Faculty of Economics and Business
Final Thesis MarketingOnline Customer Experience
The impact of online customer service on offline purchase intentions
moderated by the user-generated content
by
IOANNA MALANDRINOU
Research Supervisor
Dr. Hüseyin Güngör
In partial fulfillment of the requirements for the degree of MASTER OF SCIENCE IN BUSINESS ADMINISTRATION
with Marketing Specialization
Name: Ioanna Malandrinou Student Number: 11375191
[2]
This document is written by Ioanna Malandrinou 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]
Acknowledgements
A rigorous but rewarding academic year finally comes to an end. It was a year full of knowledge, tight deadlines, teamwork and mixed emotions. This bunch of experiences urged me to challenge myself many times during the year but hopefully, I managed to turn them into courage and faith. Although hard work helped me develop myself a step further, success is never an individual achievement. Therefore, I would like to thank all people that helped me make a success of it.
In the first place, I would like to thank my supervisor Dr.Hüseyin Güngör, who has always been supportive and has given me insightful feedback and valuable guidance throughout the thesis process. His positivity instilled into me the willingness to keep up hard work while at the same time he encouraged me to work independently and rely on my own powers. In addition, I would like to thank my family, which has always stood by me at any cost, and has been my bulwark at every important step this year. Last but not least, I would like to thank all people that took part in the experiment setting, since their input contributed to the successful conduct of the current research.
[4]
Abstract
This study brings together the online and offline context and combines the online experience with purchase intentions in a brick and mortar environment. To investigate this connection, the impact of online customer service on offline purchase intentions is explored. This research introduces a new conceptual framework and examines the user-generated content as potential moderator in this relationship. The comparison of purchase intentions between online and offline context gives a better understanding of customers’ incentives during the decision-making process. By incorporating insights from 360 online users, five hypotheses are formed and a lab experiment is conducted. The results show that the online customer serviceaffects the offline purchase intentions significantly. This effect depends on how people perceive the quality of the service provided on a specific site. The conclusions engender useful contributions to both scientific and managerial world.
Keywords: customer experience, purchase intentions, online reviews, product type,
[5]
Contents
Acknowledgements………..………..3 Abstract………..4 1. Introduction………7 2. Literature Review………..……….92.1 Need Identification………...9 2.2 Customer Journey..………10 2.3 Customer Experience...………..11 2.4 Customer Insights………..….………...13 2.4.1 Preliminary Survey………...………13 2.4.2 Customer Responses.….………...14
2.5 Online Customer Experience (OE)………...15
2.5.1 Online Customer Service………...15
2.5.2 User-generated Content………...18
2.6 Product Type………21
2.7 Store Experience………..22
3. Theoretical Framework and Hypotheses………..……….24
3.1 Conceptual Framework……….24
3.2 Constructs & Hypotheses……...………...25
3.2.1 Online Customer Service & UGC……….25
3.2.2 Product Type & UGC………26
3.2.3 UGC & Purchase Intentions………..27
3.2.4 Product Type & Purchase Intentions……….28
4. Methodology……….…29
[6] 4.2 Variable Operationalization………..………31 4.3 Preliminary Analysis……….36 4.4 Correlation Analysis……….37 5. Results………..39 6. Discussion……….………47
7. Summary & Conclusions………..49
8. Contributions………53
9. Limitations & Further Research………55
10. References………..57
11. Appendices.………64
11.1 Appendix A: Tables……….………..……….64
11.2 Appendix B: Figures and Charts....……….84
[7]
1. Introduction
The continuous technology proliferation urges people to transform into active users day to day. As new channels become available to customers, they get access to new searching tools and acquire more power (Labrecque, vor dem Esche, Mathwick, Novak, & Hofacker, 2013). A full scope of new possibilities is now unfolding further for both companies and customers. On one hand, customers experience a new “journey” by means of different touchpoints in a complex multichannel retail environment (Maechler, 2016). Also, their channel choices differ through the buying process and over time (Gensler, Verhoef, & Böhm, 2012). Therefore, customers seek for a seamless experience whilst the traditional channels have not been obsolete (Pantano & Laria, 2012). On the other hand, companies have to exploit the potential of omnichannel marketing and carefully leverage the offered possibilities of different channels (Zhang et al., 2010).
