The impact of webrooming and showrooming behaviors on customer purchases, spending and revenues

(1)

The impact of Webrooming and Showrooming Behaviors

on Customer Purchases, Spending and Revenues

Master’s Thesis

Supervisor: Dr. Umut Konuş Student Number: 11653825 Name: Li Liu

Track: Marketing

(2)

Statement of Originality

This document is written by Student Li Liu who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.

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

(3)

Abstract

The shopping pattern has been enormously changing in the digital age. To maximize their benefits, customers nowadays are inclined to leverage both online and offline channels to search for information and make the final purchase decision, and often switch the channel during shopping. The phenomenon is called research shopping and has become a hot and front-end research area in multichannel marketing. However, previous research rarely took into consideration the diverse impact of research shopping behaviors (webrooming & showrooming) on customers’ purchases and spending. This research will evaluate the relationship between research shopping behavior, customers’ spending, and purchase frequency, considering the possible moderators, such as demographics, product category, online duration, the number of devices and shopping enjoyment. The research design is correlational, and the paper will collect the data from an online survey. The results indicate that webrooming is positively associated with revenue and showrooming has a positive impact on purchase frequency for consumer electronics. Also, female, online duration and shopping enjoyment can negatively moderate the relationship between research shopping behaviors and customers’ purchases in this category. Given the results, marketers can formulate the corresponding marketing strategies and allocate the marketing resources accordingly, to increase the total revenue and accelerate customers’ purchase frequency. They can also pay attention to these factors influencing the relationship between research shopping behavior, purchases, and revenues.

(4)

With the rapid development of information technology in the digital era, the shopping pattern has changed a lot over the last few decades. Customers have quickly learned how to shop online, and it has become a popular trend (Luo et al., 2014). The Online-to-Offline market is developing tremendously due to the penetration of the internet and mobile service (Shi et al., 2017). For companies, they strive to encourage customers to carry on both online and offline transactions, as they believe it can be more profitable (Venkatesan and Kumar, 2007); For customers, they can have more channel choices during shopping. The online channel is in widespread use as the offline one (Kumar and Venkatesan, 2005).

Both online and offline channels have their own strengths (Mike de Vere, 2013). Customers might consider online channel convenient to search information but perilous to place an order; while offline channel safe to purchase but time-consuming (Verhoef, Neslin and Vroomen, 2007). Because of their distinct benefits, customers start to utilize multiple channels to gather information, compare prices and purchase products (Hsiao et al., 2012; Yu et al., 2011). Neslin et al. (2016) define this phenomenon as multichannel shopping, as customers engage more than one channel during search and purchase stages. What’s more, they often switch the channel, which makes them research shoppers (Nunes and Cespedes, 2003).

There are two types of research shopper—webroomer and showroomer. Webroomers use the online channel to search info before purchasing at brick-and-mortar stores (Flavián, Gurrea, and Orús, 2016; Nesar et al., 2016; Andrews et al., 2016). Showroomers search products offline first and then ordering them online (Mehra, Kumar, and Raju 2013; Quint, Matthew, Rogers

(7)

and Ferguson, 2013). It becomes an urgent issue to concentrate on them because they occupy a significant proportion of customers. As the shopping process is divided into search and purchase phases, in addition to the research shoppers, this paper will also factor in single-channel shoppers, that is, entirely online and offline shoppers.

When it comes to companies, one of the primary objectives is to be profitable, so it is crucial for them to increase customers’ total spending and purchase frequency. As the research shopping becomes the mainstream phenomena, companies are likely to lose or gain customers during the shopping course (Nunes, Cespedes, 2003), and find it difficult to retain customers in the search stage and identify new customers in the purchase stage. This might have a major impact on the multichannel purchases and revenues, which has become one of the main motivations for this topic.

Additionally, if managers can notice the significant difference in the total revenue and purchase frequency, among webrooming, showrooming, fully online and offline shoppers1,

they can motivate customers to exhibit the most profitable behavior, while curtailing other ones. If it is difficult to modify their behavior, managers can at least implement corresponding marketing tactics and campaigns, choose the right marketing tools and allocate their resources efficiently to stimulate more consumption among the most profitable behavioral group to gain more profits and achieve long-term sustainable development. For instance, on condition that showrooming accelerates more purchase volume, salespeople can offer the store visitors more online coupons to encourage them shopping online; If it’s the other way around, companies

1

It should be noted that although the paper compares fully online/offline shoppers with the research shopper in terms of purchase amount and frequency, the research shopper is still the principal research object.

(8)

can enhance the in-store experience, or create an online countdown to the offline sale. As a result, companies can expand the most lucrative customer base, maximize the utility of their resources, save a lot of time and efforts, and reach their sales targets rapidly. By doing so, companies can take a big step ahead of their competitors.

Beauty giant Sephora offers an ‘in-store’ mode in its mobile app to combine the online and offline shopping experience to embrace the research shoppers. To serve webroomers, Walmart provides online shopping/in-store pick-up services without an extra delivery fee (Roman, 2017). Offline retailer Target attempts to impede showrooming by giving special prices, which prevents customers from comparing the prices online (Zimmerman 2012).

If showroomer is assumed to spend more, are there other variables moderating their relationship? Therefore, the study will also look at a few moderators, including product category, demographics and online duration which might influence the association between the research shopping behavior and customers’ purchases. If the research finds that female webroomers spend more, marketers can tailor their marketing communication strategies accordingly, such as conveying more female-oriented messages to those webroomers.

The topic is not only managerially and academically significant, but also exciting and unexploited, which appeals to researchers, managers, and marketing practitioners. Hopefully, they can refer to this paper for decision-making and future study.

This paper will first discuss the relevant literature. Then, it will deliberate the research gap, questions, and contributions followed by the conceptual framework and hypotheses. Subsequently, the paper will present the methodology, data, and results. In the end, implications and limitations will be summarized.

(9)

2. Literature Review

Before focusing on the relationship between research shopping behaviors and consumers’ total spending on average and their purchase frequency, the relevant literature will be elaborated in this chapter. First, we will briefly review the literature on multichannel marketing. Secondly, the research shopping behavior (webrooming and showrooming), its consequence and potential impact on customers’ purchases and revenues will be discussed. After that, the paper will demonstrate several potential moderating factors, such as product category, demographics (age and gender), online duration, number of devices and shopping enjoyment. In the end, the research gap, questions, and contributions will be proposed.

2.1. Multichannel Marketing

The number of shopping channels has soared in the early 21st century. Hence, companies have more opportunities to interact with prospective customers (Rangaswamy et al., 2005). They are progressively depending on multiple channels to exchange goods, deliver customer value, and communicate with customers, which has given birth to the term ‘multichannel marketing’ (Valos et al., 2010). Multichannel marketing allows customers to close the deal in many ways, such as through a physical store, a website, or a mobile app. It is becoming more and more popular due to the diffusion of the mobile application (Rangaswamy et al., 2005). Weinberg et al. (2007) claim that almost 70% of customers are multichannel shoppers.

