Insurance market research: The determinants of price sensitivity and the key role played by intermediaries.

Insurance market research: The determinants of price sensitivity and the key role played by intermediaries By Sérgio Dominique Ferreira Lopes PhD Thesis in Business and Management Studies (Branch of Marketing and Strategy) Supervisors: Prof. Helder Vasconcelos Prof. João Proença 2015 i BIOGRAPHICAL NOTE Sérgio Dominique was born on the 29th January 1984, in Paris, France. He is graduated in Economic and Consumer Psychology, holds a postgraduate course in Brand Management, and a Ph.D. in Social Psychology. He has been as Ph.D. student at the School of Management and Economics of the University of Porto since 2011. Currently he is Adjunct Professor at the School of Management of the Polytechnic Institute of Cavado and Ave. Previously, he was a researcher at the PSICOM-USC research center (Consumer Psychology Research Center of the University of Santiago de Compostela), Department of Methodology, Faculty of Psychology of the University of Santiago de Compostela. He was also a researcher at the Faculty of Economic and Business Sciences (Department of Business Organization and Marketing) of the University of Vigo. ii ACKNOWLEDGMENTS I would like to thank my supervisors, Professor Helder Vasconcelos for his endless support and encouragement and to Professor João F. Proença for his support. I am also very grateful to the Editors (Professor Steve Baron and Professor Rebekah Russell-Bennett), to the Associate Editor (Professor Javier Reynoso) and to three anonymous referees of the Journal of Services Marketing for their thorough and thoughtful reports concerning the paper Determinants of customer price sensitivity: An empirical analysis. Thanks to the moderators and authors’ feedback who attended my presentations: i) Salesmen, when should they talk about price? at the 25th Portuguese-Spanish Conference of Scientific Management, Ourense, Spain; ii) Consumer behavior in the insurance sector: A qualitative approach at the 23th Portuguese-Spanish Conference of Scientific Management, Malaga, Spain. I thank all the teachers of the doctoral programme, namely, Professor Carlos Cabral Cardoso and Professor Pedro Quelhas Brito. Finally, I would like to thank my wife and my parents for all the support. iii ABSTRACT The insurance sector plays a very important role worldwide, providing stability in markets. Insurers, premium, intermediaries, bundling strategies, customers’ satisfaction, customers’ price sensitivity and claims management are some of the most important issues in the insurance sector. However, the number of insurance studies from the perspective of customers is very little, especially in services marketing literature. Therefore, purpose of this investigation is threefold: i) to study the importance that insurance customers give to premium, insurers, intermediary recommendations, and bundling strategies, as well as the relationship between attributes and consumer price sensitivity and price elasticity of demand; ii) to identify the strategic importance of attributes’ order presentation, identifying the right moment to present premium, bundling strategy and intermediaries’ recommendation to insurance customers; iii) to study the insurance supply management through customer’s satisfaction with intermediaries and insurers, as well as the preferences of customers in the purchase decision-making process. In order to study the attributes’ importance, we used Conjoint Analysis with Full Profile. A two-stage cluster analysis was performed to segment the market. Regarding the study of the strategic importance of attributes’ order presentation, Kruskal-Wallis test was performed. Finally, in order to study customers’ satisfaction, structural equation modelling was performed and Multidimensional Scaling unfolding model was applied to understand customers’ preferences in chapter 3. Research findings indicate that price sensitivity is affected by the level of purchase involvement, bundled discounts, and brand loyalty. Also, brand loyalty has a strong influence on customer acceptance of bundled discounts. Price bundling increases firm's revenues and profits. Regarding the effect of attributes’ order of presentation, primacy and recency effects were detected, as well as a transfer effect related with the level of importance of attributes that precede and succeed attributes. Finally, results show that insurance customers’ satisfaction is statically related to intermediaries and not to insurers, and that intermediaries play a central role in the management of customers claims, as well as in premium acceptance. iv This investigation presents theoretical and practical contributions and managerial suggestions regarding customers price sensitivity, bundling strategies, salesperson approach to customers and the strategic importance of insurance intermediaries. Key words: Insurance, Price sensitivity, Price elasticity, Price bundling, Intermediaries, Strategic order of product attribute presentation, Customers satisfaction, Customers preferences, Supply chain. v RESUMO O setor de seguros desempenha um papel fundamental a nível internacional, proporcionando estabilidade nos mercados financeiros. As seguradoras, os distribuidores (mediadores), os prémios, a sensibilidade dos consumidores face ao preço, as estratégias de bundling, a gestão de sinistros e a satisfação dos clientes são aspetos críticos no sector segurador. Neste contexto, o objetivo desta investigação é triplo: i) estudar a importância que os clientes do sector segurador atribuem ao prémio, às seguradoras, às recomendações dos distribuidores, a estratégias de bundling, bem como a relação entre atributos e sensibilidade dos consumidores face ao preço e a elasticidade do procura face ao preço; ii) compreender qual a importância estratégica da ordem de apresentação dos atributos no sector segurador, identificando qual o melhor momento para apresentar cada atributo; iii) estudar a gestão da cadeia logística através da satisfação dos clientes com os distribuidores e com as seguradoras, bem como as suas preferências no processo de tomada de decisão compra. No âmbito do estudo da importância atribuída aos atributos, aplicou-se a Análise Conjunta com Full Profile. Para segmentar o mercado, recorreu-se a uma Análise Cluster de duas fases. Relativamente ao estudo da ordem estratégica da apresentação de atributos, utilizou-se a estatística de Kruskal-Wallis. Finalmente, para estudar a satisfação dos consumidores recorreu-se a modelos de equações estruturais, e a análise das preferências dos consumidores no capítulo 3 foi obtida através da aplicação do Escalonamento Multidimensional através do modelo unfolding. Os resultados indicam que a sensibilidade dos consumidores face ao preço é influenciada pelo nível de envolvimento financeiro dos consumidores na compra, estratégias de bundling e por comportamentos de lealdade. Além disso, a lealdade influencia fortemente a aceitação de estratégias de bundling. As estratégias de bundling permitem aumentar as receitas e os lucros no sector segurador. No que concerne ao efeito da ordem de apresentação dos atributos, foram detetados efeitos de recência e de primazia, bem como um efeito de transferência ou efeito âncora. Finalmente, os resultados mostram que os mediadores/distribuidores desempenham um papel vi preponderante na gestão da cadeia logística do sector segurador. Concretamente, constata-se que a satisfação dos clientes do mercado segurador depende em maior medida dos distribuidores do que das seguradoras. Paralelamente, os distribuidores desempenham, também, um papel central na gestão de sinistros, bem como numa melhor aceitação do prémio. Esta investigação apresenta claras contribuições teóricas e práticas, bem como sugestões para uma gestão otimizada do sector segurador. Palavras-chave: Setor segurador, Sensibilidade ao preço, Elasticidade da procura, Price bundling, Mediadores, Ordem estratégica da apresentação de atributos, Satisfação de consumidores, Preferências de consumidores, Cadeia de fornecimento vii Next, a graphical abstract is also provided for a more intuitive understanding. GRAPHICAL ABSTRACT viii TABLE OF CONTENTS BIOGRAPHICAL NOTE ............................................................................................... i! ACKNOWLEDGMENTS .............................................................................................. ii! ABSTRACT .................................................................................................................... iii! RESUMO ......................................................................................................................... v! GRAPHICAL ABSTRACT ......................................................................................... vii! TABLES .......................................................................................................................... xi! FIGURES ...................................................................................................................... xiii! GRAPHS ....................................................................................................................... xiv! INTRODUCTION .......................................................................................................... 1! Research scope ......................................................................................................................... 1! Methodology ............................................................................................................................. 2! Structure ................................................................................................................................... 3! CHAPTER I – DETERMINANTS OF CUSTOMER PRICE SENSITIVITY: AN EMPIRICAL ANALYSIS .............................................................................................. 6! 1. INTRODUCTION ............................................................................................................. 8! 2. THEORETICAL FRAMEWORK .................................................................................. 12! 2.1. Pricing and Price Sensitivity ................................................................................................. 12! 2.2. Bundling ................................................................................................................................ 16! 3. METHODOLOGY AND DATA .................................................................................... 20! 3.1. Sample ................................................................................................................................... 20! 3.2. Data collection ....................................................................................................................... 20! 3.3. Attributes’ selection ............................................................................................................... 21! 3.4. Procedure ............................................................................................................................... 21! 3.5. Methods and results ............................................................................................................... 24! 4. RESULTS ....................................................................................................................... 25! 4.1. Results of Conjoint Analysis ................................................................................................. 25! 4.2. Customer buying decision process ........................................................................................ 26! 4.3. Analysis of the hypotheses .................................................................................................... 28! 4.3.1. Purchase involvement .................................................................................................... 28! 4.3.2. Customer loyalty ............................................................................................................ 30! 4.3.3. Price bundling and price perception .............................................................................. 33! 1 INTRODUCTION RESEARCH SCOPE The insurance sector plays a major role in leveraging the economies of many countries, providing stability and confidence in markets. Yet, the number of insurance studies from the perspective of customers is very little, especially in services marketing literature. Premium is one of the most important elements in the insurance market (Barroso and Picón, 2012; Rai and Medha, 2013) but literature does not identify the concrete importance it has for customers. Pricing strategies has been a much-discussed issue in the management, marketing and economy literature (Goldsmith and Newell, 1997; Ramirez and Goldsmith, 2009; Li, Green, Farazmand, and Grodzki, 2012; Roy, 2012; Brophy, 2013a). Bundling strategies are very important in terms of business management, being price bundling strategies or product bundling strategies (Ferrell and Hartline, 2005, p. 286; Rao and Kartono, 2009, p. 15; Gerdeman, 2013; Brito and Vasconcelos, 2015). Involvement is another important element for customers (Zaichkowsky, 1988; Datta, 2003; Russell-Bennett, McColl-Kennedy and Coote 2007), whether with advertisements (Krugman, 1977), products (Hupfer and Gardner, 1971), with purchase decisions (Clarke and Belk, 1978), or pricing decisions (Rao and Kartono, 2009, p.30). Insurance distribution channels are quite particular in the insurance sector. Banks, postal, brokers, intermediaries and insurers are the main distribution channels in Portugal (Portuguese Association of Insurance, 2014). However, insurance marketing literature does not seem very clear about how insurance salesperson should approach customers. For example, the order in which products’ characteristics are presented to customers play an important role in terms of sales optimization (Buda and Zhang, 2000; 2 Gatzert, Huber and Schmeiser, 2010; Horgarth and Einhorn, 1992). Also, intermediaries’ recommendation can strongly influence customers’ behaviors (O’Loughlin and Szmigin, 2007; Eckardt and Rathke-Doppner, 2010; Robson and Sekhon, 2011; Brophy, 2013a). Therefore, the supply chain management plays an important role in the insurance industry, namely, in customers’ satisfaction and preferences. Furthermore, customers’ satisfaction is one of the most studied concepts in management and marketing literature (Bernhardt, Donthu and Kennet, 2000; Edvardsson, Johnson, Gustafson and Strandvik, 2000; Orsingher, Valentini and Angelis, 2010), being important to investigate the role of intermediaries in customers’ satisfaction. METHODOLOGY In this investigation, a mixed approach based on qualitative and quantitative methodologies was used. In order to identify the most important characteristics of the insurance industry, three focus groups were conducted. Two focus groups composed by eighteen (18) auto insurance consumers of the B2C market were conducted and third focus group composed by six insurance intermediaries (B2B) were conducted. The quantitative approach was used for the other analyzes. Concerning chapter 1, several quantitative methods were used. Specifically, Conjoint Analysis was performed in order to measure the concrete importance of key attributes of the insurance business, such as premium, brand (insurer), bundling strategy and intermediaries’ recommendation, as well as the relationship between attributes and consumer price sensitivity. Regarding market segmentation, a two stage post-hoc segmentation was performed through Cluster Analysis. Finally, the traditional formula to estimate price elasticity of demand was used. In chapter 2, Conjoint Analysis and bivariate methods such as Shapiro-Wilk and Kruskal-Wallis were performed in order to study the importance of the strategic order of products’ attributes presentation in the insurance market. 