That being said, multichannel management constitutes a new unexplored field that seems a good prospect. In addition, many companies are struggling with the new challenges they face up to in the current digital context. Decisions like opening a new physical store or investing money in online channels bring about key issues that executives have to deal with. Consider for example a company that promotes its products both offline and online. As customers are searching online for information they “generate and consume” content that affect other prospective customers upon their attitudes and purchase intentions. If at the same time managers desired to increase the company’s sales in its physical stores and knew what the real impact of user-generated content is, they could probably invest more money in online customer service depending on the product type. So, what managers can do to influence customers’
[8]
perceptions towards the product online that in turn could boost offline sales constitutes a challenging question. The current research aims to interpret the relationship between the perceived online customer experience and the offline purchase intentions in a multichannel environment.
To have a better understanding of this relationship, this study examines one important aspect of online customer experience: the online customer service. In particular, it focuses on its impact on customers’ purchase intentions in a brick-and-mortar environment. Since previous literature has paid much attention to store experience and how several atmospherics can influence offline sales, there is still little evidence about how the online customer experience can affect the offline sales. However, the offline activity is becoming susceptible to contemporary digital context and thus the comparison to purchase intentions online sheds light on the current research. This study also introduces a new variable- the user-generated content- that has not been examined yet as moderator in this relationship and hence it aims to extend the existing literature a step further.
Thus, the objective of this project is twofold. Search not only for new scientific insights but also for outcomes applicable to marketing strategies. Firstly, it develops a literature approach to point out the important dimensions of the current research and what previous studies have examined. Then it presents a theoretical framework and explains its main constructs and the relationship between them. Consequently, it proposes a research design method and a quantitative data analysis that will help the proposed hypotheses to be tested. Finally, it reaches statistical results and points out the potential scientific and managerial implications.
[9]
2. Literature Review
To begin with, this section sets the theoretical foundation for the core aspects of the current research. Since in a multichannel environment everything changes rapidly and customers’ demands increase sharply, firms have to understand the latent customer needs and find out the optimal way to serve them efficiently. People, at the same time, get involved in a multifaceted customer journey that encompasses a lot of choices and hard decisions. They desire not only a seamless experience but also maximum satisfaction and value over time. Therefore, the notions of need identification and customer journey are discussed. Consequently, the idea of customer experience is defined and how this is differentiated through the decision-making process. In order to elicit customer insights about people’s motives through the customer journey, a short preliminary survey is presented. The originated customer responses contribute to the formation of the proposed hypotheses, which will be tested via an experiment setting at a later stage of the research. Next, based on customer responses, two important aspects of the online customer experience are performed: the online customer service and the user-generated content. These in turn, will constitute the core object of the research. In addition, the product type seems to affect people’s purchase decisions in the online context based on previous literature, and hence it is discussed further. Finally, a comparison between the online and offline context takes place and the main research question is formed.
2.1 Need Identification
Undeniably people have multitude of needs that strive to fulfil in everyday life. In 1940, Abraham Maslow developed the Hierarchy of Needs model and its contributions have
[10]
remained solid over time. He categorizes people’s needs into five categories starting with the most important needs (biological and physiological needs) and through intermediate levels concludes to the upper level of needs called self-actualization needs (Taormina & Gao, 2013). This classification offers not only a deep understanding of customer’s motivation but also a useful tool for practitioners when training their workforce or developing marketing strategies. Need fulfilment is perceived by people as the way a task can be accomplished and involves a whole sensory experience that drives people’s priorities and actions (Strandvik, Holmlund, & Edvardsson, 2012). If someone considers the complexity of current age, the identification of the true customer needs is not a simple issue for companies. However, it is imperative need for firms to gain “customer needs knowledge”. This term refers to the degree that companies recognize the hierarchical needs of their customers and hence, try to predict customer satisfaction or ideally an efficient value delivery method (Homburg, Wieseke, & Bornemann, 2009).
