefficient units. A score which is less than one means that the linear combination of other units from the sample could produce the same vector of 59 outputs using a smaller vector of inputs. The score reflects the radial distance from the estimated production frontier to the DMU under consideration. There are a number of equivalent formulations for DEA. The most direct formulation of the exposition I gave above is as follows: Let Xi be the vector of inputs into DMU i. Let Yi be the corresponding vector of outputs. Let X 0 be the inputs into a DMU for which we want to determine its efficiency and Y 0 be the outputs. So the X's and the Y 's are the data. The measure of efficiency for DMU 0 is given by the following linear program: Min 0 Subject to _ i X i _ 0 _ i Yi _ 0 _ 0 where i is the weight given to DMU i in its efforts to dominate DMU 0 and 0 is the efficiency of DMU 0. So the 's and 0 are the variables. Since DMU 0 appears on the left hand side of the equations as well, the optimal 0 cannot possibly be more than 1. When we solve this linear program, we get a number of things: 1. The efficiency of DMU 0 (0), with 0 = 1 meaning that the unit is efficient. 2. The unit's comparables" (those DMU with nonzero ). 3. The goal" inputs (the difference between X 0 and _iXi) 4. Alternatively, we can keep inputs fixed and get goal outputs (1/ 0_iYi) DEA assumes that the inputs and outputs have been correctly identified. Usually, as the number of inputs and outputs increase, more DMUs tend to get an efficiency rating of 1 as they become too specialized to be evaluated with respect to 60 other units. On the other hand, if there are too few inputs and outputs, more DMUs tend to be comparable. A few of the characteristics that make it powerful are: DEA can handle multiple input and multiple output models. It doesn't require an assumption of a functional form relating inputs to outputs. DMUs are directly compared against a peer or combination of peers. Inputs and outputs can have very different units. For example, X1 could be in units of lives saved and X2 could be in units of dollars without requiring an a priori tradeoff between the two. The same characteristics that make DEA a powerful tool can also create problems. An analyst should keep these limitations in mind when choosing whether or not to use DEA. Since DEA is an extreme point technique, noise (even symmetrical noise with zero mean) such as measurement error can cause significant problems. DEA is good at estimating "relative" efficiency of a DMU but it converges very slowly to "absolute" efficiency. In other words, it can tell you how well you are doing compared to your peers but not compared to a "theoretical maximum." Since DEA is a nonparametric technique, statistical hypothesis tests are difficult and are the focus of ongoing research. Since a standard formulation of DEA creates a separate linear program for each DMU, large problems can be computationally intensive. 61 3.9 PROCESS OF ANALYSI S The analyses included a profile of the respondents, preliminary data analyses and hypothesis tests. The response rate and the profile of respondents would be first provided, followed by the preliminary data analyses. The preliminary analyses are essential because they can ensure that the multivariate methods chosen are applied in appropriate situations (Hair et al., 1998). They also lend credibility to all subsequent analyses and increase accuracy in the results (Sekaran, 2003). The analytical process in this initial stage would firstly provide descriptive statistics of the mean and standard deviation, followed by examining the missing data and test for normality of the data. The normality test was performed at the beginning stage by using EXCEL software in order to determine whether the variables are normally distributed. Preliminary tests includes assessing reliability of measurement scales analyzed using Cronbachs Alpha to ensure that all the variables are reliable. SEM and Regression analysis were performed in order to test the relationships of the dependent variables and independent variables. In this study, simple linear regression 1was used. 3.10 SUMMARY This chapter provides a detailed description of the research design used in the current study. Six hypotheses are developed on the basis of the literature review provided in the previous chapter. Research methods including scale development, questionnaire design, data collection and sample size are discussed. Based on these methods, data are collected for this study to examine the hypotheses proposed. After the data collection, data editing, coding and categorising are described. The procedures of data analysis and statistical techniques chosen for the analyses are also outlined. Structural Equation Modeling, regression analyses would be employed to test the hypotheses. The reason for using these techniques for the testing the hypotheses is also justified. The next chapter reports analyses of the data. 62 CHAPTER 4 THE I NDI AN TELECOM I NDUSTRY: STATUS ANALYSI S This chapter presents a comprehensive note on the path of growth and trends and transitions in the Indian Telecommunication scenario. It begins with an overview of the service providers and the types of services provided by them. The chapter then moves on to examine the status of the Indian telecom industry by providing vital figures and statistics on various economic indicators in the telecom sector pre and post privatization, the role of the private sector, and the government policies that had initiated the communication revolution in the country. Further it covers a section to analyze the growth of mobile telecom in India. It also presents a note on the Mobile Service Providers in Chennai, with a comprehensive picture of their respective strategic competencies. 4.1 I NDI AN TELECOM I NDUSTRY AN OVERVI EW There are two types of service providers in the Indian Telecom Industry:- 1. Basic Service Providers 2. Value added Service Providers 4.1.1 Basic Service Providers Basic service providers are those who provide mainly voice communication. The subscribers connection to the telecom network is called a Direct Exchange Line (DEL) and people use it for voice communication. Basic services can be differentiated as per call destination into domestic and international. Domestic calls, both local and long-distance are routed through cables/r46wireless links. International calls are routed overseas, mainly through satellite links. The 63 international carriers of various nations liaison with one-another to ensure smooth operations and efficient call transfer. Major global carriers, mainly private operators from the developed nations, determine international call tariffs. Revenue sharing agreements exist between various international carriers. 4.1.2 Value added Service Providers Value added service providers are those who provide services, such as cellular telephony, paging, e-mail and VSAT network, which provide the subscriber greater ease of communication and enhance the utility of basic services network. The following are the various types of services provided by both the service providers put together:- Telephone Services NSD / ISD Services Computerized Trunk Services Pay Phones National & International Leased Lines Circuits Telex Telegraph Services (Manual & Automatic) X-25 based Packet Switched Data Network (INET) Gateway Packet Switched Data Services (GPSS) Gateway Electronic Data Interchange Services (GEDIS) Gateway E-Mail and Store & Forward FAX Service (GEMS-400) Concert Packet Services (CPS) Satellite-based Remote Area Business Message Network 64 Electronic Mail Voice Mail Audio-Text Radio Paging Cellular Mobile Telephone Public Mobile Radio Trunked Service Video-Tex Video Conferencing V-SAT Internet ISDN INMARSAT Mobile Service INMARSAT Data Service Home Country Direct Service Intelligent Network (IN) Service 4.2 THE I NDIAN TELECOM I NDUSTRY: HI STORI CAL PERSPECTI VE The Indian Telecom industry has witnessed several events since the day on which the first operational landline was laid in 1851. For about 140 years afterwards nothing much happened in the industry to excite the world around, till the liberalization initiative in 1991, followed by telecom reforms that lead to the announcement of the Telecom policy in 1994. The Table 4.1 shows the important dates in the history of the Indian Telecom Sector. 65 Table 4.1 Important Dates and Events in the Indian Telecom Industry Year Events 1851 First Telephones in India 1943 Nationalization of telephone companies 1985 DoT was created 1986 Creation of MTNL and VSNL 1991 Telecom equipment liberalized 1994 Licenses for paging 1994 Telecom Policy announced September 1994 Guidelines for private sector participation in basic services November 1994 Cellular licenses issued J anuary 1995 Issue of tenders for 2 nd operator in basic services August 1995 VSNL launches internet services J anuary 1996 TRAI formed February 1996 Supreme court allows multiple players in Basic services October 1999 New Telecom Policy announced October 2000 BSNL formed December 2001 33 Licences to private palyers in Basic services February 2002 Disinvestment of VSNL May 2002 Bharti offers ILD Services with sharp cuts in tariffs September 2002 TRAI decides forbear from regulating cellular tariffs March 2006 WPC set subscriber thresholds for GSM and CDMA operators for spectrum allocation March 2007 9 dinstinct operators had been allocated GSM spectrum. Out of these, only Bharti has pan-India presence August 2007 Subscriber thresholds were revised by TRAI as operators could support more subscribers with lower spectrum as compared to WPC allocation J anuary 2008 Govt. of India allocated start-up spectrum to all prior licenses awaiting spectrum. These include Aircel (14 circles), Idea (2 circles), RComm (14 circles) and Vodafone (6 circles) J anuary 2009 TRAI plans to introduce MNP on a pan-India basis J anuary 2011 MNP implemented 66 4.3 GROWTH OF THE I NDI AN TELECOM SECTOR The study on the growth of the Indian telecom sector is presented under two heads: 1. Indian Telecom Industry Pre privatization 2. Indian Telecom Industry Post privatization 4.3.1 I ndian Telecom I ndustry Pre privatization The Indian telecom sector like any other infrastructure sector was owned and controlled by the Government of India, since independence. The nationalization of the telecom companies to form Post, Telephone and Telegraphs in 1943 was an important milestone in the history of Indian telecommunications. There was a gradual growth during the five-year plans in the sector and the entire sector was brought under the control of the Department of Telecommunications (DoT) in 1985. The status of the telecom sector for the period 1993-94, which is stated as the year of the monopoly reign, is as follows: Table 4.2 Status of the I ndian Telecom I ndustry in 1993-94 Items Unit Status Equipped Capacity in 000s 9795.17 DELs* in 000s 8025.6 Waiting List in 000s 2496.8 Registered Demand in 000s 10522.4 Metered Calls In crores 4671.8 Call per DEL in 00s 52.88 Faults Registered (Per hundred stations per month) numbers 18.3 * DEL Direct Exchange Lines 67 The telecom policy announced in 1994 significantly changed the Indian Telecom scenario. The Cellular Service providers broke the monopoly in the sector initially, with the government issuing licences to private players to take part in contributing towards linking the country through mobile network. The sector started experiencing tremendous growth from this period onwards, not only in terms of number of lines provided, but also in terms of investments, revenue and employment. This period of transition is analyzed in detail in the following section. 4.3.2 I ndian Telecom I ndustry Post Privatization India is perceived to have a special comparative advantage in IT and IT- enabled services. However, sustaining this advantage depends critically on high quality infrastructure. The telecom sector has witnessed the impact of major reforms since 1994, when the government liberalized the sector. The fast expansion of the telecom network is a consequence of friendlier economic environment created by positive policy thrust for promoting private investment, including foreign direct investment in private sector enterprises. The Indian Government played a major role in effecting the transition in the telecom sector as an initiator of reforms. When the reformists saw a need for increased investment in the infrastructure sector for a comprehensive economic development, they thought it necessary for liberalizing the licensing structure and enabling the private entry into the sector, thus demonopolising the sector. The two major policy reform announcements made by the Government of India that lead to the development of communication infrastructure in the country are: 1. National Telecom Policy (NTP 1994) 2. New National Telecom Policy (NTP 1999) 68 National Telecom Policy (NTP 1994) The liberalization reforms initiated in 1991, followed by Eighth Plan (1992-97), objectives to open value added services in the telecom sector for private sector, necessitated a need for a well-laid operating guidelines for the sector. This led to the announcement of National telecom Policy (NTP) 1994.This was one of the major initiatives on the part of the government, in support of the liberalization and privatization reforms for boosting private entry and foreign investment into the country. The following were some of the specific objectives of the NTP 1994 based on the broad objectives of providing affordable and accessible means of communication for all. 1. To achieve installation of 9.5 million additional Direct Exchange Lines (DEL) by 1997. 2. To achieve Telephone on Demand by 1997. 3. To achieve a target of 1 PCO (Public Call Office) for 500 units of population by 1999. 4. Every village to be provided by at least one PCO by 1999. To focus on the objectives, the following were the policy decisions taken by the Government of India, on the licensing and operational front. The basic premise on which competition has been introduced is that every circle will have one private operator apart from the existing public sector service providers in the fixed line segment, and two private operators for cellular. The public sector players had the option to become the third cellular operator in the future. The private players were invited to operate in both basic and cellular telephony. To facilitate licensing, the nation was divided into 20 telecom circles for basic and 21 circles for cellular telephony. 69 As per the terms of license agreement, private operators have to provide a minimum of 10 percent Direct Exchange Lines (DELs) as Village Public Telephones (VPTs). To set up an independent regulatory authority to monitor and regulate the operations of the players, based on the terms of licensing. The following were the achievements in the Indian Telecom Sector as a result of the announcement of NTP 1994: An independent telecom regulator called TRAI (Telecom Regulatory Authority of India) was formed in J anuary 1996. Eight licenses were issued for Cellular services initially in the 4 metro cities in 1994. Subsequently in 1995, 34 licenses were issued to 14 companies to operate in 18 circles. There were about 13.5 lakh cellular phones in the country on December 30, 1999. As on September 1999, the PCOs in the country were at a ratio of 1:453, on all India average bases. Out of 6 lakh villages in the country 3.4 lakh were provided with VPTs as on December 1999. Though NTP 1994 was hailed as an initiator of privatization in the economy, it was criticized on the following grounds: It lacked focus, since it sought to achieve a huge task in minimum period of time and ended up not achieving the targets objected. There was utter confusion as to licensing in the basic services, with one bidder dominating the whole scene, subsequently saw the other players filing a case in the supreme court for recommending multiple entry. 70 It was also affected by huge resource and finance crunch, due to improper planning and implementation. The powers of operation of the regulatory authority were not clearly defined, so DoT and TRAI ended up in legal disputes to be resolved by the court. The above criticisms made it essential for the Indian Government to announce a new telecom policy with a better framework of licensing, regulatory and promotional operations to develop the Indian Telecom Sector. This paved way for the announcement of the New National Telecom Policy (NTP) in 1999. Better focused objectives of this policy accelerated the pace of growth of telecom network in the country, after 1999. New National Telecom Policy (NTP 1999) The New Telecom Policy (NTP) announced in 1999 modified the NTP 1994 to take into account the far-reaching technological developments taking place in the telecom sector globally and to implement the Governments resolve to make India a global IT superpower. NTP 1999 also seeks to solve problems arising out of the implementation of NTP 1994. The objectives of the NTP 1999 are 1. Access to telecommunications is of utmost importance for achievement of the countrys social and economic goals. Availability of affordable and effective communications for the citizens is at the core of the vision and goal of the telecom policy. 2. Strive to provide a balance between the provision of universal service to all uncovered areas, including the rural areas, and the provision of high- level services capable of meeting the needs of the countrys economy. 3. Encourage development of telecommunication facilities in remote, hilly and tribal areas of the country. 71 4. Create a modern and efficient telecommunications infrastructure taking into account the convergence of IT, media, telecom and consumer electronics and thereby propel India into becoming an IT superpower. 5. Convert PCOs, wherever justified, into Public Teleinfo Centers having multimedia capability like ISDN services, remote database access, government and community information systems etc. 6. Transform in a time bound manner, the telecommunications sector to a greater competitive environment in both urban and rural areas providing equal opportunities and level playing field for all players. 7. Strengthen research and development efforts in the country and provide an impetus to build world-class manufacturing capabilities. 8. Achieve efficiency and transparency in spectrum management. 9. Protect defense and security interests of the country. 10. Enable Indian Telecom Companies to become truly global players. In order to achieve the above objectives, the following targets were fixed: 1. Make available telephone on demand by the year 2002 and sustain it thereafter so as to achieve a teledensity of 7 by the year 2005 and 15 by the year 2010. 2. Encourage development of telecom in rural areas making it more affordable by suitable tariff structure and making rural communication mandatory for all fixed service providers. 3. Increase rural teledensity from 0.4 (in 1999) to 4 by the year 2010 and provide reliable transmission media in all rural areas. 4. Achieve telecom coverage of all villages in the country and provide reliable media to all exchanges by the year 2002. 5. Provide Internet access to all district head quarters by the year 2000. 72 6. Provide high speed data and multimedia capability using technologies including ISDN to al towns with a population greater than 2 lakhs by the year 2002. The following are the policy decisions advocated to achieve the above objectives and targets in a smooth way. In order to separate the service providing from the policy making function of the DoT, two separate departments were formed. Department of Telecom Services (DTS), to provide service and Department of Telecom Operations (DTO), to take care of operations. DoT would take care of the policies, treaties and agreements in the Telecom Sector. DTS and DTO were subsequently corporatised to form Bharat Sanchar Nigam Limited (BSNL), the presently functioning government sector service provider. In the basic services segment, multiple players were permitted to operate in a telecom circle, as per the Supreme Court recommendations and also met the investment requirements to achieve the set objectives. Telecom Regulatory Authority of India (TRAI) was given independent powers to act as a price fixer and regulator of licensing issues. A separate Telecom Dispute Settlement and Appelliate Tribunal (TDSAT), was formed to adjudicate any dispute between the service providers, licensor and the user groups. Tariff rebalancing measures are taken to make communication affordable to all. This decision is taken to increase the volume of traffic and also the revenues. Opening National long Distance Service (NLDS) and International Long Distance Service (ILDS) to private players. 73 Creating a provision of Universal Service Fund (USF) under the Universal service Obligation (USO) for all the service providers to enable them to meet high costs of providing services to remote / hilly rural areas. Implementing Unified Licensing System, whereby a service provider in the Telecom Sector, is permitted to provide a variety of services with a single license. The NTP 1999 is clear in its objectives, comparing to NTP 1994. Built on these objectives is the phenomenal growth, the Indian Telecom Sector is witnessing today. The major achievements of NTP 1999 are: In November 2003, Unified Access Licensing is implemented with a conversion of 27 out of 31 licenses issued in the basic services, by adding the country into 23 Service Areas consisting of 19 Telecom Circle Service / Metro Service Areas for providing Unified Access Services (UAS). The Universal Service Fund of Rs.200 crore is distributed in 2003, for providing additional VPTs ( Village Public Telephones) and also to improve the quality of service of the existing VPTs in rural and remote areas. 4.4 I NDI AN TELECOM I NDUSTRY : A COMPETI TI VE ANALYSI S The Indian Telecom Industry of the 21 st century is witnessing competition due to liberalization, privatization and demopolization initiatives taken by the Government of India. While liberalization sowed the seeds for competition, it was privatization followed by disinvestment and demopolisation that led to a competitive environment in the sector. Apart from the PSU (Public Sector Unit) service providers, licenses were given to several private players to provide a variety of communication services to 74 the people. Hence the monopoly in the sector is broken and the users are now given with a wide variety to choose in almost all the services starting from the basic voice services, value-added services to data and satellite services. With the major policy governance as regards the pricing, spectrum usage, mobility, entry legislations and other licensing issues still at the hands of the government regulatory body, the competitive environment in the industry needs to be analyzed to have a clear understanding of the factors influencing the operations of the players. In this section, the micro and macro environmental factors are analysed by the application of PEST analysis and Micael Porters Five Forces Competitive Advantage Model. 