3 In chapter 3, a structure equation modeling (SEM) was developed in order to understand the specific impact of intermediaries and insurers on consumer satisfaction. Multidimensional Unfolding was performed in order to compare the structure of consumers’ preferences and the insurance supply chain process. This brief methodological presentation is only a summary of the methods and techniques employed in data analyzes. A detailed and specific description is presented in each of the papers (chapters 1, 2 and 3) of this thesis. STRUCTURE Chapter 1 analyzes the determinants of price sensitivity in the insurance sector, trying to fill this gap in the literature. This chapter also analyzes the importance that insurance customers give to premiums, insurers, intermediary recommendations, and bundling strategies. This study shows how it is possible to decrease price sensitivity. Chapter 2 focuses on identifying the most strategic order of products’ attributes in the insurance sector. Literature has highlighted the effects of using different attributes’ order of presentation. However, literature does not provide empirical results of this issue in the insurance sector. Primacy and recency effects were detected, as well as a transfer effect related with the level of importance of attributes that precede and succeed attributes. But more important, it is possible to identify a specific attribute presentation order that decreases price importance and increases the impact of bundling strategies and intermediary’s recommendation. Chapter 3 emphases the structure of the insurance market from the customers’ perspective, both in terms of customers’ satisfaction, as well as in the purchase decision-making process. 4 Chapter 4 presents general conclusions, theoretical implications in terms of Marketing and management, managerial implications concerning the insurance sector. Limitations regarding the research are also presented, as well as some further research questions. 5 Paper 1 Paper 2 Paper 3 General implications Purpose This paper investigates the importance that insurance customers give to premiums, insurers, intermediary recommendations, and bundling strategies. The relationship between attributes and consumer price sensitivity is also studied. Sales management plays an important role in firms’ profit. The main goal of this work is to identify the right moment to present price to insurance customers, as well as the insurer bundling strategy and intermediary’s recommendation. Insurance market has enormous churn rates because customers’ purchase decision-making process and claims management relies heavily on intermediaries. The purpose of this study is to investigate the role played by intermediaries in customer’s satisfaction, as well as in the preferences of customers regarding the purchase decisionmaking process. Intermediaries play a central role in the insurance market dynamics and insurance supply chain, as well as in customers’ satisfaction. For intermediaries prefer particular insurers, they must be aware of the suggestions and feedback from intermediaries. This would be especially important in terms of claims management services. Price bundling increases insurances and intermediaries’ revenues and profits. So, they should me used more often. In order to decrease the importance of premium, salespeople should present first the insurer, followed by the bundling strategy, the intermediary’s recommendation and, finally, the premium. Methodology Conjoint Analysis was performed in order to study the importance of the attributes. Cluster analysis was applied to segment the market. Conjoint Analysis was applied in order to measure attributes’ importance of each series. Kruskal-Wallis test was performed in order to study possible effects of order of product attribute presentation. Structural Equation Modeling was used in order to study the impact of insurers and intermediaries in consumers’ satisfaction. The Multidimensional Scaling unfolding model was used to analyze consumer preferences. Findings Price sensitivity is affected by the level of purchase involvement, bundled discounts, and brand loyalty. Also, brand loyalty has a strong influence on customer acceptance of bundled discounts. Price bundling increases a firm's revenues and profits. Primacy and recency effects were detected, as well as a transfer effect related with the level of importance of attributes that precede and succeed attributes. Intermediaries play a key role in the insurance market, concretely, in customers’ satisfaction, in the management of customers’ claims, and in the purchasing process (premium acceptance). Implications There is very little evidence regarding studies on price sensitivity in the insurance sector, mostly because, in many countries, premiums are strongly regulated. This study shows how it is possible to decrease price sensitivity. Salesperson can improve their approach to customers, decreasing the importance given to price and increasing the positive impact of bundling strategies and intermediary’s recommendation in sales. Intermediaries play a key role in the insurance market, concretely, in customers’ satisfaction, in the management of customers’ claims, and in the purchasing process (premium acceptance). Originality The study contributes to the service marketing literature and marketing of the insurance sector by providing empirical evidence of the impact of price bundling on insurance customer sensitivity, with the use of a methodological and experimental approach. It was possible to identify a specific order of attributes presentation in the insurance sector, considering other attributes that not only the price. This study analyzes the insurance supply chain management including three different players: i) customers; ii) intermediaries; iii) insurers. Consumers’ preferences in terms of purchasing behavior and satisfaction rely more in intermediaries than in insurers. An original and brief questionnaire to measure insurance customers’ satisfaction is tested with acceptable psychometrics properties. Findings can be used by insurers and intermediaries to improve the efficiency of the insurance market supply chain. TABLE 1 – RESUME OF THE THREE PAPERS 6 CHAPTER I – DETERMINANTS OF CUSTOMER PRICE SENSITIVITY: AN EMPIRICAL ANALYSIS 7 ABSTRACT (1) Purpose: Consumer price sensitivity has become a major issue over the past few decades. This paper investigates the importance that insurance customers give to premiums, insurers, intermediary recommendations, and bundling strategies. The relationship between attributes and consumer price sensitivity is also studied. Methodology: To calculate the importance of attributes and part-worth utilities, we performed a Conjoint Analysis with Full Profile. To segment the market, we performed a two-stage Cluster Analysis. The traditional formula for estimating price elasticity of demand was also used. Findings: Price sensitivity is affected by the level of purchase involvement, bundled discounts, and brand loyalty. Also, brand loyalty has a strong influence on customer acceptance of bundled discounts. Price bundling increases a firm's revenues and profits. Theoretical implications: There is very little evidence regarding studies on price sensitivity in the insurance sector, mostly because, in many countries, premiums are strongly regulated. This study shows how it is possible to decrease price sensitivity. Practical implications: Insurers and intermediaries can benefit from price bundling strategies in order to increase sales and profit. Originality: The study contributes to the service marketing literature and marketing of the insurance sector by providing empirical evidence of the impact of price bundling on insurance customer sensitivity, with the use of a methodological and experimental approach. Keywords: Insurance sector, Consumer preferences, Market elasticity of demand, Price sensitivity, Bundling strategies. 8 1. INTRODUCTION In late 2007, a subprime crisis was triggered in the United States of America, creating one of the most severe financial crises. Globalization quickly brought the crisis to the European economies, creating problems in financial markets and enormous mistrust due to the uncertainty and incapacity to develop medium- to long-term action plans. In this sense, the insurance sector plays a major role in leveraging the economies of many countries, providing stability and confidence in markets (e.g., buying sovereign debt). In the specific context of the industries operating within services (e.g., utilities, healthcare, financial services, insurance, etc.), the relationship between consumers and organizations is very dynamic (Bolton and Lemon, 1999). Many factors influence the buying decisions of customers and price sensitivity in the insurance industry, including premiums (Barroso and Picón, 2012; Rai and Medha, 2013), intermediary recommendations (O’Loughlin and Szmigin, 2007; Robson and Sekhon, 2011; Brophy, 2013a), involvement (Zaichkowsky, 1988; Datta, 2003), and pricing strategies such as price bundling (Weston, 2007, as cited in Brophy, 2014b). Related to premiums, Barroso and Picón (2012) found that the price paid for insurance products is very important with regard to a Spanish insurance customer’s perception of time, money, or the effort involved in switching. Related to that, Rai and Medha (2013) found that premiums play an important role in the loyalty of insurance customers. Along these lines, the present article aims to measure the importance that Portuguese insurance customers give to premiums. Related to insurance distribution, an intermediary's recommendation plays an important role in insurance sales (see O’Loughlin and Szmigin, 2007; Robson and Sekhon, 2011; Brophy, 2013a). Because in Portugal insurance intermediaries are market leaders in terms of sales (see Table 2), this investigation analyzes the importance of intermediary recommendations in a customer's buying decision process. 9 Structure of distribution channels Non-life (%) Life (%) Intermediaries 89.7 95.3 Tied and captive agents 17.1 77 Brokers 17.6 1 Multi-brand intermediaries 54 17.3 Reinsurance 0 0 Of which: banks 16.1 76.7 Of which: postal 0 8.3 Direct Sell 9.8 4.5 Office 8.1 4.5 Internet 0.3 0 Phone 1.5 0 Others 0.5 0.2 TABLE 2: STRUCTURE OF INSURANCE DISTRIBUTION CHANNELS IN 2013 (PORTUGUESE ASSOCIATION OF INSURANCE, 2014) Concerning consumer involvement, the literature states that customers having a greater involvement with a product are less sensitive to price (see Zaichkowsky, 1988; Datta, 2003). In this context, this investigation analyzes how different levels of financial involvement (low or below average vs. high or above average) actually affect consumer price sensitivity. Another factor that affects consumer price sensitivity is loyalty. Several studies show that loyal customers are very important because they contribute to increasing corporate profits (Reichheld and Sasser, 1990; Bennett and Rundle-Theile, 2005; Rauyruen and Miller, 2007), they spend more than nonloyal customers (Russell-Bennett, McColl- Kennedy and Coote, 2007), and also because they tend to be less sensitive to price (e.g., Ramirez and Goldsmith, 2009; Yoon and Tran, 2011; Roy, 2012), with special relevance in the insurance industry (O’Loughlin and Szmigin, 2007; Robson and Sekhon, 2011; Brophy, 2011; Rai and Medha, 2013; Brophy, 2013a; Brophy, 2013b). Because the insurance sector has one of the highest churn rates (Jacada, 2008; Deloitte, 10 2012; Soeini and Rodpysh, 2012), the present paper also investigates and compares the price sensitivity of loyal customers vs. nonloyal consumers. Finally, the present paper also explores the possible benefits of implementing a pricebundling strategy in the insurance sector; specifically, combining home insurance (noncompulsory) with auto insurance (compulsory). Therefore, the main contribution of this study is to bridge a gap in the service marketing literature related to the insurance industry that is less studied, i.e., consumer price sensitivity. For example, the role played by customer loyalty behaviors or even the role of bundling strategies in consumer price sensitivity is a less-studied issue in the insurance industry. One of the reasons leading to the low number of studies that focus on this issue is the strict regulation insurance premiums (especially motor insurance) in some countries (see Cummins and Tennyson, 1992; Tennyson, 1997; Weiss, Tennyson and Regan, 2010; Derrig and Tennyson, 2011; Brophy, 2012). Regulation of premiums for automobile insurance Automobile insurance is compulsory in countries such as the United States of America, United Kingdom, Germany, France, Spain, and Portugal. This being the case, and using the words of Weiss, Tennyson, and Regan (2010): Automobile insurance is a compulsory purchase for most drivers in the United States and represents a significant expense for many. Partly because of this, many states regulate automobile insurance prices. Although there are several stated goals of automobile insurance regulation, the objective of much rate regulation is premium affordability. In this sense, regulators intend to achieve adequate automobile insurance rates, i.e., “that insurance is readily available in the market, but not so high that insurance is unaffordable to drivers” (Weiss, Tennyson, and Regan, 2010). However, it is frequent that this regulation process produces a significant adverse impact on insurance costs (see Tennyson, Weiss, and Regan, 2002; Derrig and Tennyson, 2011). But, according to 17 framing effects (Kahneman and Tversky, 1979). According to Sheikhzadeh and Elahi (2013), bundling strategies are mainly used in three situations: • As a tool for price discrimination; • As a cost-saving mechanism; and • As a means of entry deterrence. Many sectors are using bundling strategies, such as telecoms, machine tools, electronic components, chemical substances, and travel companies bundling flights, rental cars, accommodations, and events (to Johnson, Herrmann and Bauer, 1999). It is a strategy that is increasingly utilized (Dolan and Simon, 1996; Naylor and Frank, 2001). Stremersch and Tellis (2002) presented two different bundling strategies: a) Product bundling – based on the principle of products that are complementary. For example, Microsoft sells the Microsoft Office software as a bundle, including Word, Excel, and PowerPoint (Gerdeman, 2013). In the economic literature the terms frequently used are “tying strategy” or “tying arrangements” (see Ferrell and Hartline, 2005, p. 286). b) Price bundling – selling at least two products separately without integration (see also Rao and Kartono, 2009, p. 15). As mentioned by Brito and Vasconcelos (2015), bundled discounts provide purchasers with the opportunity to pay less for a bundle than the sum of the prices of the bundled products when purchased separately. Consumers are therefore faced with the choice between meeting all their requirements by buying a package at a discounted price, or purchasing items individually à la carte. 