MASTER THESIS
Effects of Loyalty Programs transparency on Value
Perception
Floris Jansen 11132299 | Thesis-coach Ed Peelen
Master thesis Executive program in Managements studies | University of
Amsterdam | Amsterdam business school | September 17
3
Statement of Originality
This document is written by Floris Jansen who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.Acknowledgements
I would like to thank Prof Dr. Ed Peelen for all the time he spends in providing support, valuable feedback throughout the thesis writing process. Also, all participants that took the time to provide valuable feedback during the pre-tests and tests.Index
STATEMENT OF ORIGINALITY 3
ACKNOWLEDGEMENTS 3
ABSTRACT 6
INTRODUCTION 7
AN INTRODUCTION TO THE DEVELOPMENT OF LOYALTY PROGRAMS AND TRANSPARENCY 7
PROBLEM DEFINITION 7
MANAGERIAL CONTRIBUTION 8
THEORETICAL CONTRIBUTION 8
STRUCTURE 8
THEORY AND HYPOTHESIS 9
CUSTOMER’S BRAND LOYALTY 9
LOYALTY PROGRAMS 10
TRANSPARENCY IN LOYALTY 12
VALUE PERCEPTION 12
HYPOTHESES AND CONCEPTUAL FRAMEWORK 13
THE HYPOTHESIS 13 CONCEPTUAL MODEL 13 METHODS 15
RESEARCH DESIGN 15
MEASUREMENT 16
TRANSPARENCY OR NOT 16 BEHAVIORAL LOYALTY AND ATTITUDINAL LOYALTY 16 CONTROL VARIABLES 17 VARIABLES 17 STATISTICAL PROCEDURE AND ASSIGNING OF RESPONDENT TO CONDITIONS 18 PRE-TEST 19 SAMPLE 19
STATISTICAL PROCEDURE 19
RESULTS AND HYPOTHESIS TESTING 19
RESPONDENTS PROFILE AND RELIABILITY CHECKS 19
HYPOTHESIS TESTING 20
HYPOTHESIS 1A: TRANSPARENCY OF THE REWARD STRUCTURE HAS A POSITIVE EFFECT ON THE VALUE PERCEPTION OF THE LOYALTY PROGRAM 20 HYPOTHESIS 1B: TRANSPARENCY OF THE TIMING OF REWARD REDEMPTION HAS A POSITIVE EFFECT ON THE VALUE PERCEPTION OF THE LOYALTY PROGRAM 21 HYPOTHESIS 2A: A HIGHER VALUE PERCEPTION HAS A POSITIVE IMPACT ON THE ATTITUDINAL LOYALTY. 23 HYPOTHESIS 2B: A HIGHER VALUE PERCEPTION HAS A POSITIVE IMPACT ON THE BEHAVIORAL LOYALTY. 23 HYPOTHESIS 3A: TRANSPARENT REWARD STRUCTURE OF THE LOYALTY PROGRAM HAS A DIRECT POSITIVE EFFECT ON THE ATTITUDINAL LOYALTY. 24 HYPOTHESIS 3B TRANSPARENCY OF THE TIMING IN THE LOYALTY PROGRAM HAS A DIRECT POSITIVE EFFECT ON THE BEHAVIORAL LOYALTY. 24 REGRESSION ANALYSIS 26 SUMMARY OF FINDINGS 28
5 DISCUSSION 32
INTERPRETING THE RESULTS AND HYPOTHESES TESTS 32
THEORETICAL CONTRIBUTIONS 33
MANAGERIAL CONTRIBUTIONS 33
LIMITATIONS AND FUTURE RESEARCH SUGGESTIONS 34
CONCLUSIONS & RECOMMENDATIONS 35
REFERENCES 37
TABLE 40
IMAGES 40
APPENDIX 40
TRANSPARENT VERSION SURVEY 40
NON- TRANSPARENT VERSION SURVEY 48
Abstract
The aim of this study was to research how transparency influences the change in behavioral and attitudinal loyalty in loyalty program design. A survey was designed in different conditions with or without exposing the participants to transparency in timing and rewards structures of the loyalty program. In this thesis, no support was found that transparency influences the change in behavioral and attitudinal loyalty. Compared to the version without transparency. But strong links were found that can lead to further research in the field of loyalty program design. The plots of the effects on loyalty and value perception indicated that the effects that were hypothesized in this study might be present, but it was not possible to show their statistical significance. The rejection of the hypothesis could be caused by the limitation of the study or the focus on the wrong elements in a loyalty program design. A participant of a loyalty program has different levels of majority while being a participant. Making the transparency dynamic during the majority could lead to more positive results. Also, one of the limitations was that intentions were measured in a survey and not the actual behavior. Follow-up research could be done with an A/B test to measure actual behavior instead of intentions. This study contributes to the unexplored field of the influence of transparency on both attitudinal and behavioral loyalty and shows possible limitations as well as suggestions on how to overcome these limitations when researching this influence.7
Introduction
Loyalty describes the state of being loyal; faithfulness to commitments or obligations. Companies benefit from loyal customers as loyal customers stay longer because they prefer to buy from a known and trusted vendor (Rego, 2004). Loyal customers buy more and more often, price sensitivity decreases when customers are truly loyal and loyal customers act as the company’s ambassadors and while they are the ambassador they provide high quality feedback to the company (Rego, 2004). Companies design and introduce loyalty programs, often called a reward program, to build customer loyalty through the planned reward scheme based on a customer's purchase history. With a prime reason to ‘reward, and therefore encourage, loyal behavior’ (Byron Sharp, 1997). Loyalty programs are designed to establish higher levels of customer retention in beneficial segments by providing more satisfaction and value to a specific group of customers (Bolton, 1998).An introduction to the development of loyalty programs and transparency
Organizations are always looking to extend and improve their loyalty programs. There are new developments in loyalty programs where organizations add gamification characteristics to loyalty programs. Games are made to entertain people, the developers of the games created something from the ground up to captivate people. With gamification, the builder starts with something that isn’t built for the intent of gaming but is applying game characteristics on the non-game application. This game-thinking and game-mechanics to solve issues and engage people can’t be called new. For generations parents try to let their little children eat broccoli and other vegetables by making a game out of it (e.g., the “airplane” landing). But gamification is a word that is more and more mentioned in the combination with marketing tactics. Gamification can help in “engagement” which, in a business sense, indicates the connection between a consumer and a product or service. Engagement can be defined as the deep emotional connection with the brand, high level of participation and a long-term relationship and is therefore the most important ingredient in loyalty. Gamification can provide a powerful accelerant to the organizations efforts of the loyalty program. As the parents did with the broccoli and children, if given enough time and incentive, we can overcome our natural programming (Cunningham, 2011). Gamification is build out of 10 game mechanics (Paharia, 2013), one of which is transparency. Transparency is an interesting component of gamification in relation to loyalty. Because Loyalty programs commonly are vaguer about the benefits of the program the absolute values. It could also be said that transparency is the starting point of gamification because it enables the other game mechanics: Fast feedback, Goals, Badges, Leveling up, Onboarding, Competition, Collaboration, Community and Points. All mechanics somehow depend on the transparency factor. Would the average frequent flyer participant know what the earned mile would equivalent in benefits? The benefits of loyalty programs are usually not transparent and crystalline communicated in a program. This research investigates the moderating effect of transparency to the success of loyalty programs. The extend of the transparency of the benefits will be tested. It is known that loyalty is influenced by value perception, the higher the value perception, the more loyal a customer is (Hallberg, 2003).Problem definition
To address the issues discussed in the previous text this thesis focuses on researching the possibility of using the transparency to enable companies to create a higher loyalty. There is a lack of research papers that researched the impact of transparency on customer’s loyalty. The main research question for this thesis is therefore: • “How can transparency of the benefits of a loyalty program impact the value perception and therefore increase the behavioral and altitudinal loyalty of a customer?” The problem will be researched by analyzing the following issues:• How can Transparency of the reward structure & timing of reward have a positive effect on the value perception of the loyalty program • How can a higher value perception have a positive impact on the attitudinal & behavioral loyalty? • How can the reward structure of the loyalty program have a direct positive effect on the attitudinal & behavioral loyalty? This study aims to research the impact of transparency elements on the attitudinal & behavioral loyalty. By answering the sub questions the main research question will be answered.