A wide range of benefits arises from multichannel marketing. First, it helps companies to increase the revenue. Multichannel shoppers spend more, purchase more products and show more loyalty than single-channel ones (Myers et al., 2004; Weinberg et al., 2007; Kumar and

(10)

Venkatesan, 2005; Muller-lankenau et al., 2005). Besides, multichannel marketing allows companies to maintain a long-term relationship with shoppers by providing products or services via different channels (Rangaswamy et al., 2005). They can boost the customer retention rate because the diverse and readily-available channels prohibit customers from switching to other rivals. The article also states that companies can better understand customers’ demands by receiving more feedback from various channels. Moreover, multichannel marketing comes along with convenience. Customers have more touchpoints and can interact with the channel anytime and anywhere (Muller-lankenau et al., 2005; Rangaswamy et al., 2005). On the other hand, multichannel marketing causes many challenges for companies, such as managing internal resources among different channels and raising the distribution cost, (Rosenbloom, 2007). They must monitor customers’ online and offline movement simultaneously as well. (van Dijk et al., 2007).

In short, the emergence of new channels has overturned the traditional model of marketing and has brought about pros and cons. Companies can gain a sustainable competitive advantage if they carry out multichannel marketing efficiently (Valos et al., 2010). However, companies still have difficulties in finding the optimal channel combination in the multichannel environment (Sharma and Mehrotra, 2007). Identifying the most profitable channel is an urgent and crucial issue for marketing practitioners.

(11)

2.2. Research Shopping (Webrooming versus Showrooming)

2.2.1. Research Shopping Phenomenon

Research shopping phenomenon is the customers’ propensity to search via one channel and purchase via another (Verhoef, Neslin, and Vroomen, 2007; Nesar et al., 2016). Verhoef et al. (2007) also find that about 76% of the respondents are research shoppers. Marketers intend to embrace research shoppers as they create more value than single-channel users (Neslin, Scott and Shankar, 2009). Understanding the potential behavioral mechanism of research shopping turns into an urgent issue.

The three dominating causes for research shopping are attribute-driven decision-making, lack of channel lock-in and cross-channel synergy (Verhoef, Neslin and Vroomen, 2007). First, customers compare the attributes of various channels for search and purchase for a cost and benefit trade-off (Balakrishnan, Sundaresan, and Zhang 2014). When it comes to the online channel, it has advantages in collecting and comparing the product information, but it is regarded as an unsafe channel regarding personal information disclosure, untouchability, and unattainability (McKnight et al., 2002). Besides, the low lock-in indicates that search and purchase stages are not positively correlated in a specific channel, while the synergistic mechanism demonstrates that using one channel to search has a robust spill-over effect on using another to purchase (Verhoef, Neslin and Vroomen, 2007; Gensler, Verhoef, and Böhm, 2012).

In addition, customers’ psychographic characteristics can also have an impact on choosing different channels to search and purchase (Kumar and Venkatesan, 2005; Ailawadi, Neslin, and Gedenk 2001), such as price sensitivity, shopping gratification, innovativeness,

(12)

brand loyalty etc. (Konus, Verhoef, and Neslin, 2008). For example, customers are more inclined to alter the channels if they enjoy shopping a lot. Besides, doing research shopping might virtually give customers a sense of superiority and intelligence. (Balasubramanian et al., 2005; Chandon, Wansink and Laurent, 2000).

Channel characteristics influence customers’ proclivity to show various channel choices as well (Baker et al., 2002; Macik et al. 2012). Wang Yu-Min (2016) analyzes the impact of channel characteristics on research shoppers’ attitudes towards these channels.

2.2.2. Different Types of Research Shopping

When it comes to the classification of research shopper, competitive research shoppers refer to people who use one company’s online (offline) channel for search but use another competitor’s offline (online) channel for purchase. Loyal research shoppers refer to people who use one company’s online and offline channels for search and purchase throughout the whole shopping process (Gensler, Neslin and Verhoef, 2017). This paper concentrates explicitly on competitive research shoppers since they might pose a major threat to companies and they are of great significance in multichannel marketing.

As is mentioned above, research shopping consists of webrooming and showrooming. Webroomers search online and shop offline, while showroomers search offline and purchase online (Flavián et al., 2016; Andrews et al., 2016; Mehra, Kumar, and Raju 2013; Quint et al., 2013). The below figure shows the components of two types of research shopping behaviors.

(13)

Figure 1: Two types of research shopping behaviors

This chapter will first review the literature about webrooming. It is found that the most extended research shopping behavior is the combination of internet search and store purchase (Verhoef, Neslin and Vroomen, 2007; Weathers, Sharma, and Wood, 2007). Approximately 69% of the online shoppers surf the internet to search the product and then make the purchase decision in the brick-and-mortar stores (Harris Poll, 2014). It is common that customers opt for the ideal product and negotiate the price offline according to the first-hand information gathered online to pursue the maximum benefits (Morton, Zettelmeyer and Silva 2001; Van Bruggen et al., 2010).

Previous studies address several drivers of the webrooming phenomenon. Perceived online risk and lack of faith yield in online customers’ intention to buy offline (Forsythe et al., 2006; Chiu et al., 2011; McKnight et al. 2004). Also, it is concluded that customers perceive a high level of confidence and contentment towards the offline purchase if they search the information online before (Flavián, Gurrea, and Orús, 2016). Customers may prefer the product more and reinforce their purchase intention if they are engaged in a webrooming process (Keng et al., 2012). Flavián et al. (2016) mention that the availability of good online reviews and motivation to touch can diminish the uncertainty of shopping offline, which facilitates

(14)

webrooming. Besides, different product characteristics have varying impact on showing webrooming behavior (Arora and Sahney, 2017).

When it comes to showrooming, more than 30% of worldwide customers have exhibited showrooming behavior for once (Statista, 2014). It causes a considerable threat for retailers if a customer patronizes their offline stores first but purchase on their rival’s website. Hence, it is crucial to understand the underlying drivers of showrooming (Verhoef, Kannan and Inman, 2015).

Previous empirical researchers pinpoint several channel-related factors leading to the showrooming behavior, such as the perceived online average saving amount and the dispersion of the online prices (Branco, Sun, and Villas-Boas 2012; Arora et al., 2017), about which price-conscious customers care. Besides the price factors, many non-price factors are affecting the possibility of showrooming (Gensler, Neslin, and Verhoef, 2017). For instance, the expected sound online product quality, the waiting time in the brick-and-mortar store and the online service quality are positively related to showrooming. Online search effort, time pressure and the availability of the sales staff are negatively associated with showrooming (Konuş, Verhoef and Neslin 2008; Gensler, Neslin and Verhoef, 2017; Arora et al., 2017). Moreover, the contextual variables such as customer and product characteristics are also discussed (Gensler, Neslin and Verhoef, 2017; Daunt and Harris, 2017).