2.2 Customer Journey
In order companies to meet customer needs they have to understand the customer journey and to identify the customers’ incentives that make them develop specific behavior through decision-making process. The emergence of more and more channels makes this process dynamic and the choice of the proper channel becomes an important decision for customers. This process includes two main stages: search and purchase phase (Konuş, Verhoef, & Neslin, 2008). If we extend these shopping phases to a more holistic view, we will understand that the customer journey illustrates more than just two phases in a row. In particular, it represents several steps whereby customers become attached to the company by means of an online or retail experience or a combination of
[11]
both (Richardson, 2010). In addition, this journey includes the attitudinal change of customers through multiple channels and thus can last for days (Konuş et al., 2008). Those companies that will be able to provide customers with a seamless experience across the available touchpoints, will finally manage to increase customers’ satisfaction and gain financial profits in the long run (Maechler, 2016). However, not all touchpoints are equally valuable for both customers and businesses. Among others, it depends on the product type or the offered service. For that reason firms have to locate and leverage the distinctive possibilities that some channels offer and bridge the gap between customers’ expectations and the delivered experience in those specific channels (Meyer & Schwager, 2007).
Although people get acquainted with the digital environment, they still invest time in visiting physical stores seeking for personal product evaluation or personal service. Customers anticipate an “omnichannel” experience when moving from channel to channel while at the same time the biggest gamble for companies remains the delivery of a frictionless experience in the current multichannel environment (Bianchi, 2016). Therefore, managers should address critical challenges to enhance customers’ value and accomplish a smooth channel synergy (Neslin et al., 2006). How people perceive this sort of experience will be discussed further down.
2.3 Customer Experience
“Customer experience” is a term that is commonly used by businesses and executives in the contemporary marketing context. But does that seem to be an abstract notion or a concrete idea in our minds? How can we really define it and understand what this phrase embraces? Customer experience is generated when firms intervene in the different phases of “customer journey” and try to create value for customers. This value
[12]
can be produced when companies understand this journey and what motivates people from one stage to another (Duncan, Fanderl, Maechler, & Neher, 2016). However, customers’ incentives vary between channels and customers’ perceived value might deviate from firms’ expectations (Kushwaha & Shankar, 2013). This sort of experience could also be the overall mindset of a company; a strategic tool that expresses how the whole business functions and why the company’s culture could be an important intermediate in value delivery (Pennington, 2016). Other definitions relate customer experience to the interactions between customers and products or services. This interaction includes some degree of involvement from the customers’ perspective and could be cognitive or affect based (Verhoef et al., 2009).
The value coming from this interaction is considered to be a calculus of customer’s expectations and what he or she finally receives from the company; not only the offering but also the stimulus being communicated by the firm (Gentile, 2007). Nevertheless, the final outcome for the customer is not always what the company expected to be, since companies do not have full control over the experiences they offer (Richardson, 2010). Customers’ perceptions are shaped by different actors, like their own product evaluation, their individual characteristics (heterogeneity) and the environment (Moe & Schweidel, 2012). Therefore, people inevitably react differently to different stimuli and adopt their own opinions for products, services and brands. Yet, what actually motivates people to get involved in that interaction is a reasonable question that demands a lot of attention. Digging for consumers’ incentives before the product purchase and for what makes peoples’ decisions fluctuate depending on context, a short preliminary research was conducted.