18 In this context, Guiltinan (1987) presents two different types of price bundling: i. Mixed-joint bundling – there is a reduction when at least two products are purchased simultaneously but customers do not know to which one the reduction has been applied (see also Avlonitis and Indounas, 2006; Gilbride, Guiltinan and Urbany, 2008). ii. Mixed-leader bundling – there is a reduction on a leader product's price if one customer buys another product (see also Gilbride, Guiltinan and Urbany, 2008). As pointed out by to Johnson, Herrmann and Bauer (1999), bundled discounts increase consumer willingness to recommend and repurchase intention, i.e., loyalty behaviors. According to Harris and Blair (2012), from the retailer perspective, if consumers fail to process information about a bundle discount, optimal bundle pricing may be affected. So, why in our study did we choose car insurance as a more relevant product over home insurance? According to Yadav (1994), consumers evaluate bundled products based on an anchoring and adjustment model. In practice, customers anchor their evaluations by analyzing which product is the most important, and then they adjust their preference considering the less important product(s). In the specific case of the Portuguese insurance sector, car insurance is the product most relevant to customers (APS, p.4, 2013) and it is compulsory. In this context, insurers make a great effort regarding the sale of home insurance. Similarly, Weston (2007, as cited in Brophy, 2014b) used motor and health insurance as the anchor products, and home insurance had a significant discount. Therefore, the authors argue that bundling strategies could play an important role as an integrated strategy (see O’Loughlin and Szmigin, 2005), as well as increasing sales, especially to loyal customers. As Berry (2000, as cited in O’Loughlin and Szmigin, 2005) indicated, service companies should consciously pursue distinctiveness in performing and communicating service, connect emotionally with customers and 19 internalize the brand for service providers in order to build retention and loyalty with customers. Berry also states that although the study of financial services has been more studied in the last few decades, it continues to pose challenges for marketers as an academic area of research. In this context, the authors argue that bundling strategies allow insurers and intermediaries to increase customer retention (loyalty) by increasing their satisfaction. Morwitz, Greenleaf, and Johnson (1998) analyzed the effect of prices on price perceptions and repurchase intentions.1 For other examples in this field see Brough and Chernev (2012). It is also interesting to note that consumers present different reactions between partitioned and nonpartitioned or combined prices (Guiltinan, 1987; Chakravarti, Rajan, Pallab, and Srivastava, 2002; Janiszewki and Cunha, 2004; Xia and Monroe, 2004; Bertini and Wathieu, 2008). This paper then also studies the following additional hypotheses: Hypothesis 3: Loyal customers are more sensitive to price bundling strategies than nonloyal customers. Hypothesis 4: Partitioned prices have better acceptance than combined prices. 1 Morwitz, Greenleaf, and Johnson (1998) presented products to consumers as follows: i) combined price – telephone for $82.90, including shipping and handling; ii) partitioned price – telephone for $69.95 plus $12.95 surcharge for shipping and handling. The results showed that when using partitioned price, repurchase intentions were higher and price perceptions were lower. 20 3. METHODOLOGY AND DATA 3.1. SAMPLE According to the Portuguese Association of Insurance (APS, 2013), in 2013, there were 79 insurance companies operating in Portugal, 11,180 employees, and 24,624 insurance intermediaries. The top 10 most representative brands operating in Portugal are Fidelidade-Mundial, Ocidental Vida, BES Vida, Santander Totta Seguros, BPI Vida, Império Bonança, Allianz Portugal, Açoreana, AXA Portugal, and Tranquilidade. Data were collected from 455 insurance customers (60.2% men; 39.8% women2), ages between 19 and 80 years (mean=43.79; standard deviation=12.159). A simple random sample was performed and the sample error was ±4.59% (p=q=50), with a confidence level of 95% (k=2 sigma). Analyzing the sample by age group: • 11.6% of the sample was between 18 and 29 years old; • 43.7% was between 30 and 44 years old; • 37.7% was between 45 and 64 years old; • 5.1% was between 65 and 74 years old; • 1.9% was 753 years old or more. 3.2. DATA COLLECTION The procedure for collecting data for this study encompassed two important stages as follows: • Stage 1: The information was collected through personal interviews, using an ad hoc questionnaire developed specifically for this research. Interviews took approximately 20 minutes each to be completed and they were conducted during July 2013. These data were used to test hypotheses 1, 2 and 3. 2 According to the European Commission (2004, 2011), there should be no gender discrimination in insurance pricing. 3 In Portugal, there is no age limit to buy car insurance. The unique condition is to have a driving license. 21 • Stage 2: In order to test hypothesis 4, we returned to 42 of the 455 respondents of Stage 1, asking them if they would buy a bundled product (price bundling). From those: i. We presented a bundling strategy with partitioned price to 22 individuals. ii. And a combined price to 20 other individuals. In both stages, the authors received the support of several multibrand insurance intermediaries as far as data collection was concerned. In addition, and in order to prevent any bias in data, we trained all the managers responsible for collecting data, especially concerning Conjoint Analysis. This way, (multibrand) intermediaries knew how to correctly collect data through a simulated sale with Conjoint Analysis. 3.3. ATTRIBUTES’ SELECTION In order to select the most relevant attributes for Portuguese insurance customers, we performed a pilot study based on a qualitative approach (we conducted three focus groups with both customers and intermediaries). The results obtained show that “the intermediaries’ recommendation,” “price,4” and “insurer/brand” were the most relevant attributes for Portuguese customers. 3.4. PROCEDURE A Conjoint Analysis with Full Profile (FP) was performed in order to achieve the conditions most similar to a selling environment (other investigations used the same logic, e.g., Gareth, Levin, Chakraborty, and Levin, 1990). According to Green and Srinivasan (1978), Conjoint Analysis is defined as “a decompositional method that estimates the structure of a consumer’s preferences given his/her overall evaluations of a set of alternatives that are previously specified in terms of levels of different attributes”. Conjoint analysis is a very interesting technique for evaluating and 4"Respondents were informed about covers associated with each level of the attribute premium. 22 analyzing consumer preferences regarding products or services (Varela, Picón and Braña, 2004; Dominique-Ferreira, Rial and Varela, 2012). Authors considered the possibility of using a choice-based conjoint. However, intermediaries who participated in data collection indicated that the FP option would mimic in a better way the decision-making process of customers. Also, other studies support good performance from FP predicting consumer preferences (Molin, Oppewal, and Timmermans 2000; Oppewal and Klabbers, 2003). In the specific case of pricing studies, Conjoint Analysis is one of the most popular methods in marketing for measuring willingness to purchase (Jedidi and Jaspal, 2009, p. 42). Therefore, the subjects were asked to sort the cards based on their preferences. This procedure is called Full Profile, with simulated stimuli and sort cards – sequence. The selected attributes were: 23 Attribute Levels Recommended by intermediaries • Yes • Opinion omitted Price (Premium)5 6 • 150€ - Standard product through regulation (after the decree-law no. 72/2008, April 16th) • 200€ - the same coverage as the option of 150€ and vehicle occupants insurance • 250€ - the same coverage as the option of 200€ and auto glass insurance • 300€ - the same coverage as the option of 250€ and theft coverage Brand (insurer) • Brand A (Fidelidade-Mundial) • Brand B (Açoreana) • Brand C (Allianz) • Brand D (Tranquilidade) Price bundling Home insurance with a promotional discount (for just 30€) • Yes • No TABLE 3: ATTRIBUTES AND CORRESPONDING LEVELS To achieve the Conjoint Analysis, we selected these four attributes with different levels for each (2×4×4×2). From the 64 possible combinations, we used an orthogonal fractional factorial design, selecting 16 and two holdout cards, which were eventually used in the data collection (with an Orthoplan procedure of the SPSS software). We built 18 cards, each one representing one of the 18 combinations of attribute levels. Because we performed a post hoc segmentation (see Green, 1977; Wind 1978; Picón, Varela, and Real, 2005), the Clustering Algorithm was applied to the output of the Conjoint Analysis. Therefore, we carried out a two-stage clustering, starting with a hierarchical method (Euclidean distance) and Ward’s (1963) linkage method (the most 5 Premium includes salesperson compensation (national standard) and standard claims handling costs 6 No deductible (except for the theft coverage) 24 popular method in the social sciences; see Picón, Varela, and Real, 2005, p. 430). Then we used the iterative k-means clustering, which is considered more reliable than the conventional single-stage procedures (see Picón, Varela, and Real, 2005). 3.5. METHODS AND RESULTS The study of consumer preferences was performed through Conjoint Analysis. These results are presented in Section 4.1. Market segmentation was performed through Cluster Analysis and analysis of variance (ANOVA), presented in Section 4.2. Testing of Hypotheses 1, 2, and 3 was performed using the Mann-Whitney U test, whereas testing of Hypothesis 4 was performed using Fisher’s Exact Test (due to sample size). Consequently, consumer price sensitivity is the dependent variable. In Section 4.3.2., the authors used the Variation Attributed to the Change (based on the ideal product and the anti-ideal product obtained from Conjoint Analysis results) in order to estimate the gain or loss when changing levels of attributes. This methodology (see Rial, Dominique-Ferreira and Varela, 2011; Dominique-Ferreira, Rial and Varela, 2012) consists in: i) first, “calculating the overall utility for all profiles from the most preferred option to the least preferred one; ii) next, “from the global utilities, it is necessary to estimate the gain or loss when changing a particular stimulus as a proportion of the Maximum Loss of Utility (MLU), that is, the difference between the overall utility of the ideal stimulus (the most preferred) and the anti-ideal (least preferred) one”. Finally, the traditional formula to estimate price elasticity of demand was used in Section 4.4. 25 4. RESULTS 4.1. RESULTS OF CONJOINT ANALYSIS The model fit was very high, so we can conclude that validity of the results is high (Pearson’s R=0.999; Kendall’s Tau=0.983). The most important attribute was the price, with an importance of 77.901%. The second most relevant attribute was the bundled discount with an importance of 8.496%. Recommendation had an importance of 7.523%, and brand seemed to be the least important attribute of the four (6.081%). GRAPH 1: IMPORTANCE OF ATTRIBUTES Concerning the levels of the price attribute, the preferred one was, as expected, 150€ (u=4.448). However, we would like to note that paying 50€ more, i.e., 200€ (u=1.560) presents a positive part-worth. The levels 250€ and 300€ present negative part-worths (u=−1.377 and −4.631, respectively). Bundled discounts are important for customers (u=0.495). Regarding the recommendation attribute, customers actually gave preference to products recommended by intermediaries (u=0.438). Concerning the brand attribute, Açoreana seemed to be the preferred brand (u=0.340). Fidelidade-Mundial is the only other brand that presented a positive utility (u=0.143). Allianz and Tranquilidade had negative part-worths (−0.113 and −0.369, respectively). Price Bundled discount Recommendation Brand Série1 77,901 8,496 7,523 6,081 0 20 40 60 80 100 26 4.2. CUSTOMER BUYING DECISION PROCESS The results of a two-stage cluster analysis show the existence of four clusters regarding the customer buying decision process. The following tables (Table 4, Table 5 ) show the initial and final centers of clusters. It seems that there are no important variations between both solutions. Cluster 1 2 3 4 Brand 14,80 11,70 58,36 14,83 Price 32,64 74,93 28,15 32,69 Intermediary’s recommendation 43,01 5,81 8,05 8,53 Price bundling 9,54 7,55 5,45 43,95 TABLE 4: FINAL CLUSTER CENTERS Iteration Change in Cluster Centers 1 2 3 4 1 4,584 ,898 2,639 5,161 2 3,907 ,476 ,755 1,054 3 ,892 0,000 1,523 1,088 4 0,000 0,000 0,000 0,000 TABLE 5: ITERATION HISTORY7 Nevertheless, clusters are clearly differentiated (see Table 6). Clusters 2 and 3 are the most different, mainly because of the importance given to price. Clusters 1 and 4 are the least different, mainly because they vary almost exclusively in bundling strategy. 7 Convergence achieved due to no or small change in cluster centres. The maximum absolute coordinate change for any centre is .000. The current iteration is 4. The minimum distance between initial centres is 59.351. 33 Brand Price Intermediary’s recommendation Price bundling Global Utility VAC Açoreana 150€ Yes No 13,169 Benchmark 1 Açoreana Example 1 150€ Yes Yes 14.308 +9.63% Fidelidade-Mundial Example 2 150€ Yes Yes 14,011 +7.12% Tranquilidade 200€ Hidden No 8,627 Benchmark 2 Açoreana Example 3 200€ Yes Yes 8,591 +24.17% Açoreana Example 4 250€ Yes Yes 8,591 -0.31% TABLE 13: CHARACTERISTICS OF THE SIMULATED PRODUCTS We present examples of simulated products for illustrative purposes only. The most interesting aspect of these analyses is that insurers could make some prediction of how specific customers would react to new products based on different criteria (gender, age, consumption patterns, loyalty behaviors, purchase involvement, etc.). 