2.3. Previous Research on Research Shopping

Prior studies on research shopping phenomenon focus on its drivers and consequences, as well as the driving factors and impediments of webrooming and showrooming respectively

(15)

(Verhoef, Neslin and Vroomen, 2007; Gensler, Neslin and Verhoef, 2017; Arora, Singha and Sahney, 2017). Multiple drivers lead to various behaviors. Drivers include customers’ psychographic and demographic characteristics, product characteristics and channel characteristics (Kumar and Venkatesan, 2005; Bhatnagar, Amit, and Ghose, 2004).

When it comes to the consequence of research shopping, not surprisingly, the dilemma arises from this phenomenon. Marketers find it difficult to manage and retain customers because customers at the search and purchase stages might be completely different (Nunes and Cespedes, 2003). For example, even if a customer visits the website, he is very likely to make a final purchase in its physical store, which prevents marketers from accurately predicting the sales performance by tracking the website traffic. In addition, that customer might visit the competitor’s physical store or website for transactions. Hence, marketers and managers must value, control and leverage the impact of webrooming and showrooming as early as possible.

For webroomers, it seems that browsing an informative website can induce them to make more purchase offline and thus increase the revenue (Van Nierop et al., 2011). However, if those webroomers are very critical, they might think the information inadequate or inauthentic, and purchase the products offline from rivals. Or they might think there is too much information, which restrains them from buying the product efficiently. Hence, the consequence is difficult to predict and monitor, which makes companies feel at a loss.

(16)

Some previous studies on research shopping and key articles about multichannel shopping are shown in the chronological order as follows.

Table 1: Overview of critical articles on research shopping and multichannel shopping2

Product Category /Characteristics Demographics Customer Psychographics /Characteristics Spending /Revenue Purchase Frequency

Kumar and Venkatesan (2005) Multichannel shopping Empirical — √ √ √ √

Van Baal and Dach (2005) Multichannel shopping Empirical √ — — — √

Verhoef, Neslin and Vroomen (2007) Reseach shopping* Empirical √ — √ — —

Venkatesan and Kumar (2007) Multichannel shopping Empirical — √ — √ √

Konus, Verhoef, and Neslin (2008) Multichannel shopping Empirical √ √ √ — —

Kushwaha & Shankar (2013) Multichannel shopping Empirical √ — — √ √

Mehra, Kumar, and Raju (2013) Showrooming Theoretical √ — — √ —

Flavián, Gurrea, and Orús (2016) Webrooming Empirical √ — √ — —

Nesar et al. (2016) Research shopping* Empirical — √ — — √

Wang Yu-Min, et al. (2016) Research shopping* Empirical — — √ — —

Arora and Sahney (2017) Webrooming Empirical √ — √ — —

Arora, Singha and Sahney (2017) Showrooming Empirical — — √ — —

Daunt and Harris (2017) Showrooming Empirical √ — √ — —

Gensler, Neslin and Verhoef (2017) Showrooming Empirical √ — √ — —

This research Research shopping* Empirical √ √ √ √ √

Drivers Monetary Value Authors (year) Type of Shopping Behavior Empirical/

(17)

2.4. The Potential Impact of Research Shopping on Purchases and Revenues

To uncover the potential impact of research shopping on revenue and purchase frequency, the author introduces the classic model proposed by Bult and Wansbeek (1995). RFM model integrates the construct of recency, frequency, and monetary. Recency means the time that has passed since a shopper’s latest purchase; Frequency is defined as the number of times a shopper purchases over a period; Monetary is the amount of money a shopper pays over a period (Chen et al., 2009). As the main purpose of this research is to compare the customers’ purchase under the circumstances of webrooming and showrooming, we select the concept of monetary and frequency to explore their spending and shopping frequency, which become the dependent variables in the research. Recency is not considered as it’s not closely related to the total revenue or frequency.

The research shopping behavior is now exerting a great influence on the revenues deriving from diverse channels. Companies should realize the importance of this phenomenon (Chiou et al., 2012; Kim et al., 2005; Kumar and Venkatesan, 2005; Kushwaha and Shankar, 2013; Wang Yu-Min et al., 2016; Chiu et

al., 2011). In the Harris Poll (2013), it is claimed that on average showroomers spend less than webroomers in US ($174 vs. $203.9 for latest purchase). However, it only clarifies the difference in volume

for the last purchase, without considering the total revenue for a specific period. Additionally,

(18)

customer’s shopping behavior has changed quickly in recent years, but the research has not been updated. The research shopping phenomenon strikingly influences the offline and offline revenue (Chiu et al., 2011). However, its revenue impact has not been fully explored.

Regarding purchase frequency, Venkatesan and Kumar (2007) suggest that research shoppers are inclined to have a higher purchase frequency. Statista (2015) shows that in Germany about 60% of webroomers and 50% of showroomers have relatively high shopping frequency. Besides, webroomer’s average shopping frequency is different from showroomer’s frequency among regular shoppers. Customers with higher purchase frequency contribute to more revenues (Chiou and Pan, 2009; Martin et al., 2015).

In sum, there is sparse academic literature focusing on the different purchase and revenue impact of research shopping. Only a few studies suggest that research shopping would have an enormous influence on companies’ revenues and customers’ purchase frequency, but they merely discuss their general relationship, rather than delving into which behavior (webrooming or showrooming) results in more revenue and purchase frequency. Therefore, more research needs to be carried out. Marketers can then refer to the outcomes and target the customers who spend more and shop more frequently. Identifying the most profitable shopping behavior helps companies allocate the resources to its optimal channel, which leads to better multichannel marketing strategies. Current research on this topic is not systematic and in-depth, and few implications have been reached for marketers. This paper will emphasize this topic as it is an imperative issue for companies to tackle.

(19)

2.5. Other Factors Affecting the Impact of Research Shopping on Purchases and Revenues

2.5.1. Product Category

Based on the product category, a single customer might exhibit different research shopping behaviors and choose different channels for search and purchase (Daunt and Harris, 2017; Heitz-Spahn and Sandrine, 2013; Bhatnagar, Amit, and Ghose, 2004; Kushwaha and Shankar, 2013; Arora and Sahney, 2017). Some products are more suitable to purchase online than others (Gensler, Dekimpe and Skiera, 2007). For example, customers prefer the online channel to buy books and the offline one to procure housing appliances and clothing, while they probably leverage more than one channel to search and buy electronics (Konus, Verhoef, and Neslin, 2008; Harris Poll, 2013). Van Baal and Dach (2005) only mention that product category influences the possibility to exhibit showrooming behavior. This implies that a single customer may showroom in a specific product type, while may not showroom when encountered other products. Accordingly, product category becomes an essential variable when studying the impact of research shopping behaviors on revenues and purchase frequency. Van Nierop et al., (2011) point out that no empirical study focuses on whether research shoppers pay more or less across different product categories. After reviewing the relevant literature, it is found that few intensive studies consider the product category as a moderating factor associating with the research shopping behavior and the purchases and revenues.