[13]
2.4 Customer Insights
2.4.1 Preliminary Survey
This research aimed to elicit customer insights and to clarify how customers do pre-purchase research when searching for specific type of products. Linking these insights to the notion of customer journey, the foundation behind customers’ decision-making process can be set. As a consequence, the upcoming customer responses constituted the base for the formation of the tested hypotheses and contributed to the current study by incorporating real data from average customers.
Seventy-seven participants took part in the survey by means of personal interviews and open questions or by answering online questionnaires. Their age ranged from minimum 18 to maximum 35 years old and the sample was composed by 34 males and 43 females. Customers were mostly asked open questions to cue in their minds general notions and then they were given the opportunity to drive the conversation by themselves based on their beliefs and past experiences. In particular, they discussed topics related to their preferences for different kind of products, their “way of thinking” during the decision-making process and any negative experiences they faced when purchased a product online in the past. Participants were asked to specify whether this incident finally affected their decision to buy the product online, offline or not at all. Also, other questions engendered data in terms of people’s consideration of online reviews and their influence on customers’ purchase intentions. All questions are available in Table 11.
[14]
Table 11: Preliminary Survey-Interview Questions
Q1: Do you prefer purchasing products online or offline?
Q2: Why do you prefer purchasing products mostly online/offline? Q3: What kind of products do you buy online (even rarely)?
Q4: Are you doing any pre-purchase research to get extra information about the product? Eg comparing prices, product characteristics, value-for-money, quality etc
Q5: Please describe briefly how you are doing this market research before purchasing the product of your interest.
Q6: Have you ever faced any bad experience when shopping online? Eg wrong order assignment, delivery delay, bad customer service etc. If yes: 1) please describe briefly and 2) did you finally purchase the product online, offline or not at all?
Q7: When searching for a product online, are you taking into account the online reviews? Q8: Under which circumstances do you trust these reviews the most?
Q9: Do they affect the channel you will choose to make your future purchase? Why? Q10: Does this effect depend on the product type you are looking for?
2.4.2 Customer Responses
The majority of interviewees admitted that before buying a product they tend to switch between channels to compare prices, to collect information and to find the optimal solution that fits with their interests. Almost every participant revealed that he or she takes into consideration the online customer reviews, which in turn affect their final decision on purchasing a product. They also claimed that the degree of this impact depends on the product type they are looking for. For example, if they are searching for a product that has specific technical characteristics and its quality depends on objective criteria (like electronics), the online reviews are more personalized and thus they are considered to have less impact on their product evaluation. On the contrary, when customers cannot be sure that the quality of a product is standard due to certain characteristics (like clothes), the impact of online reviews is greater. The reason in the last case seems to be that the product evaluation depends more on the personal
[15]
involvement in the product selection (eg a pair of trousers fits differently from customer to customer).
After a thorough analysis of the collected data, customer responses showed that the perceived online experience primarily depends on simplicity and personalized customer experience. Customers do not only seek for frictionless interactions with the retailer that will be time-efficient, but also they strive for bidirectional communication when they consider it is necessary. They know that they can easily switch between channels and find the same products or even better alternatives ensuring the best value-to-price relation. Therefore, they demand to be always at the center of the business asking for timely service delivery.
On top of that, previous scientific studies support the importance of online customer service in the perception of a positive online experience. As people become more educated and hence empowered, companies have to provide efficient customer relationship management if they want to build long-term relationships with their customers (Cetină, Munthiu, & Rădulescu, 2012). This fact is also justified by another research showing that the positive perception of service quality increases customer satisfaction and thus people experience web browsing more positively (Carlson & O'Cass, 2010).
Due to the aforementioned reasons, the online customer service remains an important dimension of online customer experience and the prevalent reasoning from customers’ perspective. As such, it is going to be developed further.
