How multi-touchpoint customer experiences are Associated with satisfaction and loyalty

How Multi-Touchpoint Customer Experiences are

Associated with Satisfaction and Loyalty

MSc. in Business Administration

Marketing Track

Jan-Aug 2015

Name: Yaeeun Han

Student Number: 10826157

Submission Date: 29

June 2015 (Final)

First Supervisor: Dr. Umut Konus

STATEMENT OF ORIGINALITY

This document is written by Yaeeun Han 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.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

TABLE OF CONTENTS

1 INTRODUCTION ... 5

2 LITERATURE REVIEW ... 8

2.1 Customer experience ... 8

2.2 Customer Touchpoints ... 10

2.3 Customer Touchpoint Classification ... 12

2.4 Effects of Multi-Touchpoint Experience on Satisfaction and loyalty ... 14

2.5 Control variables ... 16

2.6 Gaps and Research question ... 17

2.7 Contribution ... 18

3 CONCEPTUAL FRAMEWORK ... 19

4 HYPOTHESES ... 22

5 METHODOLOGY ... 25

6 DATA ANALYSIS AND RESULTS ... 27

6.1 Reliability ... 27

6.2.1 Equation model specification and results (Compared with RET data) ... 29

6.2.2 Results for hypotheses ... 31

7 DISCUSSION AND MANAGERIAL IMPLICATION ... 40

ABSTRACT

Customers experience the brand across various touchpoints and each interaction establishes a holistic customer experience during their shopping journey. Both researchers and practitioners have emphasized the importance of touchpoint management and customer experiences, yet it is not clear how multi-touchpoint customer experiences are associated with satisfaction and consumer behavior. Moreover, there was no research that studied customer experience at multi-touchpoint in a single research framework. In this study, 14 touchpoints, including firm-initiated (TV advertisement, direct mail, email, sponsorship, promotion event), customer-initiated (show-room visiting, purchase, usage, call-center, online website or SNS interactions), and other-initiated (looking at friend’s usage, word-of-mouth (WOM) and eWOM), are tested with survey data from 120 respondents on two categories (mobile telecom and soft drink). With this data, I explore (1) how positivity and frequency of the customer experiences at multi-touchpoint are associated with customer satisfaction, and (2) whether there are differences in result depending on the touchpoint types. Multiple regressions are used to test the hypotheses and the results show that customer satisfaction is mainly associated with positivity of touchpoint experiences, not with frequency. Further, other-initiated touchpoint has a stronger association with satisfaction than firm-initiated touchpoint while the effects of touchpoint experience vary greatly across touchpoint types and categories. This study compares the result with that of real-time experience tracking(RET) data from study of Jing, Macdonald, Konus, and Wilson (2015) to cope with the limitation of survey data, which relies on retrospective recall rather than instant response.

1 INTRODUCTION

Customer experience is getting more and more attention from both practitioners and academic researchers. Academic researchers describe the shift from traditional marketing towards “experiential” marketing (Schmitt 1999). Depending on industry and product category, experiential marketing seems to either complement or replace traditional marketing. That being said, there is a clear trend according to which the marketplace has changed from selling products and services to selling consumer experiences (Pine and Gilmore 1999). Simply offering a product or service that satisfies functional needs of the consumer will no longer guarantee the brand’s competitiveness.

People experience the brand in every interaction with the company and so that the customer experience is holistic in nature (Verhoef, Lemon, Parasuraman, Roggeveen, Tsiros, and Schlesinger 2009). According to Gentile, Spiller, and Noci (2007, p.397) the customer experience “originates from a set of interactions between a customer and a product, a company, or part of its organization, which provoke a reaction”. Verheof et al. (2009) advanced the definition by adding the construct that this experience encompasses the interactions that companies are not able to control and includes every phase of the shopping journey - before, during, and after the purchase. The study of customer experiences at touchpoint is getting more attention as customers increasingly encounter the brand through multi-touchpoints and technological development enables companies to contact customers in diverse ways. The impact of each touchpoint on customer satisfaction varies. As there are more and more ways to reach customers, it is strategic for practitioners to prioritize each touchpoint in order to align investment with budget constraints. In that sense, mastering an effective multi-touchpoints strategy is a topic of growing importance though the effects of a multi-touchpoint approach

remain unclear. Practitioner-oriented journals or management books, mostly discussed managerial actions and outcomes of customer experience while the studies on this topic from academic literature have been limited to suggest generic theoretical constructs and conceptual models. Most focus of the academic research is on the perception perspective of customer experience which can be broken down into four dimensions: sensory, affective, intellectual, and behavioral (Brakus, Schmitt, and Zarantonello 2009). These articles take a theoretical approach by conceptualizing customer experience as subjective and behavioral responses evoked from the brand-related stimulus. Some other articles focus on finding out strategies to maximize customer experience or behavior outcome at one specific touchpoint. For example, Assmus, Farely, and Lehmann (1984) and Ansari and Joloudar (2011) have studied the effect of TV ads on customer experience. Although the broad set of research literature studied customer experience at one specific touchpoint, there was no research that studied customer experience at multiple touchpoints in a single research framework.

The purpose of this research is to investigate how multi-touchpoint customer experiences are associated with customer satisfaction and loyalty. As customer experiences have a major influence in determining customer satisfaction (Jing et al. 2015) and satisfaction has a strong impact on post-purchase intentions and loyalty (Dabholkar and Thorpe 1994; Anderson and Sullivan 1993; Oliver 1997), it is important to examine the relationship between customer experience and satisfaction. To specify, this study will first measure whether the positivity and the frequency of customer experience at touchpoint are associated with customer satisfaction. Also, the study will test whether there is a stronger association with loyalty if the customer experiences the brand at touchpoint more often. Therefore, the main research question of the study is: How are positivity and frequency of the customer experiences at multi-touchpoint

associated with customer satisfaction and loyalty? Second, by dividing touchpoints into different

groups, the study also wants to find out whether there are differences in result depending on the group. Touchpoints will be categorized into groups by following the standard from the study of Jing et al. (2015): firm-initiated touchpoints (FITs), customer-initiated touchpoints (CITs), and other-initiated touchpoints (OITs). The ultimate goal of this study is to find out the association of multi-touchpoint experiences on customer behavior and attitudes.As no study has done research about customer experience at multiple touchpoints in a single research frame, it will be possible to find out the meaningful implication by comparing the result from this study. Finally, in this study I will use online survey data to investigate the association and then compare them with the findings from Jing et al. (2015). Their study used Real-time Experience Tracking (RET) data tracking customers 24 hours a day by gathering quick SMS-based micro survey whenever they experience an interaction with the company whereas online survey asks customers to recall their experience. Therefore, the results from the research of Jing et al. (2015) will be compared with the findings of the present study to check validity.