4.3.3. PRICE BUNDLING AND PRICE PERCEPTION Hypothesis 4: Partitioned prices have better acceptance than combined prices. To analyze the best option to present price, we used a sample of 42 customers. Our results show that 61.9% would accept the bundle option. Regarding price perception analysis, our results show that a partitioned price strategy has a little more acceptance than a combined (or nonpartitioned) price: 52.4% and 47.6%, respectively. Because our n<50, we used the Shapiro-Wilk statistic, and the P value was significant (<0.001 on 34 both), suggesting that our data are not distributed normally (see Table 14). As shown in Table 15 – using Fisher’s Exact Test – we cannot reject the null hypothesis. Statistically, Hypothesis 4 cannot be accepted. However, in terms of business strategy, these differences should not be ignored, as is clear from the results in the next section. Kolmogorov- Smirnov8 Shapiro-Wilk Statistic df Sig. Statistic df Sig. Price bundling Partitioned price ,383 22 ,000 ,628 22 ,000 Combined price ,413 20 ,000 ,608 20 ,000 TABLE 14: TESTS OF NORMALITY Chi-Square Tests Value df Asymp. Sig. (2-sided) Exact Sig. (2-sided) Exact Sig. (1-sided) Pearson Chi-Square ,1559 1 ,694 Continuity Correction10 ,006 1 ,940 Likelihood Ratio ,155 1 ,693 Fisher's Exact Test ,758 ,470 Linear-by-Linear Association ,151 1 ,697 N of Valid Cases 42 TABLE 15: FISHER’S EXACT TEST 8 Lillierfors Significance Correction 9 0 cells (0,0%) have expected count less than 5. The minimum expected count is 7.62. 10 Computed only for a 2x2 table" 35 4.4. THE EFFECT OF PRICE BUNDLING ON THE DEMAND FUNCTION In the second stage of the study, we analyzed the potential interest of a bundling strategy in the insurance sector. The 42 customers of the second sample were divided into two groups as follows: 1. Group 1 (n=24): we asked customers from this group how they would react to a bundling strategy presented by their actual insurer. 2. Group 2 (n=18): these customers were asked if they would accept a bundling strategy from another insurer. The average price paid by these customers was €30311. The bundled strategy supposes an increase of €30 over the base price, i.e., 333€12. From the 24 customers in Group 1, 18 would finally accept the bundled offer. This represents 75% acceptance. In this case, price elasticity of demand13 is 4.205, i.e., customers are sensitive to this price bundling. From the 18 customers of Group 2, eight (8) would accept the bundled offer, i.e., a 44.4% acceptance. In this case, price elasticity of demand is 4.644. In practical terms: 1. Market with the actual logic (without the price bundling): 303€×2414=7272€ 2. If customers are offered a price bundling: a. Actual customers: 333€×1815=5994€ b. Customers from other firms: 333€×816=2664€ c. And customers with previous conditions (without price bundling): 303€×617=1818€ 11 Actual price – P1 12 Price under the conditions of price bundling – P2 13 Traditional formula was used: %(/01 %(/2 % Where CQD is the percentage of change in quantity demanded and CP is the percentage of change in price 14 Actual quantity – Q1 15 Quantity accepting price bundling from one firm – Q2 16 Quantity accepting price bundling from other firms - Incorporated in Q2 17 Quantity only accepting product under actual conditions – Q3 36 Under the new conditions (price bundling), this market has a maximum potential of €10476 (RA 5994+RB 2664+RC 1818). Comparing the actual market with the new possible one (with price bundling), the loss is €4482 (10476-5994). Graphical representation offers a more intuitive understanding. FIGURE 1: DF UNDER ACTUAL CONDITIONS FIGURE 2: DF UNDER PRICE BUNDLING CONDITIONS 37 5. DISCUSSION AND CONCLUSIONS Consumer price sensitivity has been a much-discussed subject because it has a direct impact on a firm's profits as well as on consumer satisfaction and loyalty behavior. Price sensitivity is affected by many elements, such as the perceived similarity between brands (Light, 1997; Iyer and Muncy, 2005), involvement with products, brand loyalty, and bundling strategies. In this study, we only considered car insurance customers. And in this sector, it is not uncommon for Portuguese customers to forget the name of their insurer (brand). Therefore, we can say that there is a perceived similarity between brands, and this could explain why insurer brand is the less important attribute [Importance (IMP) =6.081%]. Further, intermediary recommendation is even more important for customers (Imp=7.523%), making marketing management even more difficult for insurers. In this sense, insurers should improve their support and partnership with intermediaries as mentioned by Hawksby (2015). This would benefit both – insurers and intermediaries – as a win-win solution, improving loyalty programs, e.g., through bundling strategies, which turns out to be the second most important attribute for customers (Imp=8.496%). In this study we considered a price bundling strategy, specifically, a mixed-leader bundling (Guiltinan, 1987; Brito and Vasconcelos, 2015). Results show that 61.9% (approximately two out of three customers) of our sample would accept the bundle proposed, with the main product being auto insurance and home insurance being the product that would receive the reduction. On price presentation we did not find significant differences whether the price was partitioned or whether it was combined (acceptance percentage of 52.4% and 47.6%, respectively). Our results share some similarities with those obtained in other studies (Drumwright, 1992; Mazumbar and Jun, 1993; Morwitz, Greenleaf, and Johnson, 1998). In the Portuguese insurance sector, bundling can be considered as a tool for price discrimination and as a cost-saving mechanism (see Sheikhzadeh and Elahi, 2013). Price discrimination because of this kind of bundling allows a reduction in the perceived price. It is a cost-saving mechanism because it may bring new customers and also 38 because of the indirect impact that this practice has on loyal customers. Regarding bundling as a means of entry deterrence (Nalebuff, 2004 as cited in Sheikhzadeh and Elahi, 2013), we also think that this strategy may be consistent with the reality of the Portuguese insurance sector because in the last couple of months some international players were making an effort to enter the Portuguese insurance sector. The insurance industry is currently a difficult market, i.e., higher insurance premiums; more stringent underwriting criteria, which means underwriting is more difficult; reduced capacity, which means insurance carriers write less insurance policies; and less competition among insurance carriers (PSA Insurance and Financial Services, 2013). Therefore, in this context, bundling strategies could play an important role in product and service differentiation instead of traditional price discounts. As far as price sensitivity is concerned, our results suggest that loyal customers give less importance to price when compared to nonloyal customers (Imp=75.829% and Imp=80.836%, respectively). However, we did not achieve significant statistical differences for this topic. Therefore, brands and intermediaries must be aware that, although loyal customers are “tolerant” to some price oscillation, they still give great importance to price. This may reinforce the idea that firms (insurers and intermediaries) may be more likely to be successful in increasing their revenues through bundling strategies than by simply increasing price. This result, together with the result of cluster analysis (Cluster 3 “loyal to insurers”) and with the fact that the insurance sector has one of the highest turnover rates (Jacada, 2008; Deloitte, 2012; Soeini and Rodpysh, 2012), seems to indicate that more loyalty programs could be developed. In this context, the authors suggest an increase of contractual time for loyal customers. For example, instead of the standard 12-month contract, it could be interesting to develop long-term agreements with loyal customers of 3 years with some benefits, such as more coverage, freezing the price of auto insurance during the 3 years, increase cross-selling offers (e.g., auto insurance, home insurance, and health insurance). It could also be important to treat those customers as a cluster with differentiated services. For example, having an integrated advisory service based on a dedicated online Web service platform, an account manager (as in banking), and a specialized salesperson in order to convert a standard transactional sale to a customer relationship marketing, i.e., CRM (Sheth, 2002; Baron, Warnaby, and Hunter-Jones, 2014). Customer relationship profitability 39 arises through the acquisition and retention of high-quality customers with low maintenance costs and high revenue (Anderson and Mittal, 2000). Therefore, distribution channels could play an important role knowing what strategies to apply to loyal customers (similar results were obtained by Li et al., 2012). These strategies could also be applied to customers who have a higher financial involvement because they give less importance to price and more importance to intermediary recommendation and they are much more sensitive to price bundling. Nevertheless, we suggest that insurers and intermediaries should preferably present first products with more coverage and services associated (e.g., online Web service platform, account manager) to prevent customers comparing products with low prices (in line with Gázquez-Abad and Sánchez-Pérez, 2009 work). Actually, “insurance companies face technological uncertainty that comes from how big data and analytics investments will drive revenue” (Dyer, Furr, and Lefrandt, 2014), so this online Web service platform could play an important role in customer relationship management. Cross-selling could also be a relevant strategy for these customers as a way of optimizing the wallet share of insurers and intermediaries. Our sample in the experimental stage (the above-mentioned second stage) is relatively small. In future research it would be important to have a larger sample in order to take more general conclusions, minimizing a firm's risks before any commercial program. Or it would be interesting to make a first selective approach to market based on each customer’s buying profile. Intermediaries could play an important role giving qualitative feedback on how customers react to bundling strategies and what could be other anchor products and complementary products. It would also be interesting to have the collaboration of an actuary in order to carry out more precise analysis of premium estimation of bundling strategies. In particular, it would be important to test different anchor products as well as different bundling combinations. It would also be interesting to study the ideal number of products that would compose the bundling strategy, for example, two anchor products and a price discount in the third product or different percentages of price discounts in the two associated products (e.g., home insurance and health insurance). Moreover, it could be relevant to consider life insurance products as part of a bundling strategy. It would also be interesting to study whether there is any benefit in applying 40 the bundle discount to the anchor product instead of applying it to the accessory product. Finally, and depending on the characteristics of consumers in each country, it could be important to perform a stratified random sampling. 41 CHAPTER II – HOW IMPORTANT IS THE STRATEGIC ORDER OF PRODUCTS’ ATTRIBUTES PRESENTATION IN THE INSURANCE MARKET? 42 ABSTRACT (2) Objectives: Sales management plays an important role in firms’ profit. Its main goal is to determine the best time to present insurance customers with prices, insurers, bundling strategies and the intermediary’s recommendation. Methods: We used a triangular approach. For attribute selection, three focus groups were performed with insurance customers and intermediaries. Conjoint analysis was carried out by presenting the attributes in three different orders. Results: Primacy and recency effects were detected; a transfer or anchor effect was also found related to the importance of the attributes preceding and succeeding a given attribute. Managerial implications: Salespeople can improve their approach to customers by decreasing the importance given to price and increasing the positive impact of bundling strategies and the intermediary’s recommendation in sales. Originality: Although the order of attribute presentation has previously been analyzed, this is the first study to examine this issue in non-life insurance products, providing useful information to insurance salespeople and marketing managers for a better understanding of insurance customers’ buying decision process. Keywords: Effects of the order of attribute presentation, Customer services management, Sales management, Bundling strategy, Insurance sector. 49 3. METHODOLOGY 3.1. PARTICIPANTS Data was gathered from 394 insurance customers (59.1% men; 40.9% women), aged between 19 and 80 (mean=43.92; standard deviation=12.299). The sample error was ± 4.94% (p=q=50), with a confidence level of 95% (k=2 sigma). 3.2. ATTRIBUTES’ SELECTION In order to select the most relevant attributes for Portuguese insurance customers, we performed a pilot study based on a qualitative approach (we conducted three focus groups with both customers and intermediaries). The results obtained show that “the intermediaries’ recommendation,” “premium,18 ” and “insurer/brand” were the most relevant attributes for Portuguese customers. These focus groups lasted 47-55 minutes and were performed in 2012. Literature revision has also proved that premium (Barroso and Picón, 2012; Rai and Medha, 2013), intermediary recommendations (O’Loughlin and Szmigin, 2007; Robson and Sekhon, 2011; Brophy, 2013), and pricing strategies such as price bundling (Weston, 2007, as cited in Brophy, 2014) were very important attributes. 3.3. PROCEDURE A Conjoint Analysis with Full Profile (FP) was performed in order to achieve the most similar conditions as selling environment (other investigations used the same logic, e.g., Gareth, Levin, Chakraborty and Levin, 1990). Authors considered the possibility of using choice-based conjoint. However, intermediaries who collected data indicated that the full profile option would mimic in a better way the decision-making process of customers. Also, other studies support good performance from FP predicting consumers’ preferences (Molin, Oppewal and Timmermans 2000; Oppewal and Klabbers, 2003). In the specific case of pricing studies, Conjoint Analysis is one of the 18 Respondents were informed about covers associated with each level of the attribute premium. 