2.5 Online Customer Experience
(OE)
2.5.1 Online Customer Service
[16]
delivery. According to the Service Delivery Network concept, managers have to rethink established service practices and to realize their contribution to customer journey. Then, cooperating with third parties they have to offer a coherent experience to customers (Tax, McCutcheon, & Wilkinson, 2013). High-level service could lead to higher customer satisfaction and increased customer loyalty (Chang, Wang, & Yang, 2009). The transmission from customer’s satisfaction and trust to customer loyalty is empowered by word-of-mouth (WOM) (Kassim & Abdullah, 2010). Sheng & Liu, (2010), in their study attest that the service efficiency and the timely process of information fulfil customer needs effectively and thus customer satisfaction is being increased. As long as the responsiveness rate is high and the service representatives develop a solid personal contact with customers, the latter become attached to the company and more loyal over time (Yen & Lu, 2008). Other findings suggest that such service increases the positive impressions of the web site as well (Carlson & O'Cass, 2010).
Three factors that affect the perception of a good service quality the most are: the perceived risk, the web content and the user convenience. Based on previous research findings, less perceived risk leads to more positive attitudes towards the service quality (Udo, Bagchi, & Kirs, 2010). There are several practices that contribute to risk elimination. For instance, companies use certified servers that protect personal transactions, provide transparent contracts that describe the obligations and rights of all involved parties in detail or even correspond promptly to customers’ requests (Chen, Hsu, & Lin, 2010). Thus, managers have lots of instruments at their disposal ranging from IT functionalities to online assistance executed by service representatives. Consequently, they have to accommodate these mediums properly depending on specific target groups and customized customer needs (Chen et al., 2010). However,
[17]
the less perceived risk does not necessarily correlate with positive intentions to purchase or sense of satisfaction (Udo et al., 2010).
Web content refers to the degree to which the content being delivered is reliable
and meets customers’ expectations (Broberg, Buyya, & Tari, 2009). In particular, it comprises the necessary tools to help customers accomplish their transactions (Tan, Benbasat, & Cenfetelli, 2013). In order to become that feasible, executives have to adjust the online content to individual preferences and display it accordingly. The more personalized the content becomes, the more the perceived quality is augmented (Kardaras, Karakostas, & Mamakou, 2013). Web content is considered to be a significant function that attracts people and motivates them to keep searching online and share information. Normally, web content transmission goes hand in hand with service delivery. Τhe latter constitutes the means for delivering web content functions and amplifies the value of content in mind of customers (Tan et al., 2013).
Researchers regard convenience as the set of tailored activities and practices that facilitate online searching and save time for customers (Chen et al., 2010). These actions give customers the ability to search, compare and evaluate products and services in a time-efficient way (Jiang, Yang, & Jun, 2013). Findings from previous researches state that when convenience is accompanied by ease-of-use, it activates customers, makes interaction with the retailer meaningful to them or even customers are willing to recommend the website to others (Connolly, Bannister, & Kearney, 2010). Jiang et al., (2013), in their study, specify the key dimensions of convenience in online context: accessibility, searching/transaction options, pre- and post-purchase flexibility. If managers exploit the potential of these aspects, they will be able to deliver a smooth shopping experience and increase their customer base. Potential practical implications could be the availability of various payment alternatives, the provision of return
[18]
policies, the ease of product comparison or even the range of contact options displaying on the website (Chen et al., 2010). Hence, convenience constitutes a key aspect of perceived service quality (Hu, Brown, Thong, Chan, & Tam, 2009).
As previously discussed, it can be seen that customers’ evaluation is affected not only by perceived online experience but also by environment. Meaning that people are influenced by other peoples’ decision and the content that is distributed online. All these components are filtered intuitively in customers’ minds and nudge them to different purchase decisions each time. The nature of user-generated content and its effect on customers’ purchase intentions is developed below.