(20)

2.5.2. Demographics

As far as the demographical factors are concerned, this paper includes age, gender, and education as the potential moderators. In terms of age, existing research only examines its impact on the probability of channel switch, that is, showing the research shopping behavior, and various results are concluded. Ansari et al. (2006) claim that they have a significant negative correlation, while Gupta, Su, and Walter (2004) find that age is not dramatically related to the channel switch. According to the survey conducted by MarketingProfs (2015), American customers among 18-40

shop online more frequently than older people. It indicates that younger customers have more tendency to become research

shoppers, generate more

expenditure and purchase more often, as they embrace the online channel more and combine it with the traditional offline channel. All the studies investigate the age impact on purchases and revenues among research shoppers in general, rather than investigate webroomer and showroomer respectively. In a word, not many studies look into the topic of whether age can influence customer spending and purchase frequency among webroomers and showroomers.

In terms of gender, Harris Poll (2013) mentions that men are more likely to spend more than women among showroomers. Both men and women are increasingly showing webrooming and showrooming behavior. Men tend to adopt the new channel and technology

(21)

in a shorter duration than women (Venkatesan, Kumar and Ravishanker, 2007) because they perceive the compatibility and complexity more positively (Lian et al., 2014). The mindset might increase their purchase intention and then purchase more when they encounter new online platforms. However, Akhter (2012) proves that gender has no significant impact on online purchase volume. Nonetheless, further research should pay attention that which gender generates more expenditure and shopping frequency in different research shopping behavioral groups (showroomer versus webroomer).

Moreover, people with higher education are more accessible to mastering the methods to shop across multiple channels. They are more likely to extract useful information through various channels (Kushwaha and Shankar, 2008; Strebel, Erdem, and Swait, 2004). So, they have more opportunities to switch the channel during the dynamic shopping process, instead of staying with one channel. It also explains why educated people are buying more expensive products and probably spend more than others (Konus, Verhoef, and Neslin, 2008).

2.5.3. Online Duration

Online duration stands for the total time people spend on the Internet in a certain period (Akhter, 2012). When customers spend more time online, their expected usage of the internet will dramatically increase, since they deem that Internet can benefit them in different ways, and they feel more confident and relaxed when shopping by online channel (Li and Chuan, 2010). This psychological factor explains why customers prefer online shopping rather than the traditional offline model. As a result, they treat the online channel as a trustable transactional shopping platform. Their changed shopping behavior will enhance total spending

(22)

and shopping frequency (Lohse et al., 2000; Nesar et al., 2016). Hence, the online duration possibly influences the impact of webrooming and showrooming on revenues and purchases.

2.5.4. Number of Devices

It is found that about 20% of smartphone users shop through their smartphones on a daily basis and the behavior occurs more frequently than other customers (Google, 2012). Peep Laja (2017) states that it costs 50% more per online purchase for tablet users compared with smartphone users, and 20% more compared to desktop or laptop users. Moreover, tablet users are three times more likely than smartphone users to conduct online transactions based on the data from 150 US retailers in 2011. The reason behind this surprising finding is that the large screen of the tablet is more user-friendly to carry out the online shopping. Furthermore, as a tablet is relatively expensive, its users might be much wealthier, and they are very likely to

own both smartphone and tablet (Austin Bankhead, 2012). Depending on the articles above,

apparently, customers with more devices (smartphone, tablet, laptop or desktop) have a higher possibility to generate more revenue and higher shopping frequency. Thus, the number of devices a customer is using might moderate the relationship between research shopping behavior, purchase volume, and purchase frequency as well.

2.5.5. Shopping Enjoyment

Shopping enjoyment implies customers’ perceived hedonic and entertaining value during shopping activities (Konus, Verhoef, and Neslin 2008; Daunt and Harris, 2017). Customers with shopping enjoyment inherently find shopping interesting and are more inclined to exhibit research shopping behavior (Forsythe et al., 2006; Kim et al., 2007; Verhoef, Neslin and

(23)

Vroomen, 2007), as they see shopping as a way of relaxation and entertainment. Social occasion strongly affects the channel selection for search and purchase, such as shopping with the acquaintance (Nicholson et al., 2002). Regarding the monetary impact, Hart et al. (2007) reveal that higher level of shopping enjoyment causes greater patronage intentions, which generates more spending and more frequent purchase. However, little research considers shopping enjoyment as the moderator of the relationship between research shopping, revenue, and shopping frequency, let alone discussing webrooming and showrooming separately.

2.6. Research Gap and Research Questions

It can be concluded that little has been compared so far about the influence different research shopping behaviors and total online/offline shopping behaviors might have on the customer purchase amount and purchase frequency, given the moderating factors, including product category, demographics, online duration, number of devices and shopping

enjoyment. It should be noted that the study mainly focuses on evaluating research shopper rather than fully online/offline shoppers, but they cannot be excluded. Regarding the existing literature, webrooming and showrooming should also have different financial effects on companies. This paper is designed to fill this research gap by reviewing all relevant literature. This gap leads to the research questions:

Does research shopping behavior have an impact on the revenues and purchase frequency? Do webrooming and showrooming behaviors have different consequences on the total revenue and purchase frequency, moderated by product category, demographics, online duration, device quantity and shopping enjoyment?

(24)

2.7. Contributions

This paper aims at studying the relationship between the independent variable research shopping behavior and the dependent variables total customer spending and purchase frequency, comparing the dependent variables derived from the webroomer and showroomer in different product categories, demographics, and several other moderators. Little research has been carried out to explore their correlation. Although some literature mentions that research shoppers spend more and have higher purchase frequency (Venkatesan and Kumar, 2007; Weinberg et al., 2007), they are kind of obsolete and don’t go in-depth. Furthermore, they merely regard two research shopping behaviors as one variable, to investigate its association with the revenues and purchases. Customers nowadays show different research shopping behaviors or entirely online/offline shopping behavior. There should be some research that deliberates on every single behavior and make a comparison towards its impact on purchase amount and frequency. Besides, previous research didn’t incorporate the moderating effects on the above variables. As this topic becomes a hot issue in multichannel marketing, customers are increasingly exhibiting research shopping behavior, and companies are devoted to discovering the optimal channel for pursuing high profits, new research has to be conducted at this moment to fill this knowledge gap. In a word, the academical contributions lie in:

a) Comparing different research shopping behavioral groups (webroomer versus showroomer) regarding the total purchase volume they bring and the purchase frequency they have.

