Purchasing experiences goods through mobile applications : evidence from fashion retail

Master Thesis

Purchasing experiences goods through mobile applications:

Evidence from fashion retail

University of Amsterdam Faculty of Economics and Business MSc. in Business Administration -- Entrepreneurship and

Management in the Creative Industries--track Azieza Salham 11152400

Thesis Supervisor: I. Rozentale

Statement of originality

This document is written by Azieza Salham 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.

Abstract

The fast growth of smartphone users has led to an increase in mobile commerce. Fashion retailers who want to capture market share in this new channel have developed mobile shopping applications in order to generate more revenue. This has changed the way people shop for clothes and mobile shopping is now seen as a viable shopping channel. Prior studies have emphasized the reasons for the adoption, acceptance and usage of mobile shopping. This study elaborates on the effects of mobile shopping on customer satisfaction and purchase intent. Five factors that might have an effect on customer satisfaction and purchase intent, which drive people to use mobile applications, were chosen from the literature (i.e. ease of access, ease of use, richness of information, interactivity and purchase intent). The data was collected from among 283 people who filled in an online or paper survey. The results show ease of access, ease of use and fashion innovativeness have a direct positive effect on purchase intent. The results also showed that ease of use and richness of information have an effect on purchase intent when mediated by customer satisfaction. According to our results, the ease of use has the most effect on purchase intent (β = .33). The results of this research will be useful for clothing companies who have mobile shopping applications to optimize their marketing strategies, and show them the need to promote their mobile shopping applications.

Table of content

Abstract ... 2

Introduction ... 4

2. Literature review ... 7

2.1 Mobile shopping vs online shopping ... 7

2.2 Clothing shopping in a new multichannel environment ... 9

2.3 Mobile shopping linked with customer satisfaction and purchase intent ... 10

2.3.1 Factors that influences mobile shopping channel ... 11

2.4 Conceptual framework ... 19

3. Data and method ... 20

3.1 Research design ... 20

3.2 Data collection ... 20

3.3 The sample and procedure ... 21

3.4 Measurements ... 21 4. Results ... 23 4.1 Descriptive ... 23 4.2 Reliability analysis ... 23 4.3 Correlation ... 24 4.4 Regression ... 25 4.5 Process ‘’mediation’’ ... 26 4.6 Split data ... 32 5. Discussion ... 35 5.1 Limitations ... 38 6. Conclusion ... 39 References ... 41

Appendix A Survey questionnaire ... 48

Introduction

The fast growth of smartphone users has led to an increase in availability of mobile applications, which encourage people to spend more time on their phones (Garg & Telang, 2012). This has also led to a change in the way people purchase products. Apps, as they are commonly called, are rapidly catching up to mobile browsers as a viable shopping channel; 49% of smartphone shoppers in 2013 said they would rather use apps than browsers to make a purchase and 42% of smartphone shoppers said apps strengthen brand connections (Adobe, 2013). It is also interesting to note that consumers view the same number of products on their smartphone that they do on desktop sites (criteo, 2015). Mobile commerce in the Netherlands is expected to generate approximately 19.2 billion euro in 2017, 11.1 billion euro of which is expected to be generated by retail (Forrester Research, 2012). This shows the system’s growth and the potential it offers as an independent sales channel and thus makes it an interesting channel to research.

This new channel can provide retailers with a wider variety of commercial activities than the existing shopping channels have been able to. Mobile services have unique advantages that retailers can use to enhance their position, for instance, mobile commerce (m-commerce) enables retailers to send customized information and pinpoint user location services in real-time (K. Yang, 2010a). Despite such advantages and increase in m-commerce, some firms do not know if and how they should respond. According to the Strong Mail mobile survey (2012) 37% of the businesses surveyed cite “lack of strategy” as the top reason for not launching a mobile application. Therefore, the broader objective of this thesis is to generate knowledge about m-commerce that would assist firms in making such strategies.

Previous studies have identified two behavioural outcomes of the online shopping experience: customer satisfaction and purchase intent (Rose, Clark, Samouel, & Hair, 2012). Customer satisfaction has a direct impact on the primary source of future revenue streams for most companies (Fornell, 1992) and is considered an antecedent of purchase intent (Seiders, Voss, Grewal, & Godfrey, 2005a).

This makes it interesting to research the effect of mobile shopping on customer satisfaction and purchase intent. Previous studies on mobile shopping have mainly focused on the adoption, acceptance and usage of mobile shopping (Kim, Chan, & Gupta, 2007; Kleijnen, De Ruyter, & Wetzels, 2007a; J. Wu & Wang, 2005).

In addition, the suitability of online retailing to consumers varies per product category, depending on the uniqueness of the products (Grohmann, Spangenberg, & Sprott, 2007). Most studies look at retailing in general (Wang, Malthouse, & Krishnamurthi, 2015), relatively few studies so far have examined and developed studies on mobile shopping (m-shopping) with respect to specific products or product categories. Evidence, from sectors like food retail, shows that the number of orders placed per year increases as customers adopt m-shopping (Wang et al., 2015), while in cosmetics retail customers buy the products online once overcome tradition barriers (Lian & Yen, 2013). Despite the fact that the most popular type of good purchased via mobile applications is clothing (Episerver, 2015), there is no evidence of how this works specifically. We only know that shopping for clothes is not the same as shopping for other items because attitudes towards online shopping are stronger for cognitive products (Warrington & Shim, 2000). Shopping online does not provide the same level of product information as shopping in physical stores (Burke, 1997). An individual’s judgment of a product may be affected by whether they can touch a product during evaluation. Moreover, the existing research becomes less useful as years pass; this is due to the fact that mobile commerce takes place in a fast-paced environment with continuous development.

It is necessary to obtain more information about consumer mobile shopping behaviour for clothing in order to create better strategies. This makes it interesting to look at the fashion industry in the mobile context. Hence this theses explores the following research question:

- What drivers influence customer satisfaction and purchase intent in the mobile fashion shopping context?

In order to answer the research question, we first identified the possible drivers of m-shopping in our context using a literature review. In order to empirically test the relationships, we conducted a cross-sectional survey (n=283). The survey was distributed in various ways to ensure a variety of responders (e.g. online, email and face to face in a shop). The model was tested via regression and PROCESS analyses using SPSS software.

This paper is structured as follows. In the next section, the relevant previous literature is discussed. The third section discusses the methodology. The fourth section analyses and test the research hypotheses. This will be followed by the discussion and finally the conclusions are presented.

2. Literature review

This chapter provides an overview of the existing literature on mobile shopping. First, the differences between mobile shopping and conventional online shopping channel will be discussed. Secondly, the ideas on clothing shopping in a new multichannel environment are presented. Thirdly, mobile shopping linked to customer satisfaction and purchase intent will be explained with the factors that influence mobile shopping channel. Finally, the chapter ends with presenting the conceptual framework.

