Exploring the effects of negative experiences in online channels on subsequent channel selection and usage

(1)

EXPLORING THE EFFECTS OF NEGATIVE EXPERIENCES IN

ONLINE CHANNELS ON SUBSQUENT CHANNEL SELECTION AND

USAGE

FINAL VERSION

Author: John J Cortes   Student number: 10894489

Supervisor: Javier Sese   Date of submission: June 23, 2017 Version being submitted: Final Version 

MSc. in Business  Administration – Digital Business Track 

University of Amsterdam

(2)

Statement of originality

This document is written by Student John J Cortes, 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.

(3)

Abstract

The development of new digital technologies is changing the way in which companies interact with customers through different channels. Interactions may lead to positive experiences that might establish profitable and long-term relationships or negative experiences that may lead customers to consider the use of alternative channels. Therefore, it is critical for managers to understand consumers’ multichannel behaviour to properly manage their multiple channels effectively to deliver a consistent multichannel experience to their customers. In this study, the author argues that negative experiences in online channels have an effect on subsequent channel choice and usage. Building on an adapted version of the P-P-M framework, this study provides theoretical understanding of consumers’ channel switching intentions, when dealing with service failures. This framework is examined empirically in financial services, specifically, in the context of online banking. The results reveal that negative experiences that involve security issues and poor e-service quality increase the likelihood that customers will engage in channel switching behaviour. Moreover, it was found out that more loyal customers will be more likely to switch to other channels when experiencing security problems in online channels. This research offers practical implications for organizations to manage customer experience across channels efficiently and effectively. It is critical for managers to understand consumers’ multichannel behaviour to properly deliver a consistent customer experience. Moreover, managers have to pay attention on how to integrate customer experience across channels to improve service delivery and reduce or avoid service failures that might affect the business-customer relationship.

(4)

List of tables

Table 1. Sample demographics ... 36

Table 2. Descriptive Statistics... 40

Table 3. Measurements validity and reliability... 41

Table 4. Correlation matrix ... 42

Table 5. Model summary ... 44

Table 6. Hierarchical regression, model 3 (DV: ISW) ... 44

Table 7. Effect of negative experiences ... 45

Table 8. Interaction effect: Loyalty & Experience ... 46

Table 9. Model summary ... 46

Table 10. Actual behavior ... 47

Table 11. Frequency of usage (before vs after) ... 48

Table 12. Overview of hypotheses... 48

List of figures

Figure 1. P-P-M Framework (modified version from Moon (1995)). ... 20

Figure 2. Conceptual model ... 21

(7)

1. Introduction

In the last few years the development of new technologies has changed the way people communicate and the way people do business. Internet has become an integral part of our lives since it has converted as the main source of information for millions of people around the globe (Meeker, 2015). In business, internet has opened the opportunity to create new business models and structures. As a consequence, online channels have transformed the face of different business environments (Zott, Amit, & Massa, 2011). Internet users have increased dramatically in the last decade. Internet has become an empowering tool, providing people with access to information and communication technologies. It has helped people to make better-informed purchasing decisions (Xiong, Zhao, & Fang, 2016). For this reason, many companies have reacted, taking advantage of this opportunity by implementing digital technologies and developing more sophisticated online channels to keep their position in this changing environment (Edelman & Singer, 2015).

The proliferation of new information and communication technologies, specifically the internet, has been changing the way in which customers and businesses interact. For this reason, many companies have been forced to turn themselves into multichannel businesses, doing business through different channels in parallel to offer better service to their customers (Sousa & Voss, 2006). In this regard, many companies have expanded their traditional face-to-face marketing channels, initiating online businesses. Consequently, this has encouraged consumers to turn themselves into multichannel users. The inclusion of online channels to the companies’ current marketing and distribution channels provides businesses with the opportunity to reach a broader audience, improve their operations and reduce costs. Nowadays, companies are constantly looking for options to add new marketing and distributions channels to better serve their clients and fulfil their needs. Although companies may take advantage of the benefits of a multichannel strategy, it may represent a challenge (Schibrowsky, Peltier, &

(8)

Nill, 2007). In this multichannel context, the number of business-customer interactions has increased. These interactions happen while customer are searching for information about companies, products and/or services or when using online banking, e-commerce platforms or engaging with other consumers in social networking (Rose, Hair, & Clark, 2011). However, as the number of business-customer interactions increases through different channels, it also increases the complexity of these interactions. According to Forrester Research (2004), due to the growing number of multichannel users; companies should understand the behaviour of users in order to address their specific needs and expectations; and therefore, provide consistent experiences to customers. For this reason, it is critical for firms to develop relevant strategies to deliver satisfactory experiences; particularly, in online environments. This could increase customer value through effective customer relationship management (Neslin, et al., 2006). Moreover, it could provide companies with an opportunity to obtain additional revenue from their current customer base, allocate resources more efficiently, reduce costs and maximize profits (Forrester, 2004). In general, online business-customer interactions lead to positive experiences that might establish profitable and long-term relationships. However, these experiences can also be negative. In effect, negative experiences in online channels are very frequent. For instance, when customers are using online banking or any e-commerce platform, they may experience issues such as website failure, poor online customer service, payment issues and credit card fraud. These could lead customers to frustration and dissatisfaction. Problems in any service encounter could generate negative reactions of consumers, such as intention to switch to other service providers, consider the use of alternative channels, negative word-of-mouth, or disappointment, all of which might impact negatively company’s financial position and reputation (Holloway & Beatty, 2003). Nevertheless, considering the importance of this topic, little attention has been paid to the understanding of negative customer experience and how it affects consumers’ behaviour, specifically, the effects on subsequent channel choice

(9)

and usage. This is the reason of the increasing importance to companies and scholars to explore the online customer experience in a multichannel setting (Verhoef, et al., 2009). The improvement of the customer experience has become the top priority for companies’ executives, leading it to a top management objective (Homburg, Jozic, & Kuehnl, 2015). The concept of customer experience in general is well-developed in the traditional contexts; however, exploring the online context needs more attention. According to Schibrowsky, Peltier, & Nill (2007), online customer experience is one of the most important topics for internet marketing research and it is still not completely explored. In addition, it is critical for managers to understand consumers’ multichannel behaviour to properly manage their multiple channels effectively to deliver a consistent multichannel experience to their customers. To do this, managers should understand how and why customers use and adopt the available channels and how to provide satisfactory experiences to customers, and particularly the importance of delivering satisfactory online experiences (Trueman, Cornelius, & Wallace, 2012).