50 most popular methods in marketing for measuring willingness to purchase (Jedidi and Jaspal, 2009, p. 42). To perform conjoint analysis, we selected the abovementioned four attributes with different levels for each (2x4x4x2). From the 64 possible combinations, we used an orthogonal fractional factorial design, selecting 16 and two holdout cards, which were eventually used in the data collection (with the “Orthoplan” procedure of SPSS v. 21). We built 18 cards, each of which represented one of the 18 combinations of the levels of attributes. For the specific purpose of this study, we used three orders of attribute presentations called “series” and characterized as follows: SERIES A SERIES B SERIES C Insurer Bundling Insurer Price Intermediary’s recommendation Bundling Intermediary’s recommendation Insurer Intermediary’s recommendation Bundling Price Price TABLE 17: ORDERS OF ATTRIBUTES’ PRESENTATION Due to the high number of possible combinations, we chose three series. In series A, price was placed at the beginning, mainly for the purpose of verifying the results of Bagchi and Davis (2012) (that individuals tend to anchor on the first part of information for initial judgments). In series B and series C, price was placed at the end for the same reason. The only difference between series B and series C was the order of the presentation of the attributes preceding price; in series B, we sorted the other attributes based on what we believed to be the descending order of importance. In other words, we thought that bundling would be the second most important attribute, followed by the intermediary’s recommendation and finally the insurer. In series C, we placed what we thought would be the least important attribute first, followed by the other attributes. 51 3.4. METHODS AND RESULTS the study of consumer preferences was performed through conjoint analysis. In order to analyze the possible order effect on each attribute (Section 4.2.), we performed the Kruskal–Wallis test. Finally, as outlined in Section 4.3.3., we used the variation attributed to the change (based on the ideal product and the anti-ideal product obtained from the results of the conjoint analysis) in order to estimate the gain or loss when changing the order of the attributes’ presentation. 4. RESULTS 4.1. CONJOINT ANALYSIS’ RESULTS Model fit is very high, so we can conclude that validity of the results is high (Pearson’s R=0.999; Kendall’s Tau=0.983). The most important attribute is the price, with an importance of 77.901%. The second most relevant attribute is the bundled discount with an importance of 8.496%. The recommendation has an importance of 7.523% and the brand seems to be the least important attribute of the four (6.081%). GRAPH 2: IMPORTANCE OF ATTRIBUTES Concerning levels of the price attribute, the preferred level is, as expected, €150 (u=4.448). However, it should be noted that paying €50 more, i.e., €200 (u=1.560), presents a positive part-worth. The levels €250 and €300 present negative part-worths Price Bundling"strategy Recommendation Insurer Série1 77,901 8,496 7,523 6,081 0 20 40 60 80 100 52 (u=-1.377 and -4.631, respectively). In addition, bundled discounts are a good option for customers (u=0.495). Concerning the recommendation attribute, customers actually give preference to products recommended by intermediaries (u=0.438). Concerning the brand attribute, Açoreana seems to be the preferred brand (u=0.34). Fidelidade-Mundial is the only other brand that presents a positive utility (u=0.143). 4.2. RESULTS BY SERIES 4.2.1. PRICE Results show that price has a lower importance when presented at the end of the sale (UPrice B=79.359; UPrice C=72.996). Consumers give more importance to price when it is presented at the beginning (UPrice A=83.97). The difference between the highest value (UPrice A=83.97) and the lowest one (UPrice C=72.996) is 10.974%. This difference of importance (10.974%) is higher than the importance of any of the other attributes. This could be an indication of a primacy effect when price is presented at the beginning (series A). There is also a transfer effect, i.e., when price is presented at the end and is preceded by a relevant attribute (such as the intermediary’s recommendation), its importance is lower (series C). When price is also presented at the end but is preceded by the least important attribute (insurer), the transfer effect is not so strong (series B). ATTRIBUTES SERIES A SERIES B SERIES C A-B A-C B-A INSURER 3,542 5,678 8,550 2,136 5,008 2,872 PRICE 83,97 79,359 72,996 4,611 10,974 6,363 INTERMEDIARY’S 6,116 3,841 10,678 2,275 4,562 6,837 53 RECOMMENDATION BUNDLING STRATEGY 6,371 11,122 7,777 4,751 1,406 3,345 AVERAGE --- --- --- 3.433 5.488 4.854 TABLE 18: IMPORTANCE OF EACH ATTRIBUTE BY SERIES In order to check normality of data, we performed the test of normality. Results show that data are not normally distributed (p-value <0.001). So, Kruskal-Wallis test was performed and the result show that the null hypothesis should be retained (p-value = 0.374). Tests of Normality Series Kolmogorov-Smirnova Shapiro-Wilk Statistic df Sig. Statistic df Sig. Price A ,226 103 ,000 ,838 103 ,000 B ,295 45 ,000 ,663 45 ,000 C ,186 172 ,000 ,875 172 ,000 a Lilliefors Significance Correction TABLE 19: TESTS OF NORMALITY FOR PRICE Null Hypothesis Test Sig. Decision The distribution of price is the same across of categories Independent Samples Kruskal- Wallis Test 0.374 Retain the null hypothesis Asymptotic significances are displayed. The significance level is .05 TABLE 20: KRUSKAL-WALLIS TEST FOR PRICE 54 4.2.2. INSURER In the case of the insurer, there is a more complex effect. When the insurer is presented at the beginning (series A and series C), consumers give: i) little (lowest) importance in series A (UInsurer A=3.542); but also the highest importance in series C (UInsurer C=8.550%). The difference between the highest and the lowest importance is 5.008%. So, it is not possible to argument that there is a primacy effect or a recency effect (UInsurer B=5.678%). Perhaps, the reason is that in series A, the attribute insurer (UInsurer A=3.542) is immediately succeeded by price (UPrice A=83.97%) which is the most important attribute. So, this could be explained by the enormous importance of price in series A together with the primacy effect observed in price. Again, in series B it seems that the same situation occurs. In series C the situation is different because the attribute that succeeds insurer is a less important attribute (bundling strategy) compared with price. In order to check normality of data, we performed a test of normality. Results show that data are not normally distributed (p-value <0.001). In this sense, Kruskal-Wallis test was performed and the result show that it is possible to reject the null hypothesis (p-value = 0.024). Tests of Normality Series Kolmogorov-Smirnova Shapiro-Wilk Statistic df Sig. Statistic df Sig. Insurer A ,238 103 ,000 ,720 103 ,000 B ,328 45 ,000 ,561 45 ,000 C ,232 172 ,000 ,674 172 ,000 a Lilliefors Significance Correction TABLE 21: TESTS OF NORMALITY FOR INSURER 55 Null Hypothesis Test Sig. Decision The distribution of insurer is the same across of categories Independent Samples Kruskal- Wallis Test 0.024 Reject the null hypothesis Asymptotic significances are displayed. The significance level is .05 TABLE 22: KRUSKAL-WALLIS TEST FOR INSURER 4.2.3. BUNDLING STRATEGY Regarding the attribute bundling strategy, it seems that there is a primacy effect, because when this attribute is presented at the beginning, it gets its highest importance (UBundling B=11.122%). When presented in second place, it gets the second highest importance (UBundling C=7.777%). Finally, when presented at the end, the importance of bundling strategy is the lowest one (UBundling A=6.371%). The difference between the highest and the lowest value is 4.751. It seems that there is a primacy effect in series B (presented in first place) and C (presented in second place). In order to check normality of data, we performed the Shapiro-Wilk statistic. Results show that data are not normally distributed (p-value < 0.001). So, Kruskal-Wallis test was performed and the result show that the null hypothesis should be retained (p-value = 0.794). 56 Tests of Normality Series Kolmogorov- Smirnova Shapiro- Wilk Statistic df Sig. Statistic df Sig. Bundled strategy A ,195 103 ,000 ,755 103 ,000 B ,179 45 ,001 ,721 45 ,000 C ,208 172 ,000 ,735 172 ,000 a Lilliefors Significance Correction TABLE 23: TESTS OF NORMALITY FOR BUNDLED STRATEGY Null Hypothesis Test Sig. Decision The distribution of bundling strategy is the same across of categories Independent Samples Kruskal- Wallis Test 0.794 Retain the null hypothesis Asymptotic significances are displayed. The significance level is .05 TABLE 24: KRUSKAL-WALLIS TEST FOR BUNDLING STRATEGY 4.2.4. INTERMEDIARY’S RECOMMENDATION Regarding the recommendation made by the intermediary, the highest value is observed when the attribute is presented near the end (URecommendation C=10.678%; URecommendation A=6.116%). When presented near the beginning, it has the lowest importance (URecommendation B=3.841%). The difference between the highest and the lowest values is 6.837%. In this case, it seems there is not a clear effect (possibly a recency effect). In order to check normality of data, we performed the Shapiro-Wilk statistic. Results show that data are not normally distributed (p-value < 0.001). So, Kruskal-Wallis test was performed and the result show that the null hypothesis should be retained (p-value = 0.747). 57 Finally, it is interesting to note that the variability of attributes’ importance is lower when price is presented at the beginning (series A), exactly in the circumstances in which the importance of price is the highest one. Our argument is also supported because the highest variability of attributes’ importance is found in series C (precisely our recommendation in terms of salespeople approach to customers). Tests of Normality Series Kolmogorov- Smirnova Shapiro- Wilk Statistic df Sig. Statistic df Sig. Intermediary’s recommendation A ,210 103 ,000 ,759 103 ,000 B ,204 45 ,000 ,717 45 ,000 C ,239 172 ,000 ,680 172 ,000 a Lilliefors Significance Correction TABLE 25: TESTS OF NORMALITY FOR INTERMEDIARY’S RECOMMENDATIONS Null Hypothesis Test Sig. Decision The distribution of intermediary’s recommendation is the same across of categories Independent Samples Kruskal- Wallis Test 0.747 Retain the null hypothesis Asymptotic significances are displayed. The significance level is .05 TABLE 26: KRUSKAL-WALLIS TEST FOR INTERMEDIARY’S RECOMMENDATIONS 58 4.3. STATISTICAL DIFFERENCES VERSUS SIMULATION ANALYSIS According to Bakan (1966, as cited in Cohen, 1994), “a great deal of mischief has been associated” with the test of significance. In most cases, the practical reality is sidelined because “if he tried to publish this result without a significance test, one or more reviewers might complain? It could happen.” This logic seems to place the statistical implications in contradiction with the practical implications. However, in terms of the results obtained in this study, it is questionable as to whether the non-statistically significant differences are relevant in real sales situations. Table 27 shows some simulation analyses through VAC For example, “benchmark 1” shows that it is possible to increase the attractiveness of the same product by 4.74% based on the effect of the order of attribute presentation (from series A to series C). Series Brand Price Intermediary’s Recommendation Bundling Strategy Global Utility VAC A Açoreana 200$ Yes Yes 10.990 Benchmark 1 C Açoreana 200$ Yes Yes 11.540 + 4.74% (Comparing with Benchmark 1) TABLE 27: SIMULATIONS 65 Contemporary global supply and value chains involve international networks that are built upon personal inter- and intra-organizational relationships. Understanding these relationships and their link to performance, requires an exploratory approach which gets as close as possible to the actors involved. 2. THEORETICAL FRAMEWORK 2.1. SUPPLY CHAIN MANAGEMENT According to Lambert, García-Dastugue and Croxton (2005), in 1996 The Global Supply Chain Forum developed the following definition of supply chain management: the integration of key business processes from end user through original suppliers that provides products, services, and information that add value for customers and other stakeholders (see also Lambert, Cooper and Pagh, 1998, p. 1). Lambert, García- Dastugue and Croxton 2005’ work presents eight supply management processes that were included in The Global Supply Chain Forum framework, as follows: • Customer Relationship Management – the structure that shows how relationships with customers are developed and maintained; • Customer Service Management – it is considered the firm’s face to customers, a single source of information; • Demand Management – the structure for balancing the customers’ requirements with supply chain capabilities, including reducing demand variability and increasing supply chain flexibility; • Order Fulfilment – includes all activities necessary to define customer requirements, design a network, and enable the firm to meet customer requests while minimizing the total delivered cost; • Manufacturing Flow Management – comprises the activities to obtain, implement and manage manufacturing flexibility and move products through the plants in the supply chain; 66 • Supplier Relationship Management – specifies how should be developed and maintained the relationships with suppliers; • Product Development – delivers the structure for developing and bringing to market new products jointly with customers and suppliers; • Returns Management – “includes all activities related to returns, reverse logistics, gatekeeping, and avoidance”. In order to fulfill the goals of this investigation, the authors considered the following three supply management processes: a) Customer Relationship Management (CRM) provided by insurers and intermediaries – sales are generally performed by intermediaries, while claims management are performed by intermediaries (downstream) and by insurers (upstream); b) Customer Service Management – mainly provided by intermediaries (Robson and Sekhon, 2011; Brophy, 2013; Portuguese Association of Insurance, 2014); c) Demand Management – essentially provided by intermediaries. 2.2. CUSTOMERS’ SATISFACTION According to Parasuraman, Zeithaml and Berry (1993), there are some basic principles that provide customer satisfaction, and this relationship can be translated in the following way: • The Satisfaction (S) of the client, results from the difference of the set of perceptions (P) and expectations (E), i.e.: S = P - E When there is a complaint from a consumer, the more quickly the problem is solved, the more satisfied the consumer will be (Orsingher, Valentini and de Angelis, 2010). This situation is frequently called “the paradox of satisfaction”. 