2.1 Mobile shopping vs online shopping

Mobile commerce (m-commerce) is defined by Chong (2013) as online transactions conducted through mobile devices (e.g. smartphones and tablets) using wireless telecommunication networks. M-commerce and e-M-commerce share some commonalities as they both represent a form of online commerce. Additionally, e-commerce research served as a starting point for many studies investigating similar concepts in the m-commerce context, thus making it important to include the concept in this study. E-commerce can be defined as computer-mediated shopping transactions, conducted through computers and laptops (Dholakia & Dholakia, 2004). Varshney and Vetter (2002) suggest that m‐commerce can be seen as an e‐commerce over the wireless devices. However, some researchers disagree and suggest that m‐commerce is more than e‐commerce due to its different interaction style, usage pattern and value chain (Feng, Hoegler, & Stucky, 2006). M-commerce has the advantage of interactivity, two-way communication between senders and receivers which becomes a source for vital information (Kang 2002). This is both an advantage for the customer as well as the retailer. It also offers a new business opportunity with its own unique characteristics and functions, such as mobility and broad reachability (Feng et al., 2006). Unlike the Internet, mobile applications enable retailers to send customized information and provide the application users with direct access to sale information such as discounts, offers, and coupons. When given relevant promotion information, consumers are expected to have a positive attitude toward using mobile shopping services(K. Yang, 2010b) and experience greater satisfaction.

There are two different ways to shop through a mobile device: mobile applications (closed-ended source), a specific software that can be downloaded on a mobile device, and mobile websites (open-ended source) which consist of an adapted version of the regular website (Bellman, Potter, Treleaven-Hassard, Robinson, & Varan, 2011). According to Criteo (2015) the average smartphone user can see 14 products via the app compared to four products on the browser, the report also claims that mobile apps are easier to purchase from, which leads to a purchase rate that is 1.6 times higher than via the mobile browser.

The construct m-commerce is a bit too broad, as examples include investing, banking, shopping, mobile phone service and auctions (Kleijnen, De Ruyter, & Wetzels, 2007b) therefore, a more specific concept, mobile shopping, is chosen for this research. M-shopping is a part of m-commerce and can be defined as the activities undertaken by consumers who use wireless Internet services when shopping using mobile phones or tablets (Ko, Kim, & Lee, 2009). Online shopping and mobile shopping are often seen as the same, however many researchers have argued that there is a difference between them (Dholakia & Dholakia, 2004). Various researchers have provided different definitions for m-shopping. Previous studies suggest that mobile shopping is very different from its more traditional, desktop computer–based predecessor as mobile shopping services are accessible on the move through mobile devices (e.g. smartphones and tablets). With fundamental differences in presentation, processing, and interaction modalities compared to a desktop computer, these differences enable a new set of service possibilities, for instance location‐, customer‐, personalization‐ , presence‐and context‐based services (Dholakia & Dholakia, 2004). For that reason, mobile devices are seen as an effective marketing tool because they can provide these services. M-shopping has numerous advantages over online shopping such as the fact that consumers can shop ‘anytime’ and ‘anywhere’ due to the uninterrupted connectivity of using mobile devices. This brings a benefit in terms of convenience and time saving. Additionally, the shopping flow can be supported from the initial phase of logging, searching, comparing prices, ordering, and paying to conducting after sales services. However, there are also some disadvantages with m-shopping, for example, mobile devices

have smaller screens for online transactions, mobile shopping therefore, requires more processing pages and steps than PCs do (Lu & Yu-Jen Su, 2009).

2.2 Clothing shopping in a new multichannel environment

Clothing and apparel are different in nature from other products. Clothing varies extensively on different attributes such as price, quality, fit, design etc. Previous studies have shown that there is a difference between the shopping experiences for different types of goods. Products and services can be classified as a search or experience good (Peterson, Balasubramanian, & Bronnenberg, 1997a). Search goods are products whose attributes and quality can be assessed using readily available information, and can therefore be determined before purchase (McWilliams & Siegel, 2001). For example, the quality of a computer can be determined before the purchase by looking up information about the computer. Search goods are easier to compare and evaluate prior to purchase. Experience goods are those products which consumers feel they need to directly and physically (sensory) inspect to evaluate (Weathers, Sharma, & Wood, 2007). Zhai, Coa Mokhtarian and Zhen (2016) argue that clothes can be seen as an experience good. One reason for this may be the fact that clothing is also used by many consumers as a form of self-expression and is reflective of the consumer’s self-image (Shim & Kotsiopulos, 1991). With experience goods, information about the good's features may not be sufficient for a consumer to engage in an Internet-based transaction. Despite this, clothing is one of the most common product categories purchased online (Dennis, Merrilees, Hansen, & Møller Jensen, 2009).

Consumers do spend similar amounts of time online gathering information for both search and experience goods (Huang, Lurie, & Mitra, 2009). However, there are differences in the browsing and purchase behaviour of consumers for these two types of goods. With experience goods consumers spend more time on a page compared to search goods (Huang et al., 2009). In addition, people that purchase experience goods online go through fewer pages before deciding to purchase the product than when purchasing search goods (Huang et al., 2009). Also interesting to note, consumers who

purchase experience goods usually buy it from the primary source who provide the product information. This is the opposite from search goods, where consumers do buy from other retailers that the primary source of product information.

At every stage of the customer journey, from the first orientation to the purchase and even post purchase, more and more consumers use a smartphone or a tablet. Despite the increase in popularity, the literature is very limited to the reasons why people would want to use it and there is still a lack of hard evidence regarding its impact. This makes mobile shopping an important channel to research.

In the past, consumers could only purchase clothes through a single retail channel. Over the years, retailers have employed multi‐channel retailing by combining different distribution channels to deliver products and/or services (Cho & Workman, 2011). In a multi-channel environment consumers can choose from which channel they would like to shop. They can either purchase clothes in physical stores, through the mail, catalogue, phone etc. In more recent time, consumers have gained the option of purchasing their apparel online and now there is a new channel, mobile commerce.

2.3 Mobile shopping linked with customer satisfaction and purchase intent

Two behavioural outcomes of online customer experience are identified in prior research: satisfaction and purchase intent (Rose et al., 2012). Customer satisfaction has a direct impact on the primary source of future revenue streams for most companies (Fornell, 1992) and is considered an antecedent of purchase intent (Seiders et al., 2005a). Previous literature has studied customer satisfaction in an offline and e-commerce context but there is little research done on customer satisfaction in the m-commerce context.

Customer satisfaction is a key driver of loyalty in the retail (Cronin, Brady, & Hult, 2000) and is considered as a key antecedent to loyalty and repurchase intention(Seiders, Voss, Grewal, & Godfrey, 2005b). This was also confirmed in the online context (Ha, Janda, & Muthaly, 2010). Satisfaction is described by Hunt (1977, p. 459) as “an evaluation of an emotion” meaning "satisfaction is not the

pleasurableness of the [consumption] experience, it is the evaluation rendered that the experience was at least as good as it was supposed to be" (Hunt, 1977, p. 459). Customer satisfaction is also described as the degree of overall pleasure or contentment felt by the customer, resulting from the ability of the service to fulfil the customer's desires, expectations and needs in relation to the service (Hellier, Geursen, Carr, & Rickard, 2003a). According to Lin and Wang (2006) customer satisfaction of mobile commerce is consumer’s total response to the purchase experiences in a mobile commerce environment. Therefore, in this study, customer satisfaction is defined as the degree of overall pleasure or contentment felt by the customer with the experience. In this context satisfaction is the overall level of customer pleasure and contentment resulting from experience with the service and thus shopping channel.