Considering the increasing importance of this topic and the need to explore the effects of negative experience in online context on channel adoption and usage, this study aims to contribute to literature by answering the main research question “How do negative experiences

in online channels affect subsequent channel selection and usage?” by answering this research

question, the contribution of this paper lies in its initiation of research into customer experience in online channels and its effects on subsequent channel choice and usage. Managerial implications include, providing strategic insights, which may help managers to develop solid strategies to promote greatest customer satisfaction through online channels. Moreover, it provides contributions towards new knowledge and understanding of how companies can provide effective and consistent multichannel experiences to customers in order to build long-lasting relationships, and generate high companies’ financial returns on customer online experience. In order to answer the research question, it is proposed a Push-Pull-Mooring (PPM)

(10)

framework to identify the causes of customers’ switching behaviour in online channels. This framework is tested empirically in financial services; specifically, online banking. The structure of this thesis is the following. First, I summarize prior literature pertaining to channel adoption behaviour, customer experience, and negative experience. In section 3, it is presented the conceptual framework, which includes the PPM framework, and conceptual model and hypotheses. Consequently, the results of the empirical study are reported. This paper is concluded by discussing the implications of the findings; likewise, the limitations and directions for future research.

2. Literature review

2.1 Channel choice and usage

In order to better serve their customers, companies have been constantly including new marketing channels, such as online channels (Geyskens, Gielens , & Dekimpe, 2002). In the same pace, customers are becoming multichannel users. However, due to the increasing number of multichannel users, the number of challenges have also increased. In the financial industry, banks have to deal with high competition in a complex environment. Therefore, they have the need to offer efficient and consistent service across channels. In addition, the development of new digital channels may impact traditional channels. For decades, business-customer interactions took place in physical stores. Nowadays, these interactions have extended to digital channels (Pantano & Timmermans, 2014).

Digital technologies, such as the internet, have provided customers with access to a broader range of information, facilitating their evaluation and selection process by making better-informed decisions (Rodriguez-Torrico, San Jose Cabezudo, & San-Martin, 2016). As a result, an increasing number of companies are incorporating online channels to their businesses (Yang, Lu, & Cahu, 2013). Specifically, companies’ websites have become a key

(11)

communication channel between businesses and customers. Considering the importance of web sites, companies should carefully design their online facilities. Websites characteristics and the service provided through the online channel are critical when attracting customers to this particular channel (Song, Bakker, Lee, & Wetherbe, 2011). The explosion of e-commerce has created the need to study users’ adoption and usage of digital channels (Ha & Stoel, 2009). Yet, most of the existing literature has been focused on the study of the initial adoption of online channels, while less attention has been paid to investigate consumers’ post-adoption and channel switching behaviour (Yang, Lu, & Cahu, 2013). Limayem et al. (2007), argue that initial adoption is just the first step to understand channel adoption behaviour; However, the overall success of a channel depends mainly on the sustained use of the channel. In more detail, initial adoption does not guarantee a sustained use of the channel. During the process, customers could face a service failure, resulting in channel switching behaviour (Yang, Lu, Zhao, & Gupta, 2011). Marketing literature has studied the phenomenon of digital channels from various perspectives. Some academics have focused on the adoption of online channels, identifying some drivers of channel adoption, such as channel characteristics, company’s efforts and consumers’ behavioural reaction, and consumers’ traits.

The effects of channel characteristics have been explored in different studies. In general, researchers agree that consumers’ perceptions of online channel attributes play an important role in channel selection (Verhoef, Neslin , & Vroomen, 2007). According to Song et al. (2011), website elements, such as order taking, billing and payment and information convenience influence channel adoption. Specifically, the perceived usability and usefulness related to these elements may determine the quality of the channel. Hence, higher the customers’ perceived quality, higher the intention to adopt the channel. Soorpramanien & Robertson (2007) argue that other elements, such as ease of use, image and compatibility influence channel adoption. Howcroft et al. (2002) conclude that the perceived level of security

(12)

and the system’s quality play a key role on consumers’ channel adoption. Verhoef et al. (2007) support these findings. The authors show that the perceived convenience and quality of the channel and the perceived level of security are main drivers of consumers’ channel selection. Companies may influence channel adoption and the continued use of the channel by offering customer-oriented services and products. Marketing efforts such as: price promotions, customer-oriented marketing campaigns and brand positioning are elements that might attract to and retain consumers in the online channel (Verhoef, Neslin , & Vroomen, 2007). Likewise, Polo & Sese (2016) show that marketing activities influence channel selection; however, these activities are more effective at driving channel choice for communications. In addition to marketing efforts, companies should focus on improving service delivery. Yang et al. (2011) argue that service quality is a key factor that influences the use of the channel. Likewise, the authors show that perceived quality is an important determinant of the intention to use the channel in the future. In general, companies’ efforts to create customer-oriented offerings may result in positive consumers’ behavioural reactions, such as loyalty and satisfaction. According to Verhoef et al. (2007), loyalty and satisfaction with the company are predictors of channel adoption. Customers who have a strong relationship with the firm are more likely to adopt the online channel (Estrella-Ramon, Sanchez-Perez, & Swinnen, 2016). In addition to this, Rexha et al. (2003) show that trust and satisfaction with a company’s offline channel affects subsequent online channel adoption.

Other studies suggest that consumers’ traits influence the adoption of online channels. In general, many studies have investigated users’ characteristics to predict channel selection and usage. According to these studies the most important characteristics are: demographics, users’ level of experience with online channels, and customer innovativeness(Estrella-Ramon, Sanchez-Perez, & Swinnen, 2016). Vijayasarathy (2004) concludes that having basic computer and internet skills, and a perception of the internet as a valuable tool may influence the adoption

(13)

of online channels. Moreover, research suggests that the fit between the online channel and the consumer’s personality influences the adoption of the online channel significantly (Flavian , Guinaliu, & Torres , 2006). Likewise, Yang et al. (2011) demonstrate that willingness to change is an important factor that influences the use of the online channel. Willingness to change depends on consumers’ attitudes, subjective norms and the perceived benefits of using a new channel. Liu & Forsythe (2010) argue that personal enjoyment is another predictor of channel selection and usage. Enjoyment influences attitudes toward the online channel and use of online self-service. In addition, Flavian et al. (2016) concludes that socio-demographical traits such as age and education might influence online channels adoption.