4.4.1 PEST Analysis The framework for PEST analysis is drawn from the contributions made by several eminent management practitioners and theorists like Nigel M Healey (1994), De Avison, Wa Eardley, P Powell (1998), Alistair R. Anderson & Martin H.Atkins (2001), J udith Broady-Preston & Tim Hayward (2001), J ames Manktelow (2004), in their research works on various management issues. The PEST analysis is primarily a tool used to analyse the business environment and to understand the market growth or decline. PEST template encourages proactive thinking rather than relying on habitual or instinctive reactions. PEST analysis focuses on the understanding of each of the following four factors, viz., P Political Environment E Economic Environment S Social Environment T Technological Environment 75 Analysis of Political Environment: The political environment of the business, namely the government regulations and legal issues, which govern the operations of the fim or the industry are analyzed under this template. These include the ecological issues, current legislation, Government and trading policies, funding, grants and initiatives, and so on that are prevalent at a particular point of time. With reference to the Telecom industry, the analysis is be on the TRAI regulatory and licensing procedures including tariff determination, spectrum allocation, and other policy issues as imposed and administered by the Government regulatory body from time to time. Analysis of Economic Environment: Factors like the purchasing power of the potential customers, the demand and supply factors that affect the firms operations, the consumer surplus concept, the cost of capital of the firms operating are some of the micro economic issues analyzed under this heading. Macro issues like the impact of economic growth, interest rates, exchange rates, inflation rates etc are also analyzed with respect to the firm. In the Telecom industry, these may include demand for and supply of telecom sectors by various user groups, economic growth of the IT sector and its impact on the Telecom sector and so on. Analysis of Social Environment: This includes the demographic and cultural aspects of the market in context. Factors like the demographic and psychographic profile of the customers, the groups that influence them in their purchase decisions etc are analyzed under this section. With reference to the Telecom Industry, the social analysis deals with the analysis of user profiles both demographic and personality. Analysis of Technological Environment: This concerns the technological factors that influence the operations of the business. Factors like R & D activity, automation, adoption of new technology for hard / soft processes within business etc are analyzed under this head. Variables like rate of change of technology and operational feasibilities also form an important part of this analysis. As regards the telecom industry, this deals with analyzing the technological changes and its 76 significant impact on the industry, since the industry is primarily categorized as a high-tech industry. Thus PEST analysis enables better understanding of the present position and hence provides inputs for analyzing the future potential and directions for a business. PEST can be effectively used for marketing and business development and decision-making. The scan of the Telecom Environment of India provides the following insights which enabled the researcher to understand the basic micro and macro environmental factors that influence the operations of the industry on a whole. The various factors that influence the telecom industry of India are categorized under the four heads of the PEST and summarized as under: Summary of PEST analysis on Indian Telecom Industry Political Factors The political factors that influence the operations of the telecom players in India are: Liberalized FDI : For mobilizing Foreign Direct Investment (FDI) into the sector, the cap on investment holding is increased from 49% to 74%, this is a welcome move especially for private players, who depend on foreign capital. Tariff Policy: Tariff is determined by the government body and due to the tariff re-balancing measures adopted by the government; the Indian voice line users are charged the lowest call-rate in the world, this is promising wider avenues for development as the economy primarily comprises of low and middle income groups. Licensing Policy: Stringent licensing procedure adopted by TRAI, the government regulatory body, has led to restricting the number of 77 operators in the sector. The licensing policies act as an entry barrier for new entrants aspiring to venture into the sector. Spectrum Allocation: The limitations on spectrum sharing and policies pertaining revenue sharing are hampering the smooth flow of activities for service providers mainly the basic and internet service providers. The limited resource availability is acting as a barrier to entry. Economic Factors The economical factors influencing the industry are as follows: Growth of Service Industry: The growth of the service sector, especially IT and its consequent impact on the related sectors like Business Process Outsourcing (BPO) is promising a huge growth potential for the Telecom Sector, since telecom forms the life blood for operations of these sectors. Favorable investment climate: The decline in interest in favouring the investment climate to boom. With a promising all-round growth in the telecom sector, investors see it highly lucrative to invest in the sector. Demand-Supply Gap: Due to availability of telephone on demand, the demand-supply gap is getting narrowed over the period from 2000-2004. Increased Disposable Income: The booming IT sector has created high-paying jobs, due to which the consumers are left with higher disposable income. Earning Sensitivity: 16% of the villages are still unconnected through communication links, which promises a huge market potential for the high growth telecom market. But, the earning sensitivity issues pertaining to providing village telephones restricts the private players from actively participating in linking the villages through communication network. 78 Social Factors The social factors influencing the industry are as follows: Transforming lifestyles: Due to demopolisation and privatization, there is transformation in the lifestyle of the people, which is moving westwards, as a result there is an increasing demand for a developed urban communication system, especially personal communication systems like mobile phones. Changing Demographic Combination: There is an increase in middle age group (30-40) in the population, which is decision- making population. This encourages the private players, since the group primarily comprises of innovators, achievers and makers. I nformation revolution: The IT and media boom is delivering information to the consumers at their doorsteps, due to which the consumers are adequately enlightened as regards the products and the competition. Cultural Values: India is being basically perceived as a country bound by family values. Generally, it is stated that the share of fixed- line phones will not be eaten up by the mobile phones in the long-run, unlike in developed economies of the west, where fixed-line users migrate to mobile phones. Technological Factors The technological factors influencing the industry are as follows: Capital I ntensive: Telecom is a high capital and technology intensive sector. Huge investments are pumped into technology for providing and updating a variety of basic and value-added services. The players are required to keep pace with the developments in the 79 technology for gaining competitive edge. The very fact that the industry is technology intensive acts as an entry barrier. Return on Investments (ROI) concerns : The ROI on technology investments is a matter of concern for the telecom players because the technology innovations in any form have a very short life. So players operating on high economies of scale gain in the market, while others succumb to M&A (Mergers and Acquisition) motives. I mpact of PEST on I ndian Telecom I ndustry The Indian Telecom Industry post-liberalization is experiencing transformation on all fronts. In the political front the policies framed by the government regulatory body are conducive to development of telecom network of the country. In terms of licensing and other regulatory issues, a more transparent approach from the part of the regulator is expected to reap huge benefits for the development of communication infrastructure in the country. As regards the economic issues confronting the telecom sector, the liberalized FDI policy has spurred the investment inflow and the existence of demand-supply gap throws a huge potential and favourable investment opportunities are luring foreign investors to venture into the telecom market. The surge in incomes and employment levels and improving standard of living have changed the perceptions of the people, who have started looking at telecommunication services as an essential commodity for better living. This has created a huge demand for telephones especially the personal phones. With respect to technological issues, the industry is catching up with the changes taking place in the rest of the world at a fast pace as a result innovations have a very short life. The investors are thus subjected to earning sensitivity and ROI issues. Due to highly favourable market condition and a huge growth potential posed by the sector, more and more investors are venturing into the telecom market, since it promises a long-term sustainability. 80 On the whole, there is an overall favourable climate as far as the political, economical, social and technological environment is concerned, despite certain challenges posed by the regulatory policies and ROI issues. 4.4.2 Michael Porters Five Forces Model The five forces model is widely accepted as a yardstick for measuring the industry profitability. It analyzes the various forces influencing the industrys competitive environment. This is a macro model concentrating on the external forces like rivalry between firms, bargaining power of the buyer, bargaining power of the supplier, availability of substitutes and entry barriers. In order to perform this analysis, the competitive forces in the industry need to be thoroughly understood, the attractiveness of and growth opportunities within, a new industry need to be assessed and the effective strategies to raise the profitability, power and competitive position in an industry need to be developed accordingly. Following is the detailed description on the components of the Porters Model. Bargaining power of the buyers: Buyers/customers are a moderate force in the industry. By virtue of the market becoming customer centric, the buyers have bargaining capacity and may demand better product features for the same price. Thus the buyers behaviour and their bargaining capacity impact the profitability of a firm. Their behaviour and power is dependent on the availability of alternatives, market awareness, preferences and so on. Rivalry between Firms: Rivalry among competitors in the industry is powerful. Competitors may adopt under cutting and bring down the industry. Hence the rivalry between firms affects the industry to a major extent. This factor is dependent on price wars, corporate on price wars, corporate image, adoption of new technology, value added services, and so on. 81 Bargaining power of Suppliers: Suppliers are again a moderate force in the industry. But if the supplier is a monopoly, then his powers know no bound. This is dependent on the level of switching cost and the terms and conditions laid down by the suppliers. Availability of Substitutes: A substitute product is a product similar to the relevant product but is not identical to it. Substitute products are a very weak force in the industry. Substitute products restrict industry profitability by limiting the selling price companies in the industry can charge. In the Telecom industry, this is determined by availability of very close substitutes and low customer loyalty. Entry Barriers: New entrants are potential competitors. New entrants are a weak force in the industry. The lesser the entry barriers, easier it is for new companies to enter the industry and hence, greater the competition in the industry. New entrants will often attempt to break into the industry with low prices, innovative products, or new features and benefits. When the barriers are strong, the threat of new entrants is low. Applicability of Porters Competitive Model to I ndian Telecom I ndustry Thanks to privatization and subsequent restructuring of the economy, the entry of private players in the sector, has made the environment get transformed from its previous monopoly state to its present competitive state. All the parameters of the five forces model see their due presence in the Indian Telecom Scenario. An analysis performed on the competitive scenario of the sector reveals adequately and precisely the relevance and adaptability to present a picture of the Indian Telecom Industry of today. The model analyzes the competitive environment in the Telecom Industry of India in Figure: 8 2 ENTRY BARRIERS Pricing Earning Sensitivity Issues Consolidation Capital Intensive Technology Investments Licensing & Operating Regulations RIVALRY BETWEEN FIRMS PriceWars Quick Adapters Economies of Scale Differentiated Services Disputes Legal / Operational Betting on Technology Corporate Image SUPPLI ER POWER SpectrumIssues Infrastructure Issues Monopoly Supplier High Switching Cost Licensing Issues Fixed Pricing BUYER POWER Availability of alternatives Low Switching Costs Market Awareness Standard of Living AVAIALABILITY OF SUBSTITUTES Close substitutes Undifferentiated Pricing Differentiated Services SubstitutePerformance Low-loyalty levels VAS Preferences Figure 4.1 Michael Porters Five Forces Model 83 Summary of the Five-Forces impacting I ndian Telecom Industry 1. Buyer Power: There are several factors leading to increase in the buyer bargaining power in the telecom industry. As a result of privatization and competition, the buyers are provided a variety of alternatives to choose from, which was unavailable during the pre-privatization period. Due to falling prices the switching costs are reduced and the development of IT and increasing incomes has resulted in change in the standard of living of the consumers. 2. Rivalry between firms: As a result of increasing competition, there are stiff price wars and innovations have a very short life. The firms operating on economies of scale enjoy a competitive advantage with respect to their earning sensitivity. Another form of rivalry is the legal and operational disputes that arise between firms as a way to capture the market. There are a good number of players who have their corporate image working to their advantage. 3. Supplier Power: The supplier comes in the form of government regulator, which has impact on the operations of the players. The spectrum sharing and infrastructure issues are governed by the regulatory body, which is the monopoly supplier to the telecom industry. As regards the instruments, there are a few suppliers, and the switching cost is very high. The suppliers follow a uniform pricing structure and it is fixed pricing in the case of government supplier. 4. Availability of Substitutes: There is availability of close substitutes, since there is no differentiation in terms of the basic services delivered to the customers. The differentiation comes in the form of value added services. Since VAS forms the major differentiator, the players attract the customers of other service providers; hence there are very low loyalty levels. Due to increasing market awareness, the customer preferences towards value added services have also increased. 84 5. Entry Barriers: The entry barriers are in the form of earning sensitivity issues and technology investments, which make the industry feasible to operate only for cash-rich companies. The licensing and operating regulations adopted by the government also form the entry barriers. The picture of consolidation and capital intensive nature of the industry are the other factors that function as entry barriers for the new aspirants to the industry. The competitive analysis of the Indian telecom industry and the exhaustive insights derived through the application of PEST templates and Porters Five Forces Models, thus throws light on the external factors that influence the operations of service providers in the telecom sector. 4.5 I NDI AN MOBI LE SECTOR Mobile communications systems revolutionized the way people communicate, joining together communications and mobility. A long way in a remarkably short time has been achieved in the history of wireless. Evolution of wireless access technologies is about to reach its fourth generation (4G). Looking past, wireless access technologies have followed different evolutionary paths aimed at unified target: performance and efficiency in high mobile environment. The first generation (1G) has fulfilled the basic mobile voice, while the second generation (2G) has introduced capacity and coverage. This is followed by the third generation (3G), which has quest for data at higher speeds to open the gates for truly mobile broadband experience, which will be further realized by the fourth generation (4G). The Fourth generation (4G) will provide access to wide range of telecommunication services, including advanced mobile services, supported by mobile and fixed networks, which are increasingly packet based, along with a support for low to high mobility applications and wide range of data rates, in accordance with service demands in multiuser environment. 85 The last few years have witnessed a phenomenal growth in the wireless industry, both in terms of mobile technology and its subscribers. There has been a clear shift from fixed to mobile cellular telephony, especially since the turn of the century. By the end of 2010, there were over four times more mobile cellular subscriptions than fixed telephone lines. Both the mobile network operators and vendors have felt the importance of efficient networks with equally efficient design. Driven by wireless revolution, the Indian telecommunications industry is one of the fastest growing in the world. Government policies and regulatory framework implemented by Telecom Regulatory Authority of India (TRAI) have provided a conducive environment for service providers. This has made the sector more competitive, while enhancing the accessibility of telecommunication services at affordable tariffs to the consumers. According to TRAI's report 'Telecom Sector in India: A Decadal Profile', the tele-density has increased from 4.3 in March 2002 to 78.1 in February 2012, wherein the rural areas registered an increase from 1.2 in March 2002 to 38.5 in February 2012. Also, the share of telecommunication services (excluding postal and miscellaneous services), as per cent of the total gross domestic product (GDP), has increased from 0.96 in 2000-01 to 3.78 in 2009 -10. According to the same report, international comparisons (among 222 countries) show that India has the second largest number of telephone subscribers in the world accounting for 12 per cent of the world's total telephone subscribers. Key Statistics In its recent statement issued, TRAI has revealed that the country's mobile subscriber base has reached 951.3 million wherein the operators added 8 million subscribers in March 2012. 86 The overall tele-density in India reached 78.66. The urban tele- density was recorded to be 169.55, while rural tele-density stood at 39.22. Total broadband subscriber base increased from 13.54 million in February 2012 to 13.79 million in March 2012, registering a growth of 1.86 per cent. 4.6 KEY TRENDS I N I NDIAN TELECOM The wireless segment in India is much larger than the wire line segment and is growing steadily due to the convenience and utility it offers. Wireless services hold a major market share of 94.6 per cent. The subscriber base of the wire line segment is decreasing due to its limited usage. Rural markets are expected to be the next key growth drivers for the Indian telecom sector, given rural Indias growing population and disposable income. The subscriber base in the rural market has improved significantly in 20092010, with rural tele-density at 26.4 per cent as of J une 2010. By 2012, the rural subscriber base is expected to account for nearly half of the total subscriber base, thereby fuelling sector growth. Bharti Airtel has the largest market share in the GSM segment. As of J une 2011, Bharti accounted for 25.9 per cent of the GSM market, followed by Vodafone, with a 20.7 per cent market share. Private players accounted for approximately 86.4 per cent, while public sector operators (BSNL and MTNL) accounted for the remaining share (13.6 per cent). 87 Reliance Communications dominates the Indian CDMA mobile services segment with a market share of 52.4 per cent as of June 2011. India is expected to feature among the top 10 broadband markets by 2013. The total number of Internet subscribers grew from 14.1 million subscribers in J une 2010 to 16.7 million subscribers in J une 2011. BSNL is the biggest player in this market with 9.7 million subscribers, followed by MTNL, Bharti Airtel, Reliance and Hathway Cable & Datacom. Digital subscriber line (DSL) is the most preferred technology among service providers to provide broadband services. DSL constitutes 86.6 per cent of total broadband subscribers. In India, growth in the subscriber base, which contributes to healthy revenue growth, mitigates the reduction in average revenue per user (ARPU). In addition, high MOUs compensate for declining tariffs. Operators are reducing operating costs and hiving off infrastructure elements such as towers into separate entities, thus inviting significant investments. Passive infrastructure sharing has benefitted the Indian mobile industry and its customers, reducing the cost burden of each operator and speeding the rollout of mobile services. In recent years, initiatives such as network cost optimisation, outsourcing of non-core activities, as well as low-cost business models have been focus areas. Every telecom service provider is looking beyond basic voice services by offering a wide range of bundled offerings. For example, nearly all leading operators, including incumbents, are in the testing 88 phase to launch commercial IPTV services. Indian operators are still new in terms of offering using existing network infrastructure for data, voice, video and basic communication services. Consumers can get all these services from the same telecom operator, and enterprises can also access virtual private networks (VPNs), video-conferencing, enterprise solutions, mobility and fixed telephony. 4.7 DETAI LS OF MOBI LE SERVI CE PROVI DERS CONSI DERED FOR THE STUDY This section gives the details of the mobile service providers used for the study. 4.7.1 Bharti Airtel Bharti Airtel Limited, commonly known as Airtel, is an Indian Telecommunications company that operates in 20 countries across South Asia, Africa and the Channel Islands. It operates a GSM network in all countries, providing 2G, 3G and 4G services depending upon the country of operation. Airtel is the world's third-largest mobile telecommunications company with over 261 million subscribers across 20 countries as of August 2012. It is the largest cellular service provider in India, with 200 million subscribers as of August 2012. Airtel is the third largest in-country mobile operator by subscriber base, behind China Mobile and China Unicom. Airtel is the largest provider of mobile telephony and second largest provider of fixed telephony in India, and is also a provider of broadband and subscription television services. It offers its telecom services under the airtel brand, and is headed by Sunil Bharti Mittal. Bharti Airtel is the first Indian telecom service provider to achieve Cisco Gold Certification. It also acts as a carrier for national and international long distance communication services. The company has a submarine 89 cable landing station at Chennai, which connects the submarine cable connecting Chennai and Singapore. It is known for being the first mobile phone company in the world to outsource all of its business operations except marketing, sales and finance. Its networkbase stations, microwave links, etc.is maintained by Ericsson, Nokia Siemens Network and Huawei, business support is provided by IBM and transmission towers are maintained by another company (Bharti Infratel Ltd. in India). On May 2012, Bharti Airtel awarded the three year contract to Alcatel- Lucent for setting up an Internet Protocol across the country. This would help consumers access internet at faster speed and high qualiy internet browsing on mobile handsets. 4.7.2 Aircel Aircel group is an Indian mobile network operator headquartered in Chennai, that provides wireless voice, messaging and data services in India. It is a joint venture between Maxis Communications Berhad of Malaysia and Sindya Securities & Investments Private Limited, whose current shareholders are the Reddy family of Apollo Hospitals Group of India, with Maxis Communications holding a majority stake of 74%. Aircel commenced operations in 1999 and today the leading mobile operator in Tamil Nadu, Assam and North- East. It is Indias fifth largest GSM mobile service provider & seventh largest mobile service provider (both GSM and CDMA) with a subscriber base of over 51.83 million, as of J anuary 31, 2011. It has a market share of 6.72% among the GSM operators in the country. Additionally, Aircel has also obtained permission from Department of Telecommunications (DoT) to provide International Long Distance (ILD) and National Long Distance (NLD) telephony services. It also has the largest service in Tamil Nadu. 90 4.7.3 I dea Cellular In 2000, Tata Cellular was a company providing mobile services in Andhra Pradesh. When Birla-AT&T brought Maharashtra and Gujarat to the table, the merger of these two entities was a reality. Thus Birla-Tata-AT&T, popularly known as Batata, was born and was later rebranded as IDEA. Then Idea set sights on RPGs operations in Madhya Pradesh which was successfully acquired, helping Batata have a million subscribers, and the licence to be the fourth operator in Delhi was clinched.In 2004, Idea (the company had by then been rechristened) bought over the Escorts groups Escotel gaining Haryana, Uttar Pradesh (West) and Kerala and licences for three more UP (East), Rajasthan and Himachal Pradesh. By the end of that year, four million Indians were on the companys network. In 2005, AT&T sold its investment in Idea, and the year after Tatas also bid good bye to pursue an independent telecom business. And Idea was left only with one promoter, the AV Birla group. Rs 2,700 crore adding Punjab and Karnataka circles. Modis joint venture partner, Telekom Malaysia, invested Rs 7,000 crore for a 14.99% stake in Idea. J ust around then, Ideas subsidiary, Aditya Birla Telecom sold a 20% stake to US-based Providence Equity Partners for over Rs 2,0000 crore. 