This research has important implications for both marketers and academic researchers. The findings of the study offer an overview of customer experiences across touchpoints and their impact on satisfaction and loyalty. Companies that manage multi-touchpoint customer experiences skillfully can get many benefits such as improved customer satisfaction, increased revenue, and more effective collaboration within the organization (Rawson, Duncan, and Jones 2013). Also, the result of the study may help marketing managers to identify which marketing strategy to take. For example, if the positivity of the customer experience matters more on satisfaction and loyalty than frequency, the firm should focus on creating positive experiences. On the other hand, if frequency matters more than positivity, the company should focus on

making their brand more visible. The research explores academic knowledge on the relationship between customer experience and its effect on satisfaction and loyalty. Also, this study provides empirical evidence on whether positivity and frequency of customer experience have an association with satisfaction and loyalty.

The research constructed as follows. First, relevant academic literatures will be reviewed to clarify key concepts and the relationship among them: customer experience, customer touchpoints and its classification, and the relationship between customer experience and satisfaction and loyalty. From there, the research gap and research question are presented and followed by the contribution of the study and the conceptual framework. Hypotheses and analyses are then developed, followed by discussions, limitations, and reference list.

2 LITERATURE REVIEW

2.1 Customer experience

Customer experience is a broad concept that many have attempted to define. Definitions of customer experience include: “Customer Experience is the internal and subjective response customers have towards any direct or indirect contact with a company. Direct contact generally occurs in the course of purchase, use, and service and is usually initiated by the customer. Indirect contact most often involves unplanned encounters with representatives of a company’s products, service or brands and takes the form of word-of-mouth recommendations or criticisms, advertising, news reports, reviews and so forth” (Meyer and Schwager 2007, p. 118). In addition, customer experience is holistic and encompasses the total experience from all the phases of shopping experience, which includes information search, purchase, use, and after-sale service

For instance, imagine there is someone who would like to purchase a product. This person might start to look for information online and encounter with firm-initiated content as well as with online reviews from other customers. That same person might visit an offline store and try out the actual product or simply ask their opinion to friends who are already using the product. There can also be spontaneous encounters with ads of the product on TV, newspaper, and billboard and it is possible to accidently hear people having conversations about the product. All these illustrations compose a holistic customer experience and therefore every encounter between customers and brands is a customer experience.

Several conceptual models of customer experience have been proposed and most of the academic focus is on conceptualizing and categorizing the determinants of customer experience. Verhoef et al. (2009) proposed social environment, service interface, retail atmosphere, assortment, and price as the main drivers of customer experience. They took a dynamic view, in which they insist that prior customer experience and their experience in alternative channels will impact the overall customer experience. The paper proposed a summarized overview of literature on each driver and provided a holistic portrayal of the customer experience, however it left room for further research on each driver. Brakus et al. (2009) focused on building a brand experience scale that is reliable, valid and includes four dimensions: sensory, affective, intellectual, and behavioral. There seems to be consensus on the categorization from diverse existing literature on the aforementioned four dimensions and it covers holistic response that consumer might have during their experience. Many studies have focused on the perception perspective of customer experience, however, considering that customer experiences occur at various touchpoints, it requires further research on the encounters where customers and companies meet.

2.2 Customer Touchpoints

Customer touchpoints refer to a medium or contact point where the customer and the company interact (Neslin, Grewal, Leghorn, et al. 2006). Dhebar (2013) defines customer touchpoints as “points of human, product, service, communication, spatial, and electronic interaction collectively constituting the interface between an enterprise and its customers over the course of customers’ experience cycles” (Dhebar 2013, p. 200). This means that touchpoints encompass every moment and place where interaction between the brand and customers occur. Customers increasingly experience the brand through multiple touchpoints and due to technological development, companies contact customers in diverse and new ways (Edelman 2010). For instance, marketers used to deliver their message through paid-media such as magazines, newspapers, and TV, in one-way communication without knowing specifically what types of target group they will be exposed to. The Internet brought about the possibility for such message to be spread out to a large number of people and the ability for marketers to offer targeted, customized, and sophisticated messages or experiences to customers (Edelman 2010). In addition, marketers now have to consider diverse devices such as smartphones, tablets, PCs, and wearable devices. As an increasing number of people use these devices, it is important to have suitable interactions at each touchpoint. Above all, technological development integrates diverse channels and devices in providing experiences to customers.

Among all the touchpoint experiences, some have stronger and more long-lasting impact on customers than others.From the company’s point of view, as there are more and more diverse ways to reach their customers, it becomes strategic to know which touchpoints to prioritize within the boundaries of their limited budget. Therefore, it is critical to uncover high-impact touchpoints to then allocate budget accordingly. Hogan, Alquist, and Glynn (2005) said that

many touchpoints have small positive impact on the brand, but if they fail to meet the need of customer, it will result in negative influence. Besides, the impact varies depending on the touchpoints. For instance, according to their research, interaction with electronic forms has strong negative impact when they fail and hardly satisfy customers while touchpoints involving people have huge potential in delighting customers. Employees’ unexpected friendly service, empathy, and solution on complicated problems exemplify the way companies satisfy customers and offer the experience of tremendous impact (Hogan et al. 2005). In uncovering high-impact touchpoints, the firm should be careful not to look solely at each interaction rather than considering the whole consumer decision journey (Edelman 2010). Simply perfecting touchpoint experiences is not enough – such narrow view will give company a distorted picture and it is not anymore enough to have true competitive advantage. The company should keep in mind that whole experience at multiple touchpoints should offer more than the sum of experiences at each touchpoint (Dhebar 2013). In other words, the company should focus on managing cumulative experiences from multiple touchpoints over time (Rawson et al. 2013) and the multi-touchpoint architecture should be comprehensive and holistic. Fast moving companies already have more focus on whole customer journey experiences than on single touchpoint experiences (Rawson et al. 2013). It is not easy to shift and make a decision based on journey-orientated strategies, but the reward is truly worthwhile (Rawson et al. 2013). Further, Macdonald, Wilson, and Konus (2012) present a radical new tool in marketing research, Real-time Experience Tracking (RET) which tracks customers 24 hours a day by gathering quick SMS-based micro-survey whenever they experience interaction with the company. This new mobile-phone based tool enables marketers to get a holistic view on multi-touchpoints customer experience. With this real-time

data, marketers are able to identify not only key driver touchpoints that have the most impact on consumer behavior but also the overall interrelation among those touchpoints.

2.3 Customer Touchpoint Classification

In order to have overview of customer experience at touchpoint, it is important to have proper touchpoint classification. Marketing researchers often classify between paid, owned, and earned consumer contact points (Corcoran 2009; Goodall 2009; Boncheck 2014). Paid media is a traditional advertising channel such as TV and radio commercials, direct mail, and roadside billboards that the company pays to leverage the brand. While, earned media is word-of-mouth that consumers share and is usually seen in the viral form. Consumers even create the contents so that it is not under control of the company. In contrast, owned channel is created and owned by the company so that it is under direct control of the company. Corporate web site, Facebook fan page, official blog, and brand community can be an example (Stephen and Galak 2012; Harrison 2013; Boncheck 2014; Edelman and Salsberg 2010). Harrison (2013) indicates that the brand is able to reach in an easier way to greater number of non-buyers and buyers in paid media, while earned and owned media have high-reach contact points among existing buyers. Furthermore, Nielsen’s Global Survey of Trust in Advertising (2013) reveals that the consumers trust earned media the most and the credibility of owned media is rising. The survey results show that the earned media even have an impact on consumer actions such as purchase. With respect of paid media, the previous marketing literature extensively studied the effects of it, especially advertising, on sales. Whereas recent global study from Nielsen (2012) reports that people do not trust much traditional paid advertising. Still paid media is important in increasing brand

awareness so that it is hard to say which one is the most critical touchpoint type. Above all, the company should be strategic in managing and utilizing touchpoints from different types.