Studies have showed that the intention to adopt online channels might be influenced by the perceived value, when comparing the offline and online channel. Soorpramanien & Robertson (2007) argue that evaluation of the offline channel may influence consumers’ attitudes towards the online channel. When selecting a channel, consumers compare the benefits and risks associated to both channels. Yang et al. (2013) suggest that consumers are more likely to use the online channel, when they have a positive image of the brand and a positive perception of the offline channel. Yet, a very positive perception of the offline channel may influence the perceived benefit of the online channel. This may be counterproductive. According to Yang et al. (2013), a very high positive assessment of the offline channel may lower the intention to adopt the online channel.

2.2 Customer experience

Customer experience encompasses a set of touch points at which interactions between customers and companies take place. These interactions may be related to a product, service, or information seeking (Lemon & Verhoef, 2016). The concept of customer experience has been defined from different points of view. In 1999, Pine & Gilmore made a relevant

(14)

contribution to the concept of customer experience in their book “Experience Economy”. They defined the concept from a customer perspective, as the enjoinment of a series of events where companies creates a favourable setting to engage with customers and offer memorable and unique experiences. In general, every business-consumer encounter leads to a customer experience (Lemon & Verhoef, 2016). The definition of customer experience also includes aspects of company’s offering, such as the quality of service and customer care and the product’s features and benefits (Verhoef, et al., 2009). With the proliferation of new marketing and distribution channels, customer experience may take place in offline or in digital settings (McCarthy & Wright, 2004), involving customers’ or businesses’ initiated activities at different touch points (Lemon & Verhoef, 2016). Previous research has treated the concept of experience from two perspectives: as the experience related to each touch point or as the accumulation of different touch points at different stages along the customer’s buying process (customer journey) (Verhoef, et al., 2009). Customer-business interactions lead to a different type of experiences. Customer experience may be perceived as positive or negative. A positive customer experience could increase level of satisfaction; and therefore, create strong relationships with customers (Homburg, Jozic, & Kuehnl, 2015). On the other hand, a negative experience, which is any event during the business-customer interaction that the customer perceives as undesirable, can generate a decrease in customer satisfaction and affect customers’ current and future buying behaviour (Sarkar Sengupta, Balaji, & Krishnan, 2014).

Marketing theories include the study of customer experience. Researchers have been treating the concept of customer experience as one of the most relevant ones in consumer industries in traditional settings. Yet, the development of new technologies has limited the understanding of the concepts of customer experience and customer experience management. As the number of marketing and distribution channels has increased, business-consumer interactions can take place in different contexts (Homburg, Jozic, & Kuehnl, 2015). In

(15)

particular, in online channels. The notion of online customer experience has been receiving less attention from experts and businesses (Klaus & Nguyen, 2013). Due the proliferation of digital channels, researchers have been highlighting the need to explore this topic (Verhoef, et al., 2009). According to Schibrowsky and Peltier (2007), the study of online customer experience should be the top priority of topics in marketing research. Previous research has focused extensively on the concept of customer experience in traditional channels (Rose, Hair, & Clark, 2011). However, exploring the concept in online contexts need further investigation (Trueman, Cornelius, & Wallace, 2012). Klaus & Nguyen (2013) also highlight the importance of exploring online customer experience and customer behaviour. Online customer-business interactions may have a stronger impact on companies compared to traditional channels. Positive or negative experiences can lead to immediate word-of-mouth, affecting companies’ position (Homburg, Jozic, & Kuehnl, 2015). Dholakia et al. (2012) show that online customer experience may have a stronger positive or negative impact on firms. For this reason, experts emphasize the need to develop frameworks to understand online customer experience; and consequently, attenuate its negative impacts (Lemke, Clark, & Wilson, 2011). Marketers should focus on developing strategies to improve online service delivery and enhance online customer experience. These strategies may have a positive impact on companies’ financial returns on customer online experience (McCarthy & Wright, 2004).

Previous research in online customer experience has mainly focused on exploring its effects on conversion, customer loyalty, and satisfaction (Lemon & Verhoef, 2016). Khan, Rahman & Fatma (2016) show that there is a positive relationship between online customer experience and brand satisfaction and loyalty. In addition, the authors argue that a satisfying online customer experience may have an effect on repurchase intentions. Other studies suggest that online customer experience may be affected by channel’s characteristics, such as website attributes and customer support (Dholakia, Bagozzi, & Klein Pearo, 2012). Moreover, the

(16)

design, speed, level of security and website content may have a significant effect on the overall online customer experience (Yoon, 2010). Van Doorn & Verhoef (2008) argue that past experiences with the company’ products, services or channels may affect subsequent interactions; and therefore, affecting online customer experience.

From a consumer perspective, customer experience in traditional channels may differ from online customer experience. In online channels, the interactional human elements are limited. In online environments, customers do not experience the same immediate support as in face-to-face environments. Hence, this can lead customers to frustration and dissatisfaction, which may affect companies’ position negatively. Based on this, marketers should develop strategies to reduce the impacts of negative online experiences. However, considering the importance of this topic, little attention has been paid to the effect of those negative interactions on consumers’ multichannel behaviour (Holloway & Beatty, 2003).

2.3 Negative experiences

Considering the characteristics of service, such as intangibility, inseparability and variability, negative experiences are sometimes unavoidable. These negative encounters may generate unfavourable reactions on consumers, such as intention to switch to other channels or service providers, negative word-of-mouth, frustration or dissatisfaction; at the same time, these may impact negatively company’s financial position or reputation (Holloway & Beatty, 2003). In general, service failures are the main cause of negative experiences. By definition, a service failure is a mistake or any incident that may arise when delivering service. For this reason, companies should develop strategies to overcome service failures. Likewise, companies should prepare recovery strategies to react immediately to service failures, reducing their impacts (Fan, Wu, & Wu, 2010). When service failures occur, customers tend to make companies liable for these incidents. From this perspective, companies fail to meet customers’

(17)

expectations. As a result, customers’ negative perceptions may be translated into channel or service provider switching behaviour or immediate negative word-of-mouth (Fan, Wu, & Wu, 2010).