4.7.4 Vodafone Vodafone India, formerly Vodafone Essar and Hutchison Essar, is the second largest mobile network operator in India after Airtel. It is based in Mumbai, Maharashtra and which operates nationally. It has approximately 146.84 million customers as of November 2011. On J uly 2011, Vodafone Group agreed terms for the buy-out of its partner Essar from its Indian mobile phone business. The UK firm paid $5.46 billion to its Indian counterpart to take Essar out of its 33% stake in the Indian subsidiary. It will leave Vodafone owning 74% of the Indian business, while the other 26% will be owned by Indian investors, in compliance with Indian law. On 11 February, 2007, 91 Vodafone agreed to acquire the controlling interest of 67% held by Li Ka Shing Holdings in Hutch-Essar for US$11.1 billion, piping Reliance Communications, Hinduja Group, and Essar Group, which is the owner of the remaining 33%. The whole company was valued at USD 18.8 billion. The transaction closed on 8 May, 2007. It offers both prepaid and postpaid GSM cellular phone coverage throughout India with good presence in the metros. Vodafone India provides 2.75G services based on 900 MHz and 1800 MHz digital GSM technology. Vodafone India launched 3G services in the country in the J anuary-March quarter of 2011 and plans to spend up to $500 million within two years on its 3G networks. 4.7.5 BSNL Bharat Sanchar Nigam Limited (abbreviated BSNL) is an Indian state- owned telecommunications company headquartered in New Delhi, India. It is the largest provider of fixed telephony and fourth largest mobile telephony provider in India, and is also a provider of broadband services. However, in recent years the company's revenue and market share plunged into heavy losses due to intense competition in Indian telecommunications sector. BSNL is India's oldest and largest communication service provider (CSP). It had a customer base of 95 million as of J une 2011. It has footprints throughout India except for the metropolitan cities of Mumbai and New Delhi, which are managed by Mahanagar Telephone Nigam (MTNL). 4.7.6 Reliance Communications Reliance Communications Ltd. (commonly called RCOM) is an Indian broadband and telecommunications company headquartered in Navi Mumbai, India. RCOM is the world's 15th largest mobile phone operator with over 150 million subscribers and India's 2nd largest telecom operator in India, only after Bharti Airtel. Established on 2004, a subsidiary of the Reliance Group. The company has five segments: Wireless segment includes wireless operations of the company; 92 broadband segment includes broadband operations of the company; Global segment include national long distance and international long distance operations of the company and the wholesale operations of its subsidiaries; Investment segment includes investment activities of the Group companies, and Other segment consists of the customer care activities and direct-to-home (DTH) activities. 4.8 SUMMARY The chapter provided an overview on the Indian telecom industry and analyses the status of the industry in the pre-privatization and post-privatization period. It then examined the competitive environment of India, performing PEST analysis and Porters Five Forces competitive environment analysis. A section on the study area is also presented. 93 CHAPTER 5 DATA ANALYSI S 5.1 I NTRODUCTION This chapter focuses on the detailed analysis of the data collected and the statistical results of the study. The primary purpose of this research is two fold: The first is to develop path models for establishing and confirming the relationship (if any) existing among service quality, service loyalty and loyalty indices. The second is to assess the current level of effectiveness of CRM practices of the Mobile Service Providers in order to identify the best performers among the service providers on the various latent factors and provide insights to the other service providers so that they can improve on the factors, using a linear programming based technique for relative efficiency called Data Envelopment Analysis. Structural Equation Modeling (SEM) is the main multivariate technique which combines aspects of regression and factor analysis to estimate series of interrelated dependence relationships simultaneously (Hair et al., 1998). Simple linear regression was used to analyze the data. To evaluate the effectiveness of CRM practices, there are lots of quantitative techniques available. But there is a dearth of literature focusing on the relative efficiency. One advanced operations research technique which evaluates the relative efficiency is the Frontier Analysis or Data Envelopment Analysis (DEA). This study attempts to use Data Envelopment 94 Analysis to assess the effectiveness of Mobile Service Providers, specifically a set of the providers offering services in Chennai, Tamil Nadu, India. The research has identified a set of input and output parameters, from the regressions performed on service quality, service loyalty and loyalty indices, for each of the Service Providers, from which the efficient frontiers or Decision Making Units (DMUs) are determined. The overall effectiveness on the usage of CRM practices by the service providers are measured with respect to the efficient frontier and then analyzed. 5.2 PROFI LE OF THE RESPONDENTS This section provides the response rate obtained in the survey and the characteristics of the respondents for this study. The questionnaires including covering letter, were personally distributed to customers of mobile services, during Apr-J un, 2011. The sampling for the study was based on Convenience Sampling Method, taking into account the operational difficulties associated in the process of data capturing. However, by going through this method of sampling it was ensured that all units of population were adequately represented. In determining the sample size the following factors were taken into consideration: Dispersion of the population Time taken by the respondents to complete the questionnaire Resources required for completing the survey Respondents willingness to part with information. Further, the data was specifically collected from residential mobile customers, who had been using the services for at least six months. Finally, of the 550 surveys individually administered, 523 questionnaires were received at a response rate of 95 per cent. On further filtering, 490 responses were found to be completely filled. Survey response rate is defined as the percentage of the total attempted interviews that are completed (Malhotra, 2006). Most of the respondents (about 60%) were pre-paid, while rest (40%) of the respondents accounted for post- paid services. 95 The respondents are classified based on their age, monthly expenditure, educational status and occupation 5.2.1 Classification based on Service Providers The distribution based on service providers of the respondents was analyzed with the help of the following table. Table 5.1 Classification based on Service Providers No. Name Frequency % 1 Aircel 78 15.9 2 Airtel 99 20.2 3 BSNL 91 18.6 4 Idea 57 11.6 5 Reliance 60 12.2 6 Vodafone 59 12 7 Others 46 9.4 From table 5.1, it is inferred that out of the total sample of 490 respondents using different mobile services, 15.9% used Aircel, 20.2% used Airtel, 18.6% used BSNL, 11.6% used Idea, 12.2% used Reliance, 12% used Vodafone and 9.4% used other service providers like Uninor and Videocon services. 5.2.2 Classification based on Plan The distribution based on plan of the respondents was analyzed with the help of the following table. 96 Table 5.2 Classification based on Plan No. Name Frequency % 1 Prepaid 295 60.2 2 Postpaid 195 39.8 From table 5.2, it is inferred that out of the total sample of 490 respondents using different mobile services, 60.2% used the prepaid plan and 39.8% used the postpaid plan. 5.2.3 Classification based on Age The distribution based on age of the respondents was analyzed with the help of the following table. Table 5.3 Classification based on Age No. Name Frequenc y % 1 18-20 27 5.5 2 21-30 87 17.8 3 31-40 87 17.8 4 41-50 158 32.2 5 >50 131 26.7 From table 5.3, it is inferred that out of the total sample of 490 respondents using different mobile services, 5.5% belonged to the age group of 18- 20, 17.8% belonged to the age group of 21-30, 17.8% belonged to the age group of 31-40, 32.2% belonged to the age group of 41-50 and 26.7% of the respondents are aged over 50. 97 5.2.4 Classification based on Length of Use The distribution based on the length of usage by the respondents was analyzed with the help of the following table. Table 5.4 Classification based on Length of Use No. Name Frequency % 1 <2 yrs 104 21.2 2 2 3 yrs 105 21.4 3 >3 yrs 281 57.3 From table 5.4, it is inferred that out of the total sample of 490 respondents using different mobile services, 21.2% respondents are using the services for less than 2 years, 21.4% respondents are using the services for two to three years and 57.3 % respondents are using the services for more than 3 years. 5.2.5 Classification based on Monthly Expenditure The distribution based on monthly expenditure on mobile services was analyzed with the help of the following table. Table 5.5 Classification based on Monthly Expenditure No. Name Frequency % 1 upto 500 155 31.6 2 501-1000 117 23.9 3 1001-2000 80 16.3 4 >2000 138 28.2 98 From table 5.5, it is inferred that out of the total sample of 490 respondents using different mobile services, 31.6 % respondents spend upto Rs. 500, 23.9 % respondents spend between Rs.501 and Rs.1000, 16.3 % respondents spend between Rs. 1001 and Rs. 2000 and 28.2 % respondents spend more than Rs. 2000. 5.2.6 Classification based on Education The distribution based on Educational Status of mobile services was analyzed with the help of the following table. Table 5.6 Classification based on Education No. Name Frequency % 1 UG 156 31.8 2 PG 200 40.8 3 Others 134 27.3 From table 5.6, it is inferred that out of the total sample of 490 respondents using different mobile services, 31.8 % respondents are undergraduates, 40.8 % respondents are postgraduates and 27.3 % respondents belong to others category. The Others category includes school dropouts, uneducated people and people who have just completed schooling. 5.2.7 Classification based on Occupation The employment status of the respondents was analyzed with the help of the following table. 99 Table 5.7 Classification based on Occupation No. Name Frequency % 1 Business 50 10.2 2 Professionals 94 19.2 3 Educator 28 5.7 4 Home Maker 81 16.5 5 Student 191 39.0 6 Others 46 9.4 From table 5.7, it is inferred that out of the total sample of 490 respondents using different mobile services, 10.2 % respondents are having their own business, 19.2 % are professionals, 5.7 % are educators, 16.5 % are home makers, 39 % are students and 9.4% are others. 5.3 DESCRIPTIVE STATISTICS This section describes the preliminary analysis of the data and discusses the results obtained in the study. It starts with the descriptive statistics and the test for normality 5.3.1 Mean and Standard Deviation To study the central tendency and the dispersion of the variables the mean and standard deviation were calculated. The results indicated that the responses of the variables had a good dispersion on the scales. The means of the variables ranged from 2.94 to 3.78 and standard deviation ranged from 0.652 to 1.15. 1 0 0 Table 5.8 Descriptive Statistics of the Interval Scaled Variables Construct Measure Variables Mean Std Dev Skew ness Kurtosis Contact employees performthe service right the first time 3.57 0.846 -0.283 0.066 Thecompany provides services at thepromised time 3.51 0.844 -0.240 -0.379 You arekept well-informed about theprogress of your complaints 3.26 0.962 -0.080 -0.615 Billing systemis accurateand error free 3.60 0.912 -0.219 -0.478 Reliability Bills are received in time 3.71 0.853 -0.053 -0.533 Contact employees gives you prompt service 3.55 0.849 -0.259 -0.357 Your Complaints / queries are taken seriously 3.48 0.924 -0.066 -0.376 Your complaints areresolved quickly 3.4 0.917 -0.014 -0.298 Responsiveness They are always willing to help you 3.54 0.977 -0.430 -0.165 Contact employees are friendly and polite while handling your complaints / queries 3.7 0.959 -0.165 -0.932 They havetheadequateknowledgeof tariffs and plans of serviceproviders 3.77 0.841 -0.374 -0.366 Thebehaviour of contact employees instils confidencein you 3.57 0.913 -0.354 0.151 Assurance You feel safein your transactions with your serviceprovider 3.66 0.794 -0.216 0.396 For lodging the complaints service provider is easily accessible 3.23 1.15 -0.451 -0.499 They haveyour best interests in heart 3.18 0.964 -0.259 0.123 Contact employees give you individual attention 3.36 0.962 -0.38 0.149 Empathy Theemployees understand your specific needs 3.42 0.935 -0.272 -0.066 Service provider's physical facilities are visually appealing 3.49 0.993 -0.316 -0.029 Contact employees appear neat- 3.35 0.913 -0.251 0.242 Tangibles Materials associated with theservice(such as pamplets etc) arevisually appealing 3.53 0.897 -0.092 -0.744 1 0 1 Construct Measure Variables Mean Std Dev Skew ness Kurtosis They have convenient business hours 3.51 0.775 -0.126 0.168 Ease of lodging the complaints / queries 3.46 0.838 -0.061 -0.186 Your service provider provides flexibility in the payment of bills 3.21 0.909 -0.043 -0.137 Convenience Application formalities are simple 3.54 0.845 -0.229 -0.355 Your service provider provides sufficient geographical coverage (on highways, inside the buildings and basement) 3.57 0.894 -0.068 -0.745 You experience minimumpremature termination of calls during conversation (ie call drops) 3.44 0.979 -0.118 -0.603 You get clear and undisturbed voice 3.57 0.876 -0.256 -0.115 Your call gets connected to the called person during the first attempt most of the time 3.43 1.10 -0.185 -0.462 Customer Perceived Network Quality You areableto makecalls at peak hours 3.66 1.077 -0.702 0.069 I will transact with this mobile service provider again for future needs 3.28 0.721 -0.859 0.630 I will try new services that areprovided by theserviceprovider 3.17 0.754 -0.654 0.063 I will recommend other people to patronize this service provider 3.25 0.804 -0.698 -0.486 Behavioural I will say positive things to other people about the services provided by the mobile service operator 3.34 0.758 -1.018 0.572 I will continue to patronize this mobile service provider / service even if theservicecharges are increased moderately 3.03 0.987 -0.535 -0.949 I havestrong preferenceto this serviceprovider 3.27 0.811 -0.958 0.328 I will continueto usethis serviceprovider regardless of changes in theservice 3.26 0.819 -0.806 -0.234 Attitudinal I amlikely to pay little more for the services when situation arises 3.05 0.921 -0.439 -1.015 1 0 2 Construct Measure Variables Mean Std Dev Skew ness Kurtosis To me, this serviceprovider would rank first among others 2.94 0.849 -0.343 -0.694 I would continuethis servicefor along period of time 3.30 0.780 -1.008 0.605 I will deal exclusively with the service provider 3.22 0.846 -0.842 -0.071 Cognitive Theserviceof theprovider reflect alot about who I am 3 0.841 -0.393 -0.644 I havefound this provider better than others 3.24 0.867 -0.751 -0.585 Conative Repeatedly, the performance of this provider is superior to that of its competitors 3.23 0.817 -0.821 -0.040 I dislike the terms of the service provider 3.44 0.696 -1.123 0.903 I liketheperformanceand services of theprovider 3.40 0.686 -1.069 1.197 I haveanegativeattitudetoward this provider 3.47 0.737 -1.540 2.347 Affective I amsatisfied wih my decision to stay with this provider 3.43 0.721 -1.428 2.327 Theemployees at theproviders placegiveindividual attention 3.48 0.674 -1.185 1.113 Theemployees understand my specific needs and go out of their way to help me 3.36 0.705 -0.863 0.262 Theemployees respond caringly when I sharemy problems 3.44 0.692 -1.175 1.306 Trust Thepersonnel at theserviceprovider arefilled with professionalismand dedication 3.50 0.652 -1.091 0.612 I amcommitted to theserviceprovider 3.38 0.762 -1.173 0.969 Even when I get to hear any negativeinformation about theprovider, I would still continue with this serviceprovider 3.31 0.770 -0.987 0.549 I like switching fromone service provider to another 3.25 0.821 -0.804 -0.200 Commitment My continued association with the mobile operator is important to me 3.32 0.780 -1.01 0.480 1 0 3 Construct Measure Variables Mean Std Dev Skew ness Kurtosis Likelihood to choosetheserviceagain 3.78 0.854 -0.794 0.113 Likelihood to recommend 3.58 0.959 -0.730 0.356 Likelihood to pass on positive feeling about the services 3.41 0.966 -0.420 0.048 Advocacy Likelihood to continuepurchasing thesameproducts / services 3.48 0.964 -0.611 0.1622 Likelihood to purchasedifferent products / services 3.55 0.849 -0.259 -0.357 Purchase Likelihood to increase frequency of purchasing 3.48 0.924 -0.066 -0.376 Defection Likelihood to continuethis only as asupplementary connection 3.38 0.762 -1.173 -0.969 Likelihood to switch to a different provider 3.31 0.770 -0.987 0.549 104 5.3.2 Normality Normality is the most fundamental assumption in multivariate analysis which greatly influences the validity of the results (Hair et al., 1998). If the variation from the normal distribution is sufficiently large, all the statistical tests are invalid (Hair et al., 1998). Therefore, non-normal variables identified should be handled before further examination. In this analysis, Skewness and kurtosis values were used to measure normality of the sixty three interval scaled independent and dependent variables. If a Skewness or Kurtosis value exceeds +1.00, a non-normal distribution is identified. In this study, there were 5 items which were not normally distributed [4 items of Affective Loyalty] and [1 item of Cognitive Loyalty]. To improve the normality of these 5 variables, the data were transformed individually by taking the square root (Hair et al, 1998). After transformation, the absolute values of Kurtosis were still larger than 1.00. For this reason, another transforming method, logarithm was employed for further improvement. After the second transformation, kurtosis values of all variables were less than 1.00. 5.4 TEST OF RELIABILI TY Reliability is an assessment of the degree of consistency between multiple measurements of a variable (Hair et al., 1998). The internal consistency of measures is an indication of the homogeneity of the items which measures, the same construct (Sekaran, 2003). Thus to obtain a high reliability of a measure, the items should be highly correlated with another to independently measure the construct. As indicated in chapter 3, the scales of the current study were obtained from the extant studies. In this analysis, Cronbachs alpha, is used to test reliability of the interval scaled variables. Cronbachs alpha is the most popular test of inter- item consistency, which is useful for interval scaled variables (Cronbach, 1946). From the Cronbach alphas standpoint, the closer the co-efficient alpha is to 1.0, the higher the 105 internal consistency reliability. In general, the cut-off of reliability Cronbach alpha is 0.6 (Nunnally 1978). It means that only the reliabilities _ 0.60 are acceptable. The results showed all alpha coefficients ranged from 0.75 to 0.9 indicating good consistency among the items within each dimension. Clark and Watson (1995) notes that inter-item correlations which examine scale internal consistency are sometimes better than the coefficient alpha. A rule of thumb is that these values should be at least 0.30. The item-to-total correlations is >0.4. Hence in this research the data has been found reliable on all three ways. 5.4.1 Service Quality Dimensions Many studies in the past in this area primarily focused on functional quality aspects (i.e., pertaining to service delivery process or how the services are delivered) and inadequately addressed technical quality aspects (i.e., issues concerning what is actually delivered). However, researchers in cellular mobile communication (Wang and Lo, 2002; Johnson and Sirikit, 2002), emphasize that technical quality attributes play an important role in forming service quality perceptions of customers. In light of this, Seth et al (2008) investigated service quality structure by combining both functional as well as technical quality (i.e., network quality in cellular mobile context) attributes. The instrument to measure service quality was used to understand the respondents perception about their level of agreement with respect to questions, each on a five point Likert Scale (1 Strongly Disagree and 5 Strongly Agree). The reliability of items was assessed by computing the coefficient alpha (Cronbach, 1951), that measures the internal consistency of the items. The alpha values for the service quality dimensions range from 0.775 to 0.863 and given in the table below. 106 Table 5.9 Cronbachs Alpha for Service Quality dimensions I nter item correlations (absolute value) Construct Items Alpha ) Alpha ( ) if item deleted Item to total correlations Highest Lowest Rel1 0.828 0.598 Rel2 0.818 0.640 Rel3 0.827 0.613 Rel4 0.791 0.735 Reliability Rel5 0.846 0.806 0.685 0.695 0.429 Resp1 0.822 0.722 Resp2 0.846 0.659 Resp3 0.822 0.717 Responsiveness Resp4 0.863 0.809 0.750 0.687 0.542 Assu1 0.730 0.661 Assu2 0.799 0.513 Assu3 0.739 0.643 Assurance Assu4 0.802 0.736 0.659 0.617 0.376 Emp1 0.792 0.606 Emp2 0.733 0.717 Emp3 0.757 0.665 Empathy Emp4 0.816 0.795 0.580 0.639 0.423 Tang1 0.654 0.649 Tang2 0.716 0.593 Tangibles Tang3 0.775 0.715 0.594 0.560 0.486 Conv1 0.722 0.590 Conv2 0.695 0.636 Conv3 0.714 0.602 Convenience Conv4 0.778 0.762 0.506 0.542 0.361 Cupqly1 0.774 0.444 Cupqly2 0.765 0.477 Cupqly3 0.717 0.639 Cupqly4 0.705 0.653 Customer Perceived Network Quality Cupqly5 0.781 0.731 0.584 0.607 0.302 107 5.4.2 Service Loyalty Dimensions The instrument to measure service loyalty was used to understand the respondents psychological and behavioural intentions towards the service provider, with respect to questions, each on a four point Likert Scale. A statement is given, and the endpoints correspond to agree strongly and disagree strongly. Use of the scales is practical and interesting. Scales with an even number of points do not have a midpoint and in that sense force a choice. 