Segmenting m-shoppers in a cross-cultural context : a latent class segmentation based on behavioral and attitudinal factors

(1)

Master Thesis

‘

Segmenting M-Shoppers in a Cross-Cultural Context: A Latent Class

Segmentation Based on Behavioral and Attitudinal Factors

’

University of Amsterdam

Faculty of Economics and Business

MSc. in Business Administration – Marketing Track

Supervisor: dr. Umut Konus

Student: Cinthia Barajas Pradales

Student Number: 11136901

24

th

of June 2016 | Final version

(2)

University of Amsterdam 2

Statement of Originality

This document is written by Student Cinthia Barajas Pradales 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)

University of Amsterdam 3

Abstract

Although much has been researched about online shoppers and multi-channel shoppers, mobile shoppers have rarely been studied in isolation. Thus, questions such as: (1) whether there are different types of m-shoppers, and if so, (2)how different are the segments, and (3) what defines the differences, have not yet been answered. This research investigates whether there are behavioral and attitudinal differences among M-Shoppers by focusing on a number of psychographic, demographic, device related and cultural variables.

Why do we need a multi-criteria segmentation based on behavioral and attitudinal characteristics of mobile shoppers? While the amount of time spent on mobile devices grows eleven times faster than on desktops (Statista, 2016), around 60% of consumers feel their expectations for a satisfactory mobile experience are not being met. Some retailers may be failing to realize of the impact of mobile transactions in their bottom line, while other may not be leveraging today’s possibilities to create personalized interactions. Delivering the same mobile experience to all mobile shoppers will jeopardize the opportunity to create brand bonding, realize sales or develop brand loyalty. Understanding the different motivations and attitudes that mobile shoppers have, will provide a starting point to create different marketing mix propositions, that enable delivering more effective and tailored experiences.

This research shows that there are different types of mobile-shoppers that can be segmented into groups. The clusters identified are: (1) ‘Mobile Hip Trend Followers’, (2) ‘Experientials’, (3) ‘Traditional Mobile Adopters’, (4) ‘True Mobile Lovers’ and (5) ‘Self-image Lovers’. These segments are differentiated by expected benefits, actual usage, appropriateness and/or frequency of engaging in mobile commerce across three different categories.

The key-words for this research are: M-Shoppers, M-Shopping, cross-cultural, behavioral and attitudinal segmentation.

(4)

University of Amsterdam 4

Acknowledgement

The completion of this Master Thesis is the final step to achieving the end of the Master in Science in Business Administration – Marketing Track that I am pursuing in the University of Amsterdam. This piece of work about M-shoppers has been a really enjoyable challenge that I was ready for since I knew which topic I was assigned. M-Commerce is an increasingly important topic in the business world, however it has also proved to be a very interesting research area that gave me the curiosity and motivation to complete this journey. This experience, accompanied by many challenges and good learnings , has helped me to grow personally and professionally. This would not have been possible without the support from the people around me every day, I am thankful to each of them for giving me the courage and strength to always continue.

A special thank you to dr. Umut Konus, who has helped me throughout the important process of writing my thesis. The knowledge, experience and enormous interest from dr. Umut Konus in Mobile Shopping and segmentation has been very a real source of motivation during the research and completion of the thesis.

I hope you enjoy reading my work about Mobile Shoppers.

Kind regards,

Cinthia Barajas Pradales University of Amsterdam

(5)

University of Amsterdam 5

University of Amsterdam 6

5.3.2. Main study 37

5.4. Latent-Class Cluster Analysis 37

5.4.1. Segmentation Basis: Indicator variables 39

5.4.2. Descriptive variables 40

6. Results and Analysis 41

6.1. Preliminary and Exploratory Analysis 41

6.1.1. Reliability 42

6.2. Segmentation Results 45

6.2.1. Mobile-Shopper Segments based on use and benefits 45

6.2.1.1. Analysis of the Segmentation indicators 46

6.2.1.2. Analysis of the Active Covariates 51

6.2.2. Mobile-Shopper Category Specific estimation results 55

6.2.2.1. Cross- category analysis of the segmentation indicators 57

6.2.2.2. Cross- category analysis of the active covariates 58

7. Discussion 60

7.1. Overall Conclusion 60

7.2. Theoretical Implications 63

7.3. Managerial Implications 64

7.4. Limitations and Future Research 66

6. References 67

List of Tables

Table 1. Factors influencing mobile shopping experience 14

Table 2. Review of Empirical Literature about Mobile Shoppers 23

Table 3. Segmentation indicators 27

Table 4. Benefits related to psychographics of M-shopping 30

Table 5. Sample Characteristics 35

Table 6. Scales Reliability and PCA 44

Table 7. Model Selection Log-likelihood statistics 45

Table 8. Profiling of the Segmentation Indicators 47

Table 9. Description of the profiles (segmentation indicators) 50

(7)

University of Amsterdam 7

Table 11. Description of the Segments 54

Table 12. Model Selection Log-likelihood statistics per category 56

Table 13. Distribution per Segment 58

Table 14. P-values of the Covariates across categories 59

Table 15. Comparative Segment Distribution per category 62

List of Figures

Figure 1. Mobile Share of Retail E-commerce Transactions 8

Figure 2.Global number of mobile retail commerce buyers from 2012 to 2018 13

Figure 3. Mobile growth is driven by smartphones globally 15

Figure 4. Behavioral device related variables 17

Figure 5.Average order value per device 19

(8)

University of Amsterdam 8

Figure 1: Mobile Share of Retail E-commerce Transactions

Source: Criterio , Q3 2015

1. Introduction

Mobile Commerce (M-commerce), often called the next generation of Electronic Commerce (e-commerce), has seen a steep growth worldwide in the number of transactions and users in the last years. Mobile share of transactions reached a global average of 35% in the second half of 2015 (see figure 1), with countries such as Japan

and U.K. reaching almost 50%

(Criteo, 2015) of total

E-commerce operations. Likewise

the number of m-shoppers

worldwide is expected to surpass one billion in 2018 (Monetate, 2015). These figures reflect the

importance and global

magnitude of mobile commerce for modern and traditional retailers.

In the last two decades electronic commerce has developed and expanded, creating new interaction channels as a consequence of wireless technology and social media. Nowadays smartphones are used at work and privately for listening to music, connecting to social networks, playing games or running business transactions. Mobile devices such as tablets or mobile phones have become part of people’s daily life, leading to an increasing ‘mobile lifestyle’ (Shankar et al, 2010).