Marketing literature has focused on exploring the effects of service failures on customer behaviour in offline settings. However, less attention has been paid to measure the effects of service failures in online channels (Fan, Wu, & Wu, 2010). Holloway and Beatty (2003) identify some antecedents of customer dissatisfaction. The authors have identified delivery problems (e.g. delivery date), website system failure, poor customer service support, payment issues (e.g. credit card overcharged), lack of security (e.g. credit card fraud) and system’s degree of complexity as antecedents that may affect negatively the business-customer relationship, resulting in dissatisfaction. In addition, service failures, such as service delivery system failures, unsolicited seller actions, and interface complexity have a negative effect on customer satisfaction, leading to customers’ switching behaviour (Kuo, Yen, & Chen, 2011). Zhu & Chin-Te Lin (2012) contribute to theory by identifying other antecedents of customer dissatisfaction that may lead customers to channel or service provider switching behaviour. The authors show that a poor system’s quality, a complex or unclear website design, and poor website content lead to dissatisfaction and intention to switch to other channels. Similarly, the quality of the website server and the interface ease of use may influence customers’ channel switching intentions (Piercy & Archer-Brown, 2014). Indeed, consumers are looking for user-friendly online environments to look for services and products efficiently and effectively (Sharma & Lijuan, 2015). Piercy & Archer-Brown (2014) also argue that lack of security and issues related to privacy have a significant effect on customers’ channel switching intention. In terms of security-related issues, Dai, Forsythe, & Kwon (2014) show that a monetary loss is one of the main concerns among users of online channels. In general, overcharging customers and credit card frauds are the most common monetary-related issues

(18)

online. Monetary losses are negatively correlated to the use of online channels. Therefore, experiencing a security-related issue may enocurage channel switching behaviours. Likewise, privacy-related issues are percieved as a severe service failure, which result in channel switching intention. The convenience of registering accounts on internet increases the probability of fraud. When customers face this type of problem, the likelihood to switch to other channels are very high (Dai, Forsythe, & Kwon, 2014).

Indisputably, negative experiences affect the business-customer relationship. Negative experiences cause dissatisfaction and frustration, influencing customers’ subsequent behaviour (van Doorn & Verhoef, 2008). After a negative experience, customers may evaluate whether to switch or not to other channels or service providers. However, this evaluation could be influenced by the perceived failure magnitude. As the perceived failure magnitude increases, the likelihood of a channel or service provider switching behaviour also increases (Smith, Bolton , & Wagner, 1999). Severe negative incidents may discourage customers to use a specific channel or even terminate strong business-customer relationships. Furthermore, intense negative emotions, such as irritation and disappointment, have a negative impact on satisfaction; and therefore, on customers’ subsequent channel behaviour (Fan, Wu, & Wu, 2010). Negative emotions also have an effect on customer loyalty, generating immediate negative word-of-mouth and complaining actions and affecting repurchase intentions and subsequent channel choice (Svari, Slatten, Svensson, & Edvardsson, 2011).

Experience with the channel and relationship quality may influence customers’ perceptions, when facing a service failure. Research suggests that customers tend to tolerate negative incidents when their level of the experience with the channel is high and when they have a strong relationship with the firm (Vidal , 2012) (Kuo, Yen, & Chen, 2011). Hess, Ganesan, & Klein (2003) argue that a strong customer-business relationship may “protect” companies from negative impacts, when inevitable service failures happen. Likewise,

(19)

customers who buy from or have experience with a channel are less likely to switch to other channels, when facing a negative experience (Lee, Kim, Lee, & Park, 2006)

2.4 Literature gap and research focus

In summary, the development of new technologies, specifically the internet, has been changing the way in which companies do business, forcing them to turn into multichannel businesses (Sousa & Voss, 2006). In this regard, many companies have expanded their traditional face-to-face marketing channels, initiating online businesses. Consequently, this has encouraged consumers to turn themselves into multichannel users. The increasing number of marketing and distribution channels has change the way customers and businesses communicate, leading to a number of interactions. Interactions with an organization’s channels create opportunities for either positive or negative experiences. In more detail, negative experiences can lead to dissatisfaction and channel switching behaviour (Homburg, Jozic, & Kuehnl, 2015). In particular, negative experiences in online channels may have a stronger effect than in traditional channels. In online channels, unfavourable reactions are immediate.

Yet, considering the importance of this topic, the notion of online customer experience has been receiving less attention from experts and businesses (Klaus & Nguyen, 2013). According to Schibrowsky and Peltier (2007), online customer experience is one of the most important topics for internet marketing research. However, it is still in need for further investigation (Trueman, Cornelius, & Wallace, 2012). Recent research highlights the importance of the online customer experience on the overall customer experience and customer behaviour (Klaus & Nguyen, 2013). This is motivated due to the lack of the interactional human elements of traditional channels in online environments, which in turn can increase the risk of negative experiences during the business-customer interaction (Holloway & Beatty, 2003). As a result, customers may switch to other channels or generate negative word-of-mouth. For this

(20)

reason, it is highlighted the importance to explore the effects of negative customer experience with online channels on subsequent channel choice and usage. By investigating this, it is extended the marketing research related to online customer experience and post-adoption behaviours. Therefore, this study could help managers to evaluate and develop strategies that could improve the companies’ customer interaction and manage consistent customer experiences across channels efficiently and effectively, and strategies that may have a positive impact on companies’ financial returns on customer online experience.

3. Conceptual framework and hypotheses

3.1 Theoretical framework

In this study, it is proposed a conceptual framework to manage different types of negative experiences with online channels differentially. As discussed in the literature review section, the importance of long-term relationships for businesses and the negative effects of consumer negative experiences with online channels on channel switching behaviour highlight the need to investigate the factors that may influence consumers’ subsequent channel choice and usage. To do this, the study builds on the Push-Pull-Mooring Framework to understand the effects of negative customer experience with online channel on subsequent channel choice. The push–pull–mooring (PPM) framework was originally used in human migration literature. By using this framework, researchers could explain the factors that might encourage or discourage people to move from one place to another (Bansal, Taylor, & James , 2005). Therefore, the origin of the PPM framework was linked to the theory “Laws of Migration” introduced in 1885. Consequently, it became a theoretical base for social migration studies (Lee E. , 1966). Basically, this framework proposes that there are some factors that may explain why people migrate. Negative factors at the place of origin that push people away, whereas positive factors that may pull people toward a certain destination. Furthermore, these negative

(21)

and positive factors could be moderated by mooring variables. In this case, mooring variables are social or particular factors that can attach people to their origin of ease migration to the new destination (Lee E. , 1966). Experts have extended the PPM framework to other disciplines such as Business and Economics. In Marketing, Bansal et al. (2005) began by applying this paradigm to explain consumers’ switching behaviour. In the online context, Hou et al. (2011) applied the PPM framework to explore the effects of individual-level factors in online role-playing games on switching behaviour.