1 - Disagree strongly 2 - Disagree somewhat 3 - Agree somewhat 4 - Agree strongly The reliability of items was assessed by computing the coefficient alpha that measures the internal consistency of the items. The alpha values for the service quality dimensions range from 0.78 to 0.894 Table 5.10 Cronbachs Alpha of Service Loyalty dimensions I nter item correlations (absolute value) Construct Items Alpha ) Alpha ( ) if item deleted Item to total correlations Highest Lowest Behloyl1 0.813 0.725 Behloyl2 0.852 0.627 Behloyl3 0.793 0.769 Behavioural Loyalty Behloyl4 0.859 0.822 0.701 0.725 0.476 Attloyl1 0.790 0.708 Attloyl2 0.792 0.706 Attloyl3 0.801 0.682 Attitudinal Loyalty Attloyl4 0.843 0.821 0.634 0.685 0.497 Cogloyl1 0.832 0.702 Cogloyl2 0.843 0.676 Cogloyl3 0.831 0.704 Cognitive Loyalty Cogloyl4 0.865 0.801 0.776 0.852 0.469 108 I nter item correlations (absolute value) Construct Items Alpha ) Alpha ( ) if item deleted Item to total correlations Highest Lowest Conloyl1 0.676 0.676 Conative Loyalty Conloyl2 0.806 0.676 0.676 0.676 0.676 Affloyl1 0.735 0.567 Affloyl2 0.725 0.587 Affloyl3 0.724 0.589 Affective Loyalty Affloyl4 0.780 0.721 0.594 0.721 0.334 Truloyl1 0.853 0.765 Truloyl 2 0.849 0.775 Truloyl 3 0.858 0.752 Trust Loyalty Truloyl 4 0.888 0.866 0.731 0.715 0.636 Commloyl1 0.876 0.735 Commloyl 2 0.877 0.732 Commloyl 3 0.844 0.820 Commitment Loyalty Commloyl 4 0.894 0.859 0.781 0.745 0.606 5.4.3 Loyalty I ndices The reliability of items was assessed by computing the coefficient alpha (Cronbach, 1951), that measures the internal consistency of the items. The alpha values for the service quality dimensions range from 0.750 to 0.908 Table 5.11 Cronbachs Alpha of Loyalty I ndices I nter item correlations (absolute value) Construct Items Alpha ) Alpha ( ) if item deleted Item to total correlations Highest Lowest LoyIndx1 0.906 0.711 LoyIndx2 0.856 0.860 LoyIndx3 0.864 0.839 Advocacy Loyalty Index LoyIndx4 0.908 0.891 0.766 0.810 0.578 LoyIndx5 0.602 0.602 Purchase Loyalty Index LoyIndx6 0.750 0.602 0.602 0.602 0.602 LoyIndx7 0.606 0.606 Defection Loyalty Index LoyIndx8 0.754 0.606 0.606 0.606 0.606 109 5.5 FACTOR ANALYSI S After establishing the reliability for the constructs, factor analysis was carried out to confirm the factors of Service Quality dimensions, Service Loyalty dimensions and Loyalty Indices. Factor analysis is a method to determine the dimensionality of a scale. Factor analysis is used to identify clusters of inter- correlated variables called factor. To check the factorability, either the values of correlation matrix should be greater than 0.3; values on anti-image correlation diagonals should be greater than 0.5 or Measures of Sampling Adequacy Bartletts, significant and KMO should be greater than 0.6. In this study, Factor analysis using Principal Component Analysis with Varimax rotation was carried out. Table 5.12 through 5.17 provides the summary from SPSS output of data for factor analysis. Before proceeding for the factor analysis, appropriateness of factor analysis needs to be assessed. This is done by examining sampling adequacy through Kaiser- Meyer-Olkin (KMO) statistic. The tables provide the SPSS output of data for factor analysis. KMO value of greater than 0.6 can be considered as adequate. (Kaiser and Rice, 1974). Service Quality Table 5.12 KMO and Bartlett's Test Service Quality Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .876 Approx. Chi- Square 7404.62 6 Df 406 Bartlett's Test of Sphericity Sig. .000 110 Table 5.13 Factor Analysis Service Quality Dimensions Components Reliability Responsiveness N/w Qlty Assurance ConvenienceEmpathy Tangibles Rel4 .835 Rel5 .758 Rel2 .677 Rel1 .643 Rel3 .639 Resp3 .764 Resp1 .749 Resp4 .729 Resp2 .696 Cupeneqlty3 .790 Cupeneqlty4 .752 Cupeneqlty1 .652 Cupeneqlty5 .642 Cupeneqlty2 .614 Assu4 .767 Assu1 .718 Assu3 .679 Assu2 .645 Conv1 .779 Conv3 .779 Conv2 .637 Conv4 .587 Emp2 .657 Emp4 .557 Emp1 .544 Emp3 .522 Tang3 .759 Tang1 .729 Tang2 .727 111 Service Loyalty Table 5.14 KMO and Bartlett's Test Service Loyalty Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .849 Bartlett's Test of Sphericity Approx. Chi- Square 8234.730 Df 325 Sig. .000 Table 5.15 Factor Analysis Service Loyalty Dimensions Commloyl Truloyl Attloyl Behloyl Cogloyly Affloyl Conloyl commloyl1 .835 commloyl2 .804 commloyl3 .786 commloyl4 .769 Truloyl2 .839 Truloyl1 .833 Truloyl4 .826 Truloyl3 .818 Attloyl1 .808 Attloyl4 .759 Attloyl2 .669 Attloyl3 .635 Behloyl3 .866 Behloyl1 .839 Behloyl4 .807 Behloyl2 .743 cogloyl4 .862 cogloyl1 .799 cogloyl3 .588 cogloyl2 .551 Affloyl1 .878 Affloyl2 .871 Affloyl4 .574 Affloyl3 .549 conloyl1 .828 conloyl3 .811 112 Loyalty I ndices Table 5.16 KMO and Bartlett's Test Loyalty I ndices Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .824 Bartlett's Test of Sphericity Approx. Chi- Square 2068.275 Df 28 Sig. .000 Table 5.17 Factor Analysis Loyalty I ndices Advocacy LI Defection LI Purchase LI Loyltyindex3 .899 Loyltyindex2 .871 Loyltyindex4 .825 Loyltyindex1 .810 Loyltyindex7 .880 Loyltyindex8 .854 Loyltyindex6 .881 Loyltyindex5 .831 The items having factor loadings less than 0.5 were eliminated (Hair et al., 2005). Finally, seven factors comprising twenty-nine items for service quality and seven factors comprising twenty-six items for service loyalty, all having eigen values of unity and above were extracted and the results are shown in Table 5.13 and 5.15. From table 5.17, the items that load on the first factor appear to have a strong emotional component to them, reflecting the extent to which customers advocate the company. Consequently, this factor is labeled Advocacy Loyalty Index. The items that load on the second factor reflect indications or symptoms of customers defecting to competitors and the items that load on the third factor reflect 113 specific purchasing behaviors. Consequently, the second factor is labeled Defection Loyalty Index and the third factor as Purchase Loyalty Index. Further, in order to assess the appropriateness of the data for factor analysis, the communalities derived from the factor analysis were reviewed. These were all relatively large (greater than 0.5), suggesting that the data set is appropriate (Stewart, 1981). The individual dimensions of the proposed instrument explained total variance exceeding 60 per cent, suggesting the appropriateness of the process. In this study, Principal Component Analysis with Varimax rotation is used and the values for factor analysis are given below. The KMO values are 0.876 for service quality, 0.849 for service loyalty and 0.824 for loyalty indices. Bartletts Test of Sphericity is significant at 0.000 for all the constructs. Communalities, which refer to the extracted factors that explains most of the variance in the variables being analyzed is greater than 0.5. For service quality, the lowest value of communality is 0.532; for service loyalty it is 0.599 and for loyalty indices it is 0.715. A good factor solution is one that explains most variance with few factors. Realistically it should be between 50-75% of the variance explained. In this study, the variances explained are as follows: Service Quality 66.59 % Service Loyalty 73.80% Loyalty Indices 79.84% Confirmatory factor analysis model is run through SPSS Amos 19 and the key model statistics are shown in Table 5.18 through 5.20. 114 Table 5.18 Confirmatory Factor Analysis Service Quality Factors Comparative Fit Index (CFI) Goodness of Fit Index (GFI) Reliability .971 .972 Responsiveness .991 .99 Assurance .973 .982 Empathy .935 .951 Tangibles .925 .931 Convenience .966 .981 Customer Perceived N/w Quality .956 .973 Table 5.19 Confirmatory Factor Analysis Service Loyalty Factors Comparative Fit Index (CFI) Goodness of Fit Index (GFI) Commitment Loyalty .956 .948 Trust Loyalty .997 .994 Attitudinal Loyalty .949 .956 Behavioural Loyalty 1.00 .999 Affective Loyalty .911 .918 Cognitive Loyalty .985 .980 Conative Loyalty .943 .952 Table 5.20 Confirmatory Factor Analysis Loyalty I ndices Factors Comparative Fit Index (CFI) Goodness of Fit Index (GFI) Advocacy Loyalty Index .934 .919 Purchase Loyalty Index .962 .951 Defection Loyalty Index .931 .903 From the Confirmatory Factor Analysis, it is inferred that the CFI and GFI values are greater than 0.91 which confirms the uni dimensionality of the factors. 115 Further the Confirmatory Factor Analysis of each of the multidimensional constructs was examined to test if it is statistically fit. The figures below gives the details of the CFA on each of the constructs, namely, service quality, service loyalty and loyalty indices. Figure 5.1 CFA Service Quality 116 Chi-square value for the overall model fit was 1692.327 for 356 degrees of freedom (p<0.001). Fit indices for the above model were NFI =0.918; ; CFI = 0.914; GFI =0.92, RMSEA =0.058. Figure 5.2 CFA Service Loyalty 117 Chi-square value for the overall model fit was 1721.072 for 402 degrees of freedom (p<0.001). Fit indices for the above model were NFI =0.941; CFI = 0.932; GFI =0.924; RMSEA =0.054. Figure 5.3 CFA Loyalty I ndices Chi-square value for the overall model fit was 63.976 for 17 degrees of freedom (p<0.001). Fit indices for the above model were NFI =0.969; CFI =0.977; GFI =0.969; RMSEA =0.075. 118 5.6 HYPOTHESES TESTI NG 5.6.1 ANOVA One way ANOVA is used when only one independent variable and one dependent variable are considered. It is used in this study to check the influence of each of the categorical demographic variables - length of use, age, monthly expenses, education and occupation on service quality dimensions, service loyalty dimensions and loyalty indices. An important assumption of one way ANOVA is that of homogeneity which requires equal variances of the dependent variable across categories of the independent variable. The Levene's Statistic is performed to test for this assumption. If Levene's statistic is significant at p<0.05, then it indicates condition of heterogeneity of variance for that particular independent variable - dependent variable combination, implying a violation of the homogeneity assumption. In such a situation, the Brown-Forsythe's F, or the Welch's F must be used to draw inferences from the ANOVA results instead of the regular ANOVA statistic (Field, 2005). The table 5.21 gives the details of the relationship between the demographic variables and the service quality dimensions. The F Values of Brown- Forsythe and Welch are given. The LSD Post Hoc results give the details of the values which are significantly different. H 01 = There is no difference in the user perception of Service Quality with respect to Demographic Variables 1 1 9 Table 5.21 One-Way ANOVA Demographic Variables on Service Quality Demographic Variables Service Quality Dimensions Sig. Value Brown- Forsyth F Value Welch F Value LSD Post Hoc Inference Reliability 0.013 3.551 3.487 The subscribers using the service <2 yrs find the service providers more reliable that >3 yrs L e n g t h
o f
U s e Responsiveness 0.001 7.719 7.958 The subscribers using the service >3 yrs find the service providers are more responsive than <2yrs Length of Use has influence on Reliability and Responsiveness only. A g e Reliability 0.000 14.056 12.484 The subscribers whose age is >50 find that the service provider is reliable. Age has influence only on Reliability. Reliability 0.000 13.887 15.517 The subscribers whose monthly expenses is between Rs.501-1000 find that the service provider is reliable than the subscribers whose monthly expenses are >Rs.2000. M o n t h l y
E x p e n s e s Responsiveness 0.003 6.215 6.396 The subscribers whose monthly expenses is >Rs.2000 find that the service provider is better responsive than the subscribers whose monthly expenses are upto Rs.500 and Rs.501-1000. Monthly Expenditure of the respondents influences Reliability, Responsiveness, Assurance, Empathy, Convenience and Customer Perceived Network Quality. 1 2 0 Demographic Variables Service Quality Dimensions Sig. Value Brown- Forsyth F Value Welch F Value LSD Post Hoc Inference Assurance 0.002 5.553 5.659 The subscribers whose monthly expenses is > Rs.2000 find that the service provider promises an assured service than the subscribers whose monthly expenses are upto Rs.500 and Rs.501-1000. Empathy 0.000 10.800 9.056 The subscribers whose monthly expenses are > Rs.2000 find the empathy of the service provider better than the subscribers whose monthly expenses are upto Rs.500 and Rs.501- 1000. Convenience 0.003 3.553 3.267 The subscribers whose monthly expenses is between Rs.501-1000 find the service provider convenient than the subscribers whose monthly expenses are >Rs.2000. Customer Perceived Network Quality 0.000 5.032 5.211 The subscribers whose monthly expenses are > Rs.2000 find that the service provider has a good network quality than the subscribers whose monthly expenses are upto Rs.500. E d u c a t i o n Reliability 0.000 12.635 12.323 The subscribers whose education is UG find the service provider more reliable that the subscribers whose education are PG or others. Education of the respondents influences reliability only 121 The rest of the variables were found to be insignificant. Since there are many variables that are found to be insignificant, the hypothesis H 01 is accepted. The table 5.22 gives the details of the relationship between the demographic variables and the service loyalty dimensions. The F Values of Brown- Forsythe and Welch are given. The LSD Post Hoc results give the details of the values which are significantly different. H 02 = There is no difference in the user perception of Service Loyalty with respect to Demographic Variables 1 2 2 Table 5.22 One-Way ANOVA Demographic Variables on Service Loyalty Demographic Variables Service Loyalty Dimensions Sig Value Brown- Forsyth F Value Welch F Value LSD Post Hoc I nference Cognitive 0.009 4.079 4.772 The subscribers using the service for 2-3 yrs are more cognitive loyal than <2 yrs. L e n g t h
o f U s e Trust 0.015 3.376 NA The subscribers using the service < 2yrs are more trust loyal than >3 yrs. Length of Use has influence on Cognitive Loyalty, Trust Loyalty and Commitment Loyalty. A g e Commitment 0.004 2.469 NA The subscribers in the age group 21-30 are more commitment loyal than the subscribers whose age is >50. Age has influence only on Commitment Loyalty Attitudinal 0.006 2.713 2.888 The subscribers whose monthly expenses is upto Rs.500 are more attitude loyal than the subscribers whose monthly expenses are > Rs.2000. Affective 0.001 5.685 4.211 The subscribers whose monthly expenses is Rs.1001-2000 are more affective loyal than the subscribers whose monthly expenses are upto Rs.500, Rs.501-1000 and >Rs.2000. Trust 0.004 10.462 11.648 The subscribers whose monthly expenses are < Rs.500 are more trust loyal than the subscribers whose monthly expenses are >Rs.500. M o n t h l y
E x p e n s e s Commitment 0.001 5.887 6.391 The subscribers whose monthly expenses are > Rs.2000 are more commitment loyal than the subscribers whose monthly expenses are < Rs.2000. Monthly Expenses have influence on Attitudinal Loyalty, Affective Loyalty, Trust Loyalty and Commitment Loyalty E d u . Trust 0.034. 3.471 3.364 The subscribers whose education is UG are more trust loyal than the subscribers whose education are PG. Education has influence only on Trust Loyalty 123 The rest of the variables were found to be insignificant. Since there are many variables that are found to be insignificant, the hypothesis H 02 is accepted. The table 5.23 gives the details of the relationship between the demographic variables and the loyalty indices dimensions. The F Values of Brown- Forsythe and Welch are given. The LSD Post Hoc results give the details of the values which are significantly different. H 03 = There is no difference in the user tendency towards Loyalty with respect to Demographic Variables 1 2 4 Table 5.23 One-Way ANOVA Demographic Variables on Loyalty I ndices Demographic Variables Loyalty I ndices Sig. Value Brown- Forsyth F Value Welch F Value LSD Post Hoc I nference L e n g t h
o f U s e Purchase Loyalty Index 0.032 8.156 8.892 The subscribers using the service for > 3 yrs exhibit purchase loyalty than subscribers using the services of the provider <2 yrs. Length of Use has influence on only Purchase Loyalty Index A g e Defection Loyalty Index 0.000 2.469 NA The subscribers in the age group 21-30 exhibit more symptoms of defection than the subscribers whose age is >50. Age has influence on only Defection Loyalty Index Advocacy Loyalty Index 0.001 8.947 8.397 The subscribers whose monthly expenses are >Rs.2000 exhibit more advocacy than the subscribers whose monthly expenses are <Rs.2000. Purchase Loyalty Index 0.000 12.667 12.128 The subscribers whose monthly expenses are > Rs.2000 are more purchase loyal than the subscribers whose monthly expenses are <Rs.2000. M o n t h l y
E x p e n s e s Defection Loyalty Index 0.000 4.402 5.161 The subscribers whose monthly expenses are >Rs.2000 exhibit more symptoms of defection than the subscribers whose monthly expenses are <Rs.2000. Monthly Expenses have influence on Advocacy Loyalty Index, Purchase Loyalty Index and Defection Loyalty Index 125 The rest of the variables were found to be insignificant. Since there are many variables that are found to be insignificant, the hypothesis H 03 is accepted. 5.6.2 Linear Regression The hypotheses H 01 to H 06 were tested using Linear Regression. Confirmatory path models were derived to depict the relationship between service quality on service loyalty, service quality and loyalty indices and service loyalty and loyalty indices using Structural Equation Modeling. H 01 : There is no significant relationship between demographic variables and service quality dimensions. The R value, indicating the strength of relationship between demographic variables and each of the service quality dimensions is given in the table below. The coefficient of determination (R 2 ) is used as an estimate of the predictive power of the regression model. R 2 values measure the percentage of the total variance of the dependent variable about its mean that is explained or accounted for by the independent variable (Lewis-Beck 1993). Table 5.24 Relationship between Demographic Variables and SERVQUAL Dimensions Dimension R R Square Durbin Watson F Values Beta Value Demographic Variable Sig. values 0.185 Age 0.00 0.126 Mon_Exp 0.002 0.120 Edu 0.049 Reliability 0.382 0.146 1.954 16.583 0.160 Occup 0.020 0.122 Len_use 0.030 Responsiveness 0.223 0.050 1.988 5.088 0.160 Mon_Exp 0.000 Customer Perceived Network Quality 0.199 0.040 1.213 3.973 0.138 Mon_Exp 0.002 Assurance 0.203 0.041 1.761 4.155 0.136 Mon_Exp 0.002 Convenience 0.143 0.020 1.748 2.017 0.104 Mon_Exp 0.019 Empathy 0.204 0.041 1.808 4.190 0.167 Mon_Exp 0.000 Tangibles 0.158 0.025 1.955 2.469 0.140 Edu 0.032 126 RLY =2.155 +0.185*AGE +0.126*ME +0.120*EDU +0.160*OCC+0.571 RSP =2.316 +0.122*LEN +0.160*ME +0.71 CPN =2.640 +0.138*ME +0.68 ASR =2.518 +0.136*ME +0.64 The Regression Equations are given below. The abbreviations used are Demographic Variables AGE Age ME Monthly Expenses EDU Education OCC Occupation LEN Length of Use Service Quality Dimensions RLY Reliability RSP Responsiveness CPN Customer Perceived Network Quality ASR Assurance CNV Convenience EMP Empathy TNG Tangibles 127 CNV =2.004 +0.104*ME +0.60 TNG =2.396 +0.140*EDU +0.744 EMP =2.106 +0.167*ME +0.546 The independence of error term was also met because the value of Durbin-Watson, which was close to 2 (the closer to the value to 2, the better the independence of error) (Field, 2005). From the table above, it is inferred that the demographic variable, Monthly Expenditure has significant relationship with most of the service quality dimensions, Occupational Status has significant relationship with Reliability and Educational Status has significant relationship with Tangibles. Hence H 01 is accepted for all the other variables. H 02 : There is no significant relationship between demographic variables and service loyalty dimensions Table 5.25 Relationship between Demographic Variables and SERVLOYAL Dimensions Dimension R R Square Durbin Watson F Value Beta Value Demographic Variable Sig values Commitment Loyalty 0.195 0.038 1.908 3.819 0.130 Mon_Exp 0.003 Trust Loyalty 0.255 0.065 1.963 6.706 0.197 Mon_Exp 0.000 The Regression Equations are given below. The abbreviations used are Demographic Variables AGE Age 128 COL =1.759 +0.130*ME +0.754 TRL =0.896 +0.197*ME +0.652 ME Monthly Expenses EDU Education OCC Occupation LEN Length of Use Service Loyalty Dimensions COL Commitment Loyalty TRL Trust Loyalty ATL Attitudinal Loyalty BHL Behavioural Loyalty CGL Cognitive Loyalty AFL Affective Loyalty CNL Conative Loyalty From the table above, it is evident that though the Durbin- Watson values are close to 2, the R values shows that there is not very significant relationship between Demographic Variables and SERVLOYAL dimensions. There is only one demographic variable, monthly expenditure which has significant relationship with commitment loyalty and trust loyalty. Hence H 02 is accepted for all other variables. 129 ALI =2.697 +0.131*ME +0.79 H 03 : There is no significant relationship between demographic variables and loyalty indices Table 5.26 Relationship between Demographic Variables on Loyalty I ndices Dimension R R Square Durbin Watson F Value Beta Value Demographic Variable Sig. Value Advocacy Loyalty Index 0.215 0.046 1.881 4.692 0.131 Mon_Exp 0.003 Defection Loyalty Index 0.178 0.032 1.812 3.164 0.114 Mon_Exp 0.010 0.180 Mon_Exp 0.000 Purchase Loyalty Index 0.302 0.091 1.99 9.717 0.087 Occup 0.002 The Regression Equations are given below. The terms used for abbreviations are Demographic Variables AGE Age ME Monthly Expenses EDU Education OCC Occupation LEN Length of Use Loyalty Indices ALI Advocacy Loyalty Index DLI Defection Loyalty Index PLI Purchase Loyalty Index 130 PLI =2.366 +0.180*ME +0.087*OCC +0.74 DLI =2.949 +0.114*ME +0.52 From the above table, it is inferred that demographic variable Monthly Expenditure has significant relationship with all the loyalty indices and Occupation has significant relationship with only the purchase loyalty index. From the R values, R 2 values, Durbin Watson values, it can be inferred that H 03 is accepted for all other variables. H 04 : There is no significant relationship between Service Quality and Service loyalty Table 5.27 Relationship between SERVQUAL and SERVLOYAL Dimension R value R Square Durbin Watson F Value Beta Value Demographic Variable Sig. Value 0.135 Reliability 0.001 0.233 Responsiveness 0.000 0.266 N/w Quality 0.000 0.137 Assurance 0.001 Commitment Loyalty 0.427 0.182 1.946 15.331 0.116 Empathy 0.005 0.115 Reliability 0.007 0.229 Responsiveness 0.000 0.201 Assurance 0.000 Trust Loyalty 0.381 0.145 1.803 11.688 0.167 Empathy 0.000 0.228 Reliability 0.000 0.097 N/w Quality 0.021 0.092 Convenience 0.029 Attitudinal Loyalty 0.389 0.152 1.819 12.302 0.267 Empathy 0.000 0.110 Responsiveness 0.014 Behavioural Loyalty 0.173 0.30 2.056 2.134 0.111 Empathy 0.014 0.117 Responsiveness 0.008 0.157 Convenience 0.000 0.131 Empathy 0.003 Cognitive Loyalty 0.277 0.077 1.805 5.714 0.135 Tangibles 0.002 0.131 Reliability 0.015 Affective Loyalty 0.35 0.085 1.86 7.14 0.346 Empathy 0.000 0.180 Responsiveness 0.000 0.096 N/w Quality 0.029 0.114 Empathy 0.010 Conative Loyalty 0.267 0.071 1.740 5.291 0.128 Tangibles 0.004 131 TRL =1.133 +0.115*RLY +0.229*RSP +0.201*ASR +0.167*EMP +0.531 COL =0.685 +0.135*RLY +0.233*RSP +0.266*CPN +0.137*ASR + 0.116*EMP +0.638 From the above table, it can be inferred all dimensions of SERVLOYAL have significant relationship with SERVQUAL dimensions. The Regression Equations are given below: The abbreviations are Service Quality Dimensions RLY Reliability RSP Responsiveness CPN Customer Perceived Network Quality ASR Assurance CNV Convenience EMP Empathy TNG Tangibles Service Loyalty Dimensions COL Commitment Loyalty TRL Trust Loyalty ATL Attitudinal Loyalty BHL Behavioural Loyalty CGL Cognitive Loyalty AFL Affective Loyalty CNL Conative Loyalty 132 ATL =1.001 +0.228*RLY +0.097*CPN +0.097*CNV +0.267*EMP +0.59 BHL =1.186 +0.110*RSP +0.111*EMP +0.57 CGL =1.275 +0.117*RSP +0.157*CNV +0.131*EMP +0.135*TNG +0.59 AFL =1.239 +0.131*RLY +0.346*EMP +0.50 CNL =1.459 +0.180*RSP +0.096*CPN +0.114*EMP +0.128*TNG +0.69 Thus from the R values, R 2 values, Durbin Watson values, it can be inferred that H 04 is rejected. H 05 : There is no significant relationship between service quality and loyalty indices Table 5.28 Relationship between SERVQUAL and Loyalty I ndices Dimension R R Square Durbin Watson F Value Beta Value Demographic Variable Sig. Value 0.190 Reliability 0.000 0.146 Responsiveness 0.000 0.263 N/w Quality 0.000 0.160 Assurance 0.000 Advocacy Loyalty Index (ALI) 0.434 0.188 1.829 15.956 0.188 Empathy 0.000 0.140 Reliability 0.001 0.150 Responsiveness 0.001 0.129 N/w Quality 0.003 Defection Loyalty Index (DLI) 0.285 0.081 1.848 6.073 0.108 Empathy 0.014 0.233 Reliability 0.000 0.779 Responsiveness 0.000 0.107 Assurance 0.000 0.111 Convenience 0.000 Purchase Loyalty Index (PLI) 0.852 0.727 1.817 183.066 0.200 Empathy 0.000 133 ALI = 2.79 +0.190*RLY +0.146*RSP +0.263*CPN +0.160*ASR + 0.188*EMP +0.509 DLI =-1.9 +0.140*RLY +0.150*RSP +0.129*CPN +0.108*EMP +0.602 PLI = -2.8 +0.233*RLY +0.799*RSP +0.107*ASR +0.111*CNV + 0.200*EMP +0.52 From the above table, it can be inferred that all the three loyalty indices have significant relationship with SERVQUAL dimensions. The Regression Equations are given below: The abbreviations are Service Quality Dimensions RLY Reliability RSP Responsiveness CPN Customer Perceived Network Quality ASR Assurance CNV Convenience EMP Empathy TNG Tangibles Loyalty I ndices ALI Advocacy Loyalty Index DLI Defection Loyalty Index PLI Purchase Loyalty Index 134 Thus from the R values, R 2 values, Durbin Watson values, it can be inferred that H 05 is rejected. H 06 : There is no significant relationship between service loyalty and loyalty indices Table 5.29 Relationship between SERVLOYAL and Loyalty I ndices Dimension R R Square Durbin Watson F Values Beta Value Demographic Variable Sig. Value 0.135 Commitment 0.000 0.150 Trust 0.000 0.528 Attitudinal 0.000 0.233 Cognitive 0.000 Advocacy Loyalty Index (ALI) 0.634 0.401 1.801 46.176 0.145 Affective 0.000 0.894 Commitment 0.000 0.136 Behavioural 0.000 Defection Loyalty Index (DLI) 0.907 0.822 1.613 317.601 0.047 Conative 0.014 0.107 Commitment 0.011 0.276 Trust 0.000 0.140 Behavioural 0.001 Purchase Loyalty Index (PLI) 0.386 0.149 1.826 12.054 0.185 Conative 0.000 From the above table, it can be inferred that all the three loyalty indices have significant relationship with SERVLOYAL dimensions. The Regression Equations are given below: The abbreviations are Service Loyalty Dimensions COL Commitment Loyalty TRL Trust Loyalty ATL Attitudinal Loyalty 135 ALI =0.135*COL +0.150*TRL +0.528*ATL +0.233*CGL +0.145*AFL DLI =0.894*COL +0.136*BHL +0.047*CNL PLI =0.107*COL +0.276*TRL +0.140*BHL +0.185*CNL BHL Behavioural Loyalty CGL Cognitive Loyalty AFL Affective Loyalty CNL Conative Loyalty Loyalty I ndices ALI Advocacy Loyalty Index DLI Defection Loyalty Index PLI Purchase Loyalty Index Thus from the R values, R 2 values, Durbin Watson values, it can be inferred that H 06 is rejected. 