According to statistics about 60% of the users engaging in mobile shopping think that their mobile experience was not as satisfactory as they expected it to be. At the same time mobile usage is increasing at a fast pace (Mobile Commerce Daily, 2016). Mobile commerce unlocks potential for retailers and consumers in different ways than conventional desktop computers or laptops. Some examples of the features of mobile devices are (1) convenience and connectivity: browse, search and buy anytime and anywhere; (2) the use of mobile apps: applications can contribute to brand engagement and loyalty but also to an easier mobile experience for consumers; (3) push add notifications: communicating

(9)

University of Amsterdam 9

promotions in nearby stores; (4) comparison: browse online and compare offline by visiting a physical store and comparing the prices or the product variety with other (e)retailers by searching on your mobile device.

According to statistics, more than 50% of all consumers in the U.K. check their phone between 25 and 50 times a day (Deloitte, 2015). This provides marketers the possibility to get visibility by pushing content about products and brands, offering promotions, increasing customer engagement, impacting loyalty, etc. However a clearer and richer image of consumers will help to exploit the potential of mobile marketing. An example is that, despite of the growth of global web visits using smartphones (+10 p.p. in Q4 2015 vs. Q4 2014), mobile conversion remains low with only 1,3% compared to 4,4% of traditional desktops and laptops in 2015 (Monetate, 2015).

Despite of the important managerial implications of the growth of M-commerce, current knowledge on the field is unable to answer: ‘whether there are there different types of mobile shoppers

(m-shoppers)? And if so, how different are they and why?’. Giving answers to these questions is

important in order to unlock the business potential from M-Commerce for retailers, for instance, defining whether there are relevant attitudinal and behavioral differences amongst mobile shoppers would largely help to define key elements of a solid mobile strategy.

Why is segmenting M-shoppers important? In the first place, dividing M-shoppers into clusters

will help to avoid the traditional ‘mass approach’ to attempt to exploit economies of scale by appealing the entire market. As already used in other channels, a proper segmentation in the mobile channel will help to match particular needs of a group of consumers more closely. Secondly, identifying the different groups of mobile shoppers contribute to efficiently leverage modern marketing tools such as location-based marketing, or mobile targeted notifications. Third, understanding the role behaviors and attitudes in segmenting mobile shoppers will help to understand how to create an optimal mobile experience meeting the users’ standards. For retailers, knowledge about the behavioral and attitudinal differences among mobile-shoppers could provide marketers with insights of how to optimize their digital advertising budgets, helping them to target consumers more effectively and establishing a point of differentiation versus competitors.

(10)

University of Amsterdam 10

The focus on actual M-shoppers in isolation will be the key difference of this paper with respect to prior segmentation research done about Mobile Commerce. This research will also focus on identifying differences which define M-shopper segments across countries. The increase of electronic cross-border operations boosts the potential for retailers to expand their sales and approach m-shoppers in different countries. Europe’s mobile commerce is increasing at a rapid pace. The Netherlands and Spain, present differences in internet usage or mobile device preferred, yet both are leading E-commerce countries in their respective regions. This research will collect data in both countries to in order to (1) investigate potential m-shopper differences between both countries, and (2) to provide insights that enable generalization.

Given the increasing relevance of M-commerce in the online landscape and the yet ‘unknowns’ about M-shoppers, investigating behavioral and attitudinal differences is highly important. The main research objectives are:

1. To identify M-shopper differences based on the use and expected benefits across different categories during the pre-purchase and purchase phases.

2. Explain the M-shopper typologies based on the underlying psychographic, demographical, cross-cultural and behavioral product-related covariates.

The results of the segmentation have important managerial implications for marketers in the way they develop their mobile strategies. The basic principle behind the concept of segmentation is to divide the market into identifiable groups so that marketers can create an appropriate marketing mix propositions. According to Neslin et al (2008), if there are different segments that can be identified based on their channel use, firms may need different approaches to reach these various segments across the different channels. If the same principle applies within mobile shopping, different M-shopper segments will possibly require different strategies, having a direct impact on how the companies develop and implement their mobile strategies and the complexity of those. If various strategies would be needed, marketers need to understand how different these segments are, what are the main characteristics defining the segments, and how relevant these differences are. The strategic decisions taken by companies will in turn affect the customer value and the turnover potential (Bhatnagar and Ghose, 2004).

(11)

University of Amsterdam 11

However, if no differences are observed within M-shoppers, distinct strategies will not be required. Given the complexity of this growing channel of commerce, a better understanding of the motivations and attitudes towards M-commerce engagement will provide the essence for marketers to elaborate on their strategy and understand: how to differentiate in a competitive market; how to positively impact customer satisfaction; how to drive sales and; how to bond consumers and develop brand loyalty. This research will also enrich the existing academic literature about shoppers by defining differences between exiting M-shoppers and the underlying covariates defining the different profiles.

In order to attend the research objectives, the research will be structured in five sections. Firstly the existing literature about M-shopping and segmentation will be reviewed. Secondly, the conceptual framework will explain the research model and its components (segmentation indicators, shopping phases and covariates). In the third place, the research method and sample will be explained and this will lead to the analysis of the results. To conclude, the last section will show the main conclusions of the research with a detailed explanation of the managerial and academic implications and the main research limitations.

(12)

University of Amsterdam 12

2. Literature Review

This section reviews the existing literature about mobile commerce, mobile shoppers and segmentation research. The purpose of this review is to define the current mobile commerce landscape and existing segmentations within mobile commerce to end up with the research gap and the setting for this study.

2.1. Mobile Commerce

Practitioners and academics have given multiple definitions to Mobile commerce (m-commerce). Mobile commerce has been described as any transaction of goods or services initiated or completed using mobile access to a computer network via an electronic device (Han, Nguyen and Nguyen, 2015). Earlier it has been described similarly to electronic commerce, with the exception that transactions are done in a wireless environment (Gunasekaran, 2007) using mobile devices (Varshney et al, 2007). In other words, M-commerce expands the functionalities of m-commerce by adding location and localization services (Junglas & Watson, 2008). Tiwari and Buse (2007) defined it as a form of m-business that involves monetary transactions (i.e. purchase) and non-monetary transactions (i.e. search or after-sale). Compared to e-commerce, m-commerce users are no longer bounded by geographical constraints (Alain Yee- Loong Chong, 2012). When we talk about mobile devices we refer to phones, smartphones, tablets and e-readers. Nowadays M-commerce is regarded as one of the epicenters on the ongoing digitalization phenomena (Pousttchi et al, 2015).