(25)

b) Investigating the moderating effects of product category, demographics, online duration, device quantity and shopping enjoyment on the revenue and purchase frequency impact of the research shopping behavior.

On the other hand, this paper also has its managerial contributions. First, managers have a clear direction on how to modify customer’s shopping behavior based on the study results. Considering their product categories, they can motivate their customers to be either showroomer, webroomer or fully online shoppers by marketing campaigns for maximizing the profits. Second, according to customers’ demographics, managers can develop diverse marketing strategies to reach the more considerable profit. For instance, they can encourage men to search offline and buy online, if men are likely to spend more when they showroom. This principle applies to other moderators as well. If the number of devices a specific customer owns positively affects the revenue impact of research shopping behavior, managers should build multiple transactional interfaces which are compatible to all devices, such as the laptop, smartphone, and tablet, to encourage him or her place the order via different devices. Webroomers and showroomers can complete the transaction more conveniently and therefore boost their expenditures. If research shoppers who love shopping a lot spend more and purchase more often, managers can put more emphasis on those shopping lovers, and they have a better understanding of the target group. On the grounds of the outcome, marketers will know which moderating factors they need to take into consideration for developing the multichannel marketing strategy and have a better prediction of the potential benefits.

(26)

3. Conceptual Framework

This chapter intends to illustrate the relationships among all variables intuitively in a conceptual framework. Please find the framework below.

Figure 4

As shown above, the primary objective of this study is to discover the relationship between the research shopping behavior and customers’ purchase. Given the two phases (search & purchase) during shopping, there are four behavioral groups of customers, the webroomer, and the showroomer, who are defined as the research shopper, as well as the online and the offline shopper (single channel user). The exhibition of different shopping behaviors might have various consequences on their total spending and purchase frequency. Other variables might also moderate this relationship, such as product category, demographics, how many devices they are using, how long they spend on the internet per day and to which extent they enjoy shopping.

(27)

4. Hypotheses

According to the conceptual framework and the relevant literature discussed above, several assumptions are presented below.

Revenue and purchase frequency

It is found in the previous study that showroomers spend less than webroomers in the US for the most recent purchase (Harris Poll, 2013). This finding might be obsolete, and the respondents are not entirely representative, but it can still serve as an essential reference due to the lack of relevant research. We posit that webroomers are prone to purchase more in comparison with showroomers, as they will be attracted by other similar goods or matching products in the physical stores, and sales staff will recommend more products to them. Webroomers are relatively price-sensitive, so they compare the prices online first to search for an optimal price and product. After deciding which product to buy, they suppose the offline purchase is extremely cost-effective and saves a lot of money on the psychological level. As a result, they are more likely to purchase other products which lead to higher revenues.

In terms of showroomers, they would like to touch and feel the product first in the store to minimize the potential risk and then go back to the online platform to decide whether they should purchase or not. They tend to buy goods online more frequently as nowadays buying online is much more convenient, instead of visiting a brick-and-mortar store. Showroomers are more familiar with each transactional platform of a retailer, such as mobile site and desktop site, encouraging them to carry out the deal more often. We suppose that they are likely to make small transactions with higher frequency.

(28)

H1a: Webrooming is associated with more revenues compared to showrooming.

H1b: Showrooming is associated with higher purchase frequency compared to webrooming.

Product category

In the clothing category, even though the online information is detailed and covers almost every attribute, webroomers intend to try on the clothes to check first, and they are willing to receive some advice from salespeople, for example, whether the outfit fits or suits them well, why the quality is superior, or how to do the clothing collocation. Consequently, salespeople will recommend other clothes that match the one they initially want to buy. Webroomers will listen to their advice and probably purchase more clothes beyond the original budget. It will lead to higher revenues or retailers. The more they try on, the more they want to buy.

In the consumer electronics category, showroomers will first test the performance and function of the electronics in the physical store before they go online to look for a better price. Nowadays online retailers often offer a discount for the next purchase through coupon, voucher or discount code to encourage consumption. The promotion policy might be

effective for consumer electronics as their prices are relatively high. Thus people intend to save more money, and this policy will increase the purchase frequency. Hence, the paper puts forward the following hypotheses.

H2a: The positive effect of webrooming on revenue will be more pronounced when customers purchase clothing.

H2b: The positive effect of showrooming on purchase frequency will be more pronounced when customers purchase consumer electronics.

(29)

Demographics

Older customers show less possibility to become a research shopper due to the

reluctance to switch channel (Ansari et al., 2006). Older research shoppers seem to be more intellect and show less spending impulses when they purchase products. In addition, they are more economical than younger research shoppers. Older webroomers and showroomers are not likely to make the purchase daily, and they only buy products when their basic needs cannot be met.

H3a: The positive effect of webrooming on revenue and showrooming on purchase frequency will both be weakened when age increases.

Women regard shopping as an enjoyable social activity and an excellent opportunity to create a close social bond with their friends (Alreck and Settle, 2002). It indicates that women prefer to shop offline more. Hence, if they are webroomers, shopping in the physical store with their friends is the best option. However, they will probably spend less. If there are a lot of females shopping together, they are more likely to weigh the benefits and recommend each other the best deal or the most cost-efficient products. Also, online promotion and online subscription are becoming more and more popular in the information age; females tend to shop online for more discount. Consequently, the revenue impact will be less pronounced among the female webroomers.

H3b: The positive effect of webrooming on revenue will be weakened among female customers.

(30)

Male showroomers don’t bother to visit the offline store, so they choose to place orders in front of their desktops, or via smartphones. Compared to female showroomers, they are likely to conduct the online purchase more frequently to save time and efforts.

H3c: The positive effect of showrooming on purchase frequency will be more pronounced among male customers.

Online Duration

The impact of webrooming versus showrooming on revenue and purchase frequency depends on how long the customer spends on the Internet per day. Internet enthusiasts spend more time online than other people, so they split more time on everything they do online on average, such as watching movies, chatting with friends, and shopping. As a result, among Internet enthusiasts, webroomers tend to decrease their offline spending. For those who spend a lot of time online per day, as they are showroomers, shopping might not be one of their main purposes of surfing the internet. As a result their purchase frequency might be lower. H4a: The positive effect of webrooming on revenue will be weakened when a customer spends more time online per day.

H4b: The positive effect of showrooming on purchase frequency will be weakened when a customer spends more time online per day.