5.7 PATH MODEL 5.7.1 Relationship between Service Quality and Service Loyalty The path model to examine and confirm the relationship between Service Quality and Service Loyalty, Service Quality and Loyalty Indices and Service Loyalty and Loyalty Indices are shown in figure 5.1 through 5.3. 136 Figure 5.4 Path Diagram of Service Quality on Service Loyalty Chi-square value for the overall model fit was 93.366 for 68 degrees of freedom (p<0.001). Fit indices for the above model were NFI =0.903; NNFI = 0.907; CFI =0.917; IFI =0.923; GFI =0.977; AGFI =0.978; RMSR =0.032; RMSEA =0.030. 137 Thus, overall the model can be considered to have a high level of fit, as most of the fit indices show a good fit for the model by Hu and Bentler (1999) and (1 <2/df <5) by Wheaton et al. (1977) Moreover, due to the dependency of the 2 statistic on the sample size, a higher than cut-off value of comparative fit index (CFI) and a value of the 2/df ratio between 1 and 5 indicate a good fit. Goodness of fit summary for method =Maximum Likelihood Chi-square =93.366 based on 68 degrees of freedom Probability value for the chi-square statistic is .00000 Fit indices Bentler-bonett normed fit index = .903 Bentler-bonett non-normed fit index = .907 Comparative fit index (CFI) = .917 Bollen (IFI) fit index = .923 AMOS GFI fit index = .977 AMOS AGFI fit index = .978 Root mean-square residual (RMR) = .032 Root mean-square error of approximation (RMSEA) = .030 138 5.7.2 Relationship between Service Quality and Loyalty I ndices Figure 5.5 Path Diagram of Service Quality on Loyalty I ndices Chi-square value for the overall model fit was 253.868 for 69 degrees of freedom (p<0.001). Fit indices for the above model were NFI =0.918; NNFI = 0.920; CFI =0.938; IFI =0.942; GFI =0.948; AGFI =0.956; RMSR =0.046; RMSEA =0.047. 139 Thus, overall the model can be considered to have a high level of fit, as most of the fit indices show a good fit for the model by Hu and Bentler (1999) and (1 <2/df <5) by Wheaton et al. (1977) Moreover, due to the dependency of the 2 statistic on the sample size, a higher than cut-off value of comparative fit index (CFI) and a value of the 2/df ratio between 1 and 5 indicate a good fit. Goodness of fit summary for method =Maximum Likelihood Chi-square =253.868 based on 69 degrees of freedom Probability value for the chi-square statistic is .00000 Fit indices Bentler-bonett normed fit index = .918 Bentler-bonett non-normed fit index = .920 Comparative fit index (CFI) = .938 Bollen (IFI) fit index = .942 AMOS GFI fit index = .948 AMOS AGFI fit index = .956 Root mean-square residual (RMR) = .046 Root mean-square error of approximation (RMSEA) = .048 140 5.7.3 Relationship between Service Loyalty and Loyalty I ndices Figure 5.6 Path Diagram of Service Loyalty on Loyalty I ndices Chi-square value for the overall model fit was 283.623 for 74 degrees of freedom (p<0.001). Fit indices for the above model were NFI =0.903; NNFI = 0.9; CFI =0.921; IFI =0.924; GFI =0.948; AGFI =0.916; RMSR =0.030; RMSEA =0.042. 141 Thus, overall the model can be considered to have a high level of fit, as most of the fit indices show a good fit for the model by Hu and Bentler (1999) and (1 <2/df <5) by Wheaton et al. (1977) Moreover, due to the dependency of the 2 statistic on the sample size, a higher than cut-off value of comparative fit index (CFI) and a value of the 2/df ratio between 1 and 5 indicate a good fit. Goodness of fit summary for method =Maximum Likelihood Chi-square =283.623 based on 74 degrees of freedom Probability value for the chi-square statistic is .00000 Fit indices Bentler-bonett normed fit index = .903 Bentler-bonett non-normed fit index = .900 Comparative fit index (CFI) = .921 Bollen (IFI) fit index = .924 AMOS GFI fit index = .948 AMOS AGFI fit index = .916 Root mean-square residual (RMR) = .030 Root mean-square error of approximation (RMSEA) = .042 5.8 ASSESSI NG THE EFFECTI VENESS OF CRM PRACTI CES USI NG DATA ENVELOPMENT ANALYSI S Data Envelopment Analysis (DEA), occasionally called Frontier Analysis, was first put forward by Charnes, Cooper and Rhodes in 1978. It is a performance measurement technique which, can be used for evaluating the relative efficiency of decision-making units (DMU's) in organisations. DEA is concerned with a number of alternative decision making units (DMU). Each of them is analyzed separately via a mathematical programming model which checks whether 142 the DMU under consideration could improve its performance by decreasing its input and increasing its output. The improvement is pursued until the boundary of the convex hull of the other DMUs is reached. A DMU which cannot improve its performance is efficient or non-dominated. Otherwise, it is dominated by a convex combination of other DMUs. Thus, possible improvements for a particular DMU are indicated, not in an arbitrary direction, but on the basis of the performance of the more successful and efficient DMUs. Thus, in the case of CRM practices by the Mobile Service Providers, it is essential to assess the performance of CRM in the light of three aspects: (1) relationship between Service Quality dimensions and the Service Loyalty dimensions; (2) relationship between Service Quality and Loyalty Indices and (3) relationship between Service Loyalty and Loyalty Indices. To determine the input and output parameters for each of the above, linear regression is performed on service quality, service loyalty and loyalty indices constructs. The figure below gives the details of the service providers used in the study. Table 5.30 Service Provider Details No Code Mobile Service Provider 1 SP 1 Aircel 2 SP 2 Airtel 3 SP 3 BSNL 4 SP 4 Idea 5 SP 5 Reliance 6 SP 6 Vodafone 7 SP 7 Others 5.8.1 Effectiveness of Service Quality on Service Loyalty The predictors from the regression analysis are used as input parameters and each of the service loyalty dimensions are used as output parameters for the study. Then the data pertaining to the service quality parameters are taken as input 143 for each of the Service Providers. Data pertaining to 7 Service providers are used for the study. 5.8.1.1 Behavioural Loyalty The input parameters are responsiveness and empathy. The output parameter is behavioural loyalty. These input and output parameters are derived from the regression analysis. Table 5.31 I nput and Output Parameters Behavioural Loyalty Mobile Service Provider Responsiveness Empathy Behavioural SP1 3.4135 3.1923 3.1955 SP2 3.4924 3.2904 3.3005 SP3 3.3544 3.1978 3.1291 SP4 3.4781 3.3465 3.2851 SP5 3.6417 3.4583 3.3875 SP6 3.589 3.25 3.3093 SP7 3.6413 3.5326 3.3587 For example, for the Service Provider SP3, Responsiveness averages to 3.3544 and Empathy averages to 3.1978 and the output parameter Behavioural Loyalty averages to 3.1291. To compare these service providers and measure their performance a commonly used method is ratios which takes output measure and divides it by the corresponding input measure. In this case, we analyze the effectiveness of service providers by taking inputs and converting them (with varying degrees of efficiency) into outputs. Hence there are two ratios, i.e., Behavioural Loyalty with respect to Responsiveness (Eff1) and Behavioural Loyalty with respect to Empathy (Eff2). 144 Table 5.32 Ratios for Efficiency Mobile Service Provider Responsiveness Empathy Network Qlty Behavioural Eff1 % Eff2 % SP1 3.4135 3.1923 3.5282 3.1955 93.6 100 SP2 3.4924 3.2904 3.5697 3.3005 94.5 100 SP3 3.3544 3.1978 3.5319 3.1291 93.2 97 SP4 3.4781 3.3465 3.4947 3.2851 94.4 98 SP5 3.6417 3.4583 3.4967 3.3875 93 97 SP6 3.589 3.25 3.6542 3.3093 92.2 100 SP7 3.6413 3.5326 3.487 3.3587 92.2 95 The table 5.32 depicts that SP2 and SP4 has the highest value for Eff1 when the behavioural loyalty is measured with respect to responsiveness of the service provider. But SP1, SP2 and SP6 have highest percentage for Eff2 when behavioural loyalty is measured with respect to empathy. The other service providers do not compare so well with these, so are presumably performing less well. That is, they are relatively less efficient at using their given input resources (responsiveness and empathy) to produce output (Behavioural Loyalty). One problem with comparison using ratios is that different ratios can give a different picture and it is difficult to combine the entire set of ratios into a single numeric judgement. For instance consider SP1 and SP5; SP1 is (93.6 / 93) = 1.006 times only as efficient as SP5 at the Eff1 but (100 / 97) =1.03 times as efficient at the Eff2. To combine these figures into a single judgement is very difficult. This problem of different ratios giving different pictures would be especially true if there is an increase in the number of service providers (and/or increase the number of input/output measures). Thus it is very difficult to interpret the values corresponding to service provider SP4 from these ratios. This is where it is essential to get into a better 145 technique called Data Envelopment Analysis which interprets the ratios and provides the efficient frontier. In words DEA, in evaluating any number of DMU's, with any number of inputs and outputs: requires the inputs and outputs for each DMU to be specified; defines efficiency for each DMU as a weighted sum of outputs [total output] divided by a weighted sum of inputs [total input]; where all efficiencies are restricted to lie between zero and one (i.e. between 0% and 100%). in calculating the numerical value for the efficiency of a particular DMU weights are chosen so as maximise its efficiency, thereby presenting the DMU in the best possible light. For those who are comfortable with mathematics, the mathematics for the simple set of mobile service providers example is presented below. To calculate the efficiency of SP 4 (for instance): Maximise E SP4 Subject to E SP1 =(3.1955W behloyl )/(3.4135W responsiveness +3.1923W empathy ) E SP 2 =(3.3005W behloyl )/(3.4924W responsiveness +3.2904W empathy ) E SP 3 =(3.1291W behloyl )/(3.3544W responsiveness +3.1978W empathy ) E SP 4 =(3.2851W behloyl )/(3.4781W responsiveness +3.3465W empathy ) E SP5 =(3.3875W behloyl )/(3.6417W responsiveness +3.4583W empathy ) E SP 6 =(3.3093W behloyl )/(3.589W responsiveness +3.25W empathy ) E SP7 =(3.3587W behloyl )/(3.6413W responsiveness +3.5326W empathy ) 0 <=E SP 1 <=1 0 <=E SP 2 <=1 146 0 <=E SP 3 <=1 0 <=E SP 4 <=1 0 <=E SP 5 <=1 0 <=E SP 6 <=1 0 <=E SP 7 <=1 W behloyl >=0 W responsiveness >=0 W empathy >=0 where: E SP1 to E SP7 is the efficiency of SP 1 to SP 7 respectively (expressed as a fraction) W behloy is the weight attached to output parameter Behavioural Loyalty W responsiveness is the weight attached to input parameter Responsiveness W empathy is the weight attached input parameter Empathy To calculate the efficiency of the other service providers the only step is to change what is to be maximised (e.g. maximise E SP4 to calculate the efficiency of SP4). This is exactly the association of non-negative weights with each input and output measure. The above optimization problem is a nonlinear problem and hence difficult to solve numerically. In fact it can be converted into a linear programming problem. To do this we: algebraically substitute for all efficiency variables, to give an optimisation problem expressed purely in terms of weights introduce an additional constraint setting the denominator of the objective function equal to one (technically this can be done as the above nonlinear problem has one degree of freedom - multiplying all the weights by a (positive) scale factor would leave the solution value unchanged) 147 Thus for SP 4, the corresponding LPP is Maximise (3.2851W behloyl )/ (3.4781W responsiveness +3.3465W empathy) Subject to 3.4781W responsiveness +3.3465W empathy =1 0 <=(3.1955W behloyl )/ (3.4135W responsiveness +3.1923W empathy) <=1 0 <=(3.3005W behloyl )/ (3.4924W responsiveness +3.2904W empathy) <=1 0 <=(3.1291W behloyl )/ (3.3544W responsiveness +3.1978W empathy) <=1 0 <=(3.2851W behloyl )/ (3.4781W responsiveness +3.3465W empathy) <=1 0 <=(3.3875W behloyl )/ (3.6417W responsiveness +3.4583W empathy) <=1 0 <=(3.3093W behloyl )/ (3.589W responsiveness +3.25W empathy) <=1 0 <=(3.3587W behloyl )/ (3.6413W responsiveness +3.5326W empathy) <=1 W behloyl >=0 W responsiveness >=0 W empathy >=0 The above is a linear program and after rearrangement is as follows: Maximise 3.2851W behloyl Subject to 3.4781W responsiveness +3.3465W empathy =1 (3.1955W behloyl ) - (3.4135W responsiveness +3.1923W empathy) <=0 (3.3005W behloyl ) - (3.4924W responsiveness +3.2904W empathy) <=0 (3.1291W behloyl ) - (3.3544W responsiveness +3.1978W empathy) <=0 (3.2851W behloyl ) - (3.4781W responsiveness +3.3465W empathy) <=0 (3.3875W behloyl ) - (3.6417W responsiveness +3.4583W empathy) <=0 (3.3093W behloyl ) - (3.589W responsiveness +3.25W empathy) <=0 (3.3587W behloyl ) - (3.6413W responsiveness +3.5326W empathy) <=0 W behloyl >=0 W responsiveness >=0 W empathy >=0 148 where we have substituted 3.4781W responsiveness + 3.3465W empathy = 1 into the objective and rearranged the constraints. Once this LP has been solved to generate optimal values for the weights then the efficiency of the service providers being optimized optimizing for, SP4 in this case, can be easily calculated using E SP 4 = (3.2851W behloyl )/ (3.4781W responsiveness +3.3465W empathy) . Here that the numerator of E SP 4 =(3.2851W behloyl )/(3.4781W responsiveness +3.3465W empathy ) is known as the weighted output for SP4 and the denominator is known as the weighted input for SP4 It is important to note that DEA can only give relative efficiencies - efficiencies relative to the data considered. It does not, and cannot give absolute efficiencies. In words, DEA while evaluating any number of decision making units (DMU's), and with any number of inputs and outputs requires the inputs and outputs for each DMU to be specified. It defines efficiency for each DMU as a weighted sum of outputs [total output] divided by a weighted sum of inputs [total input]; where all efficiencies are restricted to lie between zero and one (i.e. between 0% and 100%). It uses the numerical value for calculating the efficiency of a particular DMU. Weights are chosen to maximise its efficiency, thereby presenting the DMU in the best possible light. Thus DEA provides the details of the best performing DMU which is usually 100% efficient. The other DMUs are not performing upto this efficient frontier. The Table 5.33 gives the DEA Solver Output for the Effect on Behavioural Loyalty. 149 Effectiveness of SERVQUAL on Behavioural Loyalty Table 5.33 Relative Efficiency on Behavioural Loyalty Mobile Service Provide r Responsiv eness Empath y Behavioura l Weighte d O/p Weighte d I /p Efficienc y Working SP1 3.413 3.192 3.195 0.995 1 0.995 -0.004 SP2 3.492 3.290 3.300 1.027 1.027 1.000 2.41E- 08 SP3 3.354 3.197 3.129 0.974 0.994 0.9798 -0.02008 SP4 3.478 3.346 3.285 1.023 1.037 0.986 -0.01419 SP5 3.641 3.458 3.387 1.054 1.077 0.9794 -0.02217 SP6 3.589 3.25 3.309 1.030 1.030 1.000 0 SP7 3.6413 3.532 3.358 1.045 1.091 0.958 -0.04562 Weight 0.110 0.195 0.311 From the analysis using DEA, it can be seen that SP2 and SP6 are the two mobile service providers who are on the efficient frontier as the efficiency is 1.00. The other service providers SP1, SP3, SP4, SP5 and SP7 are less performing. For instance, SP7 is only 95% efficient when compared to SP2 and SP6. That means, given the current scenario, SP7 needs to improve by 5% to reach the efficient provider with respect to Behavioural Loyalty. 5.8.1.2 Attitudinal Loyalty The input parameters are Reliability, Empathy, Convenience and Customer Perceived Network Quality and the output parameter is Attitudinal Loyalty. 150 Table 5.34 I nput and Output Parameters Attitudinal Loyalty Mobile Service Provider Reliability Empathy Convenie nce N/w Quality Attitudinal SP1 3.4718 3.1923 3.359 3.5282 3.1282 SP2 3.6768 3.2904 3.4747 3.5697 3.197 SP3 3.4615 3.1978 3.4066 3.5319 3.1401 SP4 3.4035 3.3465 3.5175 3.4947 3.2018 SP5 3.5367 3.4583 3.4917 3.4967 3.15 SP6 3.4915 3.25 3.4534 3.6542 3.1186 SP7 3.6522 3.5326 3.5163 3.487 3.1413 The attitudinal loyalty depends on the aspects of reliability, empathy, convenience and good network quality by the service providers. The solver output is given below. Effectiveness of SERVQUAL on Behavioural Loyalty Table 5.35 Relative Efficiency on Attitudinal Loyalty Mobile Service Provider Reliabil ity Empa thy Convenie nce N/w Quality Attitudi nal Weighted O/p Weight ed I/p Efficie ncy Worki ng SP1 3.47 3.19 3.35 3.52 3.12 1.00 1.00 1.00 3E-08 SP2 3.67 3.29 3.47 3.56 3.19 0.97 1.01 0.96 -3E-02 SP3 3.46 3.19 3.40 3.53 3.14 0.99 0.99 1.00 -7E-08 SP4 3.40 3.34 3.51 3.49 3.20 0.97 0.97 0.99 -2E-03 SP5 3.53 3.45 3.49 3.49 3.15 0.99 0.99 1.00 0E+00 SP6 3.49 3.25 3.45 3.65 3.11 1.00 1.03 0.97 -3E-02 SP7 3.65 3.53 3.51 3.48 3.14 0.99 0.99 1.00 0E+00 Weight 0.03 0.05 -0.10 0.29 3.12 From the analysis using DEA, it can be seen that SP1, SP3, SP5 and SP7 are the mobile service providers who are on the efficient frontier as the efficiency is 1.00. The other service providers SP2, SP4 and SP6 are less 151 performing. For instance, SP2 is only 96% efficient when compared to SP1, SP3, SP5 and SP7. That means, given the current scenario, SP2 needs to improve by 4% to reach the efficient provider with respect to Attitudinal Loyalty. 5.8.1.3 Cognitive Loyalty The input parameters are Responsiveness, Convenience, Empathy and Tangibles and the output parameter is Cognitive Loyalty. Table 5.36 I nput and Output Parameters Cognitive Loyalty Mobile Service Provider Responsiveness Convenience Empathy Tangibles Cognitive SP1 3.4135 3.359 3.1923 3.4103 3.0208 SP2 3.4924 3.4747 3.2904 3.5421 3.3438 SP3 3.3544 3.4066 3.1978 3.3816 3.0625 SP4 3.4781 3.5175 3.3465 3.3392 3.0278 SP5 3.6417 3.4917 3.4583 3.5167 3.5 SP6 3.589 3.4534 3.25 3.5345 4 SP7 3.6413 3.5163 3.5326 3.4928 3.0815 The cognitive loyalty depends on the aspects of responsiveness, convenience, empathy and tangibles by the service providers. The solver output which evaluates the effectiveness is given below: 152 Effectiveness of SERVQUAL on Cognitive Loyalty Table 5.37 Relative Efficiency on Cognitive Loyalty Mobile Service Provider Respons iveness Conve nience Empat hy Tangi bles Cognit ive Weighted O/p Weighted I/p Effici ency Working SP1 3.413 3.359 3.192 3.4103 3.0208 1.000 1.000 1.000 -1E-07 SP2 3.492 3.474 3.290 3.5421 3.3438 0.903 1.136 0.795 -2E-01 SP3 3.354 3.406 3.197 3.3816 3.0625 0.986 1.141 0.865 -2E-01 SP4 3.478 3.517 3.346 3.3392 3.0278 0.998 0.998 1.000 -2E-07 SP5 3.641 3.491 3.458 3.5167 3.5 0.863 0.925 0.933 -6E-02 SP6 3.58 3.453 3.25 3.5345 4 0.755 0.830 0.910 -8E-02 SP7 3.641 3.516 3.532 3.4928 3.0815 0.980 0.990 0.990 -1E-02 Weight -2.493 0.495 1.104 1.26 3.020 From the analysis using DEA, it can be seen that SP1 and SP4 are the two mobile service providers who are on the efficient frontier as the efficiency is 1.00. The other service providers SP2, SP3, SP5, SP6 and SP7 are less performing. For instance, SP3 is only 86% efficient when compared to SP1 and SP4. That means, given the current scenario, SP3 needs to improve by 14% to reach the efficient provider with respect to Cognitive Loyalty. 5.8.1.4 Conative Loyalty The input parameters are Responsiveness, Empathy, Tangibles and customer perceived network and the output parameter is Conative Loyalty Table 5.38 I nput and Output Parameters Conative Loyalty Mobile Service Provider Responsiveness Empathy Tangibles N/w Quality Conative SP1 3.4135 3.1923 3.41 3.5282 3.2308 SP2 3.4924 3.2904 3.54 3.5697 3.3182 SP3 3.3544 3.1978 3.38 3.5319 3.1319 SP4 3.4781 3.3465 3.34 3.4947 3.2632 SP5 3.6417 3.4583 3.52 3.4967 3.275 SP6 3.589 3.25 3.53 3.6542 3.339 SP7 3.6413 3.5326 3.49 3.487 3.087 153 The conative loyalty depends on the aspects of responsiveness, empathy, tangibles and good network quality by the service providers. To evaluate the effectiveness, refer to the solver output below: Effectiveness of SERVQUAL on Conative Loyalty Table 5.39 Relative Efficiency on Conative Loyalty Mobile Service Provider Respon siveness Emp athy Tangi bles N/w Qualit y Cona tive Weighte d O/p Weighte d I/p Efficiency Working SP1 3.41 3.19 3.41 3.53 3.23 1.000 1.000 1.000 -3E-08 SP2 3.49 3.29 3.54 3.57 3.32 0.974 1.064 0.915 -9E-02 SP3 3.35 3.20 3.38 3.53 3.13 1.032 1.032 1.000 -1E-04 SP4 3.48 3.35 3.34 3.49 3.26 0.990 0.990 1.000 -7E-08 SP5 3.64 3.46 3.52 3.50 3.28 0.987 1.022 0.966 -4E-02 SP6 3.59 3.25 3.53 3.65 3.34 0.968 0.968 1.000 -9E-05 SP7 3.64 3.53 3.49 3.49 3.09 1.047 1.047 1.000 -3E-05 Weight -0.76 0.51 0.55 0.02 3.23 From the analysis using DEA, it can be seen that SP1, SP3, SP4, SP6 and SP7 are the mobile service providers who are on the efficient frontier as the efficiency is 1.00. The other service providers SP2 and SP5 are less performing. For instance, SP2 is only 91% efficient when compared to SP1, SP3, SP4, SP6 and SP7. That means, given the current scenario, SP2 needs to improve by 9% to reach the efficient provider with respect to Conative Loyalty. 5.8.1.5 Affective Loyalty The input parameters are Reliability and Empathy and the output parameter is Affective Loyalty 154 Table 5.40 I nput and Output Parameters Affective Loyalty Mobile Service Provider Reliability Empathy Affective SP1 3.4718 3.1923 3.234 SP2 3.6768 3.2904 3.6111 SP3 3.4615 3.1978 3.4313 SP4 3.4035 3.3465 3.5132 SP5 3.5367 3.4583 3.5458 SP6 3.4915 3.25 3.4322 SP7 3.6522 3.5326 3.2283 The affective loyalty depends on the aspects of reliability and empathy by the service providers. To evaluate the effectiveness, refer to the solver output below: Effectiveness of SERVQUAL on Affective Loyalty Table 5.41 Relative Efficiency on Affective Loyalty Mobile Service Provider Reliability Empathy Affective Weighted O/p Weighted I/p Efficiency Working SP1 3.47 3.19 3.23 0.94 1.00 0.935 -0.06 SP2 3.68 3.29 3.61 1.04 1.04 1.000 0.00 SP3 3.46 3.20 3.43 0.99 1.00 0.993 -0.01 SP4 3.40 3.35 3.51 1.02 1.02 1.000 0.00 SP5 3.54 3.46 3.55 1.03 1.05 0.974 -0.03 SP6 3.49 3.25 3.43 0.99 1.01 0.981 -0.02 SP7 3.65 3.53 3.23 0.93 1.08 0.864 -0.15 Weight 0.14 0.16 0.29 From the analysis using DEA, it can be seen that SP2 and SP4 are the mobile service providers who are on the efficient frontier as the efficiency is 1.00. The other service providers SP1, SP3, SP5, SP6 and SP7 are less performing. For instance, SP5 is only 97% efficient when compared to SP2 and SP4. That means, 155 given the current scenario, SP4 needs to improve by 3% to reach the efficient provider with respect to Affective Loyalty. 5.8.1.6 Trust Loyalty The input parameters are Reliability, Responsiveness, Assurance and Empathy and the output parameter is Trust Loyalty Table 5.42 I nput and Output Parameters Trust Loyalty Mobile Service Provider Reliability Responsiveness Assurance Empathy Trust SP1 3.47 3.41 3.56 3.19 3.54 SP2 3.68 3.49 3.74 3.29 3.60 SP3 3.46 3.35 3.55 3.20 3.40 SP4 3.40 3.48 3.75 3.35 3.42 SP5 3.54 3.64 3.84 3.46 3.45 SP6 3.49 3.59 3.66 3.25 3.24 SP7 3.65 3.64 3.73 3.53 3.37 Trust loyalty depends on the aspects of reliability, responsiveness, assurance and empathy by the service providers. To evaluate the effectiveness, refer to the solver output below: Effectiveness of SERVQUAL on Trust Loyalty Table 5.43 Relative Efficiency on Trust Loyalty Mobile Service Provider Relia bility Responsi veness Assura nce Emp athy Trust Weighted O/p Weighted I/p Efficiency Working SP1 3.47 3.41 3.56 3.19 3.54 1.000 1.000 1.000 -6E-08 SP2 3.68 3.49 3.74 3.29 3.60 1.018 1.133 0.898 -1E-01 SP3 3.46 3.35 3.55 3.20 3.40 0.960 1.021 0.941 -6E-02 SP4 3.40 3.48 3.75 3.35 3.42 0.966 1.138 0.848 -2E-01 SP5 3.54 3.64 3.84 3.46 3.45 0.975 1.101 0.886 -1E-01 SP6 3.49 3.59 3.66 3.25 3.24 0.915 0.974 0.939 -6E-02 SP7 3.65 3.64 3.73 3.53 3.37 0.952 0.952 1.000 7E-08 Weight 0.00 -0.73 1.21 -0.25 0.28 156 From the analysis using DEA, it can be seen that SP1 and SP7 are the mobile service providers who are on the efficient frontier as the efficiency is 1.00. The other service providers SP2, SP3, SP4, SP5 and SP6 are less performing. For instance, SP4 is only 84% efficient when compared to SP1 and SP7. That means, given the current scenario, SP4 needs to improve by 16% to reach the efficient provider with respect to Trust Loyalty. 