Since this study is specifically focused on exploring the effects of negative experiences in online channels on subsequent channel choice and usage, only push and mooring effects will be applied (Fig. 1). Push effects, represent the accumulated influence of customers’ negative experiences with online channels and mooring effects, the moderation effect of situational or contextual circumstances. These variables may attenuate or strengthen the effect of negative experiences on consumers’ channel switching behaviour. In this study the PPM framework is applied to explain consumer behaviour after experiencing a negative incident in an online channel.

(22)

3.2 Conceptual model and hypotheses

The aim of this section is to identify relevant factors that explain how negative experiences affect subsequent channel choice and usage. Using an adapted version of the PPM framework as a general guideline, it is developed a set of hypotheses to predict consumers’ subsequent channel choice and usage (Fig. 2). The push and mooring effects act as multidimensional constructs that consist of various sub-constructs. These sub-constructs are measured to determine their impact on channel selection and usage. The main question is visualized in the following conceptual model, and it can be used to understand the subject matter.

Figure 2. Conceptual model

Poor e-service quality Security problems Interface complexity Experience with the channel Loyalty Actual channel usage Channel switching intention H1b(+) H2c(-) H3a(-) H5(+) H2b(+) H2a(-) H3b(-) H3c(-) H1c(+) H1a(+)

(23)

3.2.1 Elements of push effects

The author builds on prior research and identifies three relevant negative antecedents of customer dissatisfaction with online channels: (1) Poor e-service quality, (2) Security problem, and (3) Interface complexity (Holloway & Beatty, 2003) (Dai, Forsythe, & Kwon, 2014). These variables are supported theoretically and provide relevant academic and practical insights for managing customer experience in online channels. Although prior research has identified the effect of these negatives antecedents on customer satisfaction and possible switching behaviour, I expect these factors to have a direct effect on the strength of the relationship between these negative events and subsequent channel choice and usage.

3.2.1.1 Poor e-service quality

E-service quality refers to the level of customer service support provided by firms through their online facilities. Online customer service support quality differs from the traditional service quality studied in previous research, which focuses on the personalised service provided by organizations directly related to face-to-face interaction (Balasubramanian & Shankar, 2009). Santos (2003) describes online service quality as customer’s evaluation of the experienced quality and excellence of the overall online customer support. Further, Parasuraman et al. (2005) define online service quality to include all the steps along the customer journey, in which interaction with a firm is created: the degree to which an online facility enables service delivery efficiently and effectively. Therefore, online service quality is related to a competent and satisfying experience and consistent service delivery. However, the e-service quality is determined by the consumers’ perceptions and feelings about the overall level of service when interacting with firms online (Jonas & Vida, 2011). Hence, customers will have a positive perception and attitude towards e-service that consistently facilitates and support the attainment of their goals when interacting with the firm. Previous studies on online

(24)

service quality have mainly focused on its effects on value and satisfaction. Chu et al. (2012) show that perceived poor service quality in online channels have a significant negative impact on satisfaction. From these arguments, I argued that as poor online service has a negative impact on satisfaction, this dissatisfaction might discourage consumer to use the channel; and therefore, push the consumers to other channels.

H1a. Poor e-service quality in online channels have positive effect on intention to switch to other channels

3.2.1.2 Security problem

Internet is a technology needed to support e-commerce. Since internet is an open network, security is critical when developing firms’ online facilities. In online environments, the level of perceived risk may be higher because of the limitation to access products physically and to interact with sales staff in person (Park & Stoel, 2005). As a result, this may discourage customers from buying online (Garbarino & Strahilevitz, 2004). Nowadays, new technologies such as cryptography and digital signatures have reduced online security concerns; yet, customers are still worried about the financial risks when performing monetary transactions online (Ranganathan & Ganapathy, 2002). In this regard, online financial transactions are a core business of banks and e-retailers. Therefore, consumers may see these monetary transactions as a financial risk.

Financial risk is could be defined as the possibility of experiencing a monetary loss from a transaction (Sweeney et al. 1999). One of the most common types of monetary loss is credit card fraud. This type is the main concern among customers when purchasing online (Holloway & Beatty, 2003). In addition, Dai, Forsythe, & Kwon (2014) show that the main concerns of consumers when shopping online are credit card fraud and buying products that fail to meet their expections. Both are perceived as a monetary loss. When cosumers perceive

(25)

a security risk, they inmediatley tend to abandon e-shopping carts (Egeln & Joseph, 2012). In general, security problems have been negatively associated with e-commerce and is determined to be a strong predictor of disatisfaction (Dai, Forsythe, & Kwon, 2014). Therefore, I argued that when consumers face security problems in online channels, this affects customer’ satisfaction negatively and increase monetary loss concerns; therefore, it may have a strong positive effect on consumers’ intentions to look for other channels.

H1b. Security problems in online channels have positive effect on intention to switch to other channels

3.2.1.3 Interface complexity

The degree of complexity is another important factor that may have an effect on people’s intention to use online channels (Davis, 1989). Rogers (1995) describes complexity as the extent to which an innovation is seen as either complex or easy to comprehend and use. In human-computer interaction, usability is an important factor. It facilitates task performance and increase satisfaction when using a computer (Palmer, 2002). In this study context, usability is defined as the degree to which a person can use an online channel without any problem and can communicate with the firm and perform any transaction clearly through the online interface (Benbunan-Fich, 2001). Previous research shows that a clear and good online channel design may enhance usability, influencing the success of the channel (Gelbrich & Sattler, 2014). Palmer (2002) shows that an online channel’s design has an effect on consumers’ interest in the online channels. A good and clear design attracts to and retains users in the channel. Previous research verifies the positive effect of a good online channel design. In addition, it shows that the design may also affect customers’ satisfaction (Zviran, Avni, & Glezer, 2006). Information systems studies also has found a significant effect of interface complexity on customer’s satisfaction and it has been treated as a measure of perceived system quality

(26)

(McHaney & Cronan, 1998), and a determinant of online channel adoption (Holloway & Beatty, 2003). Gelbrich and Sattler (2014) argue that consumers are more likely to use a technology, when it is perceived as easy-to-use. The authors also show that ease of use and usefulness are important determinants of customer satisfaction and online channel adoption. These findings have been supported by e-commerce research. In addition to these studies that have found the effect of interface complexity on web quality (Palmer, 2002) and customer satisfaction (Gelbrich & Sattler, 2014), Liao & Cheung (2008) explored the effects of interface ease of use on customer satisfaction in online banking. The authors found a strong significant effect on satisfaction. Hence, interface complexity has been identified as an antecedent of consumer satisfaction in online banking. Based on prior research, I argue that when an online channel is complex to understand and use, consumers are discouraged to use it. Therefore, more complex an online channel is, more likely that consumers will search for other channels.