Mobile commerce has expanded in the last years influenced by a number of factors (Chang et al. 2014). The first driver has been the increasing popularity of smart mobile devices (Chang et al. 2014) over desktops and tablets due to the availability of larger screens (Criteo, 2015). The second driver is consumer demand for applications to trade goods (buy and sell) and the intrusion of mobile banking. Another important factor is the wide acceptance due to stronger security practices (Criteo, 2015).

Figures show that web visits using mobile devices increase rapidly, as also do the number users purchasing online (see figure 2). In 2013, the number of buyers reached 379 million and expectations towards 2018 reach more than 1 billion users purchasing retail using a mobile device (Monetate, 2015).

(13)

University of Amsterdam 13

Figure 2: Global number of mobile retail commerce buyers from 2012 to

2018 millions)

Source: Statista, 2016

Dealing with

cross-device behavior represents one of the main challenges and opportunities in the retail environment (Criterio, 2015). The way in which consumers operate with mobile devices highlights the importance to

understanding the role of

mobile commerce behavior in a

cross-device context. The

increasing popularity of mobile commerce is also linked to how retailers should approach and adjust their multichannel customer management strategy. Multichannel customer management (MCM) is defined as ‘the design, deployment, and evaluation of channels to enhance customer value though effective customer acquisition, retention and development’ (Neslin, Grewal et al. 2006). In this context Neslin et al, (2006) define channels as customer contact points or mediums in which firm and customers can interact.

Multichannel has been high on the agendas of retailers due to the increase of multichannel transactions (Neslin and Shankar, 2009) and the economic value of the customers adopting this behavior for retailers. There is empirical evidence that multichannel customers buying in different categories are more valuable than single-channel customers, suggesting, for instance, that mass merchandisers should attract these customers by investing in all channels (Kushwaha and Shankar, 2013).

In this context, mobile transactions have an increasingly important role within the multichannel turnover mix for retailers, with more than 35% of total e-commerce transactions globally. Retailers are gradually realizing of the impact of mobile in their sales, with around 80% of retailers worldwide realizing the growth of mobile commerce (eMarketer, 2016). Hence Mobile Marketing is becoming increasingly important, as retailers can enter the consumers environment through mobile devices, giving the opportunity for retailers to be anywhere and anytime (Shankar et al, 2005).

(14)

University of Amsterdam 14

According to a study from early 2015, mobile generates over 20% of sales in categories like fashion, travel, health and beauty (Criteo, 2015). According to a consumer research ran by Accenture (2015), there are still more than 50% of consumers who think that buying via a mobile device is not easy, showing that the seamless retail mobile experience is yet to come and that important economic value can be captured from this channel.

Mobile shoppers’ experience is influenced by a number of factors, a list of these factors is summarized in table 1. These are all factors that could potentially contribute to creating a smoother and more relevant experience for those consumers whose requirements are not yet satisfied . Factors such as actual use and attitudes (expected benefits) have been widely researched within the multichannel and online environment (Konus et at, 2008; Rhom & Swaminathan, 2004; Schoder & Zaharia, 2008). The effect of demographic variables, such as gender, have been earlier researched within the mobile context (Okazaki and Mendez, 2012). Other variables such as the type of mobile device, operating system, the use of applications or cultural traits have not been within the framework of mobile academic research despite of the differences among m-shoppers found in statistics from practitioners.

Table 1. Factors influencing users’ Mobile experience

Variable Explanation

Actual Use This variable reflects whether the user makes actual use of mobile commerce. Attitudes This variable reflects users expected benefits from engaging in mobile commerce.

Categories Consumers may have different attitudes towards mobile commerce depending on the product or service category.

Mobile Device

Consumers can make use of different mobile devices (e.g. Smartphones and tablets) for different purposes.

Operating System

Operating systems, which are led mostly Android and iOs, may influence the user experience during the mobile journey (Chong, 2013).

Applications vs. Web

The use of applications or web (open source) may influence users mobile experience and brand/product positioning in consumers’ minds (eMarketer, 2016).

Cultural traits Behavior and expectations from mobile users differ across countries (Criteo, 2015; Chong, 2013).

Demographic Variables such as education or age have shown to affect the perception and expectations of mobile commerce (Okazaki and Mendez, 2012; Chong, 2013).

(15)

University of Amsterdam 15

Figure 3: Mobile growth is driven by smartphones globally

Source: Criteo, 2015

2.1.1. Mobile Shopping Devices and Platforms

Mobile devices have two main applications, audio (i.e. voice, conversations and music) and visual (i.e. text, data, picture and video) (Shankar et al, 2010). Technological innovation has marketed a wide variety of mobile

devices and platforms that consumers use on their daily life. Next to

the brand and

characteristics of the device, the devices will

differ in type and

operating system.

The most popular types of mobile devices are Smartphones and Tablets (Google, 2016). It’s dominance in the total amount of mobile transactions varies per country, with cases such as Japan, where 90% of mobile transactions are done from a mobile device, while in countries like U.K. or Germany Tablets take around 55% of total transactions (Criteo, 2015). Within devices in the last year Smartphones drove mobile growth globally. In figure 3, we can observe that Asian countries lead the way in this trend. Figures from 2015 show that while Smartphones are more frequently used in the morning (+28%), Tablets take over in the evening (up to +40%), showing that the popularity of the type of mobile device use will depend on the moment of the day (Criteo, Q1 2015)

The Operating systems that dominate the mobile landscape are Android (developed by Google) and iOS (developed by Apple), followed by Windows Phone (Okediran, 2014). Recent statistics stand out the increasing growth of Android within Smarphone devices in the segments of retail and travel, with an average of 7,1% in the total share of eCommerce transactions (Criteo, 2015). Yet iOs devices, or in other words, iPhone, remains on the lead within the two with an average share of 9,3%.

(16)

University of Amsterdam 16

Users can search or buy using Web Browsers (open source) or Mobile Applications. Statistics have shown that mobile users prefer apps for most of their activities, with a conversion rate 120% higher for Mobile Apps (Criteo, 2016). In particular, users view almost four times more products than Mobile Web and shoppers are almost two times more likely to add products to their basket (Criteo, 2016). Mobile users seem to download their favorite apps and stuck with them. Users also seem to place importance on loyalty club activities and promotional offers that apps can offer (eMarketer, 2016).