In this study, I will classify touchpoints based on two criteria and categorize in a meaningful way. First, consumer experiences can be divided into direct and indirect experiences. Direct experience includes physical product experience such as trial and mostly occurs when consuming the product. Indirect experience refers to the experience that occurs through other non-physical product experience such as marketing communication during an advertising campaign. When making a decision of purchase, consumers rely on both direct and indirect experiences. In most cases, they spent more time on indirect experiences such as searching product information online. However, their satisfaction with the product tends to be based on direct experience such as actual usage. Previous research explains the reason as direct product experiences provide consumers more complex, credible information and will trigger a more concrete mental construal (Hamilton and Thompson 2007). But in most cases, it is costly or impossible for companies to engage in direct experience with customers, such as product trial, although it is an effective way to form preferences. For that reason, customers more often and more easily encounter with indirect experiences and rely on them. The present study will consider customer experiences containing actual, physical product as direct experiences. Among all the touchpoints included in the study, some may contain direct product experience and others may not. Therefore, it will be interesting to comparing two different types of experience and find out which experience has more notable impact on customer satisfaction. Secondly, touchpoints can be separated into firm-initiated touchpoints (FITs), customer-initiated touchpoints (CITs), and other-initiated touchpoints (OITs). The FITs include any encounter initiated by brands as exemplified by advertisement and promotions through diverse channels. Customers might

respond passively to FITs but marketers have better control over FITs compare to CITs. On the contrary, CITs are defined as any encounter with a manufacturer that is initiated by a customer. CITs used to be limited to customer complaints because customers would initiate the interaction mostly when there is a problem with the product. But the concept has broadened as more and more customers actively engage and initiate interaction with firms due to development of technologies (Bowman and Narayandas 2001). For instance, customers aggressively search for information online and actively engage with companies via social networks services. Recent studies show that CITs have higher response rates and prove more influential than FICs. As a consequence it becomes more important to managing CITs (Wiesel, Pauwels, and Arts 2001; Shankar and Malthouse 2007). Lastly, and to complete the classification, other-initiated touchpoints (OITs) refer to interactions and encounters among customers. OITs are important as the Internet has empowered consumers to communicate with one another. On the Web, consumers can easily reach out to other consumers’ opinions, reviews, and may even create their own content such as tutorial video or parodies of existing ads. Marketers have no control over OITs while these have a tremendous effect on changing other consumer’s attitude regarding brand. Therefore, firms should strategically think about how to respond, react and make the best use of OITs (Fournier and Avery 2011). To sum up, touchpoints can be classified into three types – firm-initiated, customer-initiated, and other-initiated – or simply into two types – direct and indirect. In this study, both typologies are applied when investigating the research question.

2.4 Effects of Multi-Touchpoint Experience on Satisfaction and loyalty

Satisfaction has been one of the most important topics in the marketing so that many companies have implemented marketing strategy in a direction to maximize customer

satisfaction. Since mid-1970s researches about this topic provide valuable insight into post-purchase phenomena. Early academic literatures focused on complaining behavior of consumers while as the field progressed, the focus is moved to the questions why and how consumers became satisfied. Oliver (1997) defined satisfaction as pleasurable fulfillment meaning that consumption fulfills some need or desire of consumer and consumer senses this fulfillment as pleasurable. Satisfaction is not just linked to cognitive judgment but also to emotional and affective reactions to purchase situation. (Mano and Oliver, 1993) Therefore, satisfaction might be shaped by positive emotional reaction aroused from every encounter between customers and brands. Although consumer satisfaction is not simply formed by single experience and customer experience is not the only determinant, over time, the experience which is long-lasting and stored in consumer mind have impact on consumer satisfaction. Further, as individual superior brand experience becomes aggregated and formulates preference over other brands, it builds loyalty (Brakus et al, 2009). Oliver (1997) presented four loyalty phases as cognitive, affective, conative,

and action loyalty. To specify, consumers first have cognitive loyalty based on knowledge or

experience-based information and at the second phase, affective loyalty where an attitude toward the brand has developed and this reflects pleasurable fulfillment dimension of the satisfaction. The next phase, conative loyalty refers to behavioral intention that consumers become committed to repurchase the brand. Finally, action loyalty is so called true loyalty and this phase refers to the commitment of consumers coupled with the overcoming of obstacles to reach the brand. When we look at some articles defining the relationship between customer experiences, satisfaction, and loyalty, Brakus et al. (2009) mentioned that evoked experience from brand itself has direct effect on satisfaction and loyalty. Although many previous studies reveal direct relationship between customer experience and satisfaction, but researchers have divergent

opinions on whether there is a direct relationship between customer experience and loyalty. For example, Iglesias et al. (2011) conduct quantitative research and propose that the relationship between brand experience and brand loyalty is mediated by affective commitment which refers to “customers’ emotional attachment to particular brand or store based on their identification with that store or brand” (Iglesias et al. 2011, p. 572). Their study does not support the fact that there is direct relationship between brand experience and brand loyalty. On the contrary, field study of Sahin et al. (2011) support that brand experience has a direct positive effect on brand loyalty.

In this study, I will focus on how customer experience associated with the affective and attitudinal dimensions of both satisfaction and loyalty. This is due to the fact that this is cross-sectional study so that it makes more sense to measure affective satisfaction, and loyalty shaped by multi-touchpoint experiences rather than measuring behavioral outcomes. In addition, I regards loyalty as consequence of satisfaction, since dominant opinion from academic studies have been shown the relationship between the customer experience and loyalty appears to be mediated by satisfaction (Anderson et al. 1993; Oliver 1997). Therefore, with the premise that holistic experiences shape satisfaction, this research will analyze customer experience at multi-touchpoints to understand its association on customer satisfaction and loyalty.

2.5 Control variables

To control the influence of demographics, three common demographical factors are included in this research: age, gender, and education level. Additionally, the research includes three psychographic factors to control impact of them on customer satisfaction: innovativeness, consumer susceptibility to interpersonal influence, and buying impulsiveness. First,

innovativeness can be conceptualized as consumer’s desire to look for new product information. Manning, Bearden, and Madden (1995) term this attribute as consumer novelty seeking (CNS) and have developed highly reliable and valid scale to measure this attribute and this study use the scale to measure consumer innovativeness. Second, Bearden, Netemeyer, and Teel (1989) defined consumer susceptibility to interpersonal influence as the need to identify self with other’s opinion and the tendency to learn about product through information from others. Lastly, impulsiveness refers to cousumers’ tendency to buy spontaneously so that impulsive consumers are more likely to open to unexpected purchase (Rook and Fisher 1995). They tend not to spend much time and put lot of effort when purchasing product.