5.8.1.7 Commitment Loyalty The input parameters are Reliability, Responsiveness, Network Quality, Empathy and Assurance and the output parameter is Commitment Loyalty Table 5.44 I nput and Output Parameters Commitment Loyalty Mobile Service Provider Reliability Responsiveness Empathy Assurance N/w Quality Commitment SP1 3.47 3.41 3.19 3.56 3.53 3.32 SP2 3.68 3.49 3.29 3.74 3.57 3.42 SP3 3.46 3.35 3.20 3.55 3.53 3.17 SP4 3.40 3.48 3.35 3.75 3.49 3.29 SP5 3.54 3.64 3.46 3.84 3.50 3.39 SP6 3.49 3.59 3.25 3.66 3.65 3.31 SP7 3.65 3.64 3.53 3.73 3.49 3.36 Commitment loyalty depends on the aspects of reliability, responsiveness, empathy, convenience and customer perceived network quality by the service providers. To evaluate the effectiveness, refer to the solver output below: 157 Effectiveness of SERVQUAL on Commitment Loyalty Table 5.45 Relative Efficiency on Commitment Loyalty Mobile Service Provider Relia bility Respon siveness Emp athy Assur ance N/w Quality Commit ment Weighted O/p Weighted I/p Effici ency Worki ng SP1 3.47 3.41 3.19 3.56 3.53 3.32 1.00 1.000 1.00 1E-08 SP2 3.68 3.49 3.29 3.74 3.57 3.42 1.03 1.030 1.00 0E+00 SP3 3.46 3.35 3.20 3.55 3.53 3.17 0.95 1.026 0.93 -7E-02 SP4 3.40 3.48 3.35 3.75 3.49 3.29 0.99 1.047 0.94 -6E-02 SP5 3.54 3.64 3.46 3.84 3.50 3.39 1.02 1.019 1.00 -2E-11 SP6 3.49 3.59 3.25 3.66 3.65 3.31 1.00 0.996 1.00 0E+00 SP7 3.65 3.64 3.53 3.73 3.49 3.36 1.01 1.011 1.00 0E+00 Weight -0.05 -0.44 0.29 0.20 0.30 0.30 0.090 From the analysis using DEA, it can be seen that SP1, SP2, SP5, SP6 and SP7 are the mobile service providers who are on the efficient frontier as the efficiency is 1.00. The other service providers SP3 and SP4 are less performing. For instance, SP3 is only 93% efficient when compared to SP1, SP2, SP5, SP6 and SP7. That means, given the current scenario, SP3 needs to improve by 7% to reach the efficient provider with respect to Commitment Loyalty. 5.8.2 Effectiveness of Service Quality on Loyalty I ndices The predictors from the regression analysis are used as input parameters and each of the loyalty indices are used as output parameters for the study. The data pertaining to service quality parameters are taken as input for each of the Service Providers. Data pertaining to 7 Service providers are used for the study. 5.8.2.1 Advocacy Loyalty I ndex (ALI ) The input parameters are Reliability, Responsiveness, Empathy, Assurance and Network Quality and the output parameter is Advocacy Loyalty 158 Table 5.46 I nput and Output Parameters Advocacy Loyalty I ndex Mobile Service Provider Reliability Responsiveness Empathy Assurance Network Qlty Advocacy SP1 3.4718 3.4135 3.1923 3.4103 3.5282 3.4455 SP2 3.6768 3.4924 3.2904 3.5421 3.5697 3.6742 SP3 3.4615 3.3544 3.1978 3.3816 3.5319 3.5165 SP4 3.4035 3.4781 3.3465 3.3392 3.4947 3.5526 SP5 3.5367 3.6417 3.4583 3.5167 3.4967 3.5756 SP6 3.4915 3.589 3.25 3.5345 3.6542 3.7246 SP7 3.6522 3.6413 3.5326 3.4928 3.487 3.4565 Advocacy Loyalty Index (ALI) depends on the aspects of reliability, responsiveness, empathy, assurance and customer perceived network quality by the service providers. To evaluate the effectiveness, refer to the solver output below: Effectiveness of SERVQUAL on Purchase Loyalty Index Table 5.47 Relative Efficiency on Advocacy Loyalty Mobile Service Provide r Relia bility Responsi veness Emp athy Assur ance Network Qlty Advoc acy Weighte d O/p Weighte d I/p Effic ienc y Wor king SP1 3.47 3.41 3.19 3.41 3.53 3.45 0.97 1.00 0.97 -0.03 SP2 3.68 3.49 3.29 3.54 3.57 3.67 1.03 1.03 1.00 0.00 SP3 3.46 3.35 3.20 3.38 3.53 3.52 0.99 0.99 1.00 0.00 SP4 3.40 3.48 3.35 3.34 3.49 3.55 1.00 1.00 1.00 0.00 SP5 3.54 3.64 3.46 3.52 3.50 3.58 1.01 1.01 0.99 -0.01 SP6 3.49 3.59 3.25 3.53 3.65 3.72 1.05 1.05 1.00 0.00 SP7 3.65 3.64 3.53 3.49 3.49 3.46 0.97 0.97 1.00 0.00 Weight -2.33 -2.03 1.77 4.45 -1.37 0.29 0.97 1.00 0.97 -0.03 From the analysis using DEA, it can be seen that SP2, SP3, SP4 and SP6 and SP7 are the mobile service providers who are on the efficient frontier as the efficiency is 1.00. The other service providers SP5 are less performing. For instance, SP5 is only 99% efficient when compared to SP1, SP2, SP3, SP4, SP6 159 and SP7. That means, given the current scenario, SP5 needs to improve by only 1% to reach the efficient provider with respect to Advocacy Loyalty Index (ALI). 5.8.2.2 Purchase Loyalty I ndex (PLI ) The input parameters are Reliability, Responsiveness, Assurance, Empathy and Convenience and the output parameter is Purchase Loyalty Index. Table 5.48 I nput and Output Parameters Purchase Loyalty I ndex Mobile Service Provider Reliability Responsiveness Assurance Empathy Convenience Purchase SP1 3.47 3.41 3.56 3.19 3.36 3.41 SP2 3.68 3.49 3.74 3.29 3.47 3.54 SP3 3.46 3.35 3.55 3.20 3.41 3.42 SP4 3.40 3.48 3.75 3.35 3.52 3.46 SP5 3.54 3.64 3.84 3.46 3.49 3.63 SP6 3.49 3.59 3.66 3.25 3.45 3.61 SP7 3.65 3.64 3.73 3.53 3.52 3.70 Purchase Loyalty Index (PLI) depends on the aspects of reliability, responsiveness, assurance convenience and empathy by the service providers. To evaluate the effectiveness, refer to the solver output below: Effectiveness of SERVQUAL on Purchase Loyalty Index Table 5.49 Relative Efficiency on Purchase Loyalty Mobile Service Provider Reli abili ty Resp onsiv eness Assur ance Empa thy Conve nience Purch ase Weighte d O/p Weight ed I/p Efficiency Working SP1 3.47 3.41 3.56 3.19 3.36 3.41 0.984 1.000 0.984 -0.016 SP2 3.68 3.49 3.74 3.29 3.47 3.54 1.020 1.026 0.995 -0.005 SP3 3.46 3.35 3.55 3.20 3.41 3.42 0.988 0.988 1.000 0.000 SP4 3.40 3.48 3.75 3.35 3.52 3.46 1.000 1.007 0.993 -0.007 SP5 3.54 3.64 3.84 3.46 3.49 3.63 1.046 1.050 0.997 -0.003 SP6 3.49 3.59 3.66 3.25 3.45 3.61 1.042 1.043 0.999 -0.001 SP7 3.65 3.64 3.73 3.53 3.52 3.70 1.067 1.067 1.000 0.000 Weight 0.06 0.25 -0.08 0.01 0.05 0.29 160 From the analysis using DEA, it can be seen that SP3 and SP7 are the mobile service providers who are on the efficient frontier as the efficiency is 1.00. The other service providers SP1, SP2, SP4, SP5 and SP6 are less performing. For instance, SP1 is only 98% efficient when compared to SP3 and SP7. That means, given the current scenario, SP1 needs to improve by 2% to reach the efficient provider with respect to Purchase Loyalty Index (PLI). 5.8.2.3 Defection Loyalty I ndex (DLI ) The input parameters are Reliability, Responsiveness, Empathy and Customer perceived network quality and the output parameter is Defection Loyalty Index. Table 5.50 I nput and Output Parameters Defection Loyalty I ndex Mobile Service Provider Reliability Responsiveness Empathy Network Qlty Defection SP1 3.4718 3.4135 3.1923 3.5282 3.3333 SP2 3.6768 3.4924 3.2904 3.5697 3.4545 SP3 3.4615 3.3544 3.1978 3.5319 3.2418 SP4 3.4035 3.4781 3.3465 3.4947 3.2719 SP5 3.5367 3.6417 3.4583 3.4967 3.425 SP6 3.4915 3.589 3.25 3.6542 3.3136 SP7 3.6522 3.6413 3.5326 3.487 3.4348 Defection Loyalty Index (DLI) depends on the aspects of reliability, responsiveness, empathy and customer perceived network quality by the service providers. To evaluate the effectiveness, refer to the solver output below: 161 Effectiveness of SERVQUAL on Defection Loyalty I ndex Table 5.51 Relative Efficiency on Defection Loyalty Mobile Service Provider Reliability Res- ponsiveness Empathy Network Qlty Defection Weighted O/p Weighted I /p Efficiency Working SP1 3.47 3.41 3.19 3.53 3.33 0.89 1.00 0.89 -0.11 SP2 3.68 3.49 3.29 3.57 3.45 0.92 0.98 0.94 -0.06 SP3 3.46 3.35 3.20 3.53 3.24 0.86 1.13 0.77 -0.26 SP4 3.40 3.48 3.35 3.49 3.27 0.87 1.10 0.80 -0.22 SP5 3.54 3.64 3.46 3.50 3.43 0.91 0.91 1.00 0.00 SP6 3.49 3.59 3.25 3.65 3.31 0.88 0.88 1.00 0.00 SP7 3.65 3.64 3.53 3.49 3.43 0.91 0.98 0.94 -0.06 Weight -0.324 -1.903 1.509 1.078 0.266 From the analysis using DEA, it can be seen that SP5 and SP6 are the mobile service providers who are on the efficient frontier as the efficiency is 1.00. The other service providers SP1, SP2, SP3, SP4 and SP7 are less performing. For instance, SP3 is only 77 % efficient when compared to SP5 and SP6. That means, given the current scenario, SP6 needs to improve by 23 % to reach the efficient provider with respect to Defection Loyalty Index (DLI). 5.8.3 Effectiveness of Service Loyalty on Loyalty I ndices The predictors from the regression analysis are used as input parameters and each of the loyalty indices are used as output parameters for the study. The data pertaining to the service loyalty parameters are taken as input for each of the Service Providers. 5.8.3.1 Advocacy Loyalty I ndex (ALI ) The input parameters are Attitudinal Loyalty, Trust Loyalty, Affective Loyalty, Cognitive Loyalty and Commitment Loyalty and the output parameter is Advocacy Loyalty Index. 162 Table 5.52 I nput and Output Parameters Advocacy Loyalty I ndex Mobile Service Provider Attitudinal Trust Affective Cognitive Commitment Advocacy SP1 3.13 3.54 3.23 3.02 3.32 3.45 SP2 3.20 3.60 3.61 3.34 3.42 3.67 SP3 3.14 3.40 3.43 3.06 3.17 3.52 SP4 3.20 3.42 3.51 3.03 3.29 3.55 SP5 3.15 3.45 3.55 3.50 3.39 3.58 SP6 3.12 3.24 3.43 4.00 3.31 3.72 SP7 3.14 3.37 3.23 3.08 3.36 3.46 Advocacy Loyalty Index (ALI) depends on the aspects of service loyalty dimensions namely, Attitudinal Loyalty, Trust Loyalty, Affective Loyalty, Cognitive Loyalty and Commitment Loyalty. To evaluate the effectiveness, refer to the solver output below: Effectiveness of SERVLOYAL on Advocacy Loyalty I ndex Table 5.53 Relative Efficiency on Advocacy Loyalty Mobile Service Provider Attitudinal Trust Affective Cognitive Commitment Advocacy Weighted O/p Weighted I /p Efficiency Working SP1 3.13 3.54 3.23 3.02 3.32 3.45 0.99 1.00 0.99 -0.01 SP2 3.20 3.60 3.61 3.34 3.42 3.67 1.06 1.06 1.00 0.00 SP3 3.14 3.40 3.43 3.06 3.17 3.52 1.01 1.02 1.00 0.00 SP4 3.20 3.42 3.51 3.03 3.29 3.55 1.02 1.03 1.00 0.00 SP5 3.15 3.45 3.55 3.50 3.39 3.58 1.03 1.05 0.98 -0.02 SP6 3.12 3.24 3.43 4.00 3.31 3.72 1.07 1.07 1.00 0.00 SP7 3.14 3.37 3.23 3.08 3.36 3.46 1.00 1.00 1.00 0.00 Weight 0.18 0.06 0.06 0.08 -0.06 0.29 From the analysis using DEA, it can be seen that SP2, SP3, SP4, SP6 and SP7 are the mobile service providers who are on the efficient frontier as the efficiency is 1.00. The other service providers SP1 and SP5 are less performing. For instance, SP5 is only 98 % efficient when compared to SP2, SP3, SP4, SP6 and 163 SP7. That means, given the current scenario, SP5 needs to improve by 2 % to reach the efficient provider with respect to Advocacy Loyalty Index (ALI). 5.8.3.2 Purchase Loyalty I ndex (PLI ) The input parameters are Behavioural Loyalty, Conative Loyalty, Trust Loyalty and Commitment Loyalty and the output parameter is Purchase Loyalty Index. Table 5.54 I nput and Output Parameters Purchase Loyalty I ndex Mobile Service Provider Commitment Trust Behavioural Conative Purchase SP1 3.32 3.54 3.20 3.23 3.41 SP2 3.42 3.60 3.30 3.32 3.54 SP3 3.17 3.40 3.13 3.13 3.42 SP4 3.29 3.42 3.29 3.26 3.46 SP5 3.39 3.45 3.39 3.28 3.63 SP6 3.31 3.24 3.31 3.34 3.61 SP7 3.36 3.37 3.36 3.09 3.70 Purchase Loyalty Index (PLI) depends on the aspects of Behavioural Loyalty, Conative Loyalty, Trust Loyalty and Commitment Loyalty of the customers. To evaluate the effectiveness, refer to the solver output below. Effectiveness of SERVLOYAL on Purchase Loyalty Index Table 5.55 Relative Efficiency on Purchase Loyalty Mobile Service Provider Commitment Trust Behavioural Conative Purchase Weighted O/p Weighted I /p Efficiency Working SP1 3.32 3.54 3.20 3.23 3.41 0.81 1.00 0.81 -0.19 SP2 3.42 3.60 3.30 3.32 3.54 0.84 1.00 0.84 -0.16 SP3 3.17 3.40 3.13 3.13 3.42 0.81 1.00 0.81 -0.19 SP4 3.29 3.42 3.29 3.26 3.46 0.82 0.98 0.84 -0.16 SP5 3.39 3.45 3.39 3.28 3.63 0.86 0.95 0.91 -0.09 SP6 3.31 3.24 3.31 3.34 3.61 0.85 0.86 1.00 0.00 SP7 3.36 3.37 3.36 3.09 3.70 0.87 0.87 1.00 0.00 Weight -0.89 0.69 0.29 0.17 0.24 From the analysis using DEA, it can be seen that SP6 and SP7 are the mobile service providers who are on the efficient frontier as the efficiency is 1.00. For instance, SP1 is only 81 % efficient when compared to SP6 and SP7. That 164 means, given the current scenario, SP5 needs to improve by 19 % to reach the efficient provider with respect to Purchase Loyalty Index (ALI). 5.8.3.3 Defection Loyalty I ndex (DLI ) The input parameters are Behavioural Loyalty, Conative Loyalty and Commitment Loyalty and the output parameter is Defection Loyalty Index. Table 5.56 I nput and Output Parameters Defection Loyalty I ndex Mobile Service Provider Behavioural Conative Commitment Defection SP1 3.20 3.23 3.32 3.33 SP2 3.30 3.32 3.42 3.45 SP3 3.13 3.13 3.17 3.24 SP4 3.29 3.26 3.29 3.27 SP5 3.39 3.28 3.39 3.43 SP6 3.31 3.34 3.31 3.31 SP7 3.36 3.09 3.36 3.43 Defection Loyalty Index (DLI) depends on the aspects of Behavioural Loyalty, Conative Loyalty and Commitment Loyalty of the customers. To evaluate the effectiveness, refer to the solver output below: Effectiveness of SERVLOYAL on Defection Loyalty I ndex Table 5.57 Relative Efficiency on Defection Loyalty Mobile Service Provider Behavi Oural Conative Commitment Defection Weighted O/p Weighted I /p Efficiency Working SP1 3.20 3.23 3.32 3.33 0.98 1.00 0.98 -0.02 SP2 3.30 3.32 3.42 3.45 1.02 1.03 0.99 -0.01 SP3 3.13 3.13 3.17 3.24 0.95 0.95 1.00 0.00 SP4 3.29 3.26 3.29 3.27 0.96 0.99 0.98 -0.02 SP5 3.39 3.28 3.39 3.43 1.01 1.02 0.99 -0.01 SP6 3.31 3.34 3.31 3.31 0.97 0.99 0.98 -0.02 SP7 3.36 3.09 3.36 3.43 1.01 1.01 1.00 0.00 Weight -0.01 -0.01 0.32 0.29 165 From the analysis using DEA, it can be seen that SP3 and SP7 are the mobile service providers who are on the efficient frontier as the efficiency is 1.00. For instance, SP5 is only 99 % efficient when compared to SP3 and SP7. That means, given the current scenario, SP5 needs to improve by 1 % to reach the efficient provider with respect to Defection Loyalty Index (ALI). 5.8.4 Summary of Data Envelopment Analysis Table 5.58 Details of Best Performer on various Dimensions using Data Envelopment Analysis SERVI CE LOYALTY SERVI CE QUALI TY No. Dimensions Best Performers 1 Behavioural Loyalty SP2, SP6 2 Attitudinal Loyalty SP1, SP3, SP5,SP7 3 Cognitive Loyalty SP1, SP4 4 Conative Loyalty SP1, SP3, SP4, SP6, SP7 5 Affective Loyalty SP2, SP4 6 Trust Loyalty SP1, SP7 7 Commitment Loyalty SP1, SP2, SP5, SP6, SP7 LOYALTY I NDI CES SERVI CE QUALI TY No. Dimensions Best Performers 1 Advocacy Loyalty Index (ALI) SP2, SP3, SP4, SP6, SP7 2 Defection Loyalty Index (DLI) SP5, SP6 3 Purchase Loyalty Index (PLI) SP3, SP7 LOYALTY I NDI CES SERVI CE LOYALTY No. Dimensions Best Performers 1 Advocacy Loyalty Index (ALI) SP2, SP3, SP4,SP6,SP7 2 Defection Loyalty Index (DLI) SP3,SP7 3 Purchase Loyalty Index (PLI) SP6,SP7 5.9 SUMMARY This chapter discussed the data analysis of the research, namely profile of the respondents, preliminary data analysis and hypothesis testing. The result showed a response rate of 95.4%. Most of the respondents (about 60%) were pre- paid, while rest (40%) of the respondents accounted for post-paid services. Preliminary analyses of the data tested the data for internal consistency of the 166 constructs identified. Factor Analysis was performed to identify the dimensions of service quality, service loyalty ad loyalty indices. A confirmatory factor analysis was performed to check for the unidimensionality of the constructs. Linear regression was run to test the hypothesis. A Confirmatory path model was derived and tested for goodness of fit for each of the loyalty indices using Structural Equation Modeling. To assess the effectiveness of CRM implementation in the present scenario Data Envelopment Analysis was used to determine the efficient frontier on each of the dimensions. 167 CHAPTER 6 FI NDI NGS 6.1 FI NDI NGS FROM DESCRI PTI VE STATI STI CS About 20 % of the respondents use Airtel; 18.6% use BSNL and 15.9% use Aircel. The majority of respondents used Prepaid Mobile Services. The majority of respondents were in the age group of 41-50 years. The majority of respondents have been using the services of their mobile service providers for more than 3 years. More than 50% of respondents were spending less than Rs.1000 in a month towards mobile services. Majority of the respondents were Post Graduates and 31.8% were Under Graduates. 6.2 FINDI NGS FROM ANOVA 6.2.1 Demographics on Service Quality While analyzing the variance in service quality dimensions, it is observed that significant variance existed in monthly expenses. On investigating further we observe Age (more than 50 years), Education (UG), length of use (less than 2 years) and Monthly Expenses (between Rs.501-1000) determined Reliability dimension. 168 Monthly Expenses (more than Rs.2000) and length of use (more than 3 years) determined Responsiveness dimension. Monthly Expenses (more than Rs.2000) determined Assurance, Empathy and Customer Perceived Network Quality dimensions. Monthly Expenses (between Rs.501-1000) determined Convenience. 6.2.2 Demographics on Service Loyalty While analyzing the variance in service loyalty dimensions, it is observed that significant variance existed in monthly expenses. On investigating further we observe Age (21-30 years) and Monthly Expenses(more than Rs.2000) determined Commitment Loyalty Length of Use (less than 2 years), Monthly Expenses (less than Rs.500) and Education (UG) determined Trust Loyalty. Monthly Expenses (less than Rs.500) determined Attitudinal Loyalty Length of Use (2-3 years) determined Cognitive Loyalty Monthly Expenses (between Rs.1001 and 2000) determined Affective Loyalty. 6.2.3 Demographics on Loyalty I ndices While analyzing the variance in loyalty indices, it is observed that significant variance existed in monthly expenses. On investigating further we observe Monthly Expenses (more than Rs.2000) determined Advocacy Loyalty Index 169 Length of Use (more than 3 years) and Monthly Expenses (more than Rs.2000) determined Purchase Loyalty Index. Age (21-30 years) and Monthly Expenses (more than Rs.2000) determined Defection Loyalty Index 6.3 FINDI NGS FROM LI NEAR REGRESSI ON Linear regression was used to determine the relationship between demographic variables and Service Quality, Service Loyalty and Loyalty indices; relationship between Service Quality and Service Loyalty; Service Quality on Loyalty Indices and Service Loyalty on Loyalty Indices. 6.3.1 Relationship between Demographics and Service Quality While testing the relationship between Demographics and Service Quality, it is found that - Age, Monthly Expenses, Education and Occupation have significant relationship with Reliability. Length of Use, Monthly Expenses have significant relationship with Responsiveness. Monthly Expenses has significant relationship with Customer Perceived Network Quality. Monthly Expenses has significant relationship with Assurance Monthly Expenses has significant relationship with Convenience Monthly Expenses has significant relationship with Empathy Education has significant relationship with Tangibles. 170 6.3.2 Relationship between Demographics and Service Loyalty While testing the relationship between Demographics and Service Loyalty, it is found that - Monthly Expenses has significant relationship with Commitment Loyalty Monthly Expenses has significant relationship with Trust Loyalty 6.3.3 Relationship between Demographics and LoyaltyI ndices While testing the relationship between Demographics and Loyalty Indices, it is found that Monthly Expenses has significant relationship with Advocacy Loyalty Index Monthly Expenses and Occupation have significant relationship with Purchase Loyalty Index Monthly Expenses has significant relationship with Defection Loyalty Index. 6.3.4 Relationship between Service Quality and Service Loyalty In determining the relationship between Service Quality dimensions and Service Loyalty Dimensions, it is found that Reliability, Responsiveness, Customer Perceived Network Quality, Assurance and Empathy have significant relationship with Commitment Loyalty. Reliability, Responsiveness, Assurance and Empathy have significant relationship with Trust Loyalty 171 Reliability, Customer Perceived Network Quality, Convenience and Empathy have significant relationship with Attitudinal Loyalty Responsiveness and Empathy have significant relationship with Behavioural Loyalty Responsiveness, Convenience, Empathy and Tangibles have significant relationship with Cognitive Loyalty Reliability and Empathy have significant relationship with Affective Loyalty Responsiveness, Customer Perceived Network Quality, Empathy and Tangibles have significant relationship with Conative Loyalty 6.3.5 Relationship between Service Quality and Loyalty I ndices In determining the relationship between Service Quality dimensions and Customer Loyalty Indices, it is found that, Reliability, Responsiveness, Customer Perceived Network Quality, Assurance and Empathy have significant relationship with Advocacy Loyalty Index Reliability, Responsiveness, Customer Perceived Network Quality and Empathy have significant relationship with Defection Loyalty Index Reliability, Responsiveness, Assurance, Convenience and Empathy have significant relationship with Purchase Loyalty Index 6.3.6 Relationship between Service Loyalty and Loyalty I ndices In determining the relationship between Service Loyalty dimensions and Loyalty Indices, it is found that, 172 Commitment loyalty, Trust loyalty, Attitudinal loyalty, Cognitive loyalty and Affective loyalty have significant relationship with Advocacy Loyalty Index Commitment loyalty, Behavioural loyalty and Conative loyalty have significant relationship with Defection Loyalty Index Commitment Loyalty, Trust Loyalty, Behavioural Loyalty and Conative Loyalty have significant relationship with Purchase Loyalty Index 6.4 FINDI NGS FROM THE PATH MODEL The path models examined and confirmed the path to depict the relationships between the different dimensions, namely, between service quality dimensions and service loyalty dimensions; service quality dimensions and loyalty indices and service loyalty and loyalty indices. All the three path models were found to be statistically fit with the Fit Indices having values greater than 0.9 and the RMSEA Values less than 0.06 6.5 FI NDI NGS FROM DATA ENVELOPMENT ANALYSI S (DEA) Data Envelopment Analysis was used to examine the current level of performance of each service provider on each of the dimensions of service loyalty and loyalty indices. 6.5.1 Effectiveness of Service Quality on Service Loyalty While examining the performance of Service Providers on Service Loyalty with Service Quality as the determinants, it is observed that, Aircel is found to be performing its best on the following dimensions Attitudinal Loyalty, Cognitive Loyalty, Conative Loyalty, Trust Loyalty and Commitment Loyalty. 173 Airtel is found to be performing its best on the following dimensions Behavioural Loyalty, Affective Loyalty and Commitment Loyalty. BSNL is found to be performing its best on the following dimensions Attitudinal Loyalty, Conative Loyalty. Idea is found to be performing its best on the following dimensions Cognitive Loyalty, Conative Loyalty and Affective Loyalty. Reliance is found to be performing its best on the following dimensions Attitudinal Loyalty and Commitment Loyalty. Vodafone is found to be performing its best on the following dimensions Behavioural Loyalty, Conative Loyalty and Commitment Loyalty. Others are found to be performing their best on the following dimensions Attitudinal Loyalty, Conative Loyalty, Trust Loyalty and Commitment Loyalty. 6.5.2 Effectiveness of Service Quality on Loyalty I ndices While examining the performance of Service Providers on Loyalty Indices with Service Quality as the determinants, it is observed that, Airtel is found to be performing its best on the following dimensions Advocacy Loyalty Index. BSNL is found to be performing its best on the following dimensions Advocacy Loyalty Index and Purchase Loyalty Index. Idea is found to be performing its best on the following dimensions Advocacy Loyalty Index. Relaince is found to be performing its best on the following dimensions Defection Loyalty Index. 174 Vodafone is found to be performing its best on the following dimensions Advocacy Loyalty Index and Defection Loyalty Index.. Others are found to be performing their best on the following dimensions Advocacy Loyalty Index and Purchase Loyalty Index. 6.5.3 Effectiveness of Service Loyalty on Loyalty I ndices While examining the performance of Service Providers on Loyalty Indices with Service Quality as the determinants, it is observed that, Airtel is found to be performing its best on the following dimensions Advocacy Loyalty Index. BSNL is found to be performing its best on the following dimensions Advocacy Loyalty Index and Defection Loyalty Index. Idea is found to be performing its best on the following dimensions Advocacy Loyalty Index. Vodafone is found to be performing its best on the following dimensions Advocacy Loyalty Index and Purchase Loyalty Index. Others are found to be performing their best on the following dimensions Advocacy Loyalty Index, Purchase Loyalty Index and Defection Loyalty Index. 6.6 SUMMARY This chapter provided the findings in detail. The findings with respect to profile of the respondents, descriptive statistics, ANOVA and linear regression were presented. The findings on the path models as well as the DEA were also presented. The discussion on the findings and the implications on the research and the managerial aspects are presented in the next chapter. 