2.1.2. Mobile Shopping Attitudes

Motivations from online shoppers have been widely researched establishing different typologies and methodologies to segment. Past research defends that online shoppers’ main motivation is convenience, and that they are willing to pay a higher price than conventional offline shoppers (Szymanski & Hise, 2000) or even dislike conventional shopping (Morganosky & Cude, 2000). Among others Ganesh et al. (2010), found evidence of the similarities between online and offline shoppers, yet there are few unique shopper types attracted by the attributes of online shopping. Agrebi and Jallais (2014), investigated the antecedents of satisfaction among users and non-users of mobile shopping, confirming that perceived usefulness, enjoyment, and ease of use are positively correlated to the intention to engaging in m-commerce. Ström et al (2014), investigated the utilitarian, emotional, social and comparative benefits in the mobile environment to conclude that (1) content reliability and quality positively influence utilitarian motivations on loyalty to mobile services; (2) the relative importance of entertainment (as emotional benefit) is category specific; (3) social value has some effect on consumer loyalty in terms of higher willingness to pay or word-of-mouth, especially for entertainment services; (4) efficiency, convenience and safety are the most important benefits determining differences in customer value perceptions across laptop/desktop and mobile devices. Prior research in the Multi-channel context also investigated the role of variables such as convenience, flexibility, shopping enjoyment, price-comparison and variety seeking as expected benefits from multi-channel behavior (Heitz-Spahn, 2013).

Previous research supports that consumers who purchased online have broken the barriers to distance shopping, being more likely to engage in mobile commerce. As confirmed earlier by technology acceptance models (e.g. TAM), consumer innovation has a positive and direct influence m-commerce

(17)

University of Amsterdam 17

Figure 4: Behavioral product related variables

adoption (Eastlick & Lotz, 1999). Aldas-Manzano et al. (2009), concluded that mobile affinity (importance of mobile in people’s life) has a direct positive relation to the probability of engaging in m-commerce, and that compatibility (consistent experience) has a positive influence in engaging. Prior research has investigated mobile commerce adoption using and enriching the traditional technology adoption model (TAM) (Aldas-Manzano et al. 2009 ; Chong, Chong, Ooi, & Lin, 2011), yet the focus of the research remained on the adoption and not on the post-adoption behavior of existing m-shoppers.

According Bhatnagar and Ghose (2002), who focused on researching web shoppers, benefits sought by consumers can provide more diagnostic information than descriptive demographic profiling. Due to the similarities in associated benefits (e.g. convenience), we can hence expect that this will also be the case in the mobile environment. Furthermore Mobile Advertising research has found evidence that mobile display advertising campaigns (MDA), can influence positive attitudes towards products and increase purchase intentions. Yet this favorability is especially applicable when advertising products that were higher involvement (vs. low) and utilitarian (vs. hedonic) (Bart, Stephen and Sarvary, 2014).

2.1.3. Mobile Shopping Behavior: usage occasions, purchase phases and intensity

The reason why behavior plays an important role in segmentation was explained by Wu and Chou (2011), ‘customer classification based

on buying behavior is essential for developing a successful marketing strategy, which in turn creates and maintains competitive advantage’. The rise of M-commerce transactions and the increasing integration of mobile devices in our daily life-style, suggests that a unique and differentiated mobile-commerce strategy could become a source of competitive advantage.

(18)

University of Amsterdam 18

Mobile Shoppers behavior is influenced by the variables included in figure 4. M-shoppers could defer on multiple basis, such as the type of mobile device they use (mainly mobile or tablet), operating system (Android or iOS), the use of an app or web browser (open source such as Safari or Explorer). The following lines explain how m-shoppers could vary across different groups:

 Behavioral research could conclude that Android and iOS users have a different M-shopping behavior, these differences may suggest to managers and marketers the need to have different strategies for the users of these two different operating systems.

 Difference in behavior may also depend on the product category, where consumers may behave different depending on whether the product is a low involvement good or high involvement good. Prior research has shown that mobile commerce most often used for low involvement goods that are frequently purchased (Huang et al, 2015). The nature of the product will also influence on the ticket amount and the frequency or purchase.

 Different purchase phases, namely pre-purchase or purchase, could also lead to different types of M-shoppers. In this line of thought we could identify, for instance, a type of M-shopper who only buys tickets using a mobile device, while s/he does not use mobile banking because s/he doesn’t have enough trust. We could also identify a type of shopper who only uses a mobile device for the pre-purchase phase, or a profile of customers who only engages in m-commerce to buy products with high frequency of purchase.

 Other psychographic motives such as innovativeness, price consciousness (Konus et al, 2008) and time-pressure (Heitz- Spahn, 2013) could determine a basis for M-shopper profiles as they already did for other segmentation studies. For instance, statistics have shown that working mothers make use of mobile commerce due to time pressure (eMarketer, 2016), finding convenience one benefits of mobile commerce.

 Adding into figure 4, behavioral differences may also be established as a function of the cultural context. We could identify cultural traits determining specific profiles, since there is evidence that cultural traits have an influence is how m-commerce is adopted, these traits may also affect the usage and behavioral patterns.

(19)

University of Amsterdam 19

Figure 5: Average order value per device

Source: Monetate, 2015

This research will conduct a behavioral segmentation of M-shoppers by: (1) investigating whether there are different M-shopper profiles and (2) investigating how different these profiles are by describing what characterizes each of this profiles. The insights can offer insights to managers on how to leverage or create their mobile platforms. While M-shoppers may behave homogenously, based on previous research we could expect differences on the aforementioned variables.

2.1.4. Cultural Context (different countries and cultures)

While in the last years e-commerce cross-border operations have accelerated (E-Commerce Foundation, 2015), consumer research reports still show that mobile-commerce is adopted and used differently across countries (Criteo, 2015). Academics have widely focused on researching mobile technology adoption and motivations on a local basis (individual country studies), with limited cross-country comparisons in Asia (see table 2). Liyi Zhang et al (2012) demonstrated that culture has a moderating effect on the adoption of M-commerce. Furthermore statistics show that consumers’ consumption habits and preferences differ across countries. For instance while in countries like Romania and the Netherlands, less than 10% of users buy music using a mobile device while in the U.K. 25% of

consumers buy

this product using mobile commerce

(ING, 2015).