2.6 Gaps and Research question

Although there were literatures previously done and supported that brand experience has positive effect on customer satisfaction and loyalty, most of the previous theoretical research was conduct based on general brand experience rather than experience occurring at specific touchpoints. There are many empirical studies that examined the impact of specific touchpoint experiences on customer behavioral outcome (i.e., Ansari and Joloudar 2011; Assmus et al. 1984), but they were limited to single type of touchpoint rather than multi-touchpoints. Further research opportunity is there to fulfill the gap by investigating study based on multi-touchpoint customer experiences. In current study, the research would extend to construct holistic multi-touchpoint customer experience and focus on finding out association between each multi-touchpoint customer experience and customer satisfaction, and loyalty. Therefore, the research question will be:

How positivity and frequency of the customer experience at multi-touchpoints are associated with customer satisfaction and loyalty?

Accordingly, research objective is to identify customer experience at which touchpoint has strong impact on customer satisfaction and loyalty.

2.7 Contribution

As the present study takes into account holistic customer experiences, the findings of this study will have both academic and managerial implications. Academically, the study provides an overview of customer touchpoint experiences and their impact on customer behavior. There have been many researches investigating the impact of limited types of touchpoints on customer behavior but few studies conducted on multiple touchpoints in a single research framework. The results of the research conducted on limited types of touchpoints did contribute to both academic researchers and marketers for specific touchpoints. However, in practice, touchpoints are diverse, interrelated, and every encounter could affect the customer’s overall satisfaction so it is important to oversee the entire experiences at multiple touchpoints. To specify, there are many other studies looking at the impact of TV ads on consumer behavior (Assmus et al. 1984; Weilbacher 2003). In general, those studies assume that satisfaction or increasing sales come from specific ads. But there might be many other different activities that a given firm is engaging at the same time such as promotion at the point of sales, or spreading a viral video on YouTube. Therefore, in reality, it is hard to distinguish which touchpoint and activity are actually contributing to satisfaction and affecting consumer behavior. The present study will make this relationship visible by conducting a study on multiple touchpoints in a single research framework. The study will also provide empirical evidence whether positivity and frequency of

the customer experience are associated with satisfaction and loyalty. Above all, the result provides an overview on what type of touchpoint has stronger influence on satisfaction and loyalty among firm-initiated, customer-initiated, and others-initiated touchpoints.

For marketers, the result of this study helps to distinguish the touchpoint which has stronger impact on customer satisfaction. As the present study examines multi-touchpoints experience in a single research, it is possible to extract which of them really matter for customer satisfaction. Marketing managers will be able to better understand multi-touchpoints approaches and to strategically manage customer experience and interaction at each touchpoint. Most importantly, the result of the study provides guidelines for managers when they make a decision on which touchpoint to invest. Considering the fact that most of the company and brand have limited amount of marketing budget, giving an investment priority to the touchpoint with stronger impact is one of the key decisions. This study is expected to reveal out which touchpoint is strongly correlated in increasing satisfaction and loyalty and this result can be used as reference for marketing managers.

3 CONCEPTUAL FRAMEWORK

As customer experiences can vary in strength, long-lasting experience has stronger impact on customer satisfaction and loyalty (Oliver 1997). Building on this theoretical premise, in this study, positivity and frequency will be used as two variables to measure the strength of the customer experience. Positivity refers to whether the customer experience at a given touchpoint is positive or negative while frequency refers to how many times the customer experiences the brand at specific touchpoints. Therefore, the basic research question of current study is whether positivity and frequency of customer experience have an impact on satisfaction and loyalty.

Since the main purpose is to find out at which touchpoint this relationship is remarkable, the study will be conducted in a setting of multi-touchpoints experiences.

In order to propose an overview of customer experience at specific touchpoints, it is important to include diverse touchpoints and categorize them in a meaningful way. The study includes 14 touchpoint experiences spread across FITs, CITs, and OITs: Outdoor event, flagship/pop-up store, TV ads, direct mail, email, and sponsorship as FITs; Store visit and trial, usage occasion, call center, and online information search (official website/SNS) as CITs; Friend’s usage occasion, WOM (Word Of Mouth), and eWOM (electronic Word Of Mouth) as OITs. As mentioned in the literature review, touchpoints are firstly classified in terms of initiator and also divided into direct and indirect experiences. Since this research is not able to contain all the experiences, 14 representative touchpoints are included. Also, by including both online and offline touchpoints in the study (direct mail and email; WOM and eWOM), it is expected to reveal differences among those touchpoints. Lastly, to clarify, Friend’s usage occasion under OITs refers to the direct experience that consumer comes across the product during their friend’s product usage. We propose to include this touchpoint to reveal whether direct experience has stronger impact on satisfaction at OITs level. Classification of 14 touchpoints is summarized in table1.

Class. 1 Class. 2 Touchpoint

Firm-initiated Direct Outdoor event, pop-up store (Showroom) Indirect TV ad, Direct mail, Email, Sponsorship,

Customer-Initiated

Direct Store visit and trial, Usage occasion

Indirect Call center, Online information search (Website / SNS) Others-Initiated Direct Friend's usage occasion

Indirect WOM, eWOM (conversation)

Table1. Classification of touchpoints

brand loyalty. Oliver (1999) suggests six possible associations of satisfaction and loyalty. This includes suggestion that satisfaction is starting point of the loyalty and that satisfaction and simple loyalty are components of ultimate loyalty. The research scope of the present study is limited to investigating the effect of customer experience on satisfaction and loyalty while taking satisfaction as mediator.

Last but not least, demographic and psychographic factors are included as control variables to enrich the research. The demographic factors are age, gender, and education level and the psychographic factors are innovativeness, consumer susceptibility to interpersonal influence, and impulsiveness. The conceptual framework overview of the research is summarized in figure 1.

4 HYPOTHESES

The present study will measure two attributes impacting satisfaction: positivity and frequency of customer touchpoint experiences. Positivity relies on the customer’s expectation as well as the level of disconfirmation and it impacts the satisfaction level (Anderson and Sullivan 1993). Generally, the more customer experience is positive, the higher the customer satisfaction. Frequency is measured by how often the customer experiences the brand at specific touchpoints and it has an influence on customer satisfaction (Bolton and Lemon 1999). Previous literature already revealed that both variables have a positive relationship with satisfaction (Anderson et al. 1993; Bolton et al. 1999). With that premise, I would like to examine which variables have a stronger association with customer satisfaction. I expect positivity to have a stronger association than frequency with customer satisfaction in every touchpoint because positivity is directly related with product performance and quality perceived by consumers which are both associated with satisfaction (Westbrook et Oliver 1991). Therefore, the following hypothesis is developed:

H1: Positivity of customer experiences has a stronger association with customer satisfaction across FITs, CITs, and OITs than frequency of customer experiences.