Number of Devices

As discussed in the literature review, customers with multiple devices (smartphone, tablet, laptop, and desktop) are likely to generate more expenditures and have higher shopping frequency (Peep Laja, 2017). The finding signifies that webroomers with more devices intend to reduce their visits to offline stores in the purchase stage, while showroomers are more willing

(31)

to purchase goods online as it is more convenient for them to make the online purchase via multiple devices.

H5a: The positive effect of webrooming on revenue will be weakened when a customer is using more devices.

H5b: The positive effect of showrooming on purchase frequency will be more pronounced when a customer is using more devices.

Shopping Enjoyment

Customers with a higher level of shopping enjoyment have higher possibility to show multichannel shopping behavior (Forsythe et al., 2006; Kim et al., 2007) and tend to shop more often (Hart et al., 2007). Hence, we assume that shopping enjoyment has a positive effect on the relationship between showrooming and purchase frequency. However, the more a webroomer enjoys shopping, the more likely he/she is to deliberate and compare which option is the best online, the more likely he/she will not make the final offline purchase. The purchase intention cannot convert to the actual purchase in this case. Hence, the effect of webrooming on revenue will be less pronounced if a customer enjoys shopping.

H6a: The positive effect of webrooming on revenue will be weakened when a customer has a higher level of shopping enjoyment.

H6b: The positive effect of showrooming on purchase frequency will be more pronounced when a customer has a higher level of shopping enjoyment.

(32)

(33)

5. Research Design

To address the research questions and obtain the results, the author will detect the relationship among the multiple variables via the correlational research design, in which the data is collected from an online survey. Therefore, the paper uses the quantitative method to reach the external validity. This paper will focus on several key variables, the independent variable—research shopping behavior (webrooming & showrooming), the dependent variables—total customer spending in the last 6 months and purchase frequency, and the moderating variables—product category, some factors in demographics, online duration and the number of devices one customer is using. Two product categories would be considered, consumer electronics and clothing. Other demographical elements are used as control variables. The online survey is conducted via Qualtrics and SPSS (IBM Statistics 22.0) will be used to do the statistical regression analysis. For purchase frequency, as we set the same duration (last six months) for all respondents, they will answer the number of times they purchase. All respondents are going to fill in the same questionnaire, and it is available in English and Chinese. A pilot study increases the likelihood of success in a main study (Van Teijlingen et al., 2001). Before sending out the survey, we will carry out the online and offline pilot surveys with several people to see whether the questions are clear and intelligible and whether they can be answered within the given time slot.

5.1. Population and Sample

In this paper, sampling is used to make the data collection manageable and feasible. The online questionnaire will be handed out to fellow students, acquaintances, and family

(34)

members via social media and email. Hence, this paper uses non-probability convenience sampling as the respondents are chosen in a non-random manner, for the reason that they are convenient to contact. The respondents are mostly between 20-30 years old. Both male and female are going to take the questionnaire.

5.2. Measures

In the questionnaire, several questions will be asked, such as the approximate total amount of money the respondents spend and their total purchase frequency in the last 6 months. Respondents will answer the questions based on different product categories (consumer electronics and clothing). Regarding the independent variable, the questionnaire will investigate the most common channel the respondents rely on for search and purchase (either online or offline) as well, to figure out which behavioral groups they belong to, either webroomers, showroomers, online shoppers or offline shoppers. It could be obtained by the question: what is the approximate percentage of conducting the search/purchase stage online/offline? Besides, the questionnaire will ask the respondents about their demographics and how many devices they own by which they can place the order. Respondents’ online duration will be examined by the question: how much time do they spend online per day? Furthermore, attitude variable will be considered in the questionnaire such as the shopping enjoyment. To measure shopping enjoyment, the paper adopts a 7-point and 3-item Likert Scale (1 = strongly disagree; 7 = strongly agree) with a Cronbach’s alpha of 0.90 which originally established by Ducoffe (1996). Other items are adapted from Putrevu (1997) and Babin, et al. (1994), with a relatively high Cronbach’s alpha. The questions are designed to

(35)

figure out respondents’ hedonic degree of shopping. The variables, measures, and levels are summarized below.

Table 3: Summary of variables, measures, and levels

5.3. Recoding Variables

As the independent variable is categorical, the paper requires to recode it as the dummy variable. Four shopping behavioral groups were recoded, that is, webroomer, showroomer, online shopper and offline shopper. We divided the four groups by setting the critical point to 50%. For instance, regarding the channel usage, webroomer describes a person using more online channels (≥50%) than offline channels for the search stage, while using more offline

(36)

channels (≥50%) than online channels for the purchase stage. The offline shopper category was set as the reference group. Four categories imply three dummies. To be specific, webroomer (respondents who mainly search online and purchase offline) = 1, and others = 0; showroomer (respondents who mainly search offline and purchase online) = 1, and others = 0; offline shopper (respondents who mainly search offline and purchase offline) = 1, and others = 0; online shopper (respondents who mainly search online and purchase online) = 0. Note that 2 sets of dummy variables were created due to 2 product categories. These dummy variables can also be multiplied by the moderating variables to examine the moderating effects. The measure level of gender is nominal, so it was recoded as the dummy variable as well.

5.4. Analysis

Since the dependent variables are numerical (continuous variable), the paper will adopt multivariate linear regression analysis on the main effects between independent and

dependent variables, as well as the moderating effects, as they are all numerical. The aim is to predict the value of dependent variables in the best way by identifying a linear combination of independent variables. Since the dependent variables spending and purchase frequency are collected respectively according to different product categories (clothing/apparel and

consumer electronics), the data will be run separately via two regression-based models to conduct the econometric analysis for them. The interactions between moderating variables will be examined. In summary, the paper will construct 4 regression models, that is, one for spending and one for purchase frequency, both based on two product categories

(37)

6. Results

In total 182 respondents filled out the online survey. As the respondents were forced to answer every question, there is no uncompleted questionnaire. All of them are valid, and it yields 182 samples for analysis.

Regarding demographics, among all respondents, the mean age is around 26 years old (Median = 25), with a gender distribution of 38.5% male and 61.5% female. 93 respondents have a bachelor’s degree, which accounts for more than half of the samples.

The mean volume spent in the clothing/apparel category is 638.07 euro (SD = 701.40 euro), with the mean purchase times of 8.9 (SD = 8.29) in the last 6 months. The mean volume spent in the consumer electronics category is 605.98 euro (SD = 764.48 euro), with the mean purchase times of 2.41 (SD = 2.42) in the last 6 months. An overview of the purchase amount and frequency is placed below.

Table 4: Overview of purchase amount and frequency

6.1. Validity and Reliability

Since the online survey is anonymous and highly confidential, all respondents filled it out honestly, which enhances the internal validity and reliability. Besides, the validity is also guaranteed as all respondents were voluntary to complete the survey without external interference.