Purchase intent is the other behavioural outcome of online customer experience. Purchase intent is the willingness of a consumer to buy goods or services (Cronin et al., 2000; Zeithaml, Berry, & Parasuraman, 1996). Repurchase intention is the process of an individual purchasing goods or services from the same company (Hellier, Geursen, Carr, & Rickard, 2003b). Purchase intent in this study is defined as willingness to purchase goods through a particular channel and the intention to repurchase

again through the same channel.

2.3.1 Factors that influences mobile shopping channel

In e-retailing and m-retailing satisfaction results from the customer’s evaluation and impression of the website’s performance across a number of attributes (Jin & Park, 2006) e.g. security/privacy and the experiential aspect. Shankar, Venkatesh, Hofacker and Nail (2010) argue that it is important to know the drivers of device and service adoption in the shopping process to create a powerful mobile marketing strategy. For a mobile strategy to be effective, companies need to know why consumers choose certain devices for shopping. This suggests that drivers influence customer satisfaction and purchase intent. If consumers think the service will satisfy them, they are more likely to use the service and purchase products through that service. The perceived advantages of the channel help convert it

into an appealing way of shopping which keeps attracting more new users. These advantages are explained below and in this study those factors were used to see if they have an influence on customer satisfaction and purchase intent. The five factors that can be identified from the literature are ease of access, ease of use, richness of information, fashion innovativeness and interactivity.

Ease of access

Some advantages for using a mobile channel are stated by Nassuora (2013) as: faster access, more powerful, more effective and accessible at almost anytime, and anywhere for its users. M-commerce’s main advantage is the ability to offer convenience and accessibility to its users in comparison to e-commerce (Nassuora, 2013). With the ease of access that mobile devices offer, consumers can place and cancel orders whenever they want. This would suggest that ease of access would be a significant driving factor affecting consumer intention to use mobile shopping services, several studies have also found this assumption to be true (Schaupp & Bélanger, 2005; Szymanski & Hise, 2000). According to Kim, Chan and Gupta (2007), speed and time efficiency are often regarded as the main benefits consumers can achieve through the use of technology. Previous studies have shown that ease of access provides consumers with speed and time efficiency (Kleijnen et al., 2007b).

Ease of access is defined in this study as the degree to which a customer believes that accessing the Internet via a mobile phone will be free of effort. This not only means that the mobile device has access to the Internet but also that the mobile device is available to customers at all times as people have their mobile devices consistently with them. If provided easy access and usage of mobile shopping services, consumers can efficiently achieve their shopping goals. Yang (2010b) found that ease of access and use of mobile shopping services will enhance mobile shopping service quality, shopping enjoyment, and efficiency. Usually, consumers select a channel that will save them the time and effort this way they can maximize their utility. Mobile shopping services that can get easier access to mobile Internet will be seen as a benefit to the customer. Furthermore, the easy access of up‐to‐date information and promotions for products and services in the mobile shopping channel may enable

shoppers to find out variety of product information and services, thus, it will increase consumer satisfaction. Several studies have also found that mobile convenience can lead to purchase intentions and loyalty (Okazaki & Mendez, 2013; K. Yang, 2010b; K. Yang & Kim, 2012a). Therefore, considering these findings, the following hypotheses will be tested in this study:

H1a: Ease of access has a positive effect on purchase intent when buying clothes through a mobile application.

H1b: The positive relationship between ease of access and purchase intent is mediated by customer satisfaction when buying clothes through a mobile application.

Ease of use

Previous studies have shown that the ease of use plays an important role in determining the intention to use the technology, and hence to technology adoption(Bruner & Kumar, 2005; Holmes, Byrne, & Rowley, 2013; Mao, Srite, Bennett Thatcher, & Yaprak, 2005; Nysveen, Pedersen, & Thorbjørnsen, 2005). Davis’s Technology Acceptance Model (TAM) (1989) suggested that perceived ease of use is an important determinant for the acceptance of a system. Davis (1989, p. 322) defines ease of use as "the degree to which a person believes that using a particular system would be free of effort’’. If, for example, the ease of use through mobile applications do not outweigh customer losses (e.g. technical difficulties and impersonal experiences), then customers may simply revert back to traditional channels (Z. Yang & Peterson, 2004). This suggests that the ease of use plays a pivotal role in customer satisfaction with mobile shopping. The effect of ease of use included in TAM theory as acceptance driver for technology is also supported in studies of mobile services (Nysveen et al., 2005; Veríssimo, 2016) and therefore also used as a variable in this study. Although acceptance and satisfaction are different, some researchers have argued that customer acceptance leads customers using m‐Internet or m‐commerce and then customer satisfaction is built (Choi, Seol, Lee, Cho, & Park, 2008). Some studies have also examined the perception of the user's ease of use, this significantly influences user's intrinsic motivation such as enjoyment or playfulness(Moon & Kim, 2001). Then and Delong (1999)

suggested that consumers are willing to buy more clothing online if they discover the system is secure and easy to use, and the displayed product information are sufficient. In this study ease of use is defined as the ease to mobile apps and ease of navigating mobile site/application functions and features. Thus, the ease of use of a mobile application plays an important role in customer satisfaction and purchase intent. Therefore, the hypotheses states:

H2a: Ease of use has a positive effect on purchase intent when buying clothes through a mobile application.

H2b: The positive relationship between ease of use and purchase intent is mediated by customer satisfaction when buying clothes through a mobile application.

Richness of information

As mentioned above, clothing is, by nature, different from other products; purchasing clothes online is often perceived as more risky than offline purchasing because you do not know what to expect in terms of size, colour and quality (Grewal, Gopalkrishnan, & Levy, 2004). More information is needed overcome this perceived risk. A feature of both e-commerce and m-commerce is the ability of information to be made easily available to consumers. However, contrary to existing wired Internet websites, mobile applications can provide the customer with personalized information and services. Information plays an important role in influencing the purchase decision process of a consumer (Ranganathan & Ganapathy, 2002). Information allows the consumer to select the merchandise that best satisfies their needs, it also represents a predictor of online purchase intent (Ranganathan & Ganapathy, 2002). Previous studies have shown that the richer information (more extensive) available online lead to better buying decisions and higher levels of e-satisfaction (Peterson, Balasubramanian, & Bronnenberg, 1997b). Glazer (1991) found that that more information given increased consumers satisfaction with the purchase experience. This is consistent with other studies that show that the increased availability of information online will lead to more knowledgeable consumers, those consumers will be able to make better decisions, who will eventually experience greater satisfaction

with any purchases they make (Cook & Coupey, 1998). In this study, richness of information is defined as the quantity of information provided per product both general and personal.