Managerial contribution
This study can be described as a multiple purpose research. Because transparency (from gamification) is a new concept that hasn’t been researched in relation to loyalty and loyalty programs. From a managerial viewpoint knowing how transparency influence loyalty through value perception might create new opportunities. The ability to increase the loyalty of customers is important for managers because it has certain advantages, like a greater trade leverage and reduced marketing costs (Aaker, 1991). This makes that research in the correlation area of transparency and customer loyalty has a managerial benefit.Theoretical contribution
Research has been done in the field of loyalty and transparency but this was mainly in CSR and loyalty. Where organizations are transparent in their CSR activities, this can increase the loyalty of customers (Kim, 2016). There has been little to zero academic research addressing the impact of transparency on both attitudinal and behavioral loyalty. This thesis provides an entry into this area and ambition to create a theoretical framework on transparency in the context of both loyalties. There will be research done to find the impact of transparency on the behavioral and attitudinal loyalty. This research is limited to the online communication aspects of loyalty programs, all offline communication is not part of this thesis. For this thesis research, it is particularly important to test the influence of transparency on both behavioral and attitudinal loyalty via value perception.Structure
This thesis research follows the following structure. The introduction serves to introduce the field of research and provides a background on this. It explains the problem and purpose of the thesis and motivates the research. The theoretical background is presented in the following chapters that define the concept and give a literature review to connect it with previous research. First it defines the concepts of customer’s brand loyalty, and the difference between attitudinal loyalty and behavioral loyalty. Further all design elements of loyalty programs are explained that are created to increase the customer’s loyalty. Building upon that, the concept of being transparent is explained. The mediating role of value perception is explained and how this leads to increased loyalty. In the following chapter, the research methodology is presented in detail. It describes how the empirical research is conducted and the methods that were used to derive to the results. The following section presents the empirical study that was done and the results that came out of that research. It provides the gathered data such as the data from the surveys. In the subsequent chapter the data is analyzed. The results of the analyses are presented and interpreted using the theoretical framework. the next part discusses the results that were derived in the previous chapter. The results are reviewed to answer the research questions. The purpose is to give managerial advice and to deepen the theoretical understanding to provide both a managerial and theoretical benefit. Finally, limitations of the results are given and suggestions for further research are proposed.9
Theory and hypothesis
Customer’s brand loyalty
Loyalty can be defined as repeated purchases of products or services during a certain period (Brody, Cunnigham, & Scott, 1968). This is called brand loyalty, it explains the tendency of a consumer to continue buying the same brand of goods rather than competing brands. The elements that are often used to define loyalty are typically the following intervals of behavior during certain time intervals to operationalize loyalty in a competitive market: The first element is the percentage of customers purchasing a brand, second is the number of purchases per buyer (retention). The third is the percentage of customers who continue to buy the brand. The fourth element is the percentage of customers who are 100 percent loyal. The fifth and last element is the percentage of customers who also buy other brands (duplicate buyers) (Wrigley, 1989). Chestnut & Jacoby (1978) position Brand loyalty in their research as a deterministic subset of repeat purchase behavior. This allows for a very worthwhile structure and manageable approach to understanding brand loyalty. Their research divides brand loyalty into a behavioral, attitudinal, or composite perspective. The authors also distinguish between macro and micro approaches to behavioral loyalty. Attitudinal loyalty can be described as the ability to connect with customers at an emotional level. Attitudinal loyalty is the ultimate, where consumers will be loyal to a brand regardless of price, convenience or other factors because their personal connection to the brand (Chestnut & Jacoby, 1978). “A company doesn’t have to have the emotional DNA of Disney or Apple to succeed. Even a cleaning product or a canned food can forge powerful connections” (Magids, Zorfas, & Leemon, 2015). Behavioral loyalty is the type of loyalty where a customer is still loyal to a brand but doesn’t have the emotional connection to it (Chestnut & Jacoby, 1978). In the case of a retail, that the retailer closest to home. Or the one that has attractive prices. When certain conditions change - the loyal customer will most probably leave the retailer and will go elsewhere. 78% of consumers are not loyal to a brand, purchasing decisions are being made less and less on emotion and more on money based ratio. Just 25% of consumers consider brand loyalty as something that impacts their buying behavior (Russo, 2014). Jacoby and Chestnut’s research is a strong critique on the stream (or lack thereof) of existing behavioral loyalty research. The authors call for greater organization and development of research efforts to further knowledge in this area. To answer the question why are customers loyal, academic research found several elements that explaining factor that increase loyal behavior of customers (Chestnut & Jacoby, 1978). Price is one of the import stimuli in loyalty. The customers perceived price, and this price perception both directly and indirectly influenced customer loyalty (Han & Ryu, 2009). Price is also strongly related to value perception, this is discussed in the Value perception chapter. Discounts and special offers summarized as price fairness is considered as an antecedent of customer satisfaction and loyalty and makes it an important factor in the loyalty of a customer (Martín-Consuegra, 1992). A loyalty increasing element can be the customer service, customers will be more loyal when the service increases (Coelho & Henseler, 2012). Other increasing factor of loyalty are corporate responsible activities of an organization (Kim, 2016). The thirdly increasing factor can be nostalgia, the personal and collective memory that intermingle in collaborative and interactive processes can create a valued nostalgic experience that leads to a higher loyalty towards a brand (Hamilton & Wagner, 2014). Also trust increases the loyalty of customers. Luhman [1979] and Blau [1964] suggest that, in general, trust is build up when the trusted party behaves in a socially acceptable manner. This is in accordance with what is expected of that trusted party, and that, conversely, trust is reduced when the trusted party does not behave accordingly without good reason. this will be elaborated in the transparency chapter. In various industries firms adopt customer loyalty programs to encourage a strong customer relationship (Jun Kang, 2015).Loyalty programs