Secondly, between positivity of direct and indirect customer experiences, I expect that direct ones have a stronger association with customer satisfaction. As mentioned in the literature review, Hamilton and Thompson (2007) explained different effect of direct and indirect product experience on mental construal, saying direct experience creates a more concrete mental construal than indirect experience. According to psychology literature of Regan and Fazio (1977), the person who formed attitude from direct experience proves more confident and behaves more consistently than the person who formed attitude on the basis of indirect experience. Fazio and Zanna (1978) explained that the increased confidence comes from the fact that a direct

experience provides more information about the object than an indirect experience. Therefore, direct experiences that provide concrete association about the product is more likely to create customer preference and satisfaction. The following hypothesis is thus developed:

H2: Positivity of direct customer experiences has a stronger association with customer satisfaction across FITs, CITs, and OITs than indirect customer experiences.

Customer loyalty is derived from complex reasons and experiences, which formed over time. Therefore, it is hard to find a causal relationship between touchpoint experiences and loyalty in this cross-sectional study, which relies on customer’s experience only from prior two month. Yet I expect to observe an association between the number of touchpoints consumers encounter and loyalty as some previous research show the empirical evidence that customers interact with multichannel on average buys more than single channel customers and multi channel strategy enhance customer loyalty (Neslin and Shankar 2009; Wallace, Giese, and Johnson 2004). In addition, Keller (2001) said when customers reach resonance level at the customer-based brand equity (CBBE) model, they are not only attitudinally attached but also behaviorally engaged with the brand. On the contrary, this means that customers who actively participate in interactiing with the brand are more likely to have a higher interest or even have loyalty to the brand. Customers encountering with many different types of touchpoint - especially with CITs, which initiated by customers - tend to actively engage in the brand activities. Therefore, consumers who encounter with many different types of touchpoints are expected to have a stronger association with loyalty. Thus, the following hypothesis is developed.

H3: People who encounter with many touchpoints in general are more loyal to the brand.

Among touchpoint groups, the company has the most control on FITs yet people tend not to trust the message initiated by the firm. Conversely, Nielsen global study reports that a

majority of their respondents trust the recommendation from others, earned media. As the brand does not have control on earned media, namely OITs, customers consider those touchpoints as trustworthy. Also, they might be more influenced in forming their attitude on the brand from those touchpoints. In that sense I expect that positivity of OITs has a stronger association with satisfaction than the positivity of FITs.

Furthermore, to test whether demographic factors are associated with positivity of OITs and customer satisfaction, I developed the following two hypotheses by including gender and age variables as moderators. With respect to gender, women are more likely to share personal stories with a friend than men; meaning women are natural WOM spreaders. Also, the study of Gabarino and Strahilevitz (2004) indicates that, when receiving a site recommendation from a friend, women are more likely to increase their purchase intention online than men. Females are also more easily influenced by the online consumer reviews, both positive and negative, than male (Bae and Lee 2011). Therefore, I expect women exhibit a stronger association between positivity of customer experience at OITs and satisfaction than men. In addition, younger consumers search more for product information and are more strongly influenced by peers (Mandrik, Fern, and Bao 2005). Therefore, positivity of OITs from younger people is expected to have a stronger association with satisfaction than that of older people. Therefore, the following hypotheses are developed.

H4a: Positivity of customer experiences at OITs has a stronger association with customer satisfaction than positivity of customer experiences at FITs

H4b: Positivity of customer experiences at OITs has a stronger association with satisfaction when people are female.

H4c: Positivity of experience at OITs has a stronger association with satisfaction when people are younger.

To test whether personal characteristics are related to customer satisfaction, the following psychographic variables were tested: innovativeness, consumer susceptibility to interpersonal influence, and buying impulsiveness. As mentioned in the literature review, innovative customers enjoy and have the motivation to seek out the product information (Manning et al. 1995). They look up several channels to maximize their utilities of search (Konus, Verhoef, and Neslin 2008) and they are more likely to initiate interaction with the brand. In that sense I expect innovative consumers tend to have a stronger association between CITs and customer satisfaction than other touchpoints. On the other hand, consumers susceptible to interpersonal influence are more likely to care about other people’s opinion. The definition of Bearden et al. (1989) about the construct includes a tendency to seek information and learn about the product by observing others. In that sense, consumers susceptible to interpersonal influence are expected to be more likely to be influenced by customer experiences of OITs. Therefore, I expect thatpositivity of OITs has a stronger association with satisfaction when consumers are susceptible to interpersonal influence. The following hypotheses are thus developed:

H5: Positivity of customer experiences at CITs has a stronger association with customer satisfaction when people are innovative.

H6: Positivity of customer experiences at OITs has a stronger association with customer satisfaction when people are susceptible to interpersonal influences.

5 METHODOLOGY

differences and also to increase external validity of the result. The online questionnaire was consisted of two sections. The first section contained questions regarding frequency and positivity of customer experience at multi-touchpoint. Respondents were asked to indicate a brand of their choice respectively in the categories of mobile telecom and soft drink before starting the survey. The main question from this section was to ask how many times (frequency) they were exposed to customer experiences of the brand at 14 different touchpoints within the last 2 months and how positive (positivity) those experiences were in five-point scales (1 = very negative, 5 = very positive). 14 touchpoints including FITs, CITs, and OITs were presented in the survey. The second section of the questionnaire asked about customer satisfaction and loyalty. Respondents had to share their level of satisfaction and loyalty to the brand they choose (‘I am happy with the efforts the brand is making towards regular customers like me’, ‘I am satisfied with the relationship I have with this brand’, ‘I feel loyal towards this brand’, ‘Even if this brand is more difficult to reach, I would keep buying from them or using them’). The scales measuring satisfaction and loyalty are valid and reliable as they are commonly used in academic literatures as well as in practice. In addition, three types of psychographic scales were developed by previous academic research. Innovativeness from Manning at el. (1995), susceptibility to interpersonal influence defined by Bearden et al. (1989), and impulsiveness from Rook et al. (1995) are valid scales that have been commonly used.

The collected survey is self-reporting data and it does not rely on instant response of customer at every experienced touchpoint. Therefore, this research will compare the result from survey data with the results from the study of Jing et al. (2015), which analyzed RET data. By comparing two results, it is possible to check whether the results found with RET data confirm the results of the present study. Also, the comparison will stand out whether there is any

difference between how customers respond and behave in real-time and how they respond in self-reporting survey.

The questionnaire was formulated via the online survey software Qualtrics and distributed through email and SNS to the samples. The survey was conducted by non-probability sampling with snowball sampling technique. The receiver of the questionnaire spread it to other samples they knew on their SNS or through email.

6 DATA ANALYSIS AND RESULTS

The analyses of this study were performed using SPSS and involved regression analysis to examine whether frequency and positivity of customer touchpoint experiences could be used as predictors of satisfactionor loyalty. In total, appropriately completed questionnaires were 120 for both mobile telecom and soft drink categories. Main respondents are South Koreans (South Koreans = 73.3 percent, Dutch = 6.7 percent, other Asian country = 4.2 percent, other European country = 10.8 percent, non of above = 5 percent), including both genders (male = 34.2 percent, female = 65.8 percent). The counter-indicative item to measure psychographic impulsiveness variable, “I carefully plan most of my purchases”, were recoded.