(38)

Cronbach's alpha is a psychometric test to estimate the reliability of different items based on their correlations. It also shows the uncorrelated item, which allows the researcher to improve the overall reliability of a specific Likert scale by deleting the item. In this paper, we test the variable shopping enjoyment. The result shows that the shopping enjoyment scale is highly reliable (α = 0.908). The corrected item-total correlation demonstrates that 3 items are correlated (all > 0.30). None of the items would affect reliability

if they were deleted. Please see table 4 below and table 2 in Appendix A. The value of Cronbach’s Alpha is relatively high,

so this paper creates a combined scale by taking the average value of these 3 items to measure shopping enjoyment. The approach is agreeable and practical among previous researchers, and it’s more convenient to have only one variable for analysis.

(39)

6.2. Main Effects Purchase Amount

(N = 182) Revenue/Spending in euros (Clothing/Apparel)

Variables B SE Beta p (Intercept) -203.09 356.79 - 0.570 Webroomer 259.32 153.14 0.14 0.092* Showroomer 135.71 192.19 0.06 0.481 Online Shopper 110.61 126.61 0.08 0.384 Online Duration -18.30 11.37 -0.12 0.109 Number of Devices 81.42 63.97 0.09 0.205 Shopping Enjoyment 98.96 41.16 0.19 0.017** Gender 174.23 113.28 0.12 0.126 Age 0.94 9.29 < 0.01 0.919 R² 0.125 F 3.081

Table 6: Linear Regression Analysis Revenue/Spending (Clothing/Apparel) Note: Significance levels: **p<0.05; *p<0.10

The main effect is controlled by demographic variables and personal situation variables. For the clothing/Apparel category, 12.5% of the variance of the customer purchase amount is explained by the model. The model is statistically significant. F (8,173) = 3.081; p = 0.003. The results in Table 6 show that webroomer has an extremely weak effect on the purchase amount (p=0.092), while showroomer and online shopper do not affect the purchase amount (p>0.10). Shopping enjoyment has a significant positive effect on the purchase amount (p=0.017), which indicates that people purchase clothing a lot when they enjoy shopping. Thus, as an independent variable, shopping enjoyment is the most important variable influencing the total spending (Beta=0.190). Based on our result, males spend €325.35 less than females on average for clothing.

(40)

(N = 182) Revenue/Spending in euros (Consumer Electronics) Variables B SE Beta p (Intercept) 2.051 392.36 - 0.996 Webroomer 418.16 170.03 0.24 0.015** Showroomer 278.08 229.13 0.10 0.227 Online Shopper 434.65 151.25 0.28 0.005** Online Duration -12.01 12.72 -0.07 0.347 Number of Devices 107.33 72.86 0.11 0.143 Shopping Enjoyment 42.82 45.64 0.08 0.349 Gender -228.06 123.89 -0.15 0.067* Age 0.10 10.27 < 0.01 0.992 R² 0.093 F 2.217

Table 7: Linear Regression Analysis Revenue/Spending (Consumer Electronics) Note: Significance levels: **p<0.05; *p<0.10

For the consumer electronics category, 9.3% of the variance of the customer purchase amount is explained by the model. The model is statistically significant. F (8,173) = 2.217; p = 0.028. The results in Table 7 show that webroomer and online shopper have an extremely significant positive effect on the purchase amount (p=0.015; p=0.005), while showroomer does not affect the purchase amount (p>0.05). When it comes to the standardized partial regression coefficients, for 1 standard deviation increases in webroomer dummy variable, the purchase amount increases 0.238 standard deviations. For 1 standard deviation increases in online shopper dummy variable, the purchase amount increases 0.284 standard deviations. It suggests that online shopper is more important than webroomer considering the effect on purchase amount for consumer electronics. Gender has an extremely weak and negative effect on the purchase amount (p=0.067). Not surprisingly, males spend €143.75 more than females on

(41)

Purchase Frequency

(N = 182) Purchase Frequency in the last 6 months (Clothing/Apparel)

Variables B SE Beta p (Intercept) 2.04 4.12 - 0.621 Webroomer 3.52 1.77 0.17 0.048** Showroomer 2.21 2.22 0.08 0.321 Online Shopper 3.15 1.46 0.19 0.032** Online Duration -0.19 0.13 -0.10 0.155 Number of Devices -0.08 0.74 < -0.01 0.914 Shopping Enjoyment 1.13 0.48 0.18 0.019** Gender 3.15 1.31 0.19 0.017** Age -0.07 0.11 -0.04 0.541 R² 0.166 F 4.302

Table 8: Linear Regression Analysis Purchase Frequency (Clothing/Apparel) Note: Significance levels: **p<0.05; *p<0.10

For the clothing/Apparel category, 16.6% of the variance of the customer purchase frequency is explained by the model. The model is statistically significant. F (8,173) = 4.302; p = 0.000. The results in Table 8 show that webroomer and online shopper have a dramatically positive correlation with purchase frequency (p=0.048; p=0.032), while showroomer has no effect on the purchase frequency (p>0.10). When it comes to the standardized partial regression coefficients, for 1 standard deviation increases in webroomer dummy variable, the purchase frequency increases 0.166 standard deviations. For 1 standard deviation increases in online shopper dummy variable, the purchase frequency increases 0.190 standard deviations. It indicates that online shopper is the most important variable influencing the purchase frequency for clothing. Shopping enjoyment has a significant positive effect on the purchase frequency (p = 0.019), which signifies that people purchase clothing for a lot of times if they enjoy

(42)

shopping. Gender also has a significant positive effect (p=0.017). Not surprisingly, in the last 6 months, females purchase clothing 5 times more than males on average.

(N = 182) Purchase Frequency in the last 6 months (Consumer Electronics)

Variables B SE Beta p (Intercept) -0.04 1.22 - 0.976 Webroomer 0.93 0.53 0.17 0.080* Showroomer 1.76 0.71 0.21 0.014** Online Shopper 1.52 0.47 0.31 0.001** Online Duration 0.10 0.04 0.18 0.015** Number of Devices -0.07 0.23 -0.02 0.756 Shopping Enjoyment -0.02 0.14 -0.01 0.918 Gender -0.48 0.38 -0.10 0.214 Age 0.04 0.03 0.10 0.173 R² 0.130 F 3.222

Table 9: Linear Regression Analysis Purchase Frequency (Consumer Electronics) Note: Significance levels: **p<0.05; *p<0.10

For the consumer electronics category, 13.0% of the variance of the customer purchase frequency is explained by the model. The model is statistically significant. F (8,173) = 3.222; p = 0.002. The results in Table 9 show that showroomer and online shopper have a significant correlation with purchase frequency (p=0.014; p=0.001), and webroomer has an extremely weak effect on the purchase frequency (p=0.080). When it comes to the standardized partial regression coefficients, for 1 standard deviation increases in showroomer dummy variable, the purchase frequency increases 0.207 standard deviations. For 1 standard deviation increases in online shopper dummy variable, the purchase frequency increases 0.313 standard deviations. It implies that online shopper is the most important variable influencing purchase frequency

(43)

for consumer electronics. Online duration has a significant effect on the purchase frequency (p=0.015), which suggests that people purchase consumer electronics for more times if they spend more time online. Based on our result, in the last 6 months, males purchase consumer electronics slightly more than females on average.