Another example

is the average

expenditure per mobile device in Q4 2015 shown in figure 5 (Monetate, 2015),while Americans spent on average $112, British only spent $79.

Despite of Europe’s overall share in mobile shopping worldwide, no research has investigated yet the differences across the different countries in Europe. Understanding the differences of European m-shoppers, would help mobile retailers tailoring the strategy to pursue in different countries (e.g. research may reveal segments which only exist in a particular country within Gaining further Europe). The knowledge about m-shoppers potential differences in a cross cultural setting, would help managers

(20)

University of Amsterdam 20

with valuable insights that could help leveraging their international mobile platforms. Looking at the diversity of countries, cultures and economies within the European landscape rises the following question:

‘how different are m-shoppers within Europe?’.

As a first step in answering the question, this research seeks to bring insights about the European m-shoppers landscape by comparing Spain and the Netherlands. These two countries, located in different geographic areas (center and south), will enable to generalize potential findings and/or observe differences between m-shoppers in these two countries. Spain and the Netherlands present similar turnover e-commerce volumes, similar mobile traffic shares and mobile retail conversion shares (Criteo, 2015), yet internet access and preferred device for mobile transactions are different. The Netherlands is one of the top 5 e-commerce European mature markets in turnover with € 16 billion forecasted in 2015, and the third e-commerce economy in Western Europe. Spain is the top 1 e-commerce country in South Europe with a forecasted turnover of € 19 billion in 2015 (E-Commerce foundation, 2014). Likewise, mobile traffic share is 29,1% in Spain while in the Netherlands 28,5%. While the Netherlands has the highest amount of clicks for tablets, Spain has the higher amount of clicks for mobile (Twenga Solutions, 2015).Internet access in Spain is 76% while in the Netherlands is 95% (E-Commerce foundation, 2014).

Previous academic research has investigated the behavioral and attitudinal differences between these two countries on different subjects, such as, the honor-related or individualistic values in shaping the experience and expression of emotions proving significant differences (Mosquera, Manstead, & Fischer, 2000). Hofstede (1980) investigated organizational behavior across different countries, Spain and the Netherlands where part of this research.

Understanding the role of behavior and attitudinal motives among m-shoppers in these countries can describe further points of parity or difference. These two countries were selected above other countries because of reasons such as the difference in internet access, different locations, economic developments, different cultural traits (Aldás-Manzano, Ruiz-Mafe, & Sanz-Blas, 2009), but most importantly because of their relevance in m-commerce turnover in their respective regions in Europe.

(21)

University of Amsterdam 21

2.2. Segmentation

The following section offers an overview of the segmentation studies done in the online, multi-channel, and mobile-commerce context . Lastly, this section contains an explanation of the integrated segmentation approach used in this research to study mobile-shoppers.

2.2.1. Segmentation of Online and Multi-channel shoppers

Segmentation is accepted as a strategic marketing tool to define markets and allocate resources (Asaael and Rocoe, 1976). The idea is that the segments can be divided in order to maximize firms turnover potential. In Mobile-commerce, a basis for segmentation would suggest managers which segments are more appealing depending on the business activity and resources. It would also help them assessing and developing more efficient strategic approaches for their m-commerce platforms.

In the online context, Wu and Chou (2012) concluded a method to segment customers across multiple categories by focusing on shopping behavior, satisfaction, internet usage and demographics. Bhatnagar and Ghose (2004), segmented electronic-shoppers (e-shoppers) based on their purchase behavior across several product categories, finding that-shoppers are more concerned about web attributes associated to perceived losses rather than perceived gains. Rhom and Swaminathan (2002) developed a segmentation criteria for online shoppers based on the motivations for shopping online and divided online shoppers into four segments: convenience shoppers, variety seekers, balanced buyers and store-oriented shoppers. Brengman et al (2002), conducted a cross-cultural research in the United States and Belgium to find evidence of the split between two online shopping and non-shopping segments based on six dimensions, which included internet convenience, districts or window-shopping. Bhanagar and Ghose (2004), ran a latent class segmentation based on web shoppers purchase behavior and discovered that benefits sought can provide more diagnostic information than descriptive demographic profiling.

In the multichannel context, Konus et al (2008) were able to segment consumers based on the information search and purchase behavior in the multichannel environment. The research concluded three segments, which are multichannel enthusiast, uninvolved shoppers and store-focused consumers.

(22)

University of Amsterdam 22

This research was later extended by Konus, Schepers and De Keyser (2015), by incorporating after-sales and previously ignored covariates (such as product complexity) to predict segment membership. Schoder and Zaharia (2008), identified five shopping motives in the multichannel context (e.g. recreational, convenience and independence orientation) to describe single channel and multichannel shoppers in the retail environment. Balasubramian et al (2005), investigated consumers choice of channels based on the influence factors such as economic goals, self-affirmation or social interaction in three stages of the purchase process (consideration, choosing a product and final purchase). Previous multichannel, online and mobile literature has researched multiple variables to discover segments on their own domains, such as: multichannel use, purchase phase, demographical covariates, perceived benefits, categories, etc. Behavior research has been an important element on the aforementioned segmentation papers, hence we expect that behavioral variables will be defining in this paper.

2.2.2. Segmentation of Mobile Shoppers

The following two sub-chapters will give a picture of (1) what has been researched in prior literature and the variables used in the conceptual models, as well as (2) the research direction and approach chosen for this research and the managerial and academic implications of this study.

2.2.2.1. Existing Literature on Segmentation of Mobile Shoppers

M-shopper research has focused primarily in the drivers for adoption, using models such as the TAM or Unified Theory of Acceptance and Use of Technology model (UTAUT) . The research in the field includes, for instance, the role of demographic and motivation variables among m-shoppers in China. This research concludes that demographical and motivational variables (intrinsic and extrinsic) have significant relationships with m-commerce usage activities (Alain Yee- Loong Chong, 2012). The same author has also investigated the role of traditional adoption variables not included in the model (e.g. trust or cost) in China and Malaysia. Further research has been done on the motivational traits affecting m-shopping adoption in Spain and the United States (San-Martín, López-Catalán, & Ramón-Jerónimo, 2013; Yang & Kim, 2012).