One of the main disadvantages of m-shopping is that mobile devices have small display screens and this is seen as one of the main constraints of adopting m-commerce (Kleijnen et al., 2007b). However, Wang, Malthouse and Krishnamurthi (2015) argue that even though mobile devices’ screen size and functionalities are limited compared to PCs, given their temporal flexibilities, they are sufficient to provide convenient access when customers want to achieve specific goals. The majority of consumers use mobile devices in the pre-purchase phase to search for information (Church, Smyth, Cotter, & Bradley, 2007). Also, if there is too much information beyond a certain threshold, more customer effort is required to process the information, and thus the customers feel less satisfied and less confident regarding their purchase decision (Keller & Staelin, 1987). Therefore, it is important for retailers to offer personalized information to the customer to streamline the complex task of assessing products. Which can be done with mobile applications.

Previous studies have shown that information is an important element of customer satisfaction (Montoya-Weiss, Voss, & Grewal, 2003) and purchase intention (Shim, Eastlick, Lotz, & Warrington, 2001), and as information search is usually the first step in a customer’s purchase decision-making process, the more information a customer has, the better and more informed their decision will be. The richness of information is very important. Therefore, the hypotheses states:

H3a: Richness of information has a positive effect on purchase intent when buying clothes through a mobile application.

H3b: The positive relationship between richness of information and purchase intent is mediated by customer satisfaction when buying clothes through a mobile application.

Interactivity

Previous researchers have also highlighted the significance of interactivity to customer loyalty in electronic commerce (Alba et al., 1997). Interactivity can be seen as a two-component construct

consisting of navigation and responsiveness (G. Wu, 1999). First, the navigational process facilitated by interactivity dramatically increases the level of control experienced by the customer (Hoffman & Novak, 1996). Secondly, interactivity enables a search process that can quickly locate a desired product or service (Alba et al., 1997). This can be just through the availability of a search bar. Finally, interactivity enables an increase in the amount of information that can be presented to a customer (Srinivasan, Anderson, & Ponnavolu, 2002). For example, mobile apps can provide the customer with not only product information but also with reviews and recommendations from other shoppers. The mobile app can show different ways of styling the fashion items. It can show how other people who have bought the fashion item wearing it. Also, the apps can give recommendations based on what other people who have similar tastes have bought. Interactivity has been defined in many ways depending on the perspective of the researchers. Interactivity is the degree to which two or more communication parties can act on each other, on the communication medium, and on the messages and the degree to which such influences are synchronized (McMillan & Hwang, 2002). Previous research has mainly focused on interactivity in the online environment(Fiore, Jin, & Kim, 2005). As mentioned before, mobile commerce has features that are not available to online environment such as localization and mobility. Users can interact with companies, product, offers and services wherever they have connectivity through a mobile device. Previous research has shown that interactivity in the online environment can lead to purchase intentions (Zhang, Lu, Gupta, & Zhao, 2014). It will be interesting to test whether the result will be the same in the mobile context. Thus, leading to the hypotheses:

H4a: Interactivity has a positive effect on purchase intent when buying clothes through a mobile application.

H4b: The positive relationship between interactivity and purchase intent is mediated by customer satisfaction when buying clothes through a mobile application.

Fashion Innovativeness

Since mobile shopping can be viewed as an innovative way of shopping, innovativeness should have a relevance to its adoption and effects. Innovativeness is defined by Rogers (1995, p. 22) as “the degree to which an individual is relatively earlier in adopting new ideas than other members of his/her social system.” Previous studies have found that domain‐specific and not general innovativeness leads to online shopping (Varma Citrin, Sprott, Silverman, & Stem Jr, 2000). Innovativeness is thus domain-specific, which means that consumers tend to be innovators for a specific product or product category (Goldsmith & Hofacker, 1991). One can assume that this will also be the case in the mobile shopping context. This study only focuses on the product category clothing therefore innovativeness will be limited to the context of fashion. Fashion innovativeness is defined as the degree to which an individual adopts a new fashion early on, relative to other member in their social system. Innovativeness is possessed by all individuals at a different level (Varma Citrin et al., 2000) and general innovativeness has highly impacted fashion innovativeness (Simpson, 2006). Furthermore, one study found that innovativeness is able to predict the frequency of online shopping and future online shopping intention (Goldsmith & Flynn, 2004) as well as Internet shopping behaviour (Limayem, Khalifa, & Frini, 2000). Consumers characterized by a high degree of innovativeness are usually very open to new experiences (Leavitt & Walton, 1975). If you are looking for something specific (clothing) you are more likely to purchase through different channels. One can make the assumption that people that are fashion innovators are also people who like to be involved in new shopping experiences. Therefore, the hypothesis states:

H5: Fashion innovativeness has a positive effect on purchase intent when buying clothes through a mobile application.

Purchase intent

This study suggests that ease of access, ease of use, richness of information and interactivity will positively influence customer satisfaction. This is based on previous studies. So, the higher the three

factors score, the more likely a customer is satisfied. For example, if the application is very easy to use and navigate, the customer will be satisfied and will likely have an intention to purchase clothes through the mobile application. Zeithaml, Berry and Parasuraman (1996) argue that consumers that experience a higher level of satisfaction tend to have a stronger intention to repurchase and recommend the purchased product. Previous studies have also shown in that there is a positive relationship between customer satisfaction and post-purchase intention when it comes to service (Brady & Robertson, 2001; Cronin et al., 2000). It can be assumed that the mobile app encourages a pleasurable experience and such an experience is expected to enhance consumer satisfaction and encourages consumers to buy via mobile apps. In this study customer, satisfaction acts as a mediator. The mediator establishes the extent to which the four factors (ease of access, ease of use, richness of information and interactivity) influence purchase intent through the mediator, customer satisfaction. This study will also include three control variables. Previous studies have shown that satisfaction ratings vary on the basis of customer characteristics (Fornell, Johnson, Anderson, Cha, & Bryant, 1996). This was also found in the mobile environment context (e.g. (Wang et al., 2015). The customer characteristics that were most used were age and gender. An additional control variable will be used to measure how often the customer purchases fashion items.

2.4 Conceptual framework

Table 1 Hypotheses Hypotheses

H1a: Ease of access has a positive effect on purchase intent when buying clothes through a mobile application.

H1b: The positive relationship between ease of access and purchase intent is mediated by customer satisfaction when buying clothes through a mobile application.

H2a: Ease of use has a positive effect on purchase intent when buying clothes through a mobile application. H2b: The positive relationship between ease of use and purchase intent is mediated by customer satisfaction when buying clothes through a mobile application.

H3a: Richness of information access has a positive effect on purchase intent when buying clothes through a mobile application.

H3b: The positive relationship between richness of information and purchase intent is mediated by customer satisfaction when buying clothes through a mobile application.

H4a: Interactivity has a positive effect on purchase intent when buying clothes through a mobile application.

H4b: The positive relationship between interactivity and purchase intent is mediated by customer satisfaction when buying clothes through a mobile application.