67 Satisfaction and loyalty are usually linked because there is a positive relationship between both concepts and realities (Kim and Yoon, 2004; Bodet, 2008; Kahn, 2012). Many studies have analyzed the link between customer satisfaction and loyalty behaviors (Bernhardt, Donthu and Kennett, 2000; Edvardsson, Johnson, Gustafson and Strandvik, 2000). According to Orsingher, Valentini and de Angelis (2010), the more satisfied a customer is, the more likely they are to become loyal to a specific brand (see also Martensen, Gronholdt and Kristensen, 2000). Many studies show that satisfaction with products/services has an important impact on the loyalty of customers (see Bitner, 1990; Dick and Basu, 1994; Oliver, 1999; Nam, Ekinci and Whyatt, 2011). In a classic sense, loyalty can be defined as the repetitive sequence of purchasing a particular brand (Cunningham, 1956), or as an individual’s desire for a brand to remain stable for a period of time (Tucker, 1964). Dick and Basu (1994) define loyalty as the relationship between an attitude towards an entity (e.g. a brand, service, or sales) and a repetitive behavior (for more about repetitive behavior, see LaBarbera and Mazursky, 1983; Taylor and Baker, 1994; Zeithaml, Berry and Parasuraman, 1996; Bolton, 1998; Rauyruen and Miller, 2007; Saaty, 2011; Guillén, Nielson, Scheike and Pérez-Marin, 2012; Khan, 2012). In practice, loyal customers refer and recommend the products and services of the brand they are loyal to to their colleagues (Mcllroy and Barnett, 2000). According to Delgado-Ballester and Munuera-Aleman (2001, apud Roy, 2012), the main advantages of customers’ loyalty are: • A substantial entry barrier to competitors. • An increase in the firm’s ability to respond to competitive threats. • Greater sales and revenues. • A customer base less sensitive to the marketing efforts of competitors. Others works (e.g.: Rowley, 2005) identified other advantages from customers’ loyalty, such as lower customer price sensitivity, reduced expenditure on attracting new customers and improved organizational profitability. However, this is not a customer loyalty guarantee. Ganesh, Arnold and Reynolds (2000) refer that: 68 i) Not all customers should be targeted with retention and loyalty efforts; ii) some of the most satisfied and loyal customers might still switch for reasons beyond the control of the firm and at times even beyond the control of the customer. Durvasula, Lysonski, Mehta and Tang (2004) also find that satisfaction is positively associated with customers’ repurchase intentions, but is weakly associated with customers’ willingness to make recommendations to others. The findings of Lin and Wu (2011) show that quality, commitment, and trust are also statistically associated with customer satisfaction and customer retention. Customer’s lifetime value According to Becker, Spann and Schulze (2014), “subscriptions with minimum contract durations do indeed help companies to successfully retain customers”. They also conclude that incentives attract customers who would not be retained in other ways, suggesting that companies should pay special attention to minimum contract durations as well as incentives. According to Homburg, Droll and Totzek (2008, firms that use customer prioritization strategies improve average customer profitability as well as the return on sales, mainly because: i) it affects relationships with top-tier customers positively but does not affect relationships with bottom-tier customers; ii) it reduces marketing and sales costs. According to Homburg, Droll and Totzek (2014), other important issues are related with: i) The ability to assess customer profitability; ii) The quality of the information gave to customers; iii) Selective organizational alignment; iv) Selective senior-level involvement; v) And selective elaboration of planning and control because they positively moderate the link between a firm's prioritization strategy and actual customer prioritization. 69 Thus, customization can be an important strategy for increasing customer satisfaction and loyalty. According to Shugan (2005), many companies operating on the Internet show a level of customization such as one-to-one customer service. Given the importance of customer satisfaction in the success of supply chain management, this paper investigates the impact of intermediaries and insurers on consumer satisfaction (in line with Rai, Patnayakuni and Seth, 2006 apud Jean, Sinkovics and Kim, 2008). 2.3. INSURANCE DISTRIBUTION Intermediaries play a very important role in the insurance sector. Their recommendation has a significant impact on customers buying process, satisfaction and loyalty. Regarding these last two concepts, insurance sector has some particularities. Usually loyal behaviors exist between intermediaries and customers (B2C), and not between customers and insurers. Eventually this leads to high churn rates in the insurance sector (see Short, Graefe and Schoen, 2003). In this context, the fact that keeping an existing customer can cost six times less than bringing a new customer (Rosenberg and Czepiel, 1984) represents a major motivation for business organizations to retain customers. According to The Council of Insurance Agents & Brokers (TCIAB, 2015), insurance intermediaries facilitate the placement and purchase of insurance, and provide services to insurance companies and consumers that complement the insurance placement process. According to Eckardt and Rathke-Doppner (2010): The profound information asymmetries between consumers and insurance companies have resulted in the evolution of institutions that mediate between consumers and insurance companies. Insurance intermediaries such as exclusive agents and insurance brokers hold an important position as matchmakers between the supply and demand sides on insurance markets. 70 Rose (1999) presents different cost reductions from intermediaries’ service, such as search costs, information costs, opportunity costs (for more specific understanding, see Table 28). Transaction stages Intermediary service Cost reduction Searching and matching • Direct sales of information • Matchmaking • Market-making • Search costs • Information costs • Opportunity costs of time Availability or products and immediacy • Compensation of variances in demand and supply • Opportunity costs of time Negotiating and contracting • Strong bargaining position • Exploitation of differences in contract terms between supply and demand market side • To standardize contracts • Negotiation costs • Information costs • Administrative costs • Opportunity cost of time Monitoring and guaranteeing • Expertise in determining product and service quality • Cross-sectional and temporal reuse of information • Guaranteeing high product quality • Information costs • Monitoring and control costs • Costs resulting from uncertainty • Investment in expertise TABLE 28: TRANSACTION COST REDUCTION FROM INTERMEDIARIES (ROSE, 1999) There are different channels of insurance products distribution, such as intermediaries (e.g.: multibrand intermediaries, brokers, reinsurance, banks, postal, etc.) and direct sell 71 (insurers). The intermediary's recommendation plays an important role in insurance sales (see Doney and Cannon, 1997; Jap, 2000; Beloucif, Donaldson and Kazanci, 2004; Chang, 2006; O’Loughlin and Szmigin, 2007; Robson and Sekhon, 2011; Brophy, 2013). Because insurance intermediaries are the main insurance customer–firm interaction touch point in Portugal, and because they are market leaders in terms of sales (see Table 29), this investigation analyzes the relevance of insurers and intermediaries on the consumer experience (Jean, Sinkovics and Kim, 2008). This issue is particularly important in the insurance industry if we consider that “when performance levels and service offering become too similar within an industry, price is the the only competitive weapon that remains” (see Bolton, Gustafsson, McColl-Kennedy, Sirianni and Tse, 2014). Structure of distribution channels Non-life (%) Life (%) Intermediaries 89.7 95.3 Tied and captive agents 17.1 77 Brokers 17.6 1 Multi-brand intermediaries 54 17.3 Reinsurance 0 0 Of which: banks 16.1 76.7 Of which: postal 0 8.3 Direct Sell 9.8 4.5 Office 8.1 4.5 Internet 0.3 0 Phone 1.5 0 Others 0.5 0.2 TABLE 29: STRUCTURE OF INSURANCE DISTRIBUTION CHANNELS IN 2013 (“ASSOCIAÇÃO PORTUGUESA DE SEGUROS”, I.E., PORTUGUESE ASSOCIATION OF INSURANCE, 2014) 72 STUDY 1 3. METHODOLOGY 3.1. SAMPLE Data was collected from 366 consumers of car insurance. 60.1% were men and 39.9% women (age: mean=43.79; standard deviation=12.209; minimum=19; maximum=80). The sample error is ± 5.12 (p=q=50), with a level of confidence of 95% (k=2 sigma). 3.2. DATA COLLECTION In order to respond to the scientific and managerial challenge of insurer y (development of a simple tool that could be quickly applied to study customer satisfaction), the procedure for collecting data for this study encompassed two important stages: • Stage 1: In order to identify the most important attributes, authors carried out three focus groups. Two of these focus groups were conducted in the B2C market (customers) and one in the B2B (intermediaries). Selected attributes were: o Number of insurance proposals presented by intermediaries; o Intermediaries’ recommendation was explained; o Problems’ resolution by intermediaries; o Problems’ resolution by insurers; o Insurers’ fast response to customers’ problems; o Online services from insurers; o Contact made by insurers to know if there was any problem; o Payment facilities; o Quality of services from insurers; o Quality of services from intermediaries. 73 • Stage 2: Data were collected through personal interviews, using an ad hoc questionnaire developed specifically for this research. Interviews took approximately 20 minutes each to be completed and they were conducted during July 2013. 4. RESULTS 4.1. DESCRIPTIVE ANALYSIS As can be seen from Table 30, the higher results are mostly related with intermediaries. For example: i) the services’ quality provided by intermediaries is evaluated with a 4.12, while insurers with a 3.88; ii) resolution of problems by intermediaries gets 4, while insurers get 3.77. These results are even more relevant if we consider the fact that expectations are higher for intermediaries than for insurers (3.42 and 3.19, respectively). It is also important to highlight the level of satisfaction with intermediaries and insurers, as intermediaries have a higher rating for this attribute (4.38 and 4.01, respectively). The only negative evaluation is related to contact made by insurers (2.88). Finally, it is important to note that two elements related to intermediaries present a high standard deviation: the explanation of intermediaries’ recommendations, and the number of products presented to customers (1.012 and .988, respectively). This may mean that there are many customers who are unsatisfied with the performance of these attributes. 74 Code Items Mean Standard Deviation Bro_1 Number of insurance proposals presented by intermediaries 3,57 0,988 Bro_2 Intermediaries’ recommendation was explained 3,88 1,012 Bro_3 Problems’ resolution by intermediaries 4 0,939 Ins_1 Problems’ resolution by insurers 3,77 0,876 Ins_2 Insurers’ fast response to customers’ problems 3,71 0,848 Ins_3 Online services from insurers 3,24 0,881 Ins_4 Contact made by insurers to know if there was any problem 2,88 0,872 Ins_5 Payment facilities 3,8 0,951 Qual_i Quality of services from insurers 3,88 0,593 Qual_b Quality of services from intermediaries 4,12 0,772 Exp_i Expectation-experience with your insurer 3,19 0,516 Exp_b Expectation-experience with your intermediary 3,42 0,699 Sat_i Satisfaction with insurers 4,01 0,719 Sat_b Satisfaction with intermediaries 4,38 0,785 TABLE 30: DESCRIPTIVE ANALYSIS 4.2. MEASUREMENT MODEL The first step was to perform an exploratory factor analysis in order to understand how our data was composed in terms of dimensions. We used the maximum likelihood (ML) method for extraction and PROMAX for rotation. After analyzing communalities, we removed items with values lower than 0.3 (“ins_3”, “ins_4”, “ins_5” and “exp_i”). Our model was then able to explain 61.871% of the data. 81 FIGURE 3: CAUSAL MODEL 82 STUDY 2 5. METHODOLOGY 5.1. SAMPLE Data was collected from 433 insurance customers (59.7% men; 40.3% women), aged between 20 and 80 years old (Mean=43.95; Standard Deviation=12.195). The sample error was ± 4.71% (p=q=50), with confidence level of 95% (k=2 sigma). 5.2. ATTRIBUTES’ SELECTION In order to simplify the respondents’ task, seven attributes were finally selected: o Online services from insurers; o Contact made by insurers to know if there was any problem; o Promotion; o Insurer; o Intermediaries’ recommendation; o Premium (price); o Resolution problem. 5.3. PROCEDURE Data were gathered based on a rectangular matrix of 433 x 7 (subjects x attributes). Subjects were asked to order seven attributes based on their preferences. We used ALSCAL during this procedure because it incorporates individual differences in multidimensional scaling models with the multidimensional unfolding (MDU) model. ALSCAL uses the alternating least squares approach (see Takane, Young and de Leeuw, 1977). 83 6. RESULTS Model’s fit is good (Stress =0.077; RSQ =0.994). Figure 4 shows the bidimensional solution. Attributes such as “online services”, “insurers’ contacts” and “product promotion” are related, and insurers are very close to those attributes. Interestingly, intermediaries’ recommendation appears between “insurers” and “premium” resembling its role of intermediation. Also, “the resolution of problems” is closer to “intermediaries” than to insurers. If the figure is carefully analyzed, the most part of customers gives more importance to: i) premium; ii) resolution of problems; iii) intermediaries’ recommendation. FIGURE 4: PERCEPTUAL MAP FROM MULTIDIMENSIONAL UNFOLDING (MDU) 84 Through the figure 6 it is possible to observe in more detail how customers’ preferences are structured exactly as the main dynamics of the insurance market. More specifically, the the dynamics of the insurance supply chain. Concretely: • Dimension 1 (vertical) shows premium as being communicated by intermediaries but estimated by insurers (highlighted in green); • On the one hand, dimension 2 (horizontal) indicates that the resolution of problems/claims in the insurance business is closer to the intermediaries, albeit it always depends on the insurers (e.g.: insurance expertise). On the other hand, online services used for claims management are provided by insurers (highlighted in orange). FIGURE 5: CUSTOMERS’ PREFERENCES AND THE DYNAMICS OF THE INSURANCE MARKET 85 7. CONCLUSIONS AND DISCUSSION The insurance sector has some unique characteristics and functionalities. Generally, insurance customers buy products from intermediaries, and only in a few cases directly from insurers. Intermediaries have the majority of the market share, making them a key player in the insurance market (distribution). Qualitative results from the three focus groups (with intermediaries and customers) have allowed us to identify 14 items that most affect insurance customer satisfaction. After obtaining quantitative data, an exploratory factor analysis was performed, identifying three constructs/factors: a) one was related to the time taken by insurers in order to solve customers’ problems (claims management) in an appropriate manner; b) another one related to the actions that customers value most in intermediaries; c) and another one related to the “overall satisfaction” of customers. Of those 14 items, four did not fit well in the model. Thus, the final model had three elements associated with intermediaries, two elements associated