175 CHAPTER 7 DI SCUSSI ON The objective of this chapter is to bring to light: i. Insights from Literature and Gaps ii. Attempts by this study to fill these gaps iii. Research Implications iv. Managerial Implications 7.1 I NSI GHTS FROM LI TERATURE AND GAPS The insights from literature review are as follows: An extended service quality scale was proposed by Seth et al (2008) by including technical quality aspects in addition to functional quality aspects. To examine the behavioural aspects which play a bigger role in understanding customer loyalty, a SERVLOYAL scale proposed by Sudhakar et al (2006) which has seven dimensions is used In order to assess the loyalty Loyalty Indices, Bob Hayes (2007) proposed loyalty indices which has three dimensions.. This would clearly link to the CRM concepts of Positive Word of Mouth, in Up Selling and Cross Selling and Customer Churn. 176 There are no combined studies to provide a holistic suggestion as to enhancing CRM by improving service quality and service loyalty. 7.2 ATTEMPTS BY THI S STUDY TO FI LL THE GAPS This study has attempted to further strengthen the current knowledge on determining the relationship among perceived service quality, service loyalty and loyalty indices. This study worked on the underlying premise that one area that remained relatively underdeveloped, is the relationship between evaluation of service quality and loyalty of service customers.(Glemer & Brown,1996). This is in spite of the fact that loyalty is essential for service business survival (Reichheld, 1993). This study proposed a conceptual framework to examine the relationships that exist between service quality, service loyalty and loyalty indices. Figure 7.1 depicts the conceptual framework proposed in this study. Figure 7.1 Conceptual Framework 177 From the conceptual framework, the study attempted to test the following relationships. Relationship between Demographic Variables and Service Quality Relationship between Demographic Variables and Service Loyalty Relationship between Demographic Variables and Loyalty Indices Relationship between Service Quality and Service Loyalty Relationship between Service Quality and Loyalty Indices Relationship between Service Loyalty and Loyalty Indices Figure 7.2 represents the relationship between demographic variables with service quality dimensions, service loyalty dimensions and loyalty indices based on data analysed. 178 Figure 7.2 Model from Observations Since service quality, service loyalty and loyalty indices are all multidimensional variables, it can be seen from the observed model that some of the demographic variables have relationship with some dimensions only. But Monthly Expenses is the demographic variable which has relationship with most of the dimensions. The confirmatory path depicted the relationship between service quality and service loyalty; service quality and loyalty indices and service loyalty and loyalty indices. The study establishes that there is a strong linkage between (a) service quality and service loyalty; (b) service quality and loyalty indices and (c) service 179 loyalty and loyalty indices. Although Zeithaml et al. (1996) report a strong association between overall service quality and service loyalty across multiple companies, the findings clearly highlights the quality-loyalty relationship with reference to the Indian Mobile segment. This underlines the importance of a multidimensional approach to service loyalty. The findings have a number of research and managerial implications. 7.3 RESEARCH I MPLI CATI ONS This study should be seen as a preliminary attempt at addressing an issue that has important implications for services marketing theory and practice. Any preliminary attempt will involve a number of limitations. However, acknowledgement of these limitations also suggests new directions for future studies. In the first place, conceptual models as well as scales for measuring service quality and loyalty have been developed and refined over the last decade. Differences in the nature of service setting might require additional dimensions of service quality (Dabholkar et al., 1996). With regards to the complaining behaviour part of the loyalty scale, the incidental nature of service problems may require incident-based measurement (such as the Critical Incident Technique) rather than service attitude-based measurement instruments. Further research should also incorporate multiple measures of the relevant constructs in order to increase the number of items that are used for the individual service loyalty dimensions. For instance, word-of- mouth should also be formulated in negative terms. Moreover, the study focused on service loyalty intentions only and these intentions are an incomplete proxy for actual behaviour (Keaveney, 1995). They should be supplemented by behavioural measures in order to develop a composite index of service loyalty (Dick and Basu, 1994). Therefore, further research should also take actual (re)actions of consumers to perceived service 180 quality into account. An area that definitely also merits further investigation is the impact of loyalty on other organisational performance measures, such as profitability (Storbacka et al., 1994). Furthermore, the use of multiple time frames allows for an investigation of the reinforcement effect of behavioural intentions on future service quality perceptions as well as other outcome variables that determine the strength of customer-organisation relationships such as commitment, trust and customer value. Finally, for the purpose of cross-validation, additional exploration of the service quality-loyalty relationship needs to be extended beyond the settings reported here to markets in which switching barriers are perceived to be high. 7.4 MANAGERI AL IMPLICATI ONS The findings have several managerial implications as well. The results enable managers of service firms to nuance the intuitive relationship between service quality, service loyalty and loyalty indices and have a richer diagnostic value because both service quality and loyalty are measured at a detailed and specific level. In addition, information on the service quality-customer loyalty link may provide actionable benchmarks that individual firms may use to guide their service policies aimed at securing customer loyalty. Furthermore, the results have specific indications for the different path models and service industries' research and budget allocations and personnel management decisions relating to the improvement of service loyalty on the basis of service quality. The study used Data Envelopment Analysis to determine the effectiveness of the mobile service providers on each of the service loyalty dimensions where the service quality dimensions are used as the predictors in regression analysis. Aircel is found to be 100% efficient on most of the service loyalty dimensions; BSNL is found to be efficient on loyalty indices. But the other service providers performed well on some of the dimensions. DEA gives relative 181 efficiencies. In this study, DEA was used to analyse the relative efficiencies which is used to benchmark the service providers on various dimensions. For the other service providers to improve their performance they need to analyze the input and output parameters of the best performing service providers. This study attempts to links many facets of business viz., CRM approach which is adopted by the telecom industry; comparison of its effectiveness using a quantitative approach like Data Envelopment Analysis. The output of this simulated model provides an insight to the industry practitioners an effective method of evaluating a technology which is being adopted by them. Though this study concentrates only on seven service providers in Chennai, it can further be extended with more parameters so that a holistic picture on the latest business trends which can be analysed and leveraged for further improvements. Indices based on loyalty may supplement measures of financial performance and market share with crucial information on the future health of a firm or industry. Finally, company- and industry-level assessment of the service quality- customer loyalty link provides useful information to shareholders on the viability of performance in the future. Especially, when tracked over time, changes in service loyalty signal changes in the value of customer assets. The identification of service loyalty as a multi-dimensional construct may help corporate decision makers in an accurate assessment of service loyalty. 7.5 CONCLUSI ON This study focuses on studying the relationship between service quality, service loyalty and loyalty indices. Though there are many studies on the service quality and loyalty, the behavioural aspects that influence loyalty had not been paid much attention with reference to the mobile service providers. With the current trend on cut throat competition and with low switching cost for the customers, the mobile service providers are facing challenges in keeping themselves steady in the competition. On the other hand, the high cost of customer acquisition is making todays businesses understand the importance of retaining the customers for long- 182 run sustainability. The study uses path models to depict the relationships between service quality, service loyalty dimensions and loyalty indices. Results of the study indicate that there exists a relationship between service quality and service loyalty; service quality and loyalty indices and service loyalty and loyalty indices. The results of DEA indicate the current set of best performers on each of the behavioural aspects. The study concludes by asserting the fact that if the behavioural aspects are studied and its influence on the loyalty indices are focused by the mobile service providers, it could improve their overall strength in the competition. 183 REFERENCES 1. kerlund, Helena. (2005), Fading Customer Relationships in Professional Services. Managing Service Quality, 15 (2), pp 56171. 2. Alzola, L M and Robaina, V P (2005), 'SERVQUAL: Its Applicability in Electronic Commerce B2C',Quality Management J ournal, 12(4), pp46-57. 3. Anders Gustafsson, Michael D. J ohnson, & I nger Roos,(2005)., 'The Effects of Customer Satisfaction, Relationship Commitment Dimensions, and Triggers on Customer Retention', J ournal of Marketing 69(October), pp210218 4. Anderson,J .G., and Gerbing, D.W. (1988), Structural Equation Modeling in Practice: A Review and Recommended Two-step Approach, Psychological Bulletin., 103(3), pp 411-423 5. Anita Seth, K Momaya, and H M Gupta (2008), 'Managing the Customer Perceived Service Quality for Cellular Mobile Telephony: An Empirical Investigation', Vikalpa, 33(1), pp20-36 6. Asubonteng, McCleary and E. Swan (1996), 'SERVQUAL Revisited: A Critical Review of Service Quality', J ournal of Services Marketing, 10(6), pp.62 81, 1996 7. Athreya MB (1996), 'India's Telecommunications Policy: A Paradigm Shift', Telecommunications Policy, 20(1), pp 11-22. 8. Bandyopahyay S., Martell M. (2007), Does Attitudinal Loyalty Influence Behavioral Loyalty? A Theoretical and Empirical Study, J ournal of Retailing and Consumer Services, 14, pp 35-44. 9. Baker M.J ., Buttery E.A., Richter-Buttery E.M.(1998), Relationship Marketing in Three Dimensions, J ournal of Interactive Marketing, 12(4), pp 47-62, 1998 10. Bart Larivie`re, Dirk Van den Poel (2005), 'Predicting Customer Retention and Profitability by using Random Forests and Regression Forests Techniques', Expert Systems with Applications29,pp 472484 11. Beerli A, J D Martin and A Quintana, (2002), A Model of Customer Loyalty in Retail Banking Market. European J ournal of Marketing, 38: 253-75. 184 12. Bentler, P M and Bonnett(1980), Significance Tests and Goodness of Fit in the Analysis of Covariance Structures, Psychological Bulletin, 88, pp.588-606 13. Bentler, P. M. (1990), Comparative Fit Indexes in Structural Models, Psychological Bulletin, 107, pp.238-246 14. Berry L.L, Parasuraman A & Zeithaml V.A. (1988), 'The Service Quality Puzzle', Business Horizon, 31(5), pp. 35- 43. 15. Berry L L, Zeithaml V A and Parasuraman A (1990), 'Five Imperatives for Improving Service Quality, Sloan Management Review, Summer, 31(4), pp. 29 - 38. 16. Berry L.L & Parasuraman A (1991), 'Marketing Services: Competing through Quality' (New York, Free Press). 17. Berry L.L. and Parasuraman A.(1992), 'Prescriptions for a Service Quality Revolution in America, Organisational Dynamics, Spring, pp. 5 - 15. 18. Berry , Leonard (1995) Relationship Marketing of Services- Growing Interest, Emerging Perspectives J ournal of the Academy of Marketing Science, 23 (4), pp. 236-245. 19. Bhattacharya, C B, and Ruth N Bolton (2000) Relationship Marketing in Mass Markets. In Handbook of Relationship Marketing, ed. Jadish N. Sheth and Atul Parvatiyar, pp. 327354 Thousand Oaks, CA: Sage. 20. Birgit Leisen, Charles Vance (2001), 'Cross-national Assessment of Service Quality in the Telecommunication Industry: Evidence from the USA and Germany', Managing Service Quality, 11(5), pp.307 317 21. Bitner M.J ., Booms B.H. and Tetreault, M.S. (1990), 'The Service Encounter: Diagnosing Favorable and Unfavorable Incidents, J ournal of Marketing, 54(J anuary), pp. 71-84. 22. Blakenship, A.B. and Breen,G.E(1993), State of the Art Marketing Research. Illinois: NTC Publishing Group. 23. Bloemer, J . M. M. and H. D. P. Kasper (1995), The Complex Relationship between Consumer Satisfaction and Brand Loyalty, J ournal of Economic Psychology, 16(2), pp. 311329. 24. Bloemer J , Ruyter K and Wetzels M. (1999), 'Linking Perceived Service Quality and Service Loyalty: A Multi-dimensional Perspective, European J ournal of Marketing, 33(11/12), pp. 1082-1106. 185 25. Bob E Hayes (2007), 'Measuring Customer Satisfaction and Loyalty: Survey Design, Use, and Statistical Analysis Methods', ASQ Quality Press,Third Edition. 26. Bollen KA, and Long J S,(1993), Testing Structural Equation Models, Newbury Park, CA:Sage. 27. Bollen KA. (1989), Structural Equations with Latent Variables, New York: Wiley. 28. Bordens , K.S., and Abbott, B.B. (1996), Research Design and Methods: A Process Approach, California USA, Mayfield Publishing Company, Third edition. 29. Bolton, Ruth N. (1998), A Dynamic Model of the Duration of the Customers Relationship with a Continuous Service Provider: The Role of Satisfaction, Marketing Science, 17 (1), 4565. 30. Bolton, Ruth N., and J ames H. Drew. (1991) A Longitudinal Analysis of the Impact of Service Changes on Customer Attitudes, J ournal of Marketing, 55 (1), pp 19. 31. Bowen J .T., Chen S-L. (2001), The Relationship between Customer Loyalty and Customer Satisfaction, International J ournal of Contemporary Hospitality Management 13 (5), pp 213-217 32. Bueren A., Schierholz R, Kolbe L. M, Brenner, W (2005), Improving Performance of Customer-Processes with Knowledge Management, Business Process Management J ournal, 11(5), pp. 573-588. 33. Brunner T. A. Stcklin, M. Opwis, K. (2008). Satisfaction, Image and Loyalty: New versus Experienced Customers. European J ournal of Marketing.(42) pp 1095-1105. 34. Burns, A. C., and R. F. Bush (2000), Marketing Research, New Jersey: Prentice Hall International Inc 35. Byrne, B.M. (2000), Structural Equation Modeling with AMOS: Basic Concepts, Applications and Programming, Mahwah, NJ: Lawrence Erlbaum. 36. Byrne, B. M. (2003), Structural Equation Modeling with AMOS. Basic Concepts, Applications, and Programming, New J ersey, Lawrence ErlbaumAssociates. 37. Cao, Yong, and Tomas S. Gruca. (2005) Reducing Adverse Selection through Customer Relationship Management. J ournal of Marketing, 69 (4) pp 219229. 186 38. Carman, J ames M, (1990), Consumer Perceptions of Service Quality: An Assessment of the SERVQUAL Dimensions, J ournal of Retailing, 66 (Spring), pp 33-55 39. Caruana, A., (2002). Service Loyalty: The Effects of Service Quality and the Mediating Role of Customer Satisfaction. European. J ournal of Marketing, 36: pp 811-828. 40. Caruana, A. (2004) The Impact of Switching Costs on Customer Loyalty: A Study among Corporate Customers of Mobile Telephone. J ournal of Targeting, Measurement & Analysis for Marketing. 12.pp 256-269. 41. Carvalho de, F A and Leite, V A (1999),'Attribute Importance in Service Quality: An Empirical Test of the PBZ Conjecture in Brazil', International J ournal of Service Industry Management, 10(5), pp 487-504. 42. Chaudhuri, A. (1995), Brand Equity or Double J eopardy? , J ournal of Product and Brand Management, 4, pp. 26-32. 43. Chaudhuri A and Holbrook M.B. (2001), The Chain of Effects from Brand Trust and Brand Affect to Brand Performance: The Role of Brand Loyalty. Journal of Marketing; 65 (April), pp81-93. 44. Charnes. D.A., Cooper, W.W. & Rhodes, E. (1978). Measuring the Efficiency of Decision Making Units. European J ournal of Operations Research, 2(6), pp 429-444. 45. Chechen Liao, J ain-Liang Chen, David C. Yen.(2007) ,'Theory of Planning Behavior (TPB) and Customer Satisfaction in the Continued Use of e-service: An Integrated Model', Computers in Human Behavior 23 (2007) pp 28042822 46. Chen, I njazz J . & Popovitch, K. (2003) Understanding Customer Relationship Management: People, Process and Technology Business Process Management J ournal, 9(5), pp. 672-688. 47. Chen J . S. Ching, R. (2006). The Study of Mobile Customer Relationship Management and Loyalty. 2006 International Conference on Service Systems and Service Management. pp 67-72 48. Chieko Minamia, J ohn Dawson,(2008), 'The CRM Process in Retail and Service Sector Firms in J apan: Loyalty Development and Financial Return", J ournal of Retailing and Consumer Services,15(5), pp375-385 49. Chih-Ping-Wei and I -Tang-Chiu (2002), 'Turning Telecommunications Call Details to Churn Prediction: A Data mining Approach, Expert Systems, 23, pp103-112. 187 50. Christine and Binks(1999), 'Impact of Participative Service Relationships on Quality, Satisfaction and Retention: An Exploratory Study', J ournal of Business Research, 46(2), pp 121-132. 51. Churchill J G A, and I acobucci D (2002), Marketing Research Methodological Foundation, Orlando US, Harcourt College publishers. 52. Clark, L.A., And Watson, D.(1995), Constructing Validity: Basic Issues in Objective Scale Development, Psychological Assessment., 7(3), pp.309- 319 53. Cronbach, L.(1990) , Essentials of Psychological Testing, New York: Harper & Row. 54. Cronin, J . J oseph, J r. and Steven A. Taylor (1992), Measuring Service Quality: A Reexamination and Extension, J ournal of Marketing, 56 (July), pp 55-68. 55. Cronin J r, J J , Brady, M K and Hult , G. T. M . ( 2000 ) Assessing the Effects of Quality, Value, and Customer Satisfaction on Consumer Behavioral Intentions in Service Environments, J ournal of Retailing76 (2) pp 193 218 56. Crosby, Lawrence A., Kenneth Evans, and Deborah Cowles (1990), Relationship Quality in Services Selling: An Interpersonal Influence Perspective, J ournal of Marketing, 54 (J uly), pp 68-81. 57. Czepiel, J . A. and R. Gilmore (1987), 'Exploring the Concept of Loyalty in Services', American Marketing Association, pp 91-94. 58. Danaher and Gallagher, (1997), 'Modeling Customer Satisfaction in Telecom of New Zealand, European J ournal of Marketing, 31(2), pp.122 133. 59. David J . Finnegan ,Wendy L. Currie (2010), 'A Multi-layered Approach to CRM Implementation: An Integration Perspective', European Management J ournal, 28(2), pp 153-167 60. David Geffen and Catherine M Ridings.(2002), Implementation Team Responsiveness and User Evaluation of Customer Relationship Management: A Quasi-Experimental Design Study of Social Exchange Theory., J ournal of Management Information Systems, 19(1), pp. 4769. 61. Day, G.S. (1969) A Two-dimensional Concept of Brand Loyalty. J ournal of Adver tising Research, 9, pp. 29 35. 188 62. Delener, N. (1994), Religious Contrasts in Consumer Decision Behaviour Patterns: Their Dimensions and Marketing Implications, European J ournal of Marketing, 28(5), pp. 36-53. 63. Diamantopoulos, A. and Siguaw, J . A. (2000), Introducing Structural Equation Modeling, SAGE Publications, London. 64. Dick, Alan S. and Kunal Basu (1994), 'Customer Loyalty: Toward an Integrated Conceptual Framework', J ournal of the Academy of Marketing Science, 22 (Spring), pp 99-113. 65. Dwyer, F. Robert, Paul H. Schurr, and Sejo Oh (1987), Developing Buyer-Seller Relationships,J ournal of Marketing, 51 (April), 11-27 66. Eunju Koa,Sook Hyun Kimb, Myungsoo Kimc, J i Young Wood (2008), 'Organizational Characteristics and the CRM Adoption Process', J ournal of Business Research, 61(1), pp 65-74 67. Evangelos Grigoroudis, Charalambos Litos, Vassilis A. Moustakis,Yannis Politis, Loukas Tsironis.(2008), 'The Assessment of User-perceived Web Quality: Application of a Satisfaction Benchmarking Approach',European J ournal of Operational Research187 pp 13461357 68. Fink, A. (1995). How to Design Surveys, London: Sage Publications Ltd. 69. Fisk, R.P., Brown, S.W., & Bitner, M.J . (1993) Tracking the Evolution of the Services Marketing Literature, J ournal of Retailing, 69(1), pp. 61 - 103. 70. Fitzsimmons, J . A., & M. J . Fitzsimmons. (2006). Service Management: Operations, Strategy, Information Technology, fifth edition, McGraw-Hill Irwin. 71. Fornell, Claes and David F. Larcker (1988), Evaluating Structural Equation Models with Unobservable Variables and Measurement Error, J ournal of Marketing Research, 18 (February) pp39-50. 72. Forza, C. (2002), Survey research in operations management: a process- based perspective. International J ournal of Operations & Production Management, 22(2), pp.152-194. 73. Fujun Laia,Mitch Griffinb,Barry J . Babin (2009), 'How Quality, Value, Image, and Satisfaction Create loyalty at a Chinese Telecom', J ournal of Business Research Volume62(10), pp 980-986 74. 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, J ournal of Marketing, 64(3), pp6587. 189 75. Gay, L. R. (1987), Third Edition, Educational Research: Competencies for Analvsis and Application, Columbus, Merrll Publishing Company. 76. Gregory, R. J . (1996),Second Edition, Psychological Testing: History, Principles and Applications, Boston, MA: Allyn and Bacon 77. Grapentine, T. (1997), Managing Multicollinearity, Marketing Research 9(3), pp10-21,. 78. Gentry, Lance and Morris Kalliny (2008), Customer Loyalty A Synthesis, Conceptual Framework, and Research Propositions, The J ournal of American Academy of Business, 14 (1) pp 1-9. 79. Gremler, D.D. & Brown, S.W. (1996), Service Loyalty: Its Nature, Importance, and Implications. In: B. EDVARDSSON et al. (Eds), Advancing Service Quality: A Global Perspective QUIS-5, Conference proceedings, pp 132-150 80. Greenberg, Paul. (2002). CRM at the Speed of Light. McGraw-Hill Osborne, NY. Fernandes, Louis. Low Cost Solutions. CRM Magazine, #29. 81. Gustafsson A, Michael D. J ohnson, and I nger Roos(2005)., 'The Effects of Customer Satisfaction, Relationship Commitment Dimensions, and Triggers on Customer Retention', J ournal of Marketing , 69 (October), pp 210218. 82. Hair, J .F., J r., Anderson, R.E., Tatham, R. L., and Black, W.C. (1998). Fifth Edition, Multivariate Data Analysis-, Prentice Hall. 83. Hair, J . F. J r., Black, C. W., Babin, J . B., Anderson, R. E. and Tatham, L. R. (2006) Multivariate Data Analysis (6th ed.), New Jersey: Pearson. 84. Harris, L. C. and Goode, M. M. H. (2004), The Four Levels of Loyalty and the Pivotal Role of Trust: A Study of Online Service Dynamics, J ournal of Retailing, 80, pp139-158. 85. Hampton, G.M., 1993, Gap Analysis of College Student Satisfaction as a Measure of Professional Service Quality, J ournal of Professional Services Marketing, 9(1), pp. 36980. 86. Hanjoon Lee,Linda M. Delene, Mary Anne Bunda and Chankon Kim, Methods of Measuring Health-Care Service Quality, J ournal of Business Research, 48(3) pp 233-246, 2000. 87. Hannikainen, M, Hamalainen, T D, Niemi, M and Saarinen, J ,(2002), 'Trends in Personal Wireless Data Communications', Computer Communications, 25(1),pp 84-99. 190 88. Harraway, J .(1995), Regression Methods Applied. Dunedin New Zealand: theUniversity of Otage Press. 89. Hinkin, T. R., Tracey, J . B. and Enz, C. A. (1997), Scale Construction: Developing Reliable and Valid Measurement Instruments ,J ournal of Hospitality and TourismResearch, 21, pp. 100-120 90. Hoelter, J . W. (1983), The analysis of covariance structures: Goodness- of-fit indices, Sociological Methods and Research, 11, pp.325 344 91. Hutcheson, G. and Sofroniou, N. (1999), The Multivariate Social Scientist, London: Sage Publications. 92. Hu, L. And Bentler, P.M. (1999),Cut off criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternative, Structural Equation Modeling, 6, pp.1-55 93. Hu, H. H. Kandampully J . J uwaheer T. D, (2009) Relationships and Impacts of Service Quality, Perceived Value, Customer Satisfaction, and Image: An Empirical Study, Service Industries J ournal, 29, pp111-126. 