H5: Fashion innovativeness has a positive effect on purchase intent when buying clothes through a mobile application.

3. Data and method

This chapter will focus on the empirical setting of this research. First, the research design will be discussed. Second, the manner of data collection in addition to the composition of the sample will be presented. Lastly, the variable measurements, along with an explanation, will be outlined.

3.1 Research design

A research design refers to a framework for collecting and analysing data (Bryman & Bell, 2015). The aim of the research and the research question are important issues to consider when choosing a research design among other issues (such as time frame for collecting data). The empirical objective of this research is to research whether purchasing clothes through a mobile application has an impact/effect on customer satisfaction and purchase intent. In order to test the proposed hypothesis, a quantitative research was conducted. This quantitative research was executed through a cross- sectional survey. A cross-sectional survey is a study of a particular phenomenon at a particular point in time (Saunders, Lewis, & Thornhill, 2007). For this study, an online survey and a paper survey was set up. The survey is chosen for several reasons. Firstly, because they enable researchers to collect a large number of respondents. Secondly, an online survey reduces geographical dependence. This way data collection has the ability to reach a larger and more diverse group of people than its offline counterpart. Lastly, anonymity is easier to guarantee. It is important to note that there are disadvantages with a cross-sectional design. The main limitation lies in the fact that you test at one point in time, the risk is that there is low internal validity.

3.2 Data collection

The survey was distributed in various ways to ensure a variety of responders. One distribution channel that was used was social media. This distribution channel provided data from respondents that have experience with online and mobile shopping channels. The survey will also be distributed through email. Finally, the survey was also distributed in a shop (Hema). The survey was given to people who were willing to give up some time to fill in the survey. It is very difficult to retrieve a probability sample from a very large population and since the population for this study is very large as this study looks at

all Dutch consumers with a smartphone and/or tablet, a convenience (non-probability) sample was chosen in this case. Furthermore, convenience sampling was chosen as there are time constraints. This study used the software program Qualtrics to create and distribute the survey.

3.3 The sample and procedure

The research population consists of all Dutch consumers who have a mobile device on which the consumer can download a retail application. The mobile devices can either be a smartphone or a tablet. In the Netherlands, 81% of the population aged between 18 and 80 own a smartphone and 65%of the population starting from 13 years old own a tablet (GFK, 2015). Due to the fact that it is not possible to examine the entire population, a sample was chosen. For this study, the non-probability sampling technique to select the respondents was used. This research strived for as many respondents as possible in order to have a more representative sample and generalizable outcome. To ensure a big sample multiple distribution methods were used. Over a period of three week, 327 respondents filled in the survey. A number of people were approached face to face and 73 people filled the survey in completely. The survey was also posted online (Facebook) and there 106 people filled in the survey. The third approach used to spread the survey was email and this approach got 70 people to fill in the survey. The final approach was to spread the survey among friends and family and asking them to spread the survey in turn. This approach got 78 people to fill in the survey. The final database consisted of 283 respondents because of the 327 respondents, 44 did not complete the survey or did not pass the ‘’test’’ item.

3.4 Measurements

Unless otherwise stated, respondents to survey items used a seven-point Likert-type scale, ranging from 1 = “strongly disagree” to 7 = “strongly agree”. The survey was administered in English.

Dependent variables. The dependent variable of this study is customer purchase intent in the mobile shopping context. Purchase intent was assessed by using 3 items derived from Pavlou and Fygenson (2006), including ‘’If there is shopping need, I intend to purchase fashion items by using mobile shopping’’.

Independent variables. The independent variables of this study are ease of access, ease of use, richness of information, fashion innovativeness and interactivity.

Ease of access was measured by using 3 items used by Ko, Kim and Lee (2009) scale. Including ‘’Mobile shopping is accessible at any time and place’’.

Ease of use, three items adopted from Gefen, Karahanna and Straub (2003) were used to measure this item, including ‘’It is easy to purchase fashion products through mobile shopping apps’’.

Richness of information was measured by using the Childers, Carr, Peck and Carson (2002) scale, shortened to 4 items. Including ‘’Mobile shopping allows me to form an impression about a fashion item similar to that from up-close examination’’.

Interactivity was measured by using three items from McMillan and Hwang (2002). These four independent variables were tested by using a five-point Likert-type scale, ranging from 1 = “not important” to 5 = “very important”.

Fashion innovativeness was measured by using 2 items from Oliver and Breaden (1985). Including ‘’I am among the first ones to try new fashion styles’’. In addition, one item was used from Manning, Bearden and Madden (1995), including ’’I frequently look for new fashion items’’.

The mediating variable. The mediator in this study is customer satisfaction and will be measured by using the Seiders, Voss, Grewal and Godfrey (2005b) scale, shortened to 3 items. The items that fit with this study were chosen.

Control variables. To rule out potentially spurious relations in the analysis, the control variables age (in years), gender (0 = male, 1 = female) and frequency of shopping tenure were used. These control variables are commonly used in studies on shopping and online shopping (Kleijnen et al., 2007b; Montoya-Weiss et al., 2003). An additional item (‘’test’’ item) was added to test whether people are actually paying attention while filling in the survey. The ‘’test’’ item in the survey was “Respond with ‘strongly agree’ for this item’’.

4. Results

In this chapter, the reliability of scales and hypotheses were test by doing data analyses in statistical software SPSS 20. The analyses were first conducted on the whole sample and then the sample was divided into 2 groups; has never shopped via a mobile application and has shopped via a mobile application.

4.1 Descriptive

Among the 283 respondents, 196 are female respondents which represent 69.3% of the total sample group and which makes them the majority. The rest of the 30.7% respondents are male with 87 respondents. In terms of age, the group aged between 18 and 24 represented 44.2% of the group, so more younger consumers filled out the survey than older consumers. Most respondents have either never shop through a mobile application (117) or use it sometime when shopping (118).

Measure Items Frequency Percentage

Gender Female 196 69.3 Male 87 30.7 Age < 18 31 11 18 -24 125 44.2 25 – 34 66 23.3 35 – 44 43 15.2 45 – 54 16 5.7 55 – 64 2 0.7

Frequency of shopping Never 117 41.3

Sometimes 118 41.7

About half of the time 27 9.5

Most of the time 18 6

Always 4 1.4

4.2 Reliability analysis

Reliability analysis was done to determine if the scale is reliable. The alpha coefficient was used as measurement. Cronbach's alpha is a measure of internal consistency, that is, how closely related a set of items are as a group. It is considered to be a measure of scale reliability. A reliability coefficient of .70 or higher is considered "acceptable" in most social science research situations (Field, 2009). All the variables have a Cronbach’s alpha higher than 0.70, ease of access (a=0,845), ease of use (a=0,747), richness of information (a=0,755), interactivity (a=0,701), fashion innovativeness (a=0,885), customer

satisfaction (a=0,857) and purchase intent (a=0, 878). The coefficients are all above 0.70, therefore the scale is reliable. The corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above .30). Also, none of the items would substantially affect reliability if they were deleted.