H1c. Interface complexity of online channels have positive effect on intention to switch to other channels

3.2.2 Elements of mooring effect

Similar to human migration, choosing between channels is a complex decision, so mooring variables might either hinder or facilitate selection behaviours. These contextual, business relationship or personal factors can moderate the influences of push effects (Bansal, Taylor, & James , 2005). The author investigates two critical mooring effects that might have a moderator effect on the relationship between push effects and the intention to switch to other channels.

(27)

3.2.2.1 Loyalty

Customer loyalty expresses the strength of a customer-business relationship and an intentional behaviour related to the business (Andreassen & Lindestad, 1998). Important indicators of customer loyalty include: continuous purchase from the firm, positive word-of-mouth, high levels of satisfaction, and long-lasting customer-business relationships (Selnes & Hansen, 2001). In service firms, customer loyalty is seen as an important predictor of financial performance (Kim & Son, 2009) and may be a stronger predictor of revenues and profit than market share (Selnes & Hansen, 2001). In addition, customer loyalty increases positive word-of-mouth. This may help companies to attract and retain new customers at no extra advertising costs (Kim & Son, 2009). In the online context, customer loyalty may have a stronger effect. In online channels, positive word-of-mouth and support from loyal customers may be spread faster than in traditional channels (Arya & Srivastava, 2015). Research in managerial experience and customer relationship management has been used customer loyalty as one of the most important transactional behaviours (Verhoef, Ou, & Wiesel, 2017). In fact, customer loyalty may be applied as a measure of customer retention and continuation (Selnes & Hansen, 2001). In multichannel organizations, customer loyalty may be a predictor of consumers’ purchase behaviour and the adoption of new digital channels (Liao, Lin, Luo, & Chea, 2016). Previous research has explored the moderating role of loyalty in the relationship of negative experiences and customer-business relationship continuance. The findings show that loyalty has indeed a significant moderator effect on this relationship. Yet, the results of these studies have generated controversy regarding the direction of the moderation effect of loyalty. According to Gregoire and Fisher (2006), there are two opposing views, when exploring the moderator role of loyalty: “love becomes hate” and “love is blind” effects. On one hand, studies that support the “love becomes hate” effect show a negative position. These studies point out that customer with a stronger relationship with the firm may feel let down, when experiencing

(28)

a service failure, causing them frustration and dissatisfaction (Gregoire & Fisher, 2006). Facing a negative incident, during the customer-business interaction, more loyal customers tend to show the longest negative reactions and resentment against the company in the form of avoidance (Gelbrich, Gathke, & Gregoire, 2016). Likewise, Gregoire & Fisher (2006) investigate how customer loyalty attenuates or increases negative reactions towards service failures on subsequent customer behaviour and attitudes. The authors found out that loyalty magnifies negative reactions. In a similar study, Mattila (2004) provides support to this theory. The author argues that high levels of loyalty indeed increases negative reactions towards service failures, leading customers to use another marketing channels or even terminate the business relationship. In a more recent study, Kuo et al. (2011), explored the role of customer loyalty in online service failures. The authors showed that loyal customer are more likely to have unfavourable reactions when facing a service failure, depending on the failure’s severity. Despite the strong relationship with the firm, more loyal customers tend to express the feeling of betrayal when facing severe failures such as financial loss or leak of personal data.

On the other hand, some studies support the “love is blind” effect. These studies show a positive position in regard to the effects of customer loyalty. Researchers argue that loyalty to the company or the brand diminishes the effects of negative experiences. Since a strong relationship with a company could be based on trust and affinity, loyal customers are more unwilling to terminate the worthwhile relationship. Therefore, they are more likely to forgive any negative experience (Gregoire & Fisher, 2006). Two theories support the “love is blind” effect’s view: assimilation bias (Putrevu, 2010) and interpretation bias (Mathews & Mackintosh, 2000). According to assimilation bias, during the deliberation process, customers tend to consider information based on their previous experiences. Therefore, when customers experience a service failure, loyal customers are more likely to ignore this failure since it is not consistent with their previous experiences (dos Santos & Basso, 2012). In addition,

(29)

interpretation bias theory suggests that even if loyal customers consider the service failure, they are more likely to ignore or reduce the negative effects to keep a balance between previous and current feelings (Vazquez-Casielles, Suarez Alvarez, & Diaz Martin, 2010). Therefore, customers who have a strong relationship with the company may tolerate or forgive any service failure in comparison to less loyal customers.

Previous research has shown the moderator effect of loyalty on the relationship between negative experiences and subsequent customer behaviour. However, there is disagreement on the direction of this effect. Gregoire and Fisher (2006) identify two opposing effects: the “love becomes hate” and “love is blind” effects. In addition, Gelbrich et al (2016) argue that the influence of these opposing effects depends on the severity of the incident. In more detail, the authors point out that service failures may be forgiven or not by loyal customers depending on the perceived severity of the failure. Based on this reasoning, it may be concluded that the direction of the moderator effect of loyalty in the relationship between negative experiences and channel switching intentions depends on the severity of the three negative experiences under investigation (Poor e-service quality, security problems and interface complexity). Previous studies show that security and privacy issues are perceived as severe failures (Kuo, Yen, & Chen, 2011); whereas poor e-service quality, website failure and the complexity of the online channel may be treated as less severe failures (Dana & Gil, 2016). Therefore, it is expected that loyal customers will show a “love becomes hate” effect, when experiencing security problems (severe failure) and a “love is blind” effect, when experiencing poor e-service quality and interface complexity (less severe failures) in the relationship between these negative experiences and channel switching intentions.