6.3. Moderating effects/Interactions

According to the hypotheses, the paper is going to examine the moderating effects by using interaction variables. The overview of all moderating effects is demonstrated below.

(N = 182) Moderating Effects Revenue/Spending & Webrooming

Variables B SE Beta p

Webroomer*Age (ca) 28.93 46.48 0.05 0.534

Webroomer*Age (ce) -8.72 92.01 <-0.01 0.925

Webroomer*Gender (ca) 21.20 51.01 0.03 0.678

Webroomer*Gender (ce) -120.83 58.58 -0.15 0.041**

Webroomer*Online Duration (ca) -33.45 58.46 -0.04 0.568

Webroomer*Online Duration (ce) 11.55 66.06 0.01 0.861

Webroomer*Number of Devices (ca) -10.89 51.93 -0.02 0.834

Webroomer*Number of Devices (ce) 67.64 62.72 0.08 0.282

Webroomer*Shopping Enjoyment (ca) 2.26 54.64 <0.01 0.967

Webroomer*Shopping Enjoyment (ce) -121.12 62.25 -0.15 0.053*

Table 10: Overview Interaction Effects Revenue/Spending

Significance levels: **p<0.05; *p<0.10; ca: clothing/apparel; ce: consumer electronics

Table 10 exhibits that gender has a significant negative moderating effect on the relationship between purchase amount and webrooming in the consumer electronics category (p=0.041, B=-120.83). Specifically, this paper will explore the varying effects of male and female on the purchase amount below. Besides, shopping enjoyment has a marginal moderating influence on the relationship between purchase amount and webrooming in the consumer

(44)

electronics category (p=0.053, B=-121.12). It indicates that the positive effect of webrooming on revenues will be slightly less pronounced when a customer enjoys shopping consumer electronics. Gender 0.153) is almost the same important as shopping enjoyment (Beta=-0.147) negatively moderating the effect of webrooming on revenues. However, the results suggest that none of the assumptive moderators (age, gender, online duration, number of devices, shopping enjoyment) can influence the relationship between revenue and webrooming in the clothing/apparel category.

(N = 182) Moderating Effects Purchase Frequency & Showrooming

Variables B SE Beta p

Showroomer*Age (ca) 0.27 1.14 0.02 0.816

Showroomer*Age (ce) -0.32 0.21 -0.11 0.131

Showroomer*Gender (ca) -0.32 0.59 -0.04 0.587

Showroomer*Gender (ce) -0.06 0.18 -0.03 0.736

Showroomer*Online Duration (ca) 0.16 0.65 0.02 0.802

Showroomer*Online Duration (ce) -0.36 0.17 -0.17 0.031**

Showroomer*Number of Devices (ca) 0.17 0.65 0.02 0.789

Showroomer*Number of Devices (ce) -0.36 0.24 -0.11 0.133

Showroomer* Shopping Enjoyment (ca) -0.28 0.64 -0.03 0.663

Showroomer* Shopping Enjoyment (ce) -0.17 0.19 -0.08 0.364

Table 11: Overview Interaction Effects Purchase Frequency

Significance levels: **p<0.05; *p<0.10; ca: clothing/apparel; ce: consumer electronics

Table 11 exhibits that only online duration has a significantly negative moderating effect on the relationship between purchase frequency and showrooming in the consumer electronics category (p=0.031, B=-0.36). To be specific, the positive effect of showrooming on purchase frequency for consumer electronics will be less pronounced when a customer spends more time online per day. Likewise, none of the hypothetical moderators (age, gender, online duration,

(45)

number of devices, shopping enjoyment) can affect the relationship between purchase frequency and showrooming in the clothing/apparel category.

This paper is going to illustrate the interaction effect in detail through the multiple line charts. Regarding the moderating effect of gender on revenues and webrooming, we can conclude that in the consumer electronics category, the effect of webrooming on revenues will be more pronounced if a customer is

male. Moreover, the effect of webrooming on revenues will be less pronounced if a customer is

female. Additionally, male

webroomers tend to purchase

consumer electronics in a more substantial amount compared to

those who don’t exhibit

webrooming behavior. Females would exert a marginally negative effect on the relationship between revenues and webrooming. In sum, gender has a significant moderating influence on the association between revenues and webrooming.

(46)

For the moderating effect of shopping enjoyment on revenues and webrooming, we can

conclude that in the consumer

electronics category, the effect of webrooming on revenues will be more pronounced if a customer has a lower level of shopping enjoyment; while the effect of webrooming on revenues will be less pronounced if a customer has a higher level of shopping enjoyment. Hence, shopping enjoyment has a

marginal moderating effect on the relationship between revenues and webrooming.

Regarding the moderating effect of online duration on purchase frequency and showrooming, we can see that in the consumer electronics category, the effect of showrooming on purchase frequency will be more

pronounced if a customer spends less time online per day; while the effect of showrooming on purchase frequency almost keeps the same if a customer spends more time online per day. Hence, online duration has a significant moderating effect on the relationship between purchase frequency and showrooming.

Figure 6: Moderating effect of shopping enjoyment on revenues and webrooming

Figure 7: Moderating effect of online duration on purchase frequency and showrooming

(47)

7. Conclusion & Discussion

Table 9: The main effect of research shopping behavior on purchases overview

As shown in table 9 above, in the clothing/apparel category, webrooming is slightly correlated with more revenues compared to showrooming (0.05<p<0.10). On the other hand, webrooming is dramatically associated with more revenues for consumer electronics (p=0.015). Thus, hypothesis 1a is supported. For both categories, the outcome implies that showrooming cannot influence revenues. Customers’ different motives for showrooming might be the primary cause of this phenomenon. Some people intend to check the quality in stores before making the online purchase (Gensler, Neslin and Verhoef, 2017), which means they are price sensitive to some extent. Others choose to purchase online as they believe they can find more relevant products or supplements, which would add the total revenue the other way around.

The result shows that the correlation between showrooming and purchase frequency is not significant in the clothing/apparel category. The reason behind this might be the wide range of prices for clothing. People trust online platforms and shop for cheap clothing frequently, such as T-shirts and shorts. But they might think twice when buying luxury clothing online due

The impact of webrooming and showrooming behaviors on customer purchases, spending and revenues