VARIABLES CRONBACH’S ALPHA EASE OF ACCESS 0,845 EASE OF USE 0,747 RICHNESS OF INFORMATION 0,755 INTERACTIVITY 0,701 FASHION INNOVATIVENESS 0,885 CUSTOMER SATISFACTION 0,857 PURCHASE INTENT 0,878

4.3 Correlation

A correlation matrix, which lists the Pearson’s correlation coefficients of all the independent variables, is presented in the table below. This was used to determine if variables are linearly related to each other. When the Sig (2-Tailed) value is less than or equal to .05, we can conclude that there is a statistically significant correlation between two variables. This means that increases or decreases in one variable significantly relates to increases or decreases in a second variable. In this data set, some independent variables are correlated. This means that there is a presence of correlation among the independent variables. The independent variables are not totally independent. E.g. if one independent variable goes up, not only does the dependent variable go up, but also another independent variable. There are several correlations between the independent variables such as ease of access which is correlated with ease of use, and interactivity which is correlated to richness of information. There are more correlations in the matrix, however, the independent variables are not highly correlated, so multicollinearity is not an issue in this research (Farrar & Glauber, 1967).

Means, standard deviation, correlation Variables M SD 1 2 3 4 5 6 7 8 9 10 1. Ease of access 3.93 0.94 - 2. Ease of use 4.22 0.72 .55** - 3. Richness of information 4.43 1.15 .29** .25** - 4. Interactivity 4.85 0.84 .60** .48** .27** - 5. Fashion innovativeness 4.51 1.36 .15* .14* .25** .21** - 6. Customer satisfaction 4.97 1.07 .16** .21** .28** .10 .40** - 7. Purchase intent 5.04 1.15 .05 .23** .288* .07 .39** .68** - 8. Gender 1.69 4.62 -.05 -.03 .02 -.06 .11 .05 .02 - 9. Frequency of shopping 1.84 0.92 -.12* -.27** .12 -.13* .31** .31** .27** .09 - 10. Age 2.63 1.089 .15* .04 .11 .19** -.10 -.02 -.02 -.12* .02 - **Correlation significant at the 0.01 level (2-tailed)

*Correlation significant at the 0.05 level (1-tailed)

4.4 Regression

The direct relation is tested with a regression analysis, this was used to predict the value of the dependent variable based on the value of other independent variables. It identifies the relative importance of each independent variable to see if the second model, where the three independent variables are added, adds significant explanatory power over and above the other three control variables. Hierarchical multiple regression was performed to investigate the ability of ease of access, ease of use, richness of information, interactivity and fashion innovativeness to predict purchase intent, after controlling for gender, age and frequency of shopping. In the first step of hierarchical multiple regression, three predictors were entered: gender, age and frequency of shopping. This model was statistically significant F ( 3.279) = 7.430; p < .001 and explained 7.4 % of variance in purchase intent. After entry of ease of access, ease of use, richness of information, interactivity and fashion innovativeness at Step 2, the total variance explained by the model as a whole was 29.90% F(5.274) = 15.586, p<.001. The introduction of ease of access, ease of use, richness of information, interactivity and fashion innovativeness explained additional 20.50% variance in purchase intent after controlling for gender, age and frequency of shopping (R2 = .205; F( 5.274) = 15.586; p < .001). In the final model,

five out of eight predictor variables were statistically significant, with ease of use recording a higher Beta value (β = .33, p < .001) than ease of access (β = .14, p <.05), frequency of shopping (β=.23, p < .001), richness of information (β = .16, p < .01) and fashion innovativeness(β = .27, p< .001). In other words, if ease of use increases by one, their purchase intent will increase by 0.33. This regression analysis shows that H5 is supported as Fashion innovativeness is significant p<0.001.

R R² R² Change B SE Β T Step 1 .27 .07*** Gender -.02 .14 -.01 -.18 Age -.03 .06 -.03 -.53 Frequency of Shopping .33 .07 .27*** 4.69 Step 2 .53 .28*** .20*** Gender -.09 .13 -.04 -.76 Age -.00 .05 -.00 -.04 Frequency of Shopping .29 .07 .23*** 4.03 Ease of access -.18 .08 .14* -2.1 Ease of use .53 .10 .33*** 5.00 Richness of information .16 .05 .16** 2.89 Interactivity -.09 .09 -.07 -1.06 Fashion innovativeness .23 .04 .27*** 4.73 Statistical significance:*p<.05 ;**p <.01;***p <.001

4.5 Process ‘’mediation’’

To test the mediator model, the PROCESS macro written by Hayes (2012) was used. 5,000 bootstrap samples were drawn and a level of confidence of 95% was used for the analyses. All the specific indirect pathways and direct effects are given in Tables below. PROCESS was used 4 times to test all the variables going through the mediator. These variables are ease of access, ease of use, richness of information and the interactivity. The mediator is customer satisfaction and the dependent variable is

purchase intent. Controlled for age, gender and frequency of shopping Ease of Access

The effect of Ease of access (EoA) on Customer satisfaction (CatSF) a1 = 0.04 means that two people that differ by one unit on EoA are estimated to differ by 0.04 units on CatSF. The sign of a1 is positive,

meaning that those relatively higher in EoA are estimated to be higher in their CatSF. This effect is not statistically significantly different from zero, t= .4903, p = .6243, with a 95% confidence interval from -0.1221 to 0.2030 which is not above 0.

The effect b1 = .6554 indicates that two people who experience the same level of EoA but that differ by one unit in their level of CatSF are estimated to differ by b1 = .65545 units in intention to purchase. The sign of b1 is positive, meaning that those relatively higher in CatSF are estimated to be higher in their purchase intent. This effect is statistically different from zero, t= 12.6965, p = .000, with a 95% confidence interval from 0.5538 to 0.7570.

The indirect effect of 0.0265, this means that two people who differ by one unit in their reported EoA are estimated to differ by 0.0265 units in their reported purchase intent as a result of the tendency of those who perceive that their mobile is always available for mobile shopping, which in turn translates into greater purchase intent. This indirect effect is statistically not different from zero, as revealed by a 95% BC bootstrap confidence interval that is not entirely above zero (-.0937 to .1545).

The direct effect of EoA, c′ = -.2114, is the estimated difference in purchase intent between two people experiencing the same level of CatSF but who differ by one unit in their reported EoA. The sign c’ is negative, meaning that the person feeling less EoA but who is equally satisfied, is estimated to be -.2114 units lower in his/her reported purchase intent. This direct effect is statistically different from zero, t= -2.992, p = .003 (p<0.05). However the 95% confidence interval is from -0.3506 to -0.-722, which is not significant because it is not above 0.

The total effect of EoA on purchase intent is c = -.1849, meaning two people who differ by one unit in EoA are estimated to differ by -0.1849 units in their reported purchase intent. The negative sign means the person perceiving greater EoA reports lower intentions to purchase through a mobile application. This effect is statistically different from zero, t = -2.0796, p = .0385 (p<.0.05), or between -0.3599 and -0.0099 with 95% confidence. The direct effect H1a is supported and the indirect effect H1b is rejected.