H2a. Loyalty has a negative moderator effect on the relationship between Poor e-service quality and the intention to select and use other channels

(30)

H2b. Loyalty has a positive moderator effect on the relationship between security problem and the intention to select and use other channels

H2c. Loyalty has a negative moderator effect on the relationship between interface complexity and the intention to select and use other channels

3.2.2.2 Experience with online channels

In this study, experience with online channels relates to the evaluation of how familiar is a user with the online channel and how well the user understands the channel procedures, such as how to perform a monetary transfer, enter credit card information or select and buy a product. According to research, experience with online channels has a positive effect on consumers buying intentions from a specific company through its website (Frambach, Roest, & Krishnan, 2007). Moreover, experience with online channels may decrease complexity and increase trust in the channel. In more detail, experience with online channels decrease complexity because users can easily understand what is happening (Gefen, Karahanna, & Straub, 2003). In online environments, experience with the channel increases trust in the company. Experience tend to increase trust since users are put in a context of understanding of what to expect (Gao & Bai, 2014). In addition, experience may help avoiding any misunderstanding about what companies are doing through their online platforms. Hence, customers with experience will not feel that they are taken advantage of, increasing trust in the firm (Gefen, Karahanna, & Straub, 2003). The role of experience in online channels has attracted the interest of researchers to understand its nature and explore its effects on consumer behaviour; however, limited knowledge is available concerning the moderating role of experience between the relationship of negative incidents and intention to select and use other channels.

(31)

particular marketing channel is affected by their level of experience. Research shows that customers high level of experience with online channels tend to be more comfortable using this channel, whereas customers with less experience may not be willing to use the online channel due to uncertainty and perceived risk (Montoya-Weiss, Voss, & Grewal, 2003). Therefore, it is suggested that experience with a channel may have a moderator role in the relationship between channel evaluation and choice. Based on perceived uncertainty and risk, users tend to evaluate potential losses when using an unknown channel. Therefore, users with less experience are more likely to use the traditional offline channel since they are more familiar with this channel and feel more in control, thus minimizing risks (Gourville, 2006). Moreover, perceived risks of using online channels for important transactions that involve security (e.g. purchases or money transfers) and privacy concerns are more probable to increase the use of a channel that users feel more familiar with. In other words, users will choose a channel that they can trust (Wolin & Korgaonkar, 1999).

Applying this reasoning to this study, experience with online channels may attenuate the effect of negative experiences on customers’ channel switching intentions. Experience with a channel reduce complexity and increase trust. Moreover, users with experience feel more comfortable using the online channel. Therefore, even when dealing with a negative situation in online channels, users will still prefer the online channel. According to Gao & Bai (2014), when facing a negative situation, it is expected the superiority of the channel which is familiar to the user. Hence, the channel that will be chosen at this stage is more likely to be the one the user can trust. Based on this, the following hypotheses are proposed:

H3a. Experience with the online channel have a negative moderator effect on the relationship between poor e-service quality and the intention to select and use other channels H3b. Experience with the online channel have a negative moderator effect on the relationship between security problem and the intention to select and use other channels

(32)

H3c. Experience with the online channel have a negative moderator effect on the relationship between interface complexity and the intention to select and use other channels

3.2.3 Channel usage

In spite that customers may use one or multiple channels at the same time, the intention to switch to other channels does not necessarily entails an actual switching behaviour from an online channel to an alternative channel. Therefore, the potential actual behaviour is measured based on the hypothetical relative frequency of usage of a channel after the negative experiences in online channels. Therefore, this study is set to predict consumers’ actual behaviour based on their intention to switch to other channels. In psychology settings, the intention-behaviour relationship is positively correlated according to the theory of reasoned action (Limayem, Hirt, & Cheung, 2007). Therefore, the author proposes that users’ channel switching intentions significantly influence their actual behaviours on choosing other channels, as formalized in the following hypothesis:

H4. Intentions to switch have positive effects on actual usage behaviour toward other channels.

4. Research methodology

This section provides information about how, where and why the data was gathered in order to empirically test the conceptual framework and its related hypotheses. The structure of this section is based on the theory proposed by Aguinis and Bradley (2014) on how to conduct an experimental vignette methodology study (Fig. 3).

(33)

Figure 3. Research methodology diagram (Aguinis and Bradley, 2014).

4.1 Research design

This research is of exploratory nature. A quantitative study, by means of an experiment and survey, was performed to answer the research question. To gather cross-sectional data, a survey was used as data collection instrument. To test hypotheses and generalise the findings, four different surveys were prepared. Participants were randomly selected and assigned to one of the four surveys. One control group and three treatment groups. Each survey represented a different scenario. First, the three treatment groups: (1) Poor e-service quality, (2) Security problems, and (3) Interface complexity. Second, the control group (1) a normal situation without any negative incident. In order to apply this research design, the author implemented the Experimental Vignette Methodology (EVM) as a suitable approach to identify causal relationships between variables. Vignette studies use descriptions or scenarios of situations that are shown to incumbents within experiments or surveys to provoke an opinion or judgment

(34)

about these situations (Atzmuller & Steiner, 2010). The main advantage of EVM is that it combines the ideas of experiments and survey approaches to balance the weaknesses of each methodology. Experiments have a high internal validity from their carefully designed plans and measurement facilitated by control intervention and surveys present a high external validity due to their claim of representativeness. In this study, EVM has been applied since this methodology fulfil the research need to have control of the independent variables to obtain evidence regarding causation; therefore, it permits researchers to include factors that might be important to answer the research question. This degree of control facilitates the examination of hypotheses that could be otherwise complex (Atzmuller & Steiner, 2010).

The type of experimental vignette methodology used in this research is Paper People Studies. This type consists of confronting participants with hypothetical scenarios, typically in written form. Consequently, participants are asked about their opinions, choices, judgments or express their preferences. In addition, the research design used in this study is Mixed Designs. In this research design participants are divided into different groups and each group receives a different vignette (scenario) (Atzmuller & Steiner, 2010). Accordingly, this design enables researchers to explore the effect of each scenario and to allow comparison across incumbents. Finally, in order to add realism to each situation, I increased the incumbents’ level of involvement by describing each scenario clearly and presenting a picture, evoking a negative situation to increase engagement and understanding of the situation (See appendix 2) (Aguinis & Bradley, 2014).