The reason there are loyalty programs is to ‘reward, and therefore encourage, loyal behavior’ (Byron Sharp, 1997). Loyal customers are more profitable for organizations and the retention costs are lower. Some academic researches discuss if loyalty programs really work and if it leads to business success. The first mover in an industry that introduce a loyalty program has a first movers advantage, introductions of programs afterwards are usually less effective. The risk of a loyalty program is that it introduces costs for an organization, but don’t lead to increased value proposition (Dowling & Uncles, 1997). But there is evidence that loyalty programs work and have led to more sales and customer engagement. For example, Lal and Bell (2003) Lal (2003) found convincing evidence that loyalty programs serve the purpose of a company competitive advantage creator. Implying that organizations can get a competitive advantage by their investments in loyalty programs. Not all loyalty programs necessarily develop higher frequencies in purchase activities by customers and loyalty is not automatically the result of a loyalty program (Meyer-Waarden, 2006). 31% of customers stated that loyalty programs have an influence on them in their purchase behavior (Howard-Brown, 1998). 53% of grocery consumers and 2% in apparel consumers are enrolled in a loyalty program in the United States (James Cigliano, 2000). 48% of these customers claimed that they spent more than they would have done without the existence of the loyalty program (James Cigliano, 2000). What companies should do to let customers adopt a loyalty program is depending on the level and nature of competition within the market they operate; many companies introduce a loyalty program as a response to a competitor’s introduction of a loyalty program. Evidence shows that this is not the best strategy, because the lack of first movers advantage (Leenheera & Bijmolt, 2008). The design of a loyalty program is studied to affect its constitutional potential. Understanding the notable preferences of the consumers to participate is an important factor (James Peltier, 1998). When customers enroll in the program this doesn’t necessarily say that they will become active users in the program but getting a significant number of consumers to enroll is a necessary first step in realizing loyalty program effectiveness. The multitude of program design dimensions that exist attests to the large number, this differs in which industry the organization operates. Loyalty programs design can be described along the following dimensions (Reinartz, 2010): • Reward Structure: Hard vs. soft rewards. Hard rewards are financial (or) tangible rewards. Soft rewards are based on psychological or emotional benefits. A loyalty program usually exists out of both elements. • Product proposition support: The rewards entailed in a program may be directly linked to the company’s product offering. For example, Starbucks offers participants of a loyalty program to redeem their earned loyalty points for Starbucks own products. Some of the programs are designed around reward products that are not related to the organizations core business. • Hedonic Value of Reward: Hedonic products are those products whose is associated with pleasure luxury and fun. An interesting result from the research of Simonson (2012) shows that consumers prefer hedonic goods over utilitarian goods when it takes them considerable effort to earn it. This can be explained by the fact that consumers feel better indulging by luxury items when the effort they had to expend to earn that luxury is relatively high, as in the case of a program reward. For example, a free flight to a luxury destination is probably much more attractive to a buyer than vouchers for a supermarket. • Rate of rewards: the rate of rewards assign the ratio of the reward value to the transaction volume (both in monetary terms). Rate of rewards given to participant, may depend on their cumulative spending with the provider of the program. The so-called ‘asset accumulation response function’ explains how points are accumulated as a function of consumers’ cumulative spending behavior.11 • Timing of reward redemption: The last element of the reward structure of the program is the timing of Reward Redemption. Suppliers of loyalty programs prefer to create redemption rules that favor long periods, which in turn consequence customer retention. The literature describes this effect as “lock-in” because customers build up capital that eventually function as switching costs, locking them in to doing business with the supplier of the program. In general programs reward are based on accumulated purchases, this creates points pressure mechanisms that encourages customer buying frequencies or volume of purchase to obtain a certain reward (Ran Kivetz, 2006). The strength of this pressure to get to the threshold of points depends on the design of the loyalty program. The perception towards the utilitarian rewards perception should be high by the customers. In other words, the customer that is enrolled should be afraid to lose the points (investment). By getting rewarded customers feel encouraged to maintain purchase levels due to the emotional attachment to the perceived value (Robert W. Palmatier, 2009). • Sponsorship: refers to the features of the program’s supplier and participant. The loyalty program has a single vs. multi-firm design. The supplier may establish programs that reflect only transactions within its own domain, such as that of frequent flyer program that only consist out earned miles flown with the airline and rewards only are provided by the same airline. The alternative functions as a way that the supplier might allow participants to accumulate assets at firms that are associated with the supplier of the program, such as Accor hotel loyalty program that let a participant save points with renting a car at AVIS. This can also be described as Within Sector/Across Sector Within Sector/Across Sector this is a multi-firm design across-sector. STAR alliance is an example of a across sector multi-firm design. Where multiple airlines are combined in a loyalty program. • Ownership: The last element of the loyalty design is the ownership of the program. In case of multi-firm programs, the ownership dimension indicates who owns the program within the network of firms. Is it the firm, another firm, or a joint venture of the participating organizations. The customers that participate in loyalty programs perceive several advantages, there are 3 types. The first is utilitarian like savings, gifts and more convenience, second advantage is hedonic like (personalized) treatments and customers get the change to try new products. The third is symbolic like being recognized by a company and improvement of a social status. Research shows that the customer attach most value to the utilitarian benefits of the loyalty programs. (Kerrie Bridsona, 2008). That utilitarian benefits are tangible and therefore easier to evaluate by customers could be the explanation why this is an important factor in the evaluation of a loyalty program (Nathalie T.M. Demoulin, 2009). The decision to enroll in a program is determined by the perceived benefits by the customer. But research show that the impact of the benefit on the consumer’s behavior declines soon after enrollment in the program (Benavent, 2007). That’s why companies with a loyalty program should tailor the utilitarian benefits to increase the attractiveness of the program and the perceived member benefits, but they should also consider hedonic and symbolic benefits to enhance and develop relationships over time. Consumers that early adopt a loyalty program are usually the heavy users within that category. For example, the first users of frequent flyer programs were the customers that flew a lot already (Allaway, Berkowitz, & D’Souza, 2003). In general loyalty programs have positive influence on performance measures, such as penetration levels, average purchase frequency. Loyalty programs enhance customer’s affective commitment or attitudinal loyalty to produce lasting effects (Hallberg, 2003). Attitudinal loyalty (a combination of commitment, satisfaction and positive word of mouth) is an important indicator and driver of the behavioral loyalty of loyalty program participants (Hallberg, 2003). Companies can increase the effects by positively manipulate customer’s perception such as by endowing them with initial loyalty points (Hallberg, 2003).