Consequent

Antecedent

Customer Satisfaction (M) Purchase intent (Y) Coeff. SE P Coeff. SE P Ease of access (X) a1 .0405 .0826 .6243 c1 ’ -.2114 .0707 .0030 Customer Satisfaction (M) - - - b1 .6554 .0516 .0000 Constant i1 1.9270 .5069 .0002 i2 .7311 .4452 .1017 R2 = .2249 R2 = .5090 F(7.275) = 11.3961, p <.001 F(8.274) = 35.5071, P<.001

Effect SE P LLCI ULCI

Direct effect c1 ’ -.2114 .0707 .0030 -.3506 -.0722

Total effect c1 -.1849 .0889 .0385 -.3599 -.0099

Boot SE Boot LLCI Boot ULCI

Indirect effect a1b1 0.0265 .0619 -.0937 .1545

Ease of use

For the second analysis, M and Y stayed the same but the independent variable (X) was ease of use (EaseUse). The effect of EaseUse on Customer satisfaction is a1 = .3857. This effect is statistically significantly different from zero, t= 3.7997, p = .0002, with a 95% confidence interval from .1858 to .5855. The effect b1 = .6554. and this effect is statistically different from zero, t= 12.6965, p = .000, with a 95% confidence interval from .5538 to .7570. The indirect effect is 0.2528 and this indirect effect is statistically different from zero, as revealed by a 95% BC bootstrap confidence interval that is entirely above zero (.1094 to .4142). The direct effect of EaseUse, c′ = .3260, this direct effect is statistically different from zero, t= 3.6965, p = .0003, with a 95% confidence interval from .1505 to

.5015. The total effect of EaseUse on purchase intent is c = .5788 which is also statistically different from zero, t = 5.2952, p = .0000 or between .3836 to .7940 with 95% confidence. Both the direct effect H2a and the indirect effect H2b are supported.

Consequent

Antecedent

Customer Satisfaction (M) Purchase intent (Y) Coeff. SE P Coeff. SE P Ease of Use (X) a1 .3857 .1015 .0002 c1 ’ .3260 .0891 .0003 Customer Satisfaction (M) - - - b1 .6554 .0516 .0000 Constant i1 1.920 .5069 .0002 i2 .7311 .4452 .1017 R2 = .2249 R2 = .5090 F (7.275) = 11.3961, P<.001 F(8.274) = 35.5071, P<.001

Effect SE P LLCI ULCI

Direct effect c1 .3260 .0891 .0030 .1505 .5015

Total effect c1 ’ .5788 .1093 .0000 .3636 .7940

Boot SE Boot LLCI Boot ULCI

Indirect effect a1b1 .2528 .0767 .1094 .4142

Richness of information

In the third analysis, M and Y stayed the same but the independent variable (X) Richness of information (RichIn). The effect of RichIn on CatSF is a1 = .1727. This effect is statistically significantly different from zero, t= 3.8017, p = .0014, with a 95% confidence interval from 0.0671 to 0.2782. The effect b1 = .6554, and this effect is statistically different from zero, t= 12.6965, p = .000, with a 95% confidence interval

from .5538 to .75. Thus, H3b is supported. The indirect effect is 0.1132 and this indirect effect is statistically different from zero, as revealed by a 95% BC bootstrap confidence interval that is entirely above zero (.0365 to .1828). The direct effect of RichIn, c′ = .0920, this direct effect is statistically not different from zero, t= 1.9676, p = .0501, with a 95% confidence interval from 0.0000 to 0.1841, therefore, H3a is rejected. The total effect of RichIn on purchase intent is c = .2052 which is also statistically different from zero, t = 3.5531, p = .0004 or between 0.0915 to 0.3189 with 95% confidence.

Consequent

Antecedent

Customer Satisfaction (M) Purchase intent (Y) Coeff. SE P Coeff. SE P Richness of information (X) a1 .1727 .0536 .0014 c1 ’ .0920 .0468 .0501 Customer Satisfaction (M) - - - b1 .6554 .0516 .0000 Constant i1 1.920 .5069 .0002 i2 .7311 .4452 .1017 R2 = .2249 R2 = .5090 F(7.275) = 11.3961, p <.001 F(8.274) = 35.5071, P<.001

Effect SE P LLCI ULCI

Direct effect c1 ’ .0920 .0468 .0501 .0915 .3189

Total effect c1 .2052 .0577 .0004 .0000 .1841

Boot SE Boot LLCI Boot ULCI

Indirect effect a1b1 .1132 .0365 .0365 .1828

Interactivity

Also in the fourth analysis, M and Y stayed the same but the independent variable (X) Interactivity (InA). The effect of InA on CatSF is a1 = -0.0399 This effect is not statistically significantly different from

zero, t= -0.4549, p = .6495, with a 95% confidence interval from -0.2126 to 0.1328. The effect b1 = .6554, and this effect is statistically different from zero, t= 12.6965, p = .000, with a 95% confidence interval from .5538 to 7570. The indirect effect is -0.0262 and this indirect effect is not statistically different from zero, as revealed by a 95% BC bootstrap confidence interval that is not entirely above zero (-0.1630 to 0.0939). The direct effect of RichIn, c′ = .0039, this direct effect is statistically not different from zero, t= -0.2360, p = .9590, with a 95% confidence interval from -0.1441 to 0.1518. The

total effect of InA on purchase intent is c = -0.0223 which is not statistically different from zero, t = -0.2360, p = .8136 or between -0.2082 to 0.1637 with 95% confidence. Both H4a and H4b were

rejected.

Consequent

Antecedent

Customer Satisfaction (M) Purchase intent (Y) Coeff. SE P Coeff. SE P Interactivity (X) a1 -.0395 .0877 .6495 c1 ’ .0039 .0751 .9590 Customer Satisfaction (M) - - - b1 .6554 .0516 .0000 Constant i1 1.9270 .5069 .0002 i2 .7311 .4452 .1017 R2 = .2249 R2 = .5090 F(7.275) = 11.3961, p <.001 F(8.274) = 35.5071, P<.001

Effect SE P LLCI ULCI

Direct effect c1 ’ .0039 .0751 9590 -0.1441 0.1518

Total effect c1 -.0223 .0945 .8136 -0.2083 0.1637

Boot SE Boot LLCI Boot ULCI

4.6 Split data

The control item ‘’Frequency of shopping’’ showed that about half of the respondents have never shopped via a mobile application (n = 117) and the other half of the respondents have (n = 167). Considering that there is a clear divide between the respondents, additional analyses were conducted to see if there are difference between the ‘’has never shopped’’ group and the ‘’has shopped’’ group. First, a dummy variable was created. The first value is 1= has never shopped and 0 (else)=has shopped. Second, the file was split and with the dummy variable the correlation was tested. All the variables are correlated in the ‘’has shopped’’ group. The variables in the ‘’has never shopped’’ group were not all correlated. For example, ease of use is not significantly correlated with purchase intent. This contrasts with the ‘’has shopped’’ group and the whole sample, in which this factor did have a correlation with purchase intent. The regression analysis showed that ease of use has a high beta which is also significant .260** as well as fashion innovativeness .410*** for the people who have shopped through a mobile application. For the people who ‘’have never shopped’’, the two variables were not significant (.092 and .051). For all the other variables, the Beta value and the significance level was about the same as for the whole data set. Additionally, the PROCESS analyses showed no differences between the two groups. Meaning that the only differences between the two groups is that the ‘’has shopped’’ group scores higher on fashion innovativeness and that if the ease of use of a mobile application increases by one, their purchase intent will increase by .260.