(23)

University of Amsterdam 23

Aldas-Manzano et al. (2008) investigated the individual personality variables of innovativeness, affinity and compatibility using the Technology Adoption Model as baseline to investigate the factors of acceptance of M-commerce. This research corroborated that perceived usefulness does not increase users’ intention to shop in mobile devices, despite of its direct effect on attitude ( contradicting the traditional TAM model), and that personality variables have a positive effect on m-commerce adoption. Further research has investigated the moderating effect of gender in mobile commerce and the importance of convenience (Okazaki & Mendez, 2013)

Table 2 offers an overview of the empirical research done about m-shoppers and the variables used in each study. While these studies have contributed to understanding the drivers for adoption or demographical discrepancies, compared to earlier studies, this research will shed a new light on M-shopper research by, firstly, focusing exclusively on consumers who have already purchased using a mobile device, in other words, existing M-shoppers. Adoption models have been widely researched within the topic and will not be applied. Secondly the research will combine a selection of covariates, which were not combined in earlier studies, namely demographic, psychographic and behavior-product related . Lastly, this research will be conducted in two countries, cross-cultural differences becomes an additional covariate that has barely been used before.

Table 2. Review of Mobile Shoppers empirical literature

Author Country Adoption Variables (including models) Multiple shopping phases Multiple categories Covariates Demographic Psychographic Behavior product related Cross-cultural study

R. Jen-Hui Wang et al.,2015 United States



San-Martín,Blanca López-Catalán &Ramón-Jerónimo, 2013

Spain



Yang and Kim , 2012 United States



Aldas- Manzano, Ruiz-Mafé

and Sanz-Blas, 2008 Spain



Alain Yee-Loong Chong,

2012 China



Alain Yee-Loong Chong,

2013 China



Alain Yee-Loong Chong et al. 2011

China &

Malaysia



Okazaki and Mendez, 2012 Spain



This research Spain & The Netherlands





 

(24)

University of Amsterdam 24

2.2.2.2. M-Shoppers Segmentation: An Integrated Approach

This research paper will enrich existing commerce literature by focusing (exclusively) on M-shoppers in order to investigate different profiles based on consumer their attitudes, behavior and a number of active covariates. The integration of all these variables, whilst very relevant within the mobile landscape, have not been executed earlier, becoming one of the main points of difference with respect to earlier studies. The research framework will be based on a number of segmentation indicators – actual use and expected benefits- and a number of active covariates – psychographics, demographics, cultural and device related – during the phases of search and purchase. Additionally the research will focus on three categories which are 1) entertainment and travel tickets, 2) apparel and 3) consumer electronics, as these have proven to be sales volume drivers across many countries worldwide. As displayed in the conceptual framework in the next section, the novelty of this research lies on the combination of variables included in the behavioral segmentation. The second new element of this research is the cross-cultural component as one of the covariates to define the profiles.

The multicriteria and multidimensional nature of the segmentation has scientific and practical relevance. For practitioners, the findings of this research bring important managerial implications as it describes how different m-shoppers are and what defines their profiles. The insights can be used by retailers as foundation for their commerce strategy to understand how to approach and connect with m-shoppers with different profiles. Additionally, if the cross-cultural comparison shows significant differences in m-shoppers between countries, the insights will help refining retailers’ mobile-strategy by providing the basis to establish the scope of different business models when approaching m-shoppers in different countries.

Since the academic literature on m-shoppers segmentation is limited and highly linked to technology adoption research, this study will give a new light by incorporating new variables of study (behavioral product related variables) to the already studied psychographic and demographic traits. This study will also enrich the knowledge of m-shoppers in Europe by running the first cross-cultural research.

(25)

University of Amsterdam 25

3. Gap and Research Question

Current Mobile Commerce literature has disregarded the importance of studying m-shopper attitudinal and behavioral differences. This research will apply an integrated approach by considering multiple variables that are critical in consumers mobile experience, and that have not been researched within the same framework in earlier studies.

The research elaborated in this paper aims to respond the following question:

‘Are there distinct m-shopper segments which can be identified based on mobile use and expected benefits? – If so, are this segments characterized by psychological, demographic and device related variables?’

4. Conceptual framework

This section visualizes the conceptual framework used for the research, explains the segmentation indicators and the covariates applied in the framework to determine M-shopper segments. Given the nature of the research, no hypothesis will be included in the framework, yet the expected relationships within the active variables will be explained throughout the thesis.

(26)

University of Amsterdam 26

4.1. Framework

The conceptual framework in figure 6 visualizes the variables of this research. The aim of the study is to identify segments based on actual use and expected benefits across three different product categories (namely tickets, clothes and electronics), during the phases of pre-purchase and purchase.

From the relationships described in the model, we could expect, for instance, a segment that only searches for information about consumer electronics using his/her mobile device, but searches and purchases clothes with his/her mobile device. We could also see a segment that follows the journey from pre-purchase to purchase only for tickets. In order to enrich the description of the profiles, the model incorporates also M-shopper other covariates, which are: psychographics, demographical, cross-cultural and product-related. This variables will contribute to describe the characteristics of the segments.

In order to being able to observing behavioral and attitudinal differences across categories and phases, the research will focus primarily in three categories: travel tickets, apparel (clothes and sportswear) and consumer electronics. These three categories have been selected because of the relevance in the total M-commerce transactions in 2015 (Postnord, 2015; Criteo, Q1 2015).

The approach chosen for the study is a Latent Class Segmentation, which will be based on three types of variables. These variables are segmentation indicators, active covariates and latent variables. The first two will be defined and explained in the sections below. The latent variables will be the output of the research and will be explained in the analysis of results.

It is important to point out that the research will not state formal hypothesis since the study is based be based on ad hoc research there is no information available about final segments. Nonetheless, I will include the rational for possible effects of the covariates on the behavior of M-shoppers.

4.2. Segmentation Indicators

The segmentation will be based on two indicators: actual use and expected benefits. Both indicators will be used to analyze potential behavioral and attitudinal differences across product

(27)

University of Amsterdam 27

categories and shopping phases. These indicators, which will be explained in the next sections, can be found in table 2 with a short definition.

Table 3. Segmentation indicators

Indicator Explanation

Use

This indicator explains consumer’s current and actual use of mobile devices across the pre-purchase, purchase for any given product category.

Expected Benefits (attitudes)

This indicator focuses on measuring the relevance of the expected benefits from engaging in M-Commerce on characterizing profiles.