with insurers and five elements related with the expectation–experience and “overall satisfaction” of customers. The final model was able to explain 55% of the data. The most interesting result indicates that customers’ satisfaction associated to insurers is not statistically significant in customers’ global satisfaction. This can be explained because customers are loyal to intermediaries and not so much with insurers (Chang, 2006; O’Loughlin and Szmigin, 2007; Eckardt and Rathke-Doppner, 2010; Robson and Sekhon, 2011; Brophy, 2013). Regarding the preferences of insurance customers, purchase decision is mainly based in three characteristics: a) the ability to solve problems (claims management); b) the premium; c) and the recommendation of intermediaries. Interestingly, the majority of the customers surveyed seems to associate more intermediaries to the resolution of problems (claims management) and premium negotiation than to insurers. There are just a few respondents who give greater importance to insurers than to intermediaries. Finally, it is also interesting to find that customers preferences present similarities with the insurance supply chain. Concretely, it is possible to compare two important processes. One is related with premium and it seems that the recommendation of 86 intermediaries plays a central role in premium acceptance from customers (in line with O’Loughlin and Szmigin, 2007; Eckardt and Rathke-Doppner, 2010; Robson and Sekhon, 2011; Brophy, 2013), although premium is estimated by insurers. In this sense, insurers should develop a closer partnership with intermediaries. In this context, there are specific marketing strategies associated with the level of customers’ financial involvement. For example, insurers and intermediaries should exchange more information about customers’ preferences; or it should be presented products with more coverage and services associated (such as online Web service platform, account manager) for customers who buy more expensive products. This way, it would be possible to prevent customers comparing products with low prices (in line with Gázquez-Abad and Sánchez-Pérez, 2009) and increase the attractiveness of products. The other process is related to the resolution of problems (claims management). In order to improve partnerships between insurers and intermediaries, it is important to improve the online claims management services provided by insurers. Currently, intermediaries spend too many resources (e.g.: time and money) in managing customers’ claims. This way, customers would more more satisfied with insurers as well as with intermediaries. Also, intermediaries would certainly prefer insurers that provide better claim management services. Because of the crucial importance of these service employees (e.g.: intermediaries), better training should be provided (see Bateson, Wirtz, Burke and Vaughan, 2014). Intermediaries (as frontline employees) can play a more important role in insurance customers experience if they have better back office services provided by insurers (see Bolton et al., 2014). Limitations and further research The questionnaire developed in this study has already received a positive feedback from two insurer operating in the Portuguese market because it easy to administrate through telephonic inquiry and results are very important in terms of services supply management. However, the questionnaire can still be improved. In this study it was not developed nor tested an IT that could improve the supply chain related to claims management. In this sense, we suggest carrying out focus groups 87 (composed by insurers, intermediaries, customers and IT developers and programmers) in order to develop an online tool that would improve claims management. 88 CHAPTER IV – CONCLUSIONS, LIMITATIONS AND FURTHER RESEARCH 89 GENERAL CONCLUSIONS The main purpose of this study was to analyze the insurance market from a marketing strategy perspective. Much of the literature on the insurance industry focuses on actuarial models, risk management, regulation. So, this study investigated three critical topics: 1. The determinants of consumer price sensitivity through four important hypotheses; 2. The importance of the strategic order of products’ attributes presentation in the insurance market; 3. The individual relevance of insurers and intermediaries on customers’ satisfaction and preferences. The first topic was analyzed based on four hypotheses. Hypothesis 1 stated that customers with a higher financial involvement with products were less price sensitive. Results confirmed this hypothesis. Hypothesis 2 stated that loyal customers were less price sensitive. Statically, this hypothesis was not confirmed but findings show some differences that cannot be neglected in terms of marketing strategies in the insurance sector such as loyal programs. Hypothesis 3 affirmed that loyal customers were more sensitive to price bundling strategies than nonloyal customers. Findings confirmed this hypothesis, so bundling strategies could play an important role in product and service differentiation instead of traditional price discounts. Considering the combined result of hypotheses 3 and 4, brands and intermediaries must be aware that, although loyal customers can accept some degree of price oscillation, they still give great importance to price. This may indicate that both insurers and intermediaries may be more likely to be successful in increasing their revenues through bundling strategies than by simply increasing price. Finally, hypothesis 4 indicated that partitioned prices had better acceptance than combined prices. Therefore, salespeople should pay special attention to these results when they use price bundling strategies. 90 In the second topic - the importance of the strategic order of products’ attributes presentation in the insurance market – some interesting results were also obtained. For instance, when price is presented at the end and is preceded by a relevant attribute (such as the intermediary’s recommendation), its importance is lower. Concerning the use of price bundling, a primacy effect was detected, because when this attribute is presented at the beginning, it gets its highest importance. Finally, the third topic - The individual relevance of insurers and intermediaries on customers’ satisfaction and preferences – important results were obtained. Only intermediaries have a statistic impact on consumers’ satisfaction. This results do not mean that insurers have no impact on consumers’ satisfaction, but it is the intermediaries that usually manage customers’ claims. This finding seems to be confirmed through the similarities that exist between consumers’ preferences and the insurance supply chain. THEORETICAL CONTRIBUTIONS This study addresses some important gaps in the insurance services marketing literature. Concretely, chapter I analyzes the specific importance of key characteristics (attributes) in the insurance sector. It also analyzes some determinants of consumers’ price sensitivity in the insurance market, as well as some determinants of price bundling strategies. This studies demonstrates that consumers’ financial involvement has a statistical impact on price sensitivity and that loyalty behaviors increase consumers’ price bundling acceptance. Finally, price elasticity of demand is elastic, customers are sensitive to this price bundling. In chapter II, the order of attributes presentation is studied in order to cover a concrete gap in the literature: how important is the strategic order of products’ attributes presentation in the insurance business. Primacy, recency and transfer effects were detected. In chapter III, the main aim was to try to understand if insurance customers’ satisfaction is more affected by insurers or intermediaries, and understand insurance customers’ 97 Goldsmith, R.E. and Hofacker, F. (1991), “Measuring consumer innovativeness”, Journal of the Academy of Marketing Science, Vol. 19, No. 3, pp. 209-221. Goldsmith, R.E. and Newell, S.J. (1997), “Innovativeness and price sensitivity: Managerial, theoretical and methodological issues”, Journal of Product and Brand Management, Vol. 6, No. 3, pp. 163-173. Goodman, P.S., Fichman, M., Lerch, F. and Snyder, P.R. (1995), “Customer-firm relationships, involvement, and customer satisfaction”, Academy of Management Journal, Vol. 38, No. 5, pp. 1310-1324. Green, P. (1977), “A new approach to market segmentation”, Business Horizons, Vol. 20, pp. 205-220. Green, P. and Srinivasan, (1978), “Conjoint Analysis in consumer research: Issues and outlook”, Journal of Consumer Research, Vol. 5, pp. 103-123. Guiltinan, J.P. (1987), “The Price Bundling of Services: A Normative Framework”, Journal of Marketing, Vol. 51, No. April, pp. 74–85. Hawksby, D. (2015), “Hard/soft market”, Talbot Validus Group, accessed 27 April, 2015, http://www.willis.com/documents/energy/David%20Hawksby.pdf Hupfer, N. and Gardner, D. (1971), “Differential involvement with products and issues: An exploratory study", paper presented at the Association for Consumer Research, Gardner D. (Ed.). MD: Association for Consumer Research, pp. 262-269. Ibrahim, H. and Najjar, F. (2008), “Relationship bonding tactics, personality traits, relationship quality and customer loyalty: Behavioral sequence in retail environment”. The Icfai University Journal of Services Marketing, Vol. 6, No. 4, pp. 1-37 Iyer, R. and Muncy, J.A. (2005), “The role of brand parity in developing loyal customers”, Journal of Advertising Research, Vol. 45, No. 2, pp. 222-228. Jacada (2008), “Insurance providers: Improving customer retention through the contact center”, White Papers, accessed March 19, 2015, http://www.jacada.com/images/WhitePapers/pdfs/45.100.1108-Insurance-Providers- Improving-Customer-Retention-through-the-Contact-Center.pdf 98 Jacoby, J. (1975), “A brand loyalty concept: comments on a comment”, Journal of Marketing, Vol. 12, No. November, pp. 484-487. Janiszewski, C. and Cunha, M. (2004), “The influence of price discount framing on the evaluation of a product bundle”, The Journal of Consumer Research, Vol. 30, No. 4, pp. 534-546. Jedidi, K. and Jaspal, S. (2009), “Willingness to pay: measurement and managerial implications”, in Rao, V. (Ed.) Handbook of Pricing Research in Marketing, Edward Elgar Publishing, Northampton, pp. 37-60. Johnson, M.D., Herrmann, A. and Bauer, H. (1999), “The effects of price bundling on consumer evaluations of product offerings”, International Journal of Research in Marketing, Vol. 16, No. 2, pp. 129-142. Kahneman, D. and Tversky, A. (1979), “Prospect theory: An analysis of decision under risk”, Econometrica, Vol. 47, pp. 263-291. Krishnamurthi, L. and Raj, S.P. (1991), “An empirical analysis of the relationship between brand loyalty and consumer price elasticity”, Marketing Science, Vol. 10, No. 2, pp. 172-183. Krugman, H. (1977), “Memory without recall, exposure without perception”, Journal of Advertising Research, Vol. 17, pp. 7-12. Li, M-L, Gree, R., Farazmand, F. and Grodzki, E. (2012), Customer loyalty: influences on three types of retail stores’ shoppers”, International Journal of Management and Marketing Research, Vol. 5, No. 1, pp. 1-19. Light, L. (1997), “Brand loyalty management: The basis for enduring profitable growth”, Direct Marketing, Vol. 59, No. 11, pp. 36-43. Mazumbar, T. and Jun, S. (1993), “Consumer evaluations of multiple versus single price change”, Journal of Consumer Research, Vol. 20, pp. 441-450. Molin, E., Oppewal, H. and Timmermans, H. (2000), “A comparison of Full Profile and hierarchical information integration conjoint methods to modeling group preferences”, Marketing Letters, Vol. 11, No. 2, pp. 165-175. 99 Morwitz, V.G., Greenleaf, E. and Johnson, E. (1998), “Divide and prosper: Consumers’ reactions to partitioned prices”, Journal of Marketing Research, Vol. 35, No. November, pp. 453-463. Naylor, G. and Frank, E. (2001), “The effect of price bundling on consumer perceptions of value”, Journal of Services Marketing, Vol. 15, No. 4, pp. 270-281. O’Loughlin, D. and Szmigin, I. (2005), “Customer perspectives on the role and importance of branding in Irish retail financial services”, International Journal of Bank Marketing, Vol. 21, No. 1, pp. 8-27. O’Loughlin, D. and Szmigin, I. (2007), “Services branding: Revealing the rhetoric within retail banking”, The Services Industries Journal, Vol. 27, No. 4, pp. 435-452. Oppewal, H. and Klabbers, M. (2003), “Compromising between information completeness and task simplicity: A comparison of self-explicated, hierarchical information integration, and Full-Profile Conjoint methods”, Advances in Consumer Research, Vol. 30, No. 1, pp. 298-304. Picón, E., Varela, J. and Real, E. (2005), “Classificación y segmentación post hoc mediante el análisis de conglomerados”, in Lévy Mangin, J-P. and Varela, J. (Eds), Análisis Multivariable para las Ciencias Sociales, Pearson Prentice Hall, Madrid, pp. 419-450. PSA Insurance & Financial Services (2013), “PSA Perspective: Hard market vs. soft market: The insurance industry’s cycle and why we’re currently in a hard market”, accessed April 27, 2015, http://www.psafinancial.com/2013/01/hard-market-vs-soft- market-the-insurance-industrys-cycle-and-why-were-currently-in-a-hard-market/ Rai, A. and Medha, S. (2013), “The antecedents of customer loyalty: An empirical investigation in life insurance context”, Journal of Competitiveness, Vol. 5, No. 2, pp. 139-163. Rao, V. and Kartono, B. (2009), “Pricing objectives and strategies: A cross-country survey”, in Rao, V. (Ed.) Handbook of Pricing Research in Marketing, Edward Elgar Publishing, Northampton, pp. 9-36. 100 Raju, J. (1992), “The effect of price promotions on variability in product category sales”, Marketing Science, Vol. 11, No. 3, pp. 207-220. Ramirez, E. and Goldsmith, R. E. (2009), “Some antecedents of brand loyalty”, Journal of Marketing Theory and Practice, Vol. 17, No 3, pp. 199-213. Rauyruen, P., and Miller, K.E. (2007), “Relationship quality as a predictor of B2B customer loyalty”, Journal of Business Research, Vol. 60, pp. 21-31. Reichheld, F.F. and Sasser, W.E. (1990), “Zero defections: Quality comes to services”, Harvard Business Review, Vol. 68, No. 5, pp. 105-111. Reichheld, F. and Teal, T. (1996), “The loyalty effect: The hidden force behind growth, profits, and lasting value”. Boston: Harvard Business School Press. Rial, A., Dominique-Ferreira, S. and Varela, J. (2011), “Conjoint Analysis: A case study of the consumers’ preferences”, Portuguese Journal of Marketing, Vol. 26, pp. 37. Richins, M.L., Bloch, P. H. and McQuarrie, E. (1992), “How enduring and situational involvement combine to create involvement responses”, Journal of Consumer Psychology, Vol. 1, pp. 143–153. Robson, J. and Sekhon, Y. (2011), “Addressing the research needs of the insurance sector”, International Journal of Bank Marketing, Vol. 29, No. 7, pp. 512-516. Rowley, J. (2005), “The four Cs of customer loyalty”, Marketing Intelligence and Planning, Vol. 23, No. 6, pp. 574-581. Roy, S. (2012), “Brand loyalty in insurance companies”, Journal of Economic Development, Management, IT, Finance and Marketing, Vol. 4, No 1, pp. 12-26. Russell-Bennett, R., McColl-Kennedy, J. and Coote, L. (2007), Involvement, satisfaction, and brand loyalty in a small business services setting”, Journal of Business Research, Vol. 60, pp. 1253-1260. Sheikhzadeh, M. and Elahi, E. (2013), “Product bundling: Impacts of product heterogeneity and risk considerations”, International Journal Production Economics, Vol. 144, pp. 209-222. 