Transparency in Loyalty
Transparency is defined as the quality or state of being transparent so being easily detected or seen through. Some management literature describes the reasons why transparency leads to loyalty of customers. The first reasons are that it builds trust, customers think that honesty is the foundation to making the relation last longer. Trust is defined as a willingness to rely on an exchange partner in whom one has confidence (Moorman, Zaltman, & Deshpande, 1992). Trust is an essential element in building a relationship between an organization and it clients (Berry, 1995). The second reason is that it cultivates word of mouth, when a brand falls short in consistently satisfying customers, or has indigent customer service, or puts out an inferior product, word gets out expeditiously through social media channels and conversations. The third reason is that It opens better communication, social media is not only the best place to communicate with customers and market products, but it’s also the best place for a brand to show transparency. Social media is the becoming the most natural place for customers to complain about an organizations failure to satisfy. This is where a brand can connect directly with customers by opening a conversation, be apologetic, and express a way to make amends. The fourth reason is that it clears the air, it happens that, a company will work hard to fix a former wrong. And coming out and being transparent about it fuels loyalty. For example, the San Francisco-based bank Wells Fargo was involved in a scandal after the disclosure of millions of fraudulent accounts. After this they started communicating about "Moving forward to make things right," in national newspapers, which specify actions like putting customer interest first, proactively communicating, fixing what went wrong and undertaking full transparency that led to a higher level of loyalty. And the final fifth reason is that customers are attracted to authenticity. Thanks to the introduction of social media and smartphones, customers are as savvier as ever when it comes to spotting a brand that is genuinely transparent and spotting one that is merely forcing it. Being authentic and real is an attractive and rare commodity. Gurrea, (2006) found an indirect effect of transparancy in loyalty. They researched the role played by perceived usability, satisfaction and consumer trust on website loyalty. Because honesty is related to information transparency. Given that greater usability favors greater transparency, they found that honesty (trust) was improved. Because of this improved trust experience website loyalty increased. The research has confirmed that perceived usability has direct and positive relationship on the degree of consumer trust and satisfaction. Moreover, it has shown that the effect of the degree of usability on the degree of consumer loyalty does not follow a direct path but is conditioned by the role that trust and satisfaction play with respect to the individual’s fidelity. (Gurrea, 2006). Kristensen (2007) aimed in his research to analyze the effect of the transparency of products and services on customer satisfaction which leads to customer loyalty (Bolton, 1998). In this article Kristensen suggested that perceived transparency of products and services influenced the customers’ perception of the value provided by the company which led to higher loyalty.Value perception
Value perception also known as customer-perceived value, is the discrepancy between a customer's evaluation of the benefits and costs of one product when compared with others. Dowling and Uncles (1997), described in their research that loyalty programs are of important in enhancement of the comprehensive value of the product or service as they motivate loyal buyers to make their next purchases. The benefit analysis by the customer is an important factor for a customer decide to participate or not in a loyalty program (Hallberg, 2003). Zaichkowsky’s (1985) research has concluded that the influence of rewards on the perceived value of the loyalty program and how this perceived value affects customer loyalty. Youjae Yi &Hoseong Jeon (2003) investigated how reward schemes in loyalty program designs influence value perception of the program and how affects customer loyalty. They found that that involvement moderates the effects of loyalty programs design on customer loyalty. In high-involvement situations, direct rewards are preferable to indirect rewards (Jeon, 2003). The direct rewards (benefit percentages) have a wide range of percentages.13 For example, U.S. supermarkets like Tesco, Sainsbury's, Kroger, Safeway, and Stop & Shop give back 1-2 percent of the total spent to their loyalty participating shoppers, nonfood usually has a larger percentage of around 4 percent (HBS Working Knowledge , 2014).
Hypotheses and conceptual framework
The hypothesis This study researches how transparency influences loyalty in loyalty programs. In the previous chapters a description is given about the different elements of loyalty programs. The theory that is discussed in the previous chapter leads here to the hypothesis. In the end of this chapter a conceptual framework is presented, which entails an overview of the hypotheses and variables. Through testing these hypotheses and therefore the conceptual framework, an answer can be given on how transparency influences the attitudinal and behavioral loyalty of consumers. Moorman, Zaltman, & Deshpande (1992) describe in their research how transparency leads to trust, which leads to loyalty. It is an essential element in building a relationship between an organization and it clients (Berry, 1995). The transparency component can be combined with the reward structure of a loyalty program. The reward structure defines what a consumer will get for being loyal. Hypothesis 1a: Transparency of the reward structure has a positive effect on the value perception of the loyalty program The reward structure defines the utilitarian, hedonic and symbolic value will be provided in a certain timing. The timing defines when a participant will get which benefit. The literature describes for example the “lock-in” effect where customers build up capital that eventually function as switching costs, locking them in to doing business with the supplier of the program (Robert W. Palmatier, 2009). Hypothesis 1b: Transparency of the timing of reward redemption has a positive effect on the value perception of the loyalty program The height of the participant’s perceived value has an impact on loyalty (Jeon, 2003). But in this research the loyalty is divided in the attitudinal part and behavioral part (Chestnut & Jacoby, 1978). Hypothesis 2a: A higher value perception has a positive impact on the attitudinal loyalty. Hypothesis 2b: A higher value perception has a positive impact on the behavioral loyalty. The previous hypothesis focused on the indirect effect of transparency on loyalty via value perception. But the effect can also be direct, where the reward structure’s and the timing’s effect are measured on the Attitudinal & Behavioral loyalty Hypothesis 3a: Structure of the reward structure of the loyalty program has a direct positive effect on the Attitudinal & Behavioral loyalty. Hypothesis 3b Transparency of the timing structure of the loyalty program has a direct positive effect on the Attitudinal & Behavioral loyalty. Conceptual model The 3 hypotheses are visualized in the following conceptual framework. The hypothesis in the conceptual framework will be tested in this study. It presents the dependent variables and independent variables. Below the conceptual framework is visualized.