4.2.1. Actual Use (behavior)

A crucial element for this segmentation study is to target M-shoppers, or in order words, consumers who actively use mobile devices during the phases of pre-purchase or purchase. This means non mobile-shoppers or future prospects (i.e. future M-commerce adopters) will not be considered as part of the segmentation. Use is this context stands for M-shoppers behavior who engage in any purchase offer for any product category.

Thomas and Sullivan (2005) already used actual use as a segmentation indicator in the multichannel environment. The effectiveness of the indicator in the context of multichannel leads to assume that actual use can be an effective variable in the context of mobile channel.

4.2.2. Expected benefits (attitudes)

Expected benefits explain the positive consequences that consumers get from engaging in a certain behavior. In 1976, Tauber researched consumer motivations for shopping, other than the simple need for products or services. Tauber’s research concluded that there is more than utilitarian motives to

consumer shopping motivations by demonstrating than hedonic motives such as diversion, role playing or interaction, drive consumer shopping intentions (Tauber, 1976). Likewise, Childers et al (2001), researched the hedonic and utilitarian dimensions of motivations to engage in online shopping. They discovered the differential importance of the hedonic aspects of media and the traditional utilitarian motivations.

(28)

University of Amsterdam 28

The expected benefits included in this framework are: savings, convenience, exploration and entertainment. Aliwadi et al (2001), stated that expected benefits affect consumers channel choice (behavior). Table 4 offers an overview of the expected benefits included in the framework and their connection to the psychographic variables. The expected benefits will be explained in relation to the psychographic variables depicted in the model in the next section.

4.3. Active Covariates

The active covariates included in the model will contribute to provide with a clearer explanation of the elements that characterize the M-shopper segments. Konus et al. (2008) argued that psychographic and demographic covariates elicit different benefits and costs that result in different channel preferences. With this statement in mind, we can infer that these covariates could lead to different preferences on how to use m-commerce (e.g. only for purchase or only for certain product categories).

Active covariates can be M-shopper motivations (e.g. benefits), personal traits or product-related traits (such as type of mobile device or operating system) . Given the lack of research of existing profiles of M-shoppers, the significance of one or various covariates over the others is unknown. Research focusing on these covariates will facilitate a basis for segmentation and provide with meaningful strategic insights to managers.

The covariates chosen for the research combine previously researched covariates, and two variables which were not earlier considered in a mobile segmentation study: cross-cultural variables (e.g. Spanish or Dutch) and behavioral product-related variables (e.g. operating system, device and open source vs. app).

Next to the ‘new’ covariates mentioned in the previous paragraph, the conceptual framework includes a selection of demographic and psychographic variables that have been previously used within the multi-channel context ( Konus et al, 2008 ; Ailawadi et al,2001; Sharma & Gutierrez, 2010). The inclusion of these variables will help concluding ‘whether’ and ‘how’ M-shopper covariates and benefit expectations differ in their mobile use and behavior across the different purchase phases (pre-purchase, purchase and post-purchase) in the categories selected.

(29)

University of Amsterdam 29

4.3.1. Psychographic covariates

As discussed by Ailawadi et al (2001) and Konus et al (2008) in the multichannel context, we can infer that psychographic variables relate to how the benefits of engaging in M-commerce will be perceived by consumers. The effect established in prior literature between these variables will serve as basis to explain the psychographic variables depicted in the model, as well as to give rational to possible behaviors that may be observed during the research.

Psychographic variables will characterize M-shopper profiles and refer to attributes of consumers’ personality, values or lifestyles. According to previous literature, psychographic variables show a strong relationship with consumer behavior (Ailawadi, 2001). Thus, the inclusion of psychographics as active covariate, becomes highly important as basis to the behavioral segmentation that this paper pursues.

Individuals’ prior experiences can lead them to perceive shopping as rewarding or stimulating (Arnorld & Reynolds, 2012; Dawson, Bloch, and Rodway, 1990). According to Arnold and Reynolds (2012), the learnings from these experiences lead to the establishment of future goals for hedonic fulfillment when shopping, arising hedonic energizing motives. The variables exploration and entertainment, as outcome benefit of innovation and enjoyment, are two of the hedonic variables that may motivate consumers two engage in mobile commerce. This research will study the effect of this variables in characterizing one or more profiles. The model also includes two utilitarian variables, that are innovativeness time pressure and price consciousness.

In table 3 we can see the benefits and psychographic covariates selected. The benefits added in the model relate one to one to the psychographic covariates selected, since these, influence directly the customer channel choice (Ailawadi et al 2001) and may affect how the mobile channel is used and for which purposes. The positive effect of psychographic variables was earlier realized by Konus et al (2008) and Ailawadi et al (2008). To the variables selected from Konus et al (2008), the model adds ‘convenience’, which I expect to be positively related to time consciousness, as one of the expected benefits from engaging in mobile commerce. In other worlds we could say that the mobile channel may help relieving time-pressure due to the accessibility and portable nature of mobile devices.

(30)

University of Amsterdam 30

Table 4. Benefits related to psychographics of M-Shopping

Psychographic covariate Expected Benefit Economic/hedonic Expected effect

Price consciousness Savings Economic Positive (+)

Time pressure Convenience Economic Positive (+)

Innovativeness Exploration Hedonic Positive (+)

Enjoyment Entertainment Hedonic Positive (+)

Price consciousness & Savings. Price consciousness is an active covariate that stands for the

degree to which the individual seeks for the best available offer (Lichtenstein, Netemyer, and Burton, 1990; Kunus et al, 2008). In this case a customer profile may be characterized by price consciousness, or in other words, the preference to paying low prices in order to save money. Prior research shows that multichannel consumers recognize savings as one of the advantages of engaging in multichannel behavior (Balausubramanian et al, 2005; Konus et al, 2008). In the case of mobile, consumers may engage in M-shopping in order to benefit from discounts from using mobile apps such as Piggy, or Scoupy in The Netherlands, and Dooplan or Tiendeo in Spain. Mobile channel may also be used for what Tauber (1972) described as comparison shopping, during the pre-purchase phase in order to browse offers and compare with what is offered in other channels. This attribute will assess the degree to which price consciousness, and hence saving as expected economic benefit, characterizes the motivation and behavior of one or more profiles of M-shoppers.