Hypothesis and results

Hypotheses Results

H1a: Ease of access has a positive effect on purchase intent when buying clothes through a mobile application.

Supported

H1b: The positive relationship between ease of access and purchase intent is mediated by customer satisfaction when buying clothes through a mobile application.

Rejected

H2a: Ease of use has a positive effect on purchase intent when buying clothes through a mobile application.

Supported

H2b: The positive relationship between ease of use and purchase intent is mediated by customer satisfaction when buying clothes through a mobile application.

Supported

H3a: Richness of information access has a positive effect on purchase intent when buying clothes through a mobile application.

Rejected

H3b: The positive relationship between richness of information and purchase intent is mediated by customer satisfaction when buying clothes through a mobile application.

H4a: Interactivity has a positive effect on purchase intent when buying clothes through a mobile application.

Rejected

H4b: The positive relationship between interactivity and purchase intent is mediated by customer satisfaction when buying clothes through a mobile application.

Rejected

H5: Fashion innovativeness has a positive effect on purchase intent when buying clothes through a mobile application.

5. Discussion

This study researches whether shopping for clothing through a mobile shopping channel has any effect on customer satisfaction and purchase intent. Five factor were chosen from literature that might have an effect on this phenomenon. These factors are ease of access, ease of use, richness of information, interactivity and fashion innovativeness.

Ease of access has a positive direct relationship with purchase intent when people use mobile shopping to purchase clothing so H1a is supported. This is supported by several studies that also found that mobile convenience can lead to purchase intention and loyalty (Okazaki & Mendez, 2013; K. Yang, 2010b; K. Yang & Kim, 2012a). With ease of access people always have the opportunity to shop for fashion items. Retailers can create an application that sends out notifications similar to other applications such as WhatsApp and Facebook. This way retailers can notify people when a new collection is available or when a sale has started. To remind people that they can shop anytime they want. Ease of access and purchase intent is however not mediated by customer satisfaction and thus H1b is rejected. This means that the effect of ease of access on customer satisfaction was insignificant. This is probably due to the properties of the research sample. The respondents of this study were young people (e.g. students); in this era of information and technology, they would be frequently exposed to computer and mobile related matters. Therefore, they would consider the ease of access of mobile shopping not to have a significant influence over their customer satisfaction. After splitting the data between young and old responders, the result show that indeed customer satisfaction is not significant for the young responders.

One of the main reasons for technology adaptation is ease of use (Davis, 1989) and this was also supported for mobile services (Nysveen et al., 2005). If, for example, the ease of use through mobile applications does not outweigh customer loss, then customers may simply revert back to traditional channels (Z. Yang & Peterson, 2004). This suggests that the ease of use plays a pivotal role in customer satisfaction with mobile shopping. This is supported by the results, ease of use has a positive direct link to purchase intent, thus H2a is supported. H2b, ease of use and purchase intent is

mediated by customer satisfaction, is also supported. This finding is equally supported by Ko et al (2009), which suggest that ease of use appears to have a direct impact on the perceived value of the technology which in turn motivates the actual use. This indicates that people who use m-shopping apps pay attention to the effort they have contributed in order to use the device. It also suggests that companies that offer m-shopping should create an app that is user-friendly so that users can contribute less effort but still frequently use the technology. This is especially the case with people who have purchased clothing before on a mobile application.

The regression analysis showed that there is a direct relationship between richness of information and purchase intent. However the PROCESS analysis showed that there is only a mediation effect and the direct relationship disappeared completely and thus H3a is rejected. Richness of information leads to a small increase in consumer satisfaction when shopping using a mobile device. Customer satisfaction has a significant effect on purchase intent. According to Zhao, Lynch and Chen (2010) this would mean that there was an indirect-only mediation for this relationship, also known as a full mediation (Baron & Kenny, 1986). Thus, the results of this research show significant support for full mediation by customer satisfaction on the effect of richness of information on purchase intent, meaning H3b is supported. This is also consistent with previous research which have shown that information is an important element of customer satisfaction (Montoya-Weiss et al., 2003) and purchase intention (Shim et al., 2001). This is because information allows the consumer to select the merchandise that best satisfies their needs (Ranganathan & Ganapathy, 2002). Previous studies have shown that richer information (more extensive) available online leads to better buying decisions and higher levels of e-satisfaction (Peterson et al., 1997b). The items measuring this variable asked for both general and personal information. When looking at the data for both types of information, the means were high but the personalized information mean was higher. Therefore, it is important for retailer to offer personalized information to the customer to streamline the complex task of assessing products. Retailers can achieve this by providing ‘’virtual dressing rooms” and ‘’shopping galleries ‘’on the app. Shoppers can swipe through a shopping gallery of all the fashion items and can have ‘style feeds’

created base on their preferences.

In the case of interactivity, both hypothesis H4a and H4b are rejected. Interactivity does not have a positive direct relationship with purchase intent. This is also not mediated by customer satisfaction. This is not consistent with Zhang et al (2014) who suggest that interactivity in the mobile environment can lead to higher purchase intention. One possible explanation could be that just 3 items were used to measure the variable and thus the information is insufficient to prove a relevant association. Another explanation for this result could be that because consumers shop with different shopping motivations this could result in dramatically different shopping patterns (K. Yang & Kim, 2012b). One can assume that people shop mostly through a mobile application for convenience so when in a hurry or on the go and for a bigger shopping spree they choose for a different shopping channel. One of the items used to measure interactivity was ‘how important the responders find purchase recommendations’. Research has shown that a consumer’s attitude change induced by recommendations depends heavily on the individual’s trust perception of the information source (Senecal & Nantel, 2004). Retailers usually recommend fashion items that other customers have bought through the mobile application, therefore customers have to trust other customers. This is because the credibility of informers plays a significant role in the receiver’s evaluation of the message (Shin, 2013). In addition, prior research suggests that consumers discredit recommendations from endorsers if they suspect that the latter have incentives to recommend a product (Folkes, 1988). Another explanation for this result could be that given the cultural influences, the generalizability of the findings is somewhat limited. The second item that was used to measure interactivity was ‘how important they find two-way communication’. A recent study showed that culture plays a part in moderating the relationship between social interactions and purchase intention as well as affecting consumer trust (Ng, 2013). Since the purpose of this study was not to perform a longitudinal study, the results must be interpreted with caution. It is plausible that a customer’s trust in a specific recommender system would increase over time if that person was satisfied with products previously recommended by that system.