Image 1 Conceptual model
Methods
The methods section described the method of how the empirical research is conducted. The aim of this study is to research how transparency influences the change in behavioral and attitudinal loyalty and how this is mediated by value perception. This chapter is starts with a description of the research design, a description of the sample and followed by a description of the measurement of the variables. The last chapter is the statistical procedure is explained.Research design
The study researches whether transparency influences the change in behavioral and attitudinal loyalty. In this concept, the value perception works as a mediator. This study is consisted of an A/B test experiment to test the effect by manipulating the independent variable to find the effect on the dependent variable. The independent variable were the different variants of transparency that were plotted on the different loyalty design elements which were either present or absent. In this study, there were 2 conditions where participants were randomly divided into, loyalty program with a transparent reward structure and timing or not. The conditions without transparency served as control groups. Condition 1 included loyalty program design elements with clear transparent benefits and timing communicated. The condition 2 group saw the exact same loyalty program design components without the transparent benefit communicated. To make sure that the response from the respondents was not biased a fictitious retail department store’s loyalty program was designed: ‘Monogram’. The fictitious store chain is a combination of several department stores like Bijenkorf, Macys and HEMA. Image 2 Monogram Logo The reason to create this loyalty program is because of the appealing effect on respondents. There are many loyalty programs that could be designed for this research as supermarket chains, stores hotels etc. The assumption is that a full-service department store can easily be framed in the mind of the respondent.Measurement
The goal of this study is to research how transparency influences the two types of loyalty, behavioral and attitudinal and how this is mediated by value perception. There were 2 versions of the survey that provided clear transparent loyalty program benefits or non-transparent. In the following example the offered suitcases are offered to a participant of the program. In the first case the customer can see exactly how many value the program offers. The 439,- euro spend in the shop offers a ratio of 0,0025 = 10,975 in the non-transparent communicated value the 689,- euro offer the same amount in points. Which makes it a lot harder to calculate the exact value. The respondents were guided through a typical shopping journey and were asked if they were interested in participating in a loyalty program based on the benefits that were communicated. The reward timing was researched via the loyalty goal. One of the last question was to assume that they were participants in the loyalty program and that they have set a loyalty goal. This is a goal that a participant can set and when they save points/currency that they get closer to that goal. In the transparent version of the survey it was really clear to a participant where they were in reaching their goal. The participants were asked if they received benefit in knowing where they were in reaching their goal. The complete list of questions can be found in the appendix. Image 5 example non-transparent communication Transparency or not Participants were randomly assigned to the group that was exposed to transparency or not. A randomizer in Qualtrix defined if the participants saw the transparent or non-transparent version of the survey. Behavioral loyalty and attitudinal loyalty To measure the difference in behavioral and attitudinal loyalty after being exposed to the transparency of the benefits, the scale developed by Chaudhure & Holbrook (2001) was used. Participants were asked about their purchase loyalty (“I will buy this brand the next time I buy (product name)” and “I intend to keep purchasing this brand”, α = .90) and about their attitudinal Image 4 Example product17 loyalty (“I am committed to this brand” and “I would be willing to pay a higher price for this brand over other brands”, α = .83). These statements were conducted on a 7-point ratings of agreement (1 = very strong disagreement, 7 = very strong agreement) (Chaudhure & Holbrook, 2001). Control variables In the survey, the variables age, gender, nationality and participation in loyalty programs were provided to serve as control variables to filter out unusable records. If participants are under 20 years of age they were not considered because the less likelihood that they would be interested in a loyalty program. Participants were asked if they already participated in a loyalty program to exclude participant’s rejection of loyalty programs. Variables Transparency in communication Gurrea, (2006) found an indirect effect of transparancy in loyalty, the direct effect was never found. To measure the effect of transperancy in loyalty benefit communication, the scale developed by Chaudhure and Holbrook (2001) was used. The questions based on the (non)-transparent benefit communications were ‘The benefits of this loyalty program are clearly communicated’, ‘The benefits of this loyalty program are clearly communicated?’. To give an even larger proportion of transparency the concept of loyalty reward was explained to the respondents. A participant could set a goal (free product) that can be reached by earning points with a purchase. The participant saw a percentage progress bar that made it very transparent where the participant was in reaching the goal. And because of the 2 different versions of the questionnaire it was possible to measure the different between the transparent and non-transparent group. Reward structure The reward structure describes the way in which participants are rewarded for being loyal. Reward Structure can be hard and soft. Hard rewards are financial or tangible rewards. Soft rewards are based on psychological or emotional benefits. A loyalty program usually exists out of both elements. To measure the response to the transparency in the reward structure the reward structure was explained to the participants in the transparent version of the questionnaire. ‘The Rapid rewards program provides you budget for every euro spent that can be used for next purchases. In this case you will get €0,025 for every euro spend on the shop.’ Based on that it was asked to the participants if this was impacting their behavior ‘For my next purchase I'm going to consider Monogram again’. The fictitious program also contained soft benefits like exclusive membership opening hours etc. When this was communicated it was asked to the respondent if this would influence the attitudinal loyalty ‘The benefits of the program are what you are looking for in a loyalty program’. Timing of reward redemption The perception towards the utilitarian rewards perception should be high by the customers. In other words, the customer that is enrolled should be afraid to lose the points (investment). By getting rewarded customers feel encouraged to maintain purchase levels due to the emotional attachment to the perceived value (Robert W. Palmatier, 2009). This was tested by the concept of reward goals
and show the respondent the progression he or she made. To show the participant of a loyalty program how close they are to reaching their goal gives transparency to the participant. In the transparent version of the questionnaire this image of the progress was shown on both 15% and 85% of progress and it was asked if this was clearly communicated and if this was changing their attitude towards the brand. ‘The benefits of this loyalty program are clearly communicated about how close I am in reaching my goal’ and ‘For my next purchase I would consider going to Monogram again to reach my goal’. Value perception The value perception in both the transparent and non-transparent version was tested by asking what the respondent thought of the benefits of the program. In both the version of the questionnaire the benefits were shown in points of currency. ‘The Rapid rewards program provides you points for every euro spent that can be used for next purchases. In this case, you will get 1 point for every euro spend on the shop.’ This was shown in the non-transparent version and in the transparent it was shown how much actual currency was earned for every €1, - spent. All loyalty design elements that were exposed to the respondents were followed up by the question how they experienced the value of the program. Attitudinal loyalty ‘I am committed to Monogram and I would be willing to pay a higher price for this department store over other department stores’ was to question in multiple places in the questionnaire to measure the attitudinal loyalty of respondents in the questionnaire in both the transparent and non-transparent version of the questionnaire. Behavioral loyalty If the respondent would also behave differently because of the loyalty program was measured in different ways in both the non-transparent and transparent questionnaire. For example, by asking: ‘For my next purchase I’m going to consider Monogram again’. Statistical procedure and assigning of respondent to conditions An experiment has been conducted. Potential participants were reached through Facebook and Linkedin and asked to complete an online (experimental) survey. In the eventual survey anonymity was guaranteed to respondents before participating in the experiment. Because of the randomizer in the system, participants saw the transparent or not transparent version of the survey. Both version contained the same question, with the transparency as only difference. To measure the likeability if respondents were interested their current participation in loyalty programs was asked. If they participated in loyalty programs and if it was multiple or one. Gender, age, Nationality were asked to get some level of background information about the respondents. In the follow first steps that respondents had to take was fill in the scale developed by Chaudhure and Holbrook (2001) concerning behavioral and attitudinal loyalty. They were asked about their behavior and attitude towards the loyalty programs “I would be likely book these Strata tickets the next time I book a flight(of a category where this brand is present)” and “I intend to keep booking Strata (that means I intend to buy products from this brand more than once)”. Subsequently participants were asked about their attitudinal loyalty concerning Strata by selecting on the two statements, namely “I am committed to Monogram ” and “I would be willing to pay a higher price for this department store Image 6 Example transparent communication
19 over other department stores”. These statements were conducted on a 7-point Likert scale (1 = very strong disagreement, 7 = very strong agreement). Pre-test A pre-test was conducted to outlaw any clarity, mistakes to the respondents and test the feasibility. In the pre-test respondent where asked if the question were clear and if all responses could be traced. After the pretest, some questions were re-formulated to increase the likeability of understanding the question correctly. The pretest started with a loyalty design of a full-service airline, this lead to confusion and bias to the respondent, so this after the pre-test the design was changed to a department store’s loyalty program.