Time pressure & Convenience. Time pressure stimulates consumers to form more favorable

attitudes towards online rather than offline (Xu-priour, Cliquet and Fu, 2012). The term stands for consumers’ predisposition to consider time as a scarce resource that needs to be used carefully (Kleijnen, De Ruyter, and Wetzels, 2007). Time pressure is linked to convenience as it is regarded to be one of the main benefits of engaging in mobile behavior (Kleijnen et al, 2007), as mobile devices are portable, and enable immediate accessibility (Sharma & Gutierrez, 2010). Convenience stands for agility, accessibility and availability (Okazaki and Mendez, 2013). According to Keijnen et al (2007), time convenience increases consumers’ perception of the value given to their perception of time scarcity. Likewise, convenience was found to be an important predictor of online shopping attitudes (Childers et al, 2001). In the mobile context, the mobile channel may help freeing up time by conveniently allowing

(31)

University of Amsterdam 31

consumers to purchase using their mobile device and not having to physically go to the store. Additionally, previous research has found that the mobile channel attracts more often customers for low involvement goods with high purchase frequency becoming part of their routines (Wang et al; 2015). I expect time pressure, and convenience as its main benefit, to be relevant in representing one or more M-shopper profiles.

Innovativeness & Exploration. Innovativeness is the preference to engaging in new

experiences with the objective of stimulating the mind (Bearden & Netemeyer, 1999). The main hedonic benefit related to innovativeness is exploration (Konus et al., 2008). Konus, Neslin and Verhoef (2008) found that multichannel customers are enthusiastic about using different channels to explore shopping alternatives. Yang (2012) found that innovative consumers are more likely to engage in mobile shopping, as it may be seen as an interesting environment to explore products, services or mobile features. Therefore I expect innovation, and exploration as its main hedonic benefit, to be relevant in the description of one or more M-shopper profiles.

Enjoyment & Entertainment. Enjoyment is the degree to which a consumer finds shopping

pleasurable. Shopping can give consumers the opportunity to distract from routine representing a form of entertainment (Tauber, 1976). Previous research evidences that shopping enjoyment during search or purchase can influence channel selection (Verhoef et at, 2007; Konus et al, 2008). Childer et al (2001), found evidence to support that entertainment has a strong and consistent influence of enjoyment on attitudes of consumers in online retailing. Chong (2013) found that users are more likely to engage in mobile commerce activities if they find them enjoyable. Hence I expect time enjoyment, and entertainment as its main benefit, to be relevant to characterize one or more M-shopper profiles.

4.3.2. Demographical covariates

Ailawadi et al (2001) argue that although demographics do not influence behaviors directly, they have significant associations with psychographic characteristics and therefore could be used for targeting. Demographics can also profile M-shopper segments, an example can be professional mothers, who make use of mobile-commerce to buy low involvement products since it is convenient and time-efficient

(32)

University of Amsterdam 32

(Yang and Kim, 2012). The role of demographic variables within M-commerce has been previously researched by Chong (2013), or Okazaki and Mendez (2013). The following section offers a short explanation of the four active covariates included in the model.

Gender. Prior research has investigated the role of gender in mobile commerce with contradicting

results. Okazaki and Mendez (2013), found that men are more prone than women to engage in online transactions. Likewise they also found that interface aesthetics and ease of use are key to motivate females to engage in mobile commerce. On the flipside Chong (2013) concluded that men and women do not differ in how they engage in m-commerce activities. This research further will investigate whether we can establish a relationship between M-shoppers segment membership and gender.

Age. According to Chong (2013) there is a significant and negative relationship between mobile

commerce use and age. His research shows that younger users are more likely to engage in a broader spectrum of mobile activities (e.g. entertainment, content delivery, location-based activities) than older segments (Wei et al, 2009). This finding was earlier supported by Miller (1996), who described the cyberspace as the domain of young people. Contrarily Teo’s (2001) research, concludes that age does not influence internet purchase activities. This research will explore further effects of age in M-shopper membership.

Education. Some studies suggest that Internet users hold higher educational degrees (Furr and

Bonn, 1998). Likewise Konus et al. (2008) suggest that there is relationship between the level of education and multichannel behavior, which may be relate to the analytical skills and the ability to subtract benefits from the different channels. Chong (2013) and Yang and Kim (2012) also suggest such a relationship in the mobile context. Hence it is to expect that education will relate to membership in one or more M-shopper segments.

Marital status & children. Yang and Kim (2005) suggest that consumers with children find

mobile shopping a convenient solution to purchase at any time and moment. Likewise prior research finds that time pressure affects households (Bhatnagar & Ghose, 2004). This research will further investigate

(33)

University of Amsterdam 33

the relationship between mobile commerce usage and marital status and number of children. I expect that these variables may relate to segment membership.

4.3.3. Product related covariates

One of the new elements added in this M-shopper segmentation is the product or device related covariates, which stands for the device type, the type of operating system or browser type. Statistics support that the purchase and browsing patterns differ between iPhone and Android operating systems, or tablets and mobile devices (Criteo,Q3 2015; Twenga Solutions, 2015). Prior research also concluded that ease of use is important for users to engage in mobile transactions (Chong, 2013), which may reflect differences between the use of an app to the common open-source or web. This research aims to determine whether and how these variables relate or define profiles which relate to segment membership.

4.3.4. Cross-cultural covariates

The second element this research adds compared to prior research is the focus on identifying potential segment ownership differences in a cross-cultural setting (Spain and the Netherlands). Next to the mobile usage differences stated in the literature review (Criteo, 2015; Twenga Solutions, 2015), prior research has shown that culture has a moderating effect on the adoption of E-commerce (Liyi Zhang et al, 2012). Yet, there is no evidence on how the use of mobile commerce activities define segments across countries. The results could show the existence of a specific segment in the Spain, or the predominant role of another profile in the Netherlands. Likewise, if no differences are found, the results will be more generalizable than other researches that have solely focused on one country.

5. Research plan

This chapter explains the plan followed to investigate : (1) whether there are different segments of m-shoppers based on their attitude and behavior in the mobile channel , and (2) whether this segments are possibly characterized by psychographic, demographic, cultural and product related covariates. This section describes: the sample, research design, procedure and analysis method followed to investigate the existence of different mobile shopper profiles.

Segmenting m-shoppers in a cross-cultural context : a latent class segmentation based on behavioral and attitudinal factors

Master Thesis

‘