101 Sheth, J. (2002), "The future of relationship marketing", Journal of Services Marketing, Vol. 16, No. 7, pp. 590-592. Shoemaker, R., Mitra, D., Chen, Y. and Essegaier, S. (2003), “A comment on price endings when prices signal quality”, Management Science, Vol. 49, No. 12, pp. 1753-1758. Soeini, R. and Rodpysh, K. (2012), “Applying data mining to insurance customer churn management”, International Proceedings of Computer Science & Information Tech, Vol. 30, pp. 82-92. Stremersch, S. and Tellis, G. (2002), “Strategic Bundling of Products and Prices: A New Synthesis for Marketing”, Journal of Marketing, Vol. 66, No. 1, pp. 55-72. Tennyson, S. (1997), “The impact of rate regulation on state automobile insurance markets”, Journal of Insurance Regulation, Vol. 15, pp. 502-523. Thaler, R. (1985), “Mental accounting and consumer choice”. Marketing Science, Vol. 3, pp. 199-214. Tung, W., Capella, P. and Tat, P. (1997), “Service pricing: A multi-step synthetic approach”, Journal of Services Marketing, Vol. 11, No. 1, pp. 53-65. Varela, J., Picón, E. and Braña, T. (2004), “Segmentation of the Spanish domestic tourism market”, Psicothema, Vol. 16 No. 1, pp. 76-83. Ward, J. (1963), "Hierarchical Grouping to Optimize an Objective Function", Journal of the American Statistical Association, Vol. 58, pp. 236–244. Weiss, M., Tennyson, S. and Regan, L. (2010), “The effects of regulated premium subsidies on insurance costs: An empirical analysis of automobile insurance”, The Journal of Risk and Insurance, Vol. 77, No. 3, pp. 597-624. Wind, Y. (1978), “Issues and advances in segmentation research”, Journal of Marketing Research, Vol. 15, pp. 317-337. Xia, L. and Monroe, K. (2004), “Price partitioned on the internet”, Journal of Interactive Marketing, Vol. 18, No. 4, pp. 63-73. 102 Yadav, M. (1994), “How buyers evaluate product bundles: A model of anchoring and adjustment”, Journal of Consumer Research, Vol. 21, No. September, pp. 342-353. Yoon, K. and Tran, T. (2011), “Revisiting the relationship between consumer loyalty and price sensitivity: the moderating role of deal-proneness”, Journal of Marketing Theory and Practice, Vol. 19, No. 3, pp. 293-306. Yu, Y-T. and Dean, A. (2001), “The contribution of emotional satisfaction to consumer loyalty”, International Journal of Service Industry Management, Vol. 12, No. 3, pp. 234-250. Zaichkowsky, J. (1988), “Involvement and the price cue”, Advances in Consumer Research, Vol. 15, pp. 323-327. 103 REFERENCES (2) Associação Portuguesa de Seguros (APS, 2013), Seguros em Portugal: Panorama do mercado segurador 12/1. Associação Portuguesa de Seguros, Lisboa. Acito, F. (1977), “An investigation of some data collection issues in Conjoint Measurement”. American Marketing Association Educator’s Proceedings, pp. 82-85. Adams, W. and Yellen, J. (1976), “Commodity bundling and the burden of monopoly”, Quarterly Journal of Economics, Vol. 90, No. August, pp. 475-498. Atkinson, R. and Shiffrin, R. (1968), “Human memory: A proposed system and its control processes”. In: K.W. Spence and J.T. Spence (ed) The Psychology of Learning and Motivation: Advances in Research and Theory. Academic Press, New York, pp. 89-195 Bagchi, R. and Davis, D. (2012), “$29 for 70 items or 70 items for $29? How presentation order affects package perceptions”, Journal of Consumer Research, Vol. 39, No. June, pp. 62-72. Bagchi, R. and Li, X. (2011), “Illusionary progress in loyalty programs: Magnitudes, reward distances, and step-size ambiguity”, Journal of Consumer Research, Vol. 37, No. February, pp. 888-901. Barroso, C. and Picón, A. (2012), “Multi-dimensional analysis of perceived switching costs”, Industrial Marketing Management, Vol. 41, No 3, pp. 531–543. Brito, D. and Vasconcelos, H. (2015), “Inter-firm bundling and vertical product differentiation”, Scandinavian Journal of Economics, Vol. 117, No. 1, pp. 1-27. Brophy, R. (2013), “Adding value to insurance products: The AXA Irish experience”, Journal of Product & Brand Management, Vol. 22, No. 4, pp. 293-299. Brophy R, (2014), “Rebrand in crisis: How Liberty came to Ireland”, Journal of Strategic Marketing, Vol. 22, No. 2, pp. 93-103. 104 Buda, R. and Zhang, Y. (2000), “Consumer product evaluation: The interactive effect of message framing, presentation order, and source credibility”, Journal of Product & Brand Management, Vol. 9, No. 4, pp. 229-242. Cacioppo, J. and Petty, R. (1982), “The need for cognition”, Journal of Personality and Social Psychology, Vol. 42, No. 1, pp. 116-131. Chakravarti, D., Rajan, K., Pallab, P. and Srivastava, J. (2002), “Partitioned presentation of multi-component bundle prices: Evaluation, choice, and underlying processing effects”, Journal of Consumer Psychology, Vol. 12, No. 3, pp. 215-229. Chrzan, K. (1994), “Three kinds of order effect in choice-based conjoint analysis”, Marketing Letters, Vol. 5, pp. 165-172. Cohen, J. (1994), “The earth is round (p < .05)”, American Psychologist, Vol. 49, No 12, pp. 997-1003. Dhar, R., Nowlis, S. and Sherman, S. (1999), “Trying hard or hardly trying: An analysis of context effects in choice”, Journal of Consumer Psychology, Vol. 9, No. 4, pp. 189-200. Dhar, R. and Simonson, I. (1992), “The effects of the focus of comparison on consumer preference”, Journal of Marketing Research, Vol. 29, No. November, pp. 430- 440. Diehl, K. and Zauberman, G. (2005), “Searching ordered sets: Evaluations from sequences under search”, Journal of Consumer Research, Vol. 31, No. 4, pp. 824-832. DeMoranville, C. and Bienstock, C. (2003), “Question Order Effects in Measuring Service Quality”, International Journal of Research in Marketing, Vol. 20, No. 3, pp. 217-231. DeSarbo, W. and Green, P. (1984), “Concepts, theory and techniques, Choiceconstrained conjoint analysis”, Decision Sciences, Vol. 15, pp. 291-323. Drumwright, M. (1992), “A demonstration of anomalies in evaluations of bundling”, Marketing Letters, Vol. 3, No. 4, pp. 311-321. 105 Eckardt, M. and Rathke-Doppner, S. (2010), “The quality of insurance intermediary services – empirical evidence for Germany”, The Journal of Risk and Insurance, Vol. 77, No. 3, pp. 667-701. Fishbein, M. and Ajzen, I. (1975), “Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research”. Addison-Wesley, Reading. Gareth, G., Levin, I., Chakraborty, G. and Levin, A. (1990), “Consumer evaluation of multi-product bundles: An information integration analysis”, Marketing Letters, Vol. 2, No. 1, pp. 47-57. Gatzert, N., Huber, C. and Schmeiser, H. (2010), “How do price presentation effects influence consumer choice? The case of investment guarantees in unit-linked life insurance products”. Working papers on risk management and insurance, Vol. 82, pp. 1-38. Gilbride, T., Guiltinan, J. and Urbany, J. (2008), “Framing effects in mixed price bundling”, Marketing Letters, Vol. 19, pp. 125-139. Guiltinan, J. (1987), “The price bundling of services: A normative framework”, Journal of Marketing, Vol. 51, No. April, pp. 74–85. Hamilton, R. and Srivastava, J. (2008), “When 2+2 is not the same as 1+3: Variations in price sensitivity across components of partitioned prices”, Journal of Marketing Research, Vol. 45, No. 4, pp. 450-461. Harris, J. and Blair, E. (2012), “Consumer processing of bundled prices: When do discounts matter?”, Journal of Product and Brand Management, Vol. 21, No. 3, pp. 205-214. Hogarth, R.M. and Einhorn, H.J. (1992), “Order effects in belief updating the belief adjustment model”, Cognitive Psychology, January, pp. 1-55. Hsee, C. and Zhang, J. (2004), “Distinction bias: Misprediction and mischoice due to joint evaluation”, Journal of Personality and Social Psychology, Vol. 86, No. 5, pp. 680-695. 106 Jedidi, K. and Jaspal, S. (2009), “Willingness to pay: measurement and managerial implications”. In: Rao, V. (Ed.) Handbook of Pricing Research in Marketing, Edward Elgar Publishing, Northampton, pp. 37-60. Johnson, R. (1987), “Adaptive Conjoint Analysis Sawtooth Software Conference on Perceptual Mapping, Conjoint Analysis, and Computer Interviewing”. Sawtooth Software, Ketcham. Johnson, M., Herrmann, A. and Bauer, H. (1999), “The effects of price bundling on consumer evaluations of product offerings”, International Journal of Research in Marketing, Vol. 16, No. 2, pp. 129-142. Li, C. (2009), “Primacy effect or recency effect? A long-term memory test of Super Bowl commercials”, Journal of Consumer Behavior, Vol. 9, No. January- February, pp. 32-44. Molin, E., Oppewal, H. and Timmermans, H. (2000), “A comparison of Full Profile and Hierarchical Information Integration Conjoint Methods to Modeling Group Preferences”, Marketing Letters, Vol. 11, No. 2, pp. 165-175. Nowlis, S. and Simonson, I. (1997), “Attribute-task compatibility as a determinant of consumer preference reversals”, Journal of Marketing Research, Vol. 34, No. May, pp. 205-218. Okada, E. (2001), “Trade-ins, mental accounting, and products replacement decisions”, Journal of Consumer Research, Vol. 27, No March, pp. 433-446. O’Loughlin, D. and Szmigin, I. (2007), “Services branding: Revealing the rhetoric within retail banking”, The Services Industries Journal, Vol. 27, No. 4, pp. 435- 452. Oppewal, H. and Klabbers, M. (2003), “Compromising between information Completeness and task simplicity: A comparison of self-explicated, hierarchical information integration, and Full-Profile Conjoint methods”, Advances in Consumer Research, Vol. 30, No. 1, pp. 298-304. 113 Lin, J.S. and Wu, C.Y. (2011), “The role of expected future use in relationship-based service retention”, Managing Service Quality, Vol. 21 No. 5, pp. 535-551. Martensen, A., Gronholdt, L., and Kristensen, K. (2000), “The drivers of customer satisfaction and loyalty: cross-industry findings from Denmark”, Total Quality Management & Business Excellence, Vol. 11 No. 4,5,6, pp. 544-553. Mcllroy, A. and Barnett, S. (2000), “Building customer relationships: Do discount cards work?”, Managing service quality, Vol. 10 No. 6, pp. 347-355. Miguel, P., Brito, L., Fernandes, A., Tescari, F. and Martins, G. (2014), “Relational value creation and appropriation in buyer-supplier relationships”, International Journal of Physical Distribution and Logistics Management, Vol. 44 No. 7, pp. 559-576. Nam, J., Ekinci, Y. and Whyatt, G. (2011), “Brand equity, brand loyalty and consumer satisfaction”, Annals of Tourism Research, Vol. 38 No. 3, pp. 1009-1030. Nunnally, J.C. (1978). Psychometric theory (2nd ed.). New York: McGraw-Hill. O’Loughlin, D. and Szmigin, I. (2005), “Customer perspectives on the role and importance of branding in Irish retail financial services”, International Journal of Bank Marketing, Vol. 21 No 1, pp. 8-27. Oliver, R. L. (1999), “Whence consumer loyalty?”, Journal of Marketing, Vol. 63, pp. 33-44. Orsingher C., Valentini S. and de Angelis M. (2010), “A meta-analysis of satisfaction with complaint handling in services”, Journal of the Academy of Management Science, Vol. 38, pp. 169-186. Parasuraman, A., Zeithaml V. and Berry, L. (1993). Calidad total en la gestión de servicios: cómo lograr el equilibrio entre las percepciones y las expectativas de los consumidores. Madrid: Díaz de Santos. Rauyruen, P., and Miller, K.E. (2007), “Relationship quality as a predictor of B2B customer loyalty”, Journal of Business Research, Vol. 60, pp. 21-31. 114 Robson, J. and Sekhon, Y. (2011), “Addressing the research needs of the insurance sector”, International Journal of Bank Marketing, Vol. 29 No 7, pp. 512-516. Rose, F. (1999), The economics, concept, and design of information intermediaries. Physica Verlag, Heidelberg. Rosenberg, L. and Czepiel, J. (1984), “A marketing approach for customer retention”, Journal of Consumer Marketing, Vol. 1 No. 2, pp. 45-51. Rowley, J. (2005), “The four Cs of customer loyalty”, Marketing Intelligence and Planning, Vol. 23 No. 6, pp. 574-581. Roy, S. (2012), “Brand loyalty in insurance companies”, Journal of Economic Development, Management, IT, Finance and Marketing, Vol. 4 No. 1, pp. 12-26. Rugman, A., Li, J. and Oh, C.H. (2009), Are supply chains global or regional?, International Marketing Review, Vol. 26 No. 4/5, 384-395. Saaty, A. (2011), “Assessing the satisfaction of insurance customers and its impact with reference to Saudi Arabia”, Interdisciplinary Journal of Contemporary Research in Business, Vol. 2, pp. 290-305. Short, P. F., Graefe, D. R. and Schoen, C. (2003), Churn, churn, churn: How instability of health insurance shapes America’s uninsured problem. Commonwealth Fund. Shugan, S. (2005), “Are consumers rational? Experimental evidence?”, Marketing Science, Vol. 25 No. 1, pp. 1-7. Siddiqui M. H., and Sharma T. G. (2010), “Measuring the customer perceived service quality for life insurance services: An empirical investigation”, International Business Research, Vol. 3 No. 3, pp. 171-186. Stock, J. R. and Boyer, S. L. (2009), “Developing a consensus definition of supply chain management: A qualitative study”, International Journal of Physical Distribution and Logistics Management, Vol. 39 No. 8, pp. 690-711. Taylor, S. A. and Baker, T. L. (1994), “An assessment of the relationship between service quality and customer satisfaction in the formation of customer´ purchase intentions”, Journal of Retailing, Vol. 70 No. 2, pp. 163-78. 115 The Council of Insurance Agents & Brokers (2015), The role of insurance intermediaries, retrieved on February 19 2015, https://www.ciab.com/uploadedfiles/resources/roleofinsint.pdf Tucker, W.T. (1964), “The Development of Brand Loyalty”, Journal of Marketing Research, Vol. 1, pp. 32-35. Zeithaml, V.A., Berry, L.L. and Parasuraman, A. (1996), “The behavioral consequences of service quality”, Journal of Marketing, Vol. 60, pp. 31-46. 116 APPENDIX In Chapter I and Chapter II Selected Attributes and levels Ø Intermediaries recommendation: i) yes; ii) opinion omitted; Ø Premium: i) 150€ - Standard product through regulation (after the decree-law no. 72/2008, April 16th); ii) 200€ - the same coverage as the option of 150€ and vehicle occupants’ insurance; iii) 250€ - the same coverage as the option of 200€ and auto glass insurance; iv) 300€ - the same coverage as the option of 250€ and theft coverage; Ø Insurers: i) Fidelidade-Mundial; ii) Açoreana; iii) Allianz; iv) Tranquilidade; Ø Price bundling (home insurance with a promotion discount, for just 30€): i) yes; ii) no. In Chapter III Ø Number of insurance proposals presented by intermediaries; Ø Intermediaries’ recommendation was explained; Ø Problems’ resolution by intermediaries; Ø Problems’ resolution by insurers; Ø Insurers’ fast response to customers’ problems; Ø Online services from insurers; Ø Contact made by insurers to know if there was any problem; Ø Payment facilities; Ø Quality of services from insurers; Ø Quality of services from intermediaries.