6.1 Reliability

To verify whether some items should not be used for analysis or whether all the items in one scale measure the same, the Cronbach’s alpha of these variables was tested. A Cronbach’s alpha of > 0.60 is considered acceptable and in a stricter way, it should be > 0.70.

Average of satisfaction and loyalty measured in 6 points scale with two items were calculated for both industries. Cronbach’s alpha of satisfaction and loyalty for mobile telecom

“innovativeness, consumer susceptibility to interpersonal influence, and buying impulsiveness” are computed to average scale. The Cronbach’s alpha of these variables are higher than 0.60. The summary of reliability of scales can be found in the Table 2.

Table 2: Reliability of Scales

Variables Cronbach's _Alpha N of Items Mobile Telecom Satisfaction 0.81 2

Mobile Telecom Loyalty 0.74 2

Soft Drink Satisfaction 0.64 2

Mobile Telecom Loyalty 0.78 2

Innovativeness 0.83 2

Influence 0.62 2

Impulsiveness 0.65 2

For the frequency items of customer touchpoint experiences, I summed frequencies of six touchpoints of FITs into one variable MTF_FIT (Mobile Telecom Frequency of FITs) have done the same for five touchpoints of CITs and three touchpoints of OITs. In the end, I have three variables for positivity (MTP_FIT, MTP_CIT, MTP_OIT) and three variables for frequency (MTF_FIT, MTF_CIT, MTF_OIT) of mobile telecom industry. The same logic applies to the soft drink category. The summary of computed scale can be found in the Table 3.

Table 3. Summary of Computed Scale

MTP_FIT = Mean(MTP_TVad, MTP_DM, MTP_EM, MTP_spons, MTP_promo, MTP_Sroom) MTP_CIT = Mean(MTP_Purch, MTP_Use, MTP_CC, MTP_Web, MTP_SNS)

MTP_OIT = Mean(MTP_Fuse, MTP_WOM, MTP_EWOM)

MTF_FIT = Sum(MTF_TVad, MTF_DM, MTF_EM, MTF_spons, MTF_promo, MTF_Sroom, ) MTF_CIT = Sum(MTF_Purch, MTF_Use, MTF_CC, MTF_Web, MTF_SNS)

MTF_OIT = Sum(MTF_Fuse, MTF_WOM, MTF_EWOM)

SDP_FIT = Mean(SDP_TVad, SDP_DM, SDP_EM, SDP_spons, SDP_promo, SDP_Sroom) SDP_CIT = Mean(SDP_Purch, SDP_Use, SDP_CC, SDP_Web, SDP_SNS)

SDP_OIT = Mean(SDP_Fuse, SDP_WOM, SDP_EWOM)

SDF_FIT = Sum(SDF_TVad, SDF_DM, SDF_EM, SDF_spons, SDF_promo, SDF_Sroom,) SDF_CIT = Sum(SDF_Purch, SDF_Use, SDF_CC, SDF_Web, SDF_SNS)

6.2.1 Equation model specification and results (Compared with RET data)

In this section, before start to test hypothesis, multiple regression analysis was done to see the effect of the frequency and positivity of FITs, CITs, and OITs to construct  satisfaction with the following two equations. This analysis was done to compare with the results from RET data (Jing et al. 2015).

First equation explains how frequency of FITs, CITs, and OITs are associated with satisfaction:

YF = 𝜶 + 𝜷1 X1 + 𝜷2 X2 + 𝜷3 X3 + 𝜺

YF = Satisfaction (measured in scale 1 to 5) 𝛼 = Constant

𝑋1 = Frequency of FITs (measured in number), 𝛽1 = Coefficient of X1

𝑋2 = Frequency of CITs (measured in number), 𝛽2 = Coefficient of X2

𝑋3 = Frequency of OITs (measured in number), 𝛽3 = Coefficient of X3 𝜀 = Error term

The second equation model explains how positivity of FITs, CITs, and OITs are associated with satisfaction:

Yp = 𝜶 + 𝜷1 X1 + 𝜷2 X2 + 𝜷3 X3 + 𝜺

Yp = Satisfaction (measured in scale 1 to 5) 𝛼 = Constant

𝑋1 = Positivity of FITs (measured in scale 1to 5), 𝛽1 = Coefficient of X1

𝑋3 = Positivity of OITs (measured in scale 1to 5), 𝛽3 = Coefficient of X3 𝜀 = Error term

Effects of frequency variables. The survey data results show that frequency does not significantly associated with satisfaction. From the frequency equation, the standardized coefficients do not have a significant meaning (p > 0.05) in every touchpoint type and in both categories. The results of RET data confirms the findings by revealing that in most cases of their data shows the volumes of touchpoint had no effect on customer satisfaction. Therefore, the results from both self-reporting and RET data prove that frequency variables do not associated with satisfaction.

Effects of positivity variables. The survey data results show that positivity of touchpoint

has a significant association with satisfaction in all touchpoint types except for FITs (β = 0.150;

p = 0.132) of mobile telecom. In mobile telecom category, CITs have the strongest association

with satisfaction (β = 0.312; p < 0.01) while OITs are associated the strongest with satisfaction (β = 0.380; p < 0.01) in soft drink category. From RET data, positivity of CITs significant in every category but positivity of FITs, OITs varies depends on category. In both data, the results clearly show that positivity of touchpoint customer experience, not frequency has association with satisfaction. Results are shown in Table 4.

Table 4. Results of equation model

Dependent Variables Satisfaction of Mobile Telecom Satisfaction of Soft Drink Independent Independent

Frequency Positivity Frequency Positivity FITs 0.136(0.139) 0.150(0.132) 0.036(0.745) 0.240(0.028) CITs -0.107(0.269) 0.312(0.005) 0.219(0.199) 0.190(0.005) OITs 0.157(0.106) 0.190(0.047) -0.014(0.942) 0.380(0.001)

R2 0.044 0.314 0.050 0.243

F 1.772 17.667 2.045 12.421

p-value 0.156 0.000 0.111 0.000

* The value for independent variables: β (p-value)

6.2.2 Results for hypotheses

Hypothesis 1 is tested using multiple regression analyses across FITs, CITs, and OITs to investigate the ability of frequency and positivity to predict the level of satisfaction. I put positivity and frequency of FITs, CITs, and OITs respectively in different models to solely compare the size of association of positivity and frequency at each touchpoint group. In the first step, two predictors were entered: frequency of FITs and positivity of FITs. This model was statistically significant (p < 0.001) and explained 18% of the variance in satisfaction. Only positivity variable is statistically significant (β = 0.425; p < 0.01) indicating that positivity is a predictor of satisfaction while frequency is not. Further, positivity and frequency of CITs, and that of OITs were used as independent variables respectively in the following models. Positivity of CITs (β = 0.519; p < 0.01) and OITs (β = 0.455; p < 0.01) has a significant impact on the satisfaction while frequency of CITs and OITs does not. Therefore across all three types of touchpoint group, positivity is a statistically significant predictor of satisfaction while frequency does not relate to predicting satisfaction. The regression model for the soft drink industry confirms the same result. Hypothesis 1 is supported and the results for both categories are summarized in Table 5 and 6.