4.2 Data collection

The author empirically tested the conceptual framework and its related hypotheses in financial services. An online self-administered questionnaire survey, prepared in English, was carried out among people who use online banking (See appendix 1). This study followed the

(35)

criteria suggested by Parasuraman et al. (2005). These criteria suggest that respondents should: have at least one registered bank account and have made at least two transactions online. The online survey questionnaire was performed on Qualtrics, which is a survey portal recommended by the University of Amsterdam. Respondents were reached by email and via Facebook, where the link to survey was posted. To encourage survey responses, the present study used properly designed incentives to encourage responses. An online survey was employed to collect data because online surveys offer convenience, cost efficiency, and minimal time requirements; enhance the ability to reach subjects; and can be assembled more quickly, coded more flexibly, and stored electronically compared with paper surveys (Boyer, Oslon, Calantone, & Jackson, 2002). Subjects also benefit from the convenience of completing online surveys at any time and place. Duffy et al. (2005) show that online surveys appear to attract more knowledgeable, opinionated respondents than do face-to-face surveys. Therefore, the data collection method is appropriate because my target audience should have convenient internet access, are highly familiar with online contexts, and have some experience with online banking.

To test the hypotheses, this study relied on various sets of constructs and their indicators. Most indicators were derived from the items in a survey questionnaire designed with a seven-point scale from strongly disagree (1) to strongly agree (7). The items were validated in prior studies and used with minor modifications to apply to online context. The online survey comprised of four sections (See appendix 1). In the first section, participants were situated in the context of their relationship with online banking. Second, participants were asked different questions about the use of channels such as degree of online channel use, the mooring effects and demographic variables (e.g. age, education, sex). Third, the respondents were presented with a hypothetical scenario, in which a failure in the use of online channel has happened. Each independent variable tested in this study represented one scenario (1. 2. 3. 4).

(36)

finally, participants were asked about the degree of channel usage after the failure, as well as the degree of usage of other channels. To ensure the clarity of the questionnaire, a pilot test with five university students was done prior to administration of the main survey. Furthermore, two professors of University of Amsterdam reviewed the items, aiming to ensure the consistency of each item. Only few adjustments were made from the pre-testing.

4.3 Sampling and sample design

To invite the respondents to the survey, I sent a link through email and Facebook to a sample of 500 incumbents, living in the Netherlands. This country encompasses a total population of approximately 16,9 million inhabitants (CBS, 2017). Unfortunately, the population is large and the exact sampling frame was unknown. Therefore, this study used a non-probability convenience sample. This research aims to reach as many respondents as possible. Nevertheless, an initial minimum number of respondents was set. The author aimed to gather a total of 200 completed-surveys. This means, 50 surveys per group (one control group and three treatment groups). Based on the sampling technique, the response rate was difficult to predict for the surveys distributed via Facebook. However, for the surveys distributed via e-mail, earlier research on similar topics received an overall response rate between 17.7% and 27% (Piercy & Archer-Brown, 2014).

In total 297 participants filled in the survey; however, not all respondents filled in the survey completely. Incomplete questionnaires were thus unfinished questionnaires. For that reason, I decided to drop anyone who did not fill in the questionnaire completely, a total of 46 respondents. This left me with n = 251. In terms of descriptive statistics of the sample, the respondents’ age ranged from 18 to 65+ years old, being the group between 26 and 40 years old the biggest one. In terms of level of education, most respondents have a bachelor degree. Moreover, the number of males and females was approximately equally distributed. When

(37)

looking at differences between men and women’s age and level of education, Levene’s test was not significant: age (F = 3.153, p = 0.077) and level of education (F = .384, p = 536). Hence, equal variances among men and women were assumed. Both t-test between the means of men and women were not significant: age t (1.430) = .205, p = .154 and education t (-1.384) = -.149, p = .168. Therefore, there was not statistically significant differences in age and level of education between men and women. Table 1 shows the sample’s descriptive statistics.

Measure Item Number N= (251) %

Gender Female 127 50.6 Male 124 49.4 Age 18-25 74 29.5 26-40 99 39.4 41-55 33 13.1 56-65 35 13.9 65+ 10 4

Education Did Not Complete High School 4 1.6

High School/GED 55 21.9

Bachelor's Degree 117 46.6

Master's Degree 62 24.7

Advanced Graduate work or Ph.D. 13 5.2

Tenure Less than one year 12 4.8

1 - 2 years 9 3.6

3 – 5 years 54 21.5

More than five years 176 70.1

Table 1. Descriptive statistics

4.4 Variables

This section presents the description of the main variables under investigation. This includes: the independent variable, moderators, outcome variables and some possible control variables.

4.4.1 Independent variable

In this research, customer negative experience acts as the independent variable. An experiment including three negative experiences was conducted to recreate real negative incidents in online channels. Poor e-service quality (n = 63), security problems (n = 63), and

(38)

2). Presenting participants with carefully constructed and realistic scenarios to assess the dependent variable, it increases the immersion of participants. Situating in a real context is more likely to engage participants to a greater extent, allowing them to generate proper judgement (Aguinis & Bradley, 2014). The three negative experiences were selected from previous research. (Holloway & Beatty, 2003) and (Dai, Forsythe, & Kwon, 2014) identified negative antecedents of customer dissatisfaction in online channels. Therefore, these three antecedents were included in this research and presented in form of scenarios to determine the effect of real negative experiences in online channels on subsequent channel selection and usage. In addition, a control group was included (n = 60). The control group did not experience any negative incident. In this case, the scenario described a situation, where incumbents could perform a financial transaction online without any problem. The control group is used as a reference group. In more detail, the treatment groups (Participants with a negative experience) will be compared to the control group (Participants without a negative experience) to determine the effect of the negative experience in online channels on channel switching behaviour.

This research was empirically tested in financial services. Specifically, online banking. Therefore, incumbents were presented with hypothetical incidents that they may experience while performing banking transactions. In this case, each treatment group experienced a negative situation, while transferring money online. After these negative experiences participants were asked to answer a set of different questions, testing the effect on their channel selection behaviour (See appendix 1 and 2).

4.4.2 Moderators

As explained in chapter 3, moderators might either hinder or facilitate selection behaviours. Therefore, business relationship or personal factors may moderate the influences of the independent variable (Bansal, Taylor, & James , 2005). Building on previous research,

Exploring the effects of negative experiences in online channels on subsequent channel selection and usage