Sample
The sample consisted of participants that are between 18 and 80 years of age. This age is stated because the participants should be old enough to be part of a loyalty program, but also not older than 60 to assure people are used to using modern loyalty programs, such as the one designed for this research. The data collection happened during the December 2017 month. The questioners were send out via LinkedIn, Facebook, Email and intranet. The respondents were randomly assigned to a condition with transparency or without. The questionnaire consisted of N = 78 transparent and 78 non-transparent (total N = 156) respondents, of which 55% female and 45%.Statistical procedure
Participants were reached through social media and asked to participate in an online experimental survey. In the pre-test participants were first asked if they understood the questions and results were measured. Unfortunately, the results were not as expected, the participants didn’t understand the questions and the designed loyalty program wasn’t desirable. For this reason, the program was redesigned and pre-tested again. The participants were randomly assigned to either to the transparent group or non-transparent. To make sure that they were between the ages of 18-80, age was explicitly asked in the survey. Also, anonymity was guaranteed before participating in the survey.Results and hypothesis testing
In the following chapter the relation between the transparency of loyalty benefits and behavioral and attitudinal loyalty is analyzed. Transparency is the independent variable, attitudinal and behavioral loyalty are the dependent variables, value perception is the mediator. Transparency is also an independent variable on the dependent variables purchase and attitudinal loyalty directly without mediation of value perception.Respondents profile and reliability checks
First, there has been a check of frequencies to make sure that there weren’t any errors in the data. Then there was dealt with missing values by excluding cases list- wise. This means that only cases that had no missing data in any variable were analyzed for this research. If the respondent filled out that he did not participate in any loyalty program or was under 20 years of age, this respondent was deleted. The deletion was done because of the openness towards loyalty programs, if they didn’t participate in any program before it is likely that they are not interested. Responses of participants, who provided clearly futile answers such as answering with the same value for every question were not used for the analysis. Before conducting the analyses, outliers were excluded. This was done because these responses were highly different from most other responses and therefore distort the distribution of the data and consecutively would have led to a lower external validity. This could have for example been caused by respondents that just clicked through without reading and proper thoughtful answering. This would result in answers far different from the average.After the exclusion of the outliers the total sample that was used for the analyses consisted of N = 78 transparent and 78 non-transparent (total N = 156) respondents, of which 55% female and 45% male. The respondents were equally divided by the randomizer in the 2 different conditions. The conditions differed in whether transparency was present or absent in the communication of the loyalty benefits of the program. Because anyone could participate as a respondent in the research, the distribution of both age and nationality were widely distributed. Most of the participants ranged in in the age between 21 and 31 N=106 = 66,7%. The nationality of the respondents was mostly Dutch (91%), other respondents came from France, Belgium, Britain, Spain, United states and Poland. Gliem & Gliem, (2003) reported that a the Cronbach’s alpha measure is a measure of the internal consistency of the scale and is reported to validate this consistency with a score above 0.7. In this case, the internal consistency can be considered ‘good’. The values of Cronbach alpha (α) when looking at all 156 responses, can be found in the appendix.
Hypothesis testing
Hypothesis 1a: Transparency of the reward structure has a positive effect on the value perception of the loyalty program H1a predicts that transparency of the reward structure has a positive effect on the value perception of the loyalty program. First, we test if there is coherence between the reward structure and value perception with the help of the Pearson correlation coefficient, an overview of all Pearson correlation coefficients can be found in Table 11 at the end of this chapter. The Pearson correlation of r=0,68 implies there is a positive correlation between reward structure and the value perception of the loyalty program. Secondly, we test if there is a difference between transparency and non-transparency in correlation with value perception by again looking at the Pearson correlation coefficient. Both transparent and non-transparent reward structures show a significant positive correlation with the value perception. However, we do see a difference in strength of correlation where the correlation of non-transparency is much stronger with a coefficient of r=0,92 compared to that of transparency and value perception (r=0,58).And lastly, we check if there is a significant difference between means of transparency and non-transparency by conducting a one-way analysis of variance (ANOVA) test. The one-way analysis of variance (ANOVA) is used to determine whether there are any statistically significant differences between the means of two (transparent/ non-transparent) groups. Interestingly, Table 1 shows that the overall mean of transparent reward structures (4,01) is lower than the mean of non-transparent reward structures (4,34). This suggests that less transparency results in a higher perceived value. As shown in table 5b the ANOVA test indicates a result of (F=4,97; p<0,05) which implicates that this difference between the overall mean of transparent and non-transparent structures is significant. Based on these analyses one can conclude that transparency significantly impacts value perception, however in the opposite direction of our hypothesis. Less transparency results in a higher perceived value. This invalidates H1a which we therefore reject. Table 1 One-way ANOVA transparent vs. non-transparent reward structures Descriptive Reward structure
21 Table 2 One-way ANOVA transparent vs. non-transparent reward structures ANOVA Reward structure Sum of
Squares df Mean Square F Sig. Between Groups 4,123 1 4,123 4,986 ,027 Within Groups 124,037 150 ,827 Total 128,160 151 Hypothesis 1b: Transparency of the timing of reward redemption has a positive effect on the value perception of the loyalty program In order to test H1b, which states that transparency of timing of reward redemption has a positive effect on value perception, we start again by interpreting several Pearson correlation coefficients. Table 11 shows a positive moderate correlation (r=0,52) between reward redemption and value perception. Secondly, it is tested if there is a difference between "transparency" and "non-transparency" in correlation. Based on the Pearson coefficients we indeed see that there is a difference between transparency (r=0,47) and non-transparency (r=0,50), though both variables correlate moderately positive with value perception. Next step is to check whether there is a significant difference between the means of transparency and non-transparency value perception. As transparency is only measured in one of the two questionnaires, an ANOVA test is not possible. Therefore, in this case a one-sample t-test is used. The descriptive statistics in Table 3 show that in case of transparent timing the average perceived value was 4,74 with a standard deviation of 1,32. The average perceived value with non-transparent timing is higher (mean = 4,85) while the standard deviation is lower (0,98). Table 3 Descriptive statistics transparent and non-transparent timing of reward Descriptive Statistics
N Minimum Maximum Mean Std.