Table 5. Results of hypothesis 1 in mobile telecom category

Dependent Satisfaction

R2 Adj. R2 F p-value β p-value

Independent

0.184 14.444 0.000

Frequency FITs 0.095 0.258

0.270 0.257 21.630 .000 Frequency CITs -0.012 0.875 Positivity CITs 0.519 0.000 0.224 0.211 16.910 0.000 Frequency OITs 0.137 0.095 Positivity OITs 0.455 0.000

Table 6. Results of hypothesis 1 in soft drink category

Dependent Satisfaction

R R Square F p-value β p-value

Independent 0.159 0.144 11.048 0.000 Frequency FITs 0.045 0.601 Positivity FITs 0.388 0.000 0.177 0.163 12.612 0.000 Frequency CITs 0.240 0.005 Positivity CITs 0.358 0.000 0.257 0.245 20.285 0.000 Frequency OITs 0.220 0.007 Positivity OITs 0.471 0.000

To test hypothesis 2, the mean of positivity from direct customer experiences and indirect customer experiences are calculated from FITs, CITs, and OITs respectively. For instance, these variables are named as follows: MTP_FIT_D stands for mobile telecom positivity of FITs direct experiences and SDP_CIT_I refers to soft drink positivity of CITs indirect experiences. I compared direct and indirect experiences at each touchpoint type separately, i.e. comparison between MTP_FIT_D and MTP_FIT_I, and SDP_CIT_D and SDP_CIT_I. Multiple regressions are used to understand how positivity of direct and indirect experience explain the customer satisfaction across FITs, CITs, and OITs. The results of mobile telecom category show that across all touchpoint types indirect experiences appear to have a stronger effect on satisfaction than direct experiences. The results of soft drink category show the same for FITs and OITs while the result of CITs show the opposite. Interestingly, direct experiences of CITs have higher

impact on the satisfaction while indirect experiences do not have significant impact (β for direct = 0.441; p <0.01, β for indirect = 0.038; p = 0.656). Therefore although the result of soft drink category CITs shows exception, hypothesis 2 is not supported. Results are shown in Table 7.

Table 7. Results of hypothesis 2

Dependent _Satisfaction

R2 Adj. R2 F p-value β p-value

Independent

Mobile Telecom 0.192 0.178 13.879 0.000

Positivity FITs Direct 0.243 0.010

Indirect 0.272 0.004 Positivity CITs 0.274 0.261 22.025 0.000 Direct 0.192 0.030 Indirect 0.411 0.000 Positivity OITs 0.209 0.195 15.435 0.000 Direct 0.258 0.006 Indirect 0.282 0.002 Soft Drink 0.159 0.145 11.070 0.000

Positivity FITs Direct 0.211 0.046

Indirect 0.237 0.025 Positivity CITs 0.202 0.188 14.794 0.000 Direct 0.441 0.000 Indirect 0.038 0.656 Positivity OITs 0.209 0.196 15.487 0.000 Direct 0.183 0.045 Indirect 0.350 0.000

Hypothesis 3 is tested with a new variable indicating the number of touchpoints respondents encounter. If respondents said they encounter the touchpoint more than once, it counted as one while it counted as zero if they did not encounter the touchpoint. For example, the respondents answering that they watched 5 times a TV ad and attended 2 times promotion events, searched information 10 times on the Web and 7 times on the SNS, and heard WOM 4 times counts as five in the new variable as they experience five different touchpoints (TVad, Promotion, Web, SNS, and WOM). In this way, it is possible to calculate how many touchpoints

each respondent encountered among fourteen. The same logic applied to both categories and these two new variables are named as MT_many (number of touchpoints encountered by the respondent in the mobile telecom category) and SD_many (number of touchpoints encountered by the respondent in the soft drink category). The simple regression model is used to test the possible influence of the number of touchpoints (independent variable) to loyalty (dependent variable). The results show that only 2.4% of loyalty is explained by the independent variable for the mobile telecom category. But a significant level is 0.051 meaning that it is statistically not significant. On the other hand, the results of the soft drink category show that 7.2% of loyalty can be explained with the number of touchpoints and this value is statistically significant (β = 0.283; p < 0.01). Therefore, hypothesis 3 is only supported in the soft drink category. Results are shown in Table 8.

Table 8. Results of hypothesis 3

Dependent

Variables _{Mobile Telecom} Loyalty of Loyalty of Soft Drink Independent MT many 0.178(0.051) SD many 0.283(0.002) R2 0.032 0.080 Adj.R2 0.024 0.072 F 3.881 9.933 p-value 0.051 0.002

* The value for independent variables: β (p-value)

To test hypotheses 4a, 4b, and 4c, a multiple regression is used. In model 1, positivity at FITs and OITs are independent variables and satisfaction is a dependent variable to compare the effect of positivity of OITs on the satisfaction with that of FITs. In the mobile telecom industry, 25.2% of satisfaction is explained by two positivity variables (p < 0.01). Specifically, both positivity of FITs (β = 0.279; p < 0.01) and OITs (β = 0.316; p < 0.01) have a significant impact

predicted by positivity of FITs and OITs (p <0.01) and both FITs (β = 0.211; p < 0.05) and OITs (β = 0.344; p < 0.01) have a significant positive impact on the satisfaction. In both categories, standardized coefficients β of OITs is higher than that of FITs so that it is possible to conclude that positivity of OITs is more influential on customer satisfaction than that of FITs. Therefore, hypothesis 4a is supported. In model 2, to test hypothesis 4b assuming gender as a moderator in the relationship between positivity of OITs and customer satisfaction, gender as moderate variable and interaction(G) are added as independent variables. Interaction(G) variable is calculated by multiplying gender and positivity of OITs, indicating the interaction effect between two variables. The results of the mobile telecom category in model 2 show increasing variance of the dependent variable explained when gender and interaction(G) are added (Adj.R2 = 0.256, p<0.001). However, both gender (β = -0.433; p = 0.344) and interaction(G) (β = 0.547; p = 0.245) do not have a significant impact on the satisfaction of mobile telecom; thus showing there is no moderate effect of gender. The results of the soft drink category in model 2 indicate that including the independent variables gender, and interaction(G) significantly decrease the variance in the satisfaction explained by the model (Adj.R2 = 0.225, p <0.01). In addition, the standardized coefficients of gender and interaction(G) are statistically not significant, thus proving gender does not have a moderate effect in the relationship between positivity of OITs and satisfaction in the soft drink category. Therefore, hypothesis 4b is not supported. Lastly in model 3, to test hypothesis 4c assuming age as a moderator in the relationship between positivity of OITs and customer satisfaction, age as moderate variable and interaction(A) are added as independent variables. Interaction(A) variable is calculated by multiplying age and positivity of OITs, indicating the interaction effect between two variables. The hypothesis assumes that the relationship between positivity of OITs and customer satisfaction reinforces as the age of