94. Hyunchul Ahna, J ae J oon Ahnb, Kyong J oo Ohb, Dong Ha Kimb (2011), 'Facilitating Cross-selling in a Mobile Telecom Market to Develop Customer Classification Model based on Hybrid Datamining Techniques, Expert Systems with Applications, 38, Issue 5, May 2011, Pages 5005-5012 95. Hyung-Su Kima, Young-Gul Kimb(2009),'A CRM Performance Measurement Framework: Its Development Process and Application, Industrial Marketing Management, 38(4), pp 477-489 96. Hyunseok Hwang, Taesoo J ung, Euiho Suh (2004), 'An LTV model and Customer Segmentation based on Customer Value: A Case Study on the Wireless Telecommunication Industry, Expert Systems with Applications 26 pp181188 97. J acob J acoby, Robert W. Chestnut and William A. Fisher(1978), A Behavioral Process Approach to Information Acquisition in Nondurable Purchasing, J ournal of Marketing Research,15(4), pp. 532-544 98. R S J ain (2001), 'Spectrum Auctions in India: Lessons from Experience', Telecommunications Policy, 25(3), pp 671-688, 2001 99. J ain, S K and Gupta, Garima (2004). Measuring Service Quality: SERVQUAL vs. SERVPERF Scales, Vikalpa,The J ournal for Decision Makers, 29(2), pp 25-37. 100. J ames E. Richarda, Peter C. Thirkella & Sid L. Huffb(2007), 'An Examination of Customer Relationship Management (CRM) Technology 191 Adoption and its Impact on Business-to-Business Customer Relationships, Total Quality Management & Business Excellence,18(8), pp 927-945 101. J an U. Beckera, Goetz Greveb,Snke Albersc(2009) , The Impact of Technological and Organizational Implementation of CRM on Customer Acquisition, Maintenance, and Retention, International J ournal of Research in Marketing, 26(3), pp207-215 102. J ay Kandampully (1998), 'Service Quality to Service Loyalty; A Relationship which goes beyond Customer Services, Total Quality Management J ournal. 9 (6) pp 431-443. 103. J ia, S. H. Yan., H. R. (2005), Empirical Analysis on the Forming Model of Customer Loyalty: The Case Study of Mobile Communication Service, Systems and Services Management, International Conference on Services Management ,pp133-137 104. J ohn Kuada and Dorthe Serles (2006), Customer Relationship Marketing (CRM) Practices in Danish Small Businesses, International Management Review, 2(4),pp 28-45 105. J ohn Hadden,Ashutosh Tiwari, Rajkumar Roy, Dymitr Ruta.(2007),Computer Assisted Customer Churn Management:State-Of- The-Art and Future Trends, Computers & Operations Research 34(10), pp2902-2917. 106. J oreskog, K. and Sorbom, D, (1982), Recent developments in Structural Equation Modeling, J ournal of Marketing Research, 19, pp. 65-92. 107. J ohnson, William C. and Sirikit,(2002), 'Service Quality in the Thai Telecommunication Industry: A Tool for Achieving a Sustainable Competitive Advantage', Management Decision, 40(7), pp. 693-701. 108. J ohnson, R L,Tsiros, M and Lancioni R A (1995),'Measuring Service Quality: A Systems Approach', J ournal of Services Marketing, 9(5),pp 6- 19. 109. J oseph, J . & Walker, C.E. (1988), Measurement and Integration of Customer Perception into Company Performance and Quality. In: M.J . BITNER & L.A. CROSBY (Eds), Designing a Winning Service Strategy (New York, AMA). 110. Kandampully, J . (1997),Firms should give Loyalty before they can Expect it from Customers, Managing Service Quality, 7(2), pp. 9294. 111. Kashyap, R. Bojanic, D. C. (2000). A Structural Analysis of Value, Quality, and Price Perceptions of Business and Leisure Travelers, J ournal of Travel Research. 39. pp 45-52. 192 112. Karbowiak, J ohn. (2003). A Loyal Following, CRM Magazine, #31. 113. Kent Eriksson, Anna Lfmarck Vaghult.(2000),'Customer Retention, Purchasing Behavior and Relationship Substance in Professional Services', Industrial Marketing Management 29, pp 363372 114. Kerlinger, F.N and Lee, H.B. (2000), Foundations of Behavioral Research, Harcourt College Publishers, New York, NY. 115. Kim, J ., Suh, E. and Hwang, H. (2003), A Model for Evaluating the Effectiveness of CRM using the balanced scorecard, J ournal of Interactive Marketing, 17pp 519 116. Ko de Ruyter, Bloemer and Peeters. (1997), 'Merging Service Quality and Service Satisfaction, An Empirical Test of an Integrative Model', J ournal of Economic Psychology, 18(4), pp 387-406. 117. Ko, E. (1995), Impact of Quick Response Technology based Attributes on Consumer Satisfaction/Dissatisfaction among Female Apparel Consumers, Unpublished doctoral dissertation, Virginia Polytechnic Institute and State University, Blacksburg. 118. Koustuv Dasgupta, Rahul Singh, Balaji Viswanathan, Dipanjan Chakraborty, Sougata Mukherjea, Amit A. Nanavati,Anupam J oshi.(2008), 'Social Ties and their Relevance to Churn in Mobile Telecom Networks', ACM 978-1-59593-926-5/08/0003 119. Kumar, V. and Reinartz, W. (2006) Customer Relationship Management: A database approach, NewYork: J ohn Wiley and Sons Inc 120. LaBarbera, P. A. and Gurhan, Z. (1997), The Role of Materialism, Religiosity and Demographics in Subjective Well-being, Psychology and Marketing, 14(1), pp. 71-97 121. Laroche, M., J . Bergeron and C. Goutaland, (2001),A Three Dimensional Scale of Intangibility. J ournal of Service Research, 4pp 26- 38. 122. J . Lawler, P. Vandepeutte & D. Anderson.(2006), A Study of Customer Relationship Management (CRM) on Apparel European Web sites, WIT Transactions on Information and Communication Technologies, 37, 2006 123. Lehtinen, U and Lehtinen, J R (1991),'Two Approaches to Service Quality Dimensions', The Service Industries J ournal, 11(3), pp287-305. 124. Leisen, B and Vance, C (2001),Cross-National Assessment of Service Quality in the Telecommunication Industry: Evidence from the USA and Germany, Managing Service Quality, 11(5), pp 307-317. 193 125. Lewis-Beck, M. S. (1993), Applied Regression: An Introduction, In Lewis-Beck, M. S.(ed.), Regression Analysis. International Handbooks of Quantitative Applications in the Social Sciences, London: Sage Publications/Toppan Publishing, 2, pp. 1-68, 126. Li, L. (2008),Study of the Relationship between Customer Satisfaction and Loyalty in Telecom Enterprise, 4th International Conference on Wireless Communications, Networking and Mobile Computing, pp1-7. 127. Long, Y. L. J en, C. (2004). An Integrated Analysis of Relationship between Service Quality, Relationship Quality, Relationship Value and Customers' Loyalty-An Example of Consumers of Mobile Telecommunication Industry. Tamsui Oxford J ournal of Economics and Business. 12, pp 73-112. 128. Luarn, P. and H. Lin, (2003). A Customer Loyalty Model for E-Service Context, J ournal of Electronic Commerce Research., 4pp 156-167. 129. Maccallum, R. and Austin, J . (2000),Applications of Structural Equation Modeling in Psychological Research, Annual Review of Psychology, 51, pp. 201-226 130. Maccallum, R., Browne, M. and Sugawara, H. (1996), Power Analysis and Determination of Sample Size for Covariance Structure Modeling, Psychological Methods, 1(1), pp.130-149, 131. Madariaga, J . G. (2004),Customer Relations Management (CRM): A Comparative Study between U.S.A. and Spain, J ournal of International Technology and Information Management, 13(2), pp 87-96. 132. Mahapatra S.S. and Khan M.S. (2006) 'Neural Approach for Service Quality Assessment: A Case Study, Industrial Engineering J ournal, 35(7), pp.3035. 133. Malhotra, N. (2006), Fourth Edition, Marketing Research An Applied Orientation, Pearson Education 134. Malhotra, N. K. and Birks, D. F. (1999), Marketing Research: An Applied Approach. Essex: Pearson Education Ltd. 135. Mason, C. H. and Perreault, W. D. (1991), , Collinearity, Power and Interpretation of Multiple Regression Analysis, J ournal of Marketing Research, 28(3) , pp. 268-280 136. Michael K Brady, J .J oseph Cronin J r.and Richard R (2002), Brand Performance-only Measurement of Service Quality: A Replication and Extension, J ournal of Business Research, 55(1) 1, pp 17-31, 2002. 194 137. McNight, H., Choudhury, V. and Kacmar, C. (2003), Developing and Validating Trust Measures for E-Commerce: An Integrative Typology, Information Systems Research, 13 pp.334-361 138. Mithas, Sunil, M.S. Krishnan, and Claes Fornell. (2005) Why Do Customer Relationship Management Applications Affect Customer Satisfaction? J ournal of Marketing, 69 (4) pp 201209. 139. Mittal, Vikas, and Wagner A. Kamakura. (2001),Satisfaction, Repurchase Intent, and Repurchase Behavior: Investigating the Moderating Effect of Customer Characteristics, J ournal of Marketing Research, 38 (1), pp 131-142. 140. Mitchell.-W. (1996), Assessing the Reliability and Validity of Questionnaire, J ournal of Applied Management Studies, 5(2) pp. 199-217. 141. Moon-Koo Kim, Myeong-Cheol Parkb, Dong-Heon J eonga (2004), 'The Effects of Customer Satisfaction and Switching Barrier on Customer Loyalty in Korean Mobile Telecommunication Services, Telecommunications Policy, 28pp145159 142. Morgan, R.M., and Hunt, S.D. (1994), The commitment-trust theory of relationship marketing, J ournal of Marketing, 58(July), pp. 20-38. 143. Murray, K. B. (1991), 'A Test of Services Marketing Theory: Consumer Information Acquisition Activities', J ournal of Marketing, 55 (January), pp 10-25. 144. Nitin Seth, S.G. Deshmukh, Prem Vrat, (2005), 'Service Quality Models: A Review", International J ournal of Quality & Reliability Management, 22(9), pp.913 - 949 145. Oliver, R.L. (1999), Whence Customer Loyalty?, J ournal of Marketing, 63, pp. 33-44 146. Oppenheim, A. N. (1993), Questionnaire Design, Interviewing and Attitude Measurement, London: Pinter Publishers Ltd. 147. Pallant J (2001), SPSS Survival Manual Open University Press, Buckingham, UK. 148. Parasuraman A (2005),'E-S-QUAL: A Multiple-Item Scale for Assessing Electronic Service Quality', J ournal of Service Research,7(3), pp: 213-233 149. Patrick Asubonteng, Karl J . McCleary, J ohn E. Swan, (1996) 'SERVQUAL Revisited: A Critical Review of Service Quality', J ournal of Services Marketing, 10(6), pp.62 - 81 195 150. Payne, A., and Frow, P. (2005), A Strategic Framework for Customer Relationship Management, J ournal of Marketing, 69, pp167-176. 151. Pedhazur, E. J . (1997), Third Edition, Multiple Regression in Behavioral Research: Explanation and Prediction, Wadsworth- Thomson Learning. 152. Peppard, J oe (2000) Customer Relationship Management (CRM) in Financial Services European Management J ournal, 18(3), pp. 312-327. 153. Peter J . Danaher, Rodger W. Gallagher, (1997),'Modelling Customer Satisfaction in Telecom of New Zealand", European J ournal of Marketing, 31(2), pp.122 133 154. Peterson, R. (1994), A Metaanalysis of Cronbachs Coefficient Alpha, J ournal of Consumer Research, 21(September), pp.381-391. 155. Pindyck, R. S. and Rubinfeld, D. L. (1991), Third Edition, Econometric Models and Economic Forecasts, New York: McGraw-Hill. 156. Pritchard, Mark P. (1991), Development of the Psychological Commitment Instrument (PCI) for Measuring Travel Service Loyalty, doctoral dissertation, University of Oregon. 157. Qianli Xu, Roger J . J iao, Xi Yang and Martin Helander (2009), 'An Analytical Kano Model for Customer Need Analysis, Design Studies,30(1) pp 87-110 158. Rajshekhar , G . J . and Moberg, C. R. (1997) Service Loyalty: Implications for Service Providers, J ournal of Services Marketing 11 (3) pp 165 179. 159. Ran Kivetz, Oleg Urminsky, and Yuhuang Zheng (2006),'The Goal- Gradient Hypothesis Resurrected: Purchase Acceleration, Illusionary Goal Progress, and Customer Retention', J ournal of Marketing Research 39, pp 3958 160. Ranaweera C. and Neely A.D. (2003), 'Some Moderating Effects on the Service Quality-Customer Retention Link', International J ournal of Operations and Production Management, 23(2), pp 230-248. 161. Raykov, T. and Marcoulides, G. A. (2000), A Method for Comparing Completely Standardized Solutions in Multiple Groups, Structural Equation Modeling,7, pp. 292-308 162. Reinartz, Werner J ., Manfred Krafft, and Wayne D. Hoyer. (2004) The Customer Management Process: Its Measurement and Impact on Performance, J ournal of Marketing Research, 41 (3), pp 293305. 196 163. Rekha J ain (1993), 'Review of the Policy Changes in the Indian Telecom Sector: Implications for Decision Makers', J ournal of Global Information Management ,1(3),pp 33-45 164. Robert Gee (2008), Understanding and Profitably Managing Customer Loyalty, Marketing Intelligence & Planning, 26(4),pp. 359-374 165. Rosen, L D and Karwan, K R (1994), 'Prioritizing the Dimensions of Service Quality', International J ournal of Service Industry Management, 5(4), pp 39-52. 166. Rundle, T. S.(2005), Elaborating Customer Loyalty: Exploring Loyalty to Wine Retailers, J ournal of Retailing and Consumer Services,12.pg 333- 345. 167. Russell S. Winer (2001),Customer Relationship Management: A Framework, Research Directions, and the Future ,Haas School of Business, University of California at Berkeley 168. Ryals, Lynette. (2005) Making CRM Work: The Measurement and Profitable Management of Customer Relationships J ournal of Marketing, 69 (4), pp25261. 169. Sheth, J ., N. & Parvatiyar, A. (1995), Relationship Marketing in Consumer Markets: Antecedents and Consequences, J ournal of the Academy of Marketing Science, 23(Fall), pp. 255-271. 170. Siu, N Y M and Cheung, J T (2001),'A Measure of Retail Service Quality', Marketing Intelligence and Planning, 19(2), pp 88-96. 171. Snyder, R.D. (1986), Service Loyalty and its Measurement: A Preliminary Investigation (New York, AMA). 172. Soteriou, A C and Stavrinides, Y (2000). An Internal Customer Service Quality Data Envelope Analysis Model for Bank Branches, International J ournal of Bank Marketing, 18(5), pp246-252. 173. Stephen D. McDowell(2003), Indias Experiments in Mobile Licensing, Telecommunications Policy, 27(5-6),pp 371-382 174. Studenmund, A. H. (2001), Fourth Edition, Using Econometrics: A Practical Guide, Boston, MA: Addison Wesley Longman. 175. Sudhahar, J .C., D. I srael, A.P. Britto and M. Selvam, (2006), 'Service Loyalty Measurement Scale: A Reliability Assessment', American. J ournal of Applied Sciences, 3, pp 1814-1818. 197 176. Sugawara, H. M. and Maccallum, R. C. (1993),Effect of Estimation Method on Incremental Fit Indexes for Covariance Structure Models, Applied Psychological Measurement, 17, pp. 365-37 177. Sumeet Gupta, Hee W. Kim.(2008), 'Linking Structural Equation Modeling to Bayesian networks: Decision Support for Customer Retention in Virtual Communities', European J ournal of Operational Research 190 (2008), pp 818833 178. Sun, T., Youn, S., Wu, G. and Kuntaparaporn, M. (2006), Online Word-of-Mouth (or Mouse): An Exploration of Its Antecedents and Consequences, J ournal of Computer-Mediated Communication, 11, pp 1104-1127. 179. Sung Ho Ha (2007),'Applying Knowledge Engineering Techniques to Customer Analysis in the Service Industry', Advanced Engineering Informatics, 21 (2007), pp 293301 180. Tae Hyup Roha,Cheol Kyung Ahnb, I ngoo Hanc(2005), The Priority Factor Model for Customer Relationship Management System Success, Expert Systems with Applications28 (2005) pp 641654 181. Thomas, J acquelyn S., Robert C. Blattberg, and Edward J . Fox. (2004). Recapturing Lost Customers J ournal of Marketing Research, 41 (1), pp 3145. 182. Tim Coltman (2007) , 'Why build a Customer Relationship Management Capability?, J ournal of Strategic Information Systems 16 (2007) pp 301 320 183. Torsten J . Gerpott, Wolfgang Rams and Andreas Schindler (2001), 'Customer Retention, Loyalty, and Satisfaction in the German Mobile Cellular Telecommunications Market', Telecommunications Policy, 25, pp 249-269 184. Trubik, E. and M. Smith,(2000),Developing a Model of Customer Defection in the Australian Banking Industry. Managerial Auditing J ournal, 15: pp 199- 208. 185. Ueltschy, L.C., Laroche, M., Eggert, A., Bindl, U.,(2007).Service Quality and Satisfaction: An International Comparison of Professional Services Perceptions, J ournal of Services Marketing, 21,pp 410-423 186. Veith Krner, Hans-Dieter Zimmermann.(2000),Management of Customer Relationship in Business Media - The Case of the Financial 198 Industry -Proceedings of the 33rd Hawaii International Conference on SystemSciences (IEEE) 187. Verhoef, Peter C. (2003), Understanding the Effect of Customer Relationship Management Efforts on Customer Retention and Customer Share Development. J ournal of Marketing, 67(4), pp 30-45. 188. Wang, Y. and Lo, H.P. (2002), 'Service Quality, Customer Satisfaction and Behavioural Intentions: Evidence from Chinas Telecommunication Industry', Info-The J ournal of Policy, Regulation and Strategy for Telecommunications, 4(6), pp.50-60. 189. Widaman, K. F. and Thomson, J . S. (2003), On Specifying the Null Model for Incremental Fit Indices in Structural Equation Modeling, Psychological Methods, 8, pp.16-37 190. Woodside, A. G., Frey, L. L. and Daly, R. T. (1989) Linking Service Quality, Customer Satisfaction and Behavioral Intention,J ournal of Health Care Marketing, 9(December) pp 5 17 . 191. Ying-Feng Kuo a,Chi-Ming Wub, Wei-J aw Deng(2009),'The Relationships among Service Quality, Perceived Value, Customer Satisfaction, and Post-purchase Intention in Mobile Value-added Services', Computers in Human Behavior 25 (2009) pp 887896 192. Yonggui Wang, Hing-Po Lo and Yongheng Yang (2004), 'An Integrated Framework for Service Quality, Customer Value, Satisfaction: Evidence from China's Telecommunication Industry', Information Systems Frontiers, 6(4),pp 325-340 193. Yonggui W., Lei D., Shunping H., Guicheng S.,(2005), The Dimensions of Customer Loyalty and its Key Drivers: An Integrated Framework in Perspective of Customer Equity Management, Services Systems and Services Management. Proceedings of ICSSSM '05. 2005 International Conference on, 1, pp 204-210. 194. Zairi, M, (2000), Managing Customer Dissatisfaction through Effective Complaint Management Systems, The TQM Magazine, 12, pp 331-335. 195. Zeithaml, V.A. & Bitner, M.J . (1996) Services Marketing (New York, McGraw-Hill). 196. Zhilin Yanga, Shaohan Caib, Zheng Zhouc, Nan Zhoua (2005), 'Development and Validation of an Instrument to Measure User Perceived Service Quality of Information Presenting Web Portals", Information & Management,42(4), pp 575-589 199 APPENDI X I Questionnaire Dear Sir/Madam, Good morning/afternoon/evening Could you please spare a few minutes to participate in our survey. We are students from SRM University, Chennai and conducting this survey as a part of our coursework to study the influence of CRM by mobile service operators on customer satisfaction. We would be extremely grateful for your participation. 1. Which Mobile Service do you use? . BSNL . Idea . Airtel . Aircel . Vodafone . Reliance . Others 2. What is the plan that you are using? . Prepaid . Postpaid 3. How long are you using this mobile service? . <6 months . 6 months 1 yr . 1 yr 2 yrs . 2 yrs 3 yrs . >3 yrs 4. Age Group : . 18-20 . 21-30 .31-40 . 41-50 .>50 5. Monthly Expenditure per month on Mobile Services . Upto 500 . 501-1000 . 1001-2000 .2001-3000 . >3000 6. Education . Graduate . P-Graduate . Others 7. Occupation . Business . Professionals . Educator . Home-maker . Student . Others 200 SA A N DA SDA Please indicate using a mark for the following questions 5 4 3 2 1 1 Contact employees perform the service right the first time 2 The company provides services at the promised time 3 You are kept well-informed about the progress of your complaints 4 Billing system is accurate and error free 5 Bills are received in time 6 Contact employees gives you prompt service 7 Your Complaints / queries are taken seriously 8 Your complaints are resolved quickly 9 They are always willing to help you 10 Contact employees are friendly and polite while handling your complaints / queries 11 They have the adequate knowledge of tariffs and plans of service providers 12 The behaviour of contact employees instils confidence in you 13 You feel safe in your transactions with your service provider 14 For lodging the complaints service provider is easily accessible 15 They have your best interests in heart 16 Contact employees give you individual attention 17 The employees understand your specific needs 18 Retailer network of your service provider is easily located 19 Service provider's physical facilities are visually appealing 20 Contact employees appear neat 21 Materials associated with the service(such as pamplets etc) are visually appealing 22 They have convenient business hours 23 Ease of lodging the complaints / queries 24 Your service provider provides flexibility in the payment of bills 25 Application formalities are simple 26 Your service provider provides sufficient geographical coverage (on highways, inside the buildings and basement) 27 You experience minimum premature termination of calls during conversation (ie call drops) 28 You get clear and undisturbed voice 29 Your call gets connected to the called person during the first attempt most of the time 30 You are able to make calls at peak hours 201 SA A DA SD 4 3 2 1 31 I will transact with this mobile service provider again for future needs 32 I will try new services that are provided by the service provider 33 I will recoomend other people to patronize this service provider 34 I will say positive things to other people about the services provided by the mobile service operator 35 I will continue to patronize this mobile service provider / service even if the service charges are increased moderately 36 I have strong preference to this service provider 37 I will continue to use this service provider regardless of changes in the service 38 I am likely to pay little more for the services when situation arises 39 To me, this service provider would rank first among others 40 I would continue this service for a long period of time 41 I will deal exclusively with the service provider 42 The service of the provider reflect a lot about who I am 43 I have found this provider better than others 44 I always find the terms of the provider as inferior 45 Repeatedly, the performance of this provider is superior to that of its competitors 46 I dislike the terms of the service provider 47 I like the performance and services of the provider 48 I have a negative attitude toward this provider 49 I am satisfied wih my decision to stay with this provider 50 The employees at the providers place give individual attention 51 The employees understand my specific needs and go out of their way to help me 52 The employees respond caringly when I share my problems 53 The personnel at the service provider are filled with professionalism and dedication 54 I am committed to the service provider 55 Even when I get to hear any negative information about the provider, I would still continue with this service provider 56 I like switching from one service provider to another 57 My continued association with the mobile operator is important to me 202 SA A N DA SDA 5 4 3 2 1 59 There is a likelihood to choose the service again 60 There is a likelihood to recommend 61 There is a likelihood to pass on positive feeling about the services 62 There is a likelihood to continue purchasing the same products / services 63 There is a likelihood to increase purchase size 64 There is a likelihood to increase frequency of purchasing 65 There is a likelihood to purchase different products / services 66 There is a likelihood to continue this only as a supplementary connection 67 There is a likelihood to switch to a different provider We once again thank you for the opportunity and the time in providing these details. 203 PUBLI CATI ONS BASED ON THI S RESEARCH Refereed J ournals 1. Vani Haridasan and Shanthi Venkatesh, "Impact of Service Quality in Improving the Effectiveness of CRM Practices through Customer Loyalty A Study on Indian Mobile Sector" , International Journal of Management(IJ M), Volume 3, Issue 1, J anuary- April 2012 , ISSN 0976- 6502, Journal Impact Factor 3.5420 2. Vani Haridasan and Shanthi Venkatesh,"CRM Implementation in Indian Telecom Industry Evaluating the Effectiveness of Mobile Service Providers Using Data Envelopment Analysis, International Journal of Business Research and Management (IJ BRM), Volume (2), Issue (3), Sep Oct 2011, ISSN 2180-2165 Presentations in I nternational Conferences 1. Vani Haridasan and Shanthi Venkatesh, "Influence of Latent Factors in Strengthening the CRM Practices through Customer Loyalty A Study on Indian Mobile Sector" at 2011 International Research Conference and Colloquium, held at Kuala Lumpur, Malaysia. The conference was organized by Universiti Tun Abdul Razak, Kuala Lumpur, Malaysia jointly with IIM Indore, 10-11 Oct 2011. 204 CURRI CULUM VI TAE VANI HARIDASAN was born in Madurai, Tamilnadu, India, on 25th October 1974. She obtained B.E. in Electronics and Communications Engineering from Regional Engineering College, J aipur (now NIT J aipur) in 1996 and M.B.A from Department of Management Studies, Madurai Kamaraj Universty in 1999. She has 10 years of teaching experience, her area of interest being Operations Research and subjects in the Operations specialization. She is currently working as Assistant Professor in the Department of Management, SSN School of Management & Computer Applications, Chennai, and is a Research Scholar in the Department of Management Studies, SRM University. She has published 2 research papers in reputed International J ournals. She has presented one paper in International Research Conference at Kuala Lumpur, Malayasia. All her papers have been well received by the academia.