Deviation Transparent Timing of reward redemption 78 1,00 7,00 4,7436 1,32347 Non Transparent Timing of reward redemption 78 2,00 6,00 4,8462 ,98134 Valid N (list wise) 78 Via the one sample t-test we now see the influence of transparent timing on the value perception of the loyalty program. In Table 4 we see that the mean difference is negative (-0.10) but that this difference is not significant as the significance value is above 0.05%. Therefore, we must conclude 95% Confidence Interval for Mean N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum Transparent 78 4,0128 ,82737 ,09368 3,8263 4,1994 2,00 5,33 Non-Transparent 74 4,3423 ,98848 ,11491 4,1133 4,5714 1,67 7,00 Total 152 4,1732 ,92127 ,07473 4,0256 4,3209 1,67 7,00
that there is no significant difference between transparent and non-transparent timing. Herewith we reject H1b . Table 4 One-sample T Test transparent timing reward redemption on value perception One-Sample Statistics N Mean Std. Deviation Std. Error Mean Transparent Timing of reward redemption 78 4,7436 1,32347 ,14985
One-Sample Test Test Value = 4.8462 95% Confidence Interval of the Difference t df Sig. (2-tailed) Mean
Difference Lower Upper Transparent Timing of reward redemption -,685 77 ,496 -,10261 -,4010 ,1958 Hypothesis 2a: A higher value perception has a positive impact on the attitudinal loyalty. The Pearson coefficient as shown in Table 11 (r=0,70) implies a strong positive correlation between value perception and attitudinal loyalty. This positive relation is also displayed in below. As the coefficient is significant at the 0,01 level we accept hypothesis 2a. To test the relationship between value perception on attitudinal loyalty a regression analysis is performed, this can be found in the regression chapter. Hypothesis 2b: A higher value perception has a positive impact on the behavioral loyalty. Also between value perception and behavioral loyalty we see a positive relation Image 7, Image 8. The Pearson coefficient of 0,77 is even stronger than the correlation between value perception and altitudinal loyalty. As the coefficient is significant and in the hypothesized direction we accept hypothesis 2b. Image 7 Pearson correlation between 1. attitudinal loyalty and 2. behavioral loyalty and value perception
Image 8 Pearson correlation between 1. attitudinal loyalty and 2. behavioral loyalty and value perception Hypothesis 3a: Transparent reward structure of the loyalty program has a direct positive effect on the attitudinal loyalty. To test whether the reward structure has a positive effect on attitudinal behavioral loyalty we again look at the Pearson coefficient. The significant coefficient of r=0,630 implies a strong positive relationship between the two variables, therefore we accept hypothesis 3a. Hypothesis 3b Transparency of the timing in the loyalty program has a direct positive effect on the behavioral loyalty. To test hypothesis 3a, first we test if there is a positive correlation between reward redemption and behavioral loyalty. A Pearson coefficient of 0,432 reveals a moderate positive relationship between the two variables. Secondly, we test if there is a difference between transparency and non-transparency. Looking at the Pearson coefficients of both variables in relation to behavioral loyalty we do see a difference. A coefficient of 0,42 implicates a positive relationship between transparent timing of reward redemption and behavioral loyalty. Considering the Pearson coefficient of the correlation between non-transparent timing and behavioral loyalty we see a weakly positive correlation (r=0,38) Via a t-test we examine whether there is a significant difference between the means of transparency and non-transparency group. The descriptive statistics in Table 11 show that in case of transparent-timing the average behavioral loyalty was is 4,74 with a standard deviation of 1,32. The average perceived value with non-transparent timing is higher (mean = 4,85) while the standard deviation is lower (0,98). In layman's terms it could be said that being transparent in timing of reward redemption reduces the behavioral loyalty of the participant. Table 1 Descriptive statistics transparent and non-transparent timing of reward Descriptive Statistics
N Minimum Maximum Mean Std.
Deviation Transparent Timing of reward redemption 78 1,00 7,00 4,7436 1,32347 Non transparent Timing of reward redemption 78 2,00 6,00 4,8462 ,98134 Valid N (list wise) 78
25 T-Test One-Sample Statistics N Mean Std. Deviation Error Std. Mean Transparent Timing of reward redemption 78 4,7436 1,32347 ,14985 One-Sample Test Test Value = 4.8462 95% Confidence Interval of the Difference t df Sig. (2-tailed) Mean
Difference Lower Upper Transparent
By looking at the Pearson correlation coefficient in combination with ANOVA and one-sample t-tests we were able to accept hypotheses 2a, 2b, and 3a. Based on this we can conclude that a higher value perception does impact both behavioral and attitudinal loyalty and the reward structure of the loyalty program has a positive impact on attitudinal loyalty. However, based on the outcomes of the test we were unable to accept hypotheses 1a, 1b and 3b. Interestingly, the analyses show that indeed transparency significantly impacts value perception, however in the opposite direction of our hypothesis. Less transparency has a positive effect on the value perception of the loyalty program. Also, if we look specifically into transparency of timing we do not find significant evidence of a positive relationship between transparency and value perception Regression analysis So far, the different variables and their impact on both behavioral and attitudinal loyalty have been tested in isolation of one another. To test their interaction, we’ve performed an ANOVA regression analysis Regression analysis I: attitudinal loyalty First we test the impact of value perception, reward redemption and reward structure on attitudinal loyalty in one model. To measure the relationships among variables the statistical processes of regression is used. The variables that are entered were the reward structure, value perception & reward redemption to measure the effect on the dependent variable attitudinal loyalty. The R-squared of ,567 is the fraction of the variation in the attitudinal loyalty that is accounted for (or predicted by) the value perception, reward redemption and reward structure. This is the same as the square of the correlation between the dependent and independent variable.) Table 5 Model summary reward structure, Value perception and reward redemption Model Summary Model R R
Square Adjusted R Square Std. Error of the Estimate
1 ,753a ,567 ,550 ,57787 Predictors: (Constant), reward structure, Value perception, reward redemption In these results, the null hypothesis states that the mean hardness values of 3 different transparencies are equal. Because the p-value is 0.00, which is less than the significance level of 0.05, the null hypothesis can be rejected it can be concluded that some of the transparencies have different means. Table 6 ANOVA reward structure, Value perception and reward redemption ANOVA Model Sum of Squares df Mean Square F Sig 1 Regression 32,385 3 10,795 32,328 ,000b Residual 24,711 74 ,334 Total 57,096 77 a. Dependent Variable: Attitudinal loyalty b. Predictors: (Constant), reward structure, Value perception, reward redemption
27 The size of the coefficient for each independent variable gives an indication of the effect size that variable has on the attitudinal loyalty. Reward structure with a ,404 positive effect and Value perception with ,462 both have a significant effect on the attitudinal loyalty. Table 7 Coefficients reward structure, Value perception and reward redemption Coefficients Model Unstandardized Coefficients Standardized Coefficients B
Std. Error Beta t Sig.