respondents get younger. The results of mobile telecom category in model 3 show that age and interaction(A) significantly increase the amount of variance in the dependent variable explained (Adj.R2 = 0.323, p <0.001). The independent variables age (β = 1.071; p = 0.007) is positively associated meaning that as customers get older, the dependent variable satisfaction gets higher. Other independent variable interaction(A) (β = -1.485; p = 0.002) has a negative significant association with customer satisfaction indicates the relationship between positivity of OITs and customer satisfaction weakened as age gets higher. Conversely, as age gets lower, the relationship between positivity of OITs and satisfaction is fortified. The results of the soft drink category in model 3 show that Adj.R2 is increased 0.044 significantly so that the model 3 explains the better variance of dependent variable (Adj.R2 = 0.269, p < 0.001). However, the standardized coefficients of age and interaction(A) do not have a significant meaning; there is thus no moderate effect of age. Therefore, hypothesis 4c is supported only in the mobile telecom category. Results for hypotheses 4a, 4b, 4c are shown in Table 9.

Table 9. Moderate effects of gender and age (H4)

Dependent

Variables Satisfaction of Mobile Telecom Satisfaction of Soft Drink

model1 model2 model3 model1 model2 model3

Independent Positivity FITs 0.279(0.003) 0.262(0.005) 0.369(0.000) 0.211(0.029) 0.190(0.056) 0.181(0.065) Positivity OITs 0.316(0.001) 0.202(0.142) 0.983(0.000) 0.344(0.000) 0.159(0.362) 0.209(0.384) Gender -0.433(0.344) -0.703(0.230) Interaction(G) 0.547(0.245) 0.784(0.215) Age 1.071(0.007) -0.436(0.292) Interaction(A) -1.485(0.002) 0.227(0.611) R2 0.265 0.281 0.346 0.241 0.251 0.294 Adj. R2 0.252 0.256 0.323 0.228 0.225 0.269 F 21.047 11.252 15.208 18.559 9.648 11.967 p-value 0.000 0.000 0.000 0.000 0.000 0.000

Hypothesis 5 is tested using regression analysis to understand whether the relationship between positivity of CITs and customer satisfaction is moderated by innovativeness level of respondents (“I often seek out information about new products and brands”, “I am continually seeking new product experiences”). In model 1, positivity of CITs is an independent variable, satisfaction is a dependent variable, and these two variables are positively related in both the mobile telecom (β = 0.519; p < 0.001) and the soft drink (β = 0.519; p < 0.001) categories. In model 2, innovativeness and interaction variables are added as independent variables. The interaction term is calculated by multiplying positivity of CITs and innovativeness variables showing interaction effect between the two. The results of model 2 in mobile telecom show that adding independent variables innovativeness, and interaction significantly decrease the amount of variance in the customer satisfaction explained by the model (Adj.R2 = 0.253, p < 0.001). Also, the standardized coefficients for both variables do not have significant meaning indicating innovativeness does not have a moderate effect on the relationship between positivity of CITs and the customer satisfaction. On the other hand, results of model 2 in soft drink category show that model 2 explains significantly increasing amount of variance in customer satisfaction by adding two additional variables (Adj.R2 = 0.159, p < 0.001). Specifically, the independent variables innovativeness (β = 1.206; p = 0.005) and interaction (β = 1.396; p = 0.008) have a significant impact on the customer satisfaction indicating innovativeness has moderate effect on the relationship between positivity of CITs and customer satisfaction. Therefore, hypothesis 5 is supported only in the soft drink category. Results are shown in Table 10.

Table 10. Moderate effects of Innovativeness (H5) Dependent Variables Satisfaction of Mobile Telecom Satisfaction of Soft Drink

model1 model2 model1 model2

Independent Positivity CITs 0.519(0.000) 0.702(0.048) 0.346(0.000) 0.888(0.000) Innovativeness 0.264(0.637) 1.206(0.005) Interaction -0.376(0.605) -1.396(0.008) R2 0.270 0.272 0.120 0.180 Adj. R2 0.264 0.253 0.112 0.159 F 43.595 14.441 16.041 8.481 p-value 0.000 0.000 0.000 0.000

* The value for independent variables: β (p-value)

Hypothesis 6 is tested using regression analysis to understand whether the relationship between positivity of OITs and customer satisfaction is moderated by consumer susceptibility to interpersonal influence (“To make sure I buy the right product or brand, I often observe what others are buying and using”, “It is important that others like the products and brands I buy”). In model 1, positivity of OITs is an independent variable, customer satisfaction is a dependent variable, and these two variables are positively related in both the mobile telecom (β = 0.453; p < 0.001) and the soft drink (β = 0.457; p < 0.001) categories. In model 2, consumer susceptibility to interpersonal influence and interaction variables are added as independent variables. The interaction term is calculated by multiplying positivity of OITs and consumer susceptibility to interpersonal influence variables showing interaction effect between the two. The results of model 2 in mobile telecom show that adding independent variables consumer susceptibility to interpersonal influence, and interaction significantly decrease the amount of variance in the customer satisfaction explained by model (Adj.R2 = 0.187, p < 0.001). Also, the standardized coefficients for both variables do not have significant meaning, which indicates that consumer susceptibility to interpersonal influence does not have a moderate effect on the relationship

in the soft drink category show that model 2 explains significantly increasing amount of variance in customer satisfaction by adding two additional variables (Adj.R2 = 0.234, p < 0.001). However, the independent variables consumer susceptibility to interpersonal influence (β = 0.732;

p = 0.200) and interaction (β = -0.702; p = 0.349) do not have a significant impact on the

customer satisfaction indicating consumer susceptibility to interpersonal influence does not have moderate effect on the relationship between positivity of OITs and customer satisfaction. Therefore, hypothesis 6 is not supported. Results are shown in Table 11.

Table 11. Moderate effects of Influences (H6)

Dependent Variables Satisfaction of Mobile Telecom Satisfaction of Soft Drink

model1 model2 model1 model2

Independent Positivity OITs 0.453(0.000) 0.569(0.058) 0.457(0.000) 0.735(0.047) Influence 0.201(0.642) 0.732(0.200) Interaction -0.376(0.605) -0.702(0.349) R2 0.205 0.207 0.209 0.254 Adj. R2 0.199 0.187 0.202 0.234 F 30.510 10.097 31.210 13.316 p-value 0.000 0.000 0.000 0.000

* The value for independent variables: β (p-value)

Summary of findings of hypothesis testing is shown in Table 12.

Table 12. Summary of results for hypothesis testing

Hypothesis Results

H1: Positivity of customer experiences has a stronger association with customer satisfaction across FITs, CITs, and OITs than frequency of customer experiences.

Supported

H2: Positivity of direct customer experiences has a stronger association with customer satisfaction across FITs, CITs, and OITs than indirect customer experiences

Not supported

H3: People who encounter with many touchpoints in general are more

loyal to the brand. Partially supported

H4a: Positivity of customer experiences at OITs has a stronger association