1 (Constant) ,287 ,354 ,811 ,420 Value perception ,375 ,079 ,462 4,730 ,000 Reward redemption -,027 ,081 -,034 -,329 ,743 Reward structure ,421 ,113 ,404 3,725 ,000 a. Dependent Variable: Attitudinal loyalty Regression analysis II: behavioral loyalty Secondly, we test the impact of value perception, reward redemption and reward structure on behavioral loyalty in one model. To measure the relationships among variables the statistical processes of regression is used. The variables that are entered were the reward structure, value perception & reward redemption to measure the effect on the dependent variable attitudinal loyalty. The R-squared of ,714 is the fraction of the variation in the attitudinal loyalty that is accounted for (or predicted by) the value perception, reward redemption and reward structure. This is the same as the square of the correlation between the dependent and independent variable. Reward structure, Value perception, reward redemption on Dependent Variable: Behavioral loyalty Table 8 Model summary reward structure, Value perception and reward redemption Model Summary Model R R
Square Adjusted R Square Error of Std. the Estimate 1 ,845a ,714 ,702 ,49496 a. Predictors: (Constant), reward structure, Value perception, reward redemption In these results, the null hypothesis states that the mean hardness values of 3 different transparencies are equal. Because the p-value is 0.00, which is less than the significance level of 0.05, the null hypothesis can be rejected it can be concluded that some of the transparencies have different means.
Table 9 ANOVA reward structure, Value perception and reward redemption ANOVA Model Sum of Squares df Mean Square F Sig 1 Regression 45,217 3 15,072 61,524 ,000b Residual 18,129 74 ,245 Total 63,346 77 a. Dependent Variable: Behavioral loyalty b. Predictors: (Constant), reward structure, Value perception, reward redemption The size of the coefficient for each independent variable gives an indication of the effect size that variable has on the behavioral loyalty. None of the variables (Reward structure, Reward redemption and Value perception) have an effect that can be called significant on behavioral loyalty. Table 10 Coefficients reward structure, Value perception and reward redemption Coefficients Unstandardized Coefficients Standardized Coefficients Model B Std. Error Beta t Sig. 1 (Constant) ,808 ,303 2,664 ,009 ,808 Value perception ,587 ,068 ,687 8,643 ,000 ,587 Reward redemption -,111 ,070 -,133 -1,600 ,114 -,111 Reward structure ,358 ,097 ,327 3,706 ,000 ,358
a. Dependent Variable: Behavioral loyalty Summary of findings Overview of all correlation tests for the different hypothesis Table 11 Summary of findings Hypothesis 1a Reward structure Value perception Reward structure Pearson Correlation 1 ,687** Sig. (2-tailed) 0 N 152 152 Value perception Pearson Correlation ,687** 1 Sig. (2-tailed) 0 N 152 152 Transparent reward
structure Value perception Transparent reward structure Pearson
Correlation
1 ,584**
Sig. (2-tailed) 0
29 Value perception Pearson Correlation ,584** 1 Sig. (2-tailed) 0 N 78 152 Non-transparent reward
structure Value perception Non-transparent reward
structure Pearson Correlation 1 ,919**
Sig. (2-tailed) 0 N 76 76 Value perception Pearson Correlation ,919** 1 Sig. (2-tailed) 0 N 76 152 Hypothesis 1b Reward redemption Value perception Reward redemption Pearson Correlation 1 ,518** Sig. (2-tailed) 0 N 78 78 Value perception Pearson Correlation ,518** 1 Sig. (2-tailed) 0 N 78 152 Transparent timing of
reward redemption Value perception Transparent timing of reward
redemption Pearson Correlation 1 ,473**
Sig. (2-tailed) 0 N 78 78 Value perception Pearson Correlation ,473** 1 Sig. (2-tailed) 0 N 78 152 Non-transparent timing
reward redemption Value perception Non-transparent timing of
reward redemption Pearson Correlation 1 0,502
Sig. (2-tailed) 0 N 78 78 Value perception Pearson Correlation 0,502 1 Sig. (2-tailed) 0 N 78 152 Hypothesis 2a
Attitudinal loyalty Value perception Attitudinal loyalty Pearson Correlation 1 ,699** Sig. (2-tailed) 0 N 152 152 Value perception Pearson Correlation ,699** 1 Sig. (2-tailed) 0 N 152 152 Hypothesis 2b Behavioral Loyalty Value perception Behavioral loyalty Pearson Correlation 1 ,766** Sig. (2-tailed) 0 N 152 152 Value perception Pearson Correlation ,766** 1 Sig. (2-tailed) 0 N 152 152 Hypothesis 3a Reward structure Attitudinal loyalty Reward structure Pearson Correlation 1 ,630** Sig. (2-tailed) 0 N 152 152 Attitudinal loyalty Pearson Correlation ,630** 1 Sig. (2-tailed) 0 N 152 152 Hypothesis 3b Reward redemption Behavioral loyalty Reward redemption Pearson Correlation 1 ,432** Sig. (2-tailed) 0 N 78 78 Behavioral loyalty Pearson Correlation ,432** 1 Sig. (2-tailed) 0 N 78 152 Transparent timing of
reward redemption Behavioral loyalty Transparent timing of reward
31 Sig. (2-tailed) 0 N 78 78 Behavioral loyalty Pearson Correlation ,421** 1 Sig. (2-tailed) 0 N 78 152 Non-transparent timing reward redemption Behavioral loyalty Non-transparent timing of reward redemption Pearson Correlation 1 0,383 Sig. (2-tailed) 0,001 N 78 78 Behavioral loyalty Pearson Correlation 0,383 1 Sig. (2-tailed) 0,001 N 78 152