2.5.2 User-generated Content
In the current digital age, consumer behavior, internet technology advancements and marketing practices are interrelated. People increase their power to control valued resources in a digital context through production and dissemination of user-generated content (UGC) (Labrecque et al., 2013). The UGC refers to any source of information that is created and uploaded on online platforms by non-media experts (Bareau 2008). It varies from posting and online rating to visual content uploading (Presi, Saridakis, & Hartmans, 2014). This content (e.g. referrals and word-of-mouth) can significantly influence other customers’ purchase decisions (Albuquerque, Pavlidis, Chatow, Chen, & Jamal, 2012). This impact is moderated by the importance of the information given, the credibility of the source and the degree of compatibility between the user and the content generator (Herrero, San Martín, & Hernández, 2015). Seeing the user-generated content induces changes in people’s decision-making process, companies apparently have to adapt their marketing strategies and to train themselves in predicting the effect on future product sales (Liu et al., 2010). Furthermore, marketers can exploit consumer
[19]
empowerment, for example, by encouraging people to participate in the product development -that leads to a better engagement with the brand- and somewhat to digital content co-creation (Labrecque et al., 2013).
However, except for the typical users who tend to create digital content, it is also highlighted that there is a small customer share that corresponds to people who own the expertise and the experience to evaluate offerings without prejudice. Those customers are broadly recognized as experts. Both review types have a remarkable impact on customers’ evaluation and thus are being discussed in more detail below.
2.5.2.1 Expert Reviews
Experts shape their own attitudes towards products and services and perform them by means of online reviews. Their evaluation is perceived by customers as credible due to experts’ cognitive evaluation (Flanagin & Metzger, 2013). These reviews not only affect peoples’ purchase behavior, but also encourage them to increase their ratings and posts (Zhang, Zhang, & Yang, 2016). In essence, expert reviews enrich the information originated from different pipelines, like customer reviews, recommendation systems or product descriptions on websites (Mudambi & Schuff, 2010).
Although all sources contribute to customers’ evaluation, the volume of information people are exposed to, moderates this effect. People rely more on expert reviews, compared to customer reviews when the information volume is low. The opposite occurs when the information volume is high (Flanagin & Metzger, 2013). Another relationship between expert reviews and seller’s product description can be found as well. When the product is relatively cheap and/or the number of expert reviews is large, both sources of information are complements. Conversely, when the product is relatively expensive and the expert reviews are just a few, both sources of information
[20]
are substitutes (Chen & Xie, 2008). As a result, under different conditions, people switch between available information vents, favor specific review types and decide accordingly (Flanagin & Metzger, 2013).
Yet, how do people react to content generated by conventional users? How do they perceive the added value of this content, which motivates them to lean towards particular purchase decisions? The nature of online customer reviews and their impact on customer’s product evaluation are discussed below.
2.5.2.2 Online Customer Reviews
One of the most important types of user-generated content is the online customer reviews. Online reviews are either open-ended comments or in the form of numerical star ratings (from one to five stars) (Mudambi & Schuff, 2010). People post reviews to share their experiences with products and services and help other customers to evaluate them when searching online (Liu, Karahanna, & Watson, 2011). Namely, it is a means of expression of the strengths and weaknesses of a product and its overall evaluation (Decker & Trusov, 2010).
Online reviews are considered to have a solid impact on customers’ product evaluations. In particular, review extremity (when star rating is applied) and product type affect the helpfulness perception of the review and thus customers’ attitudes towards the product. The term “helpfulness” is used to describe how customers evaluate a review (Mudambi & Schuff, 2010). The degree of subjectivity and informativeness is another reason that affects the usefulness of reviews for customers. Reviews that include only subjective or only objective words are related to more online product sales. Reviews with a combination of both are considered to be more informative instead, and thus more useful to customers (Ghose & Ipeirotis, 2011). Another study shows that,
[21]
although both review valence and length have a positive impact on perceived helpfulness, the product type strengthens or loosens this effect. Therefore, the product category seems to play a moderating role (Pan & Zhang, 2011).
This effect has also been examined in many studies over time and as such, is strongly supported (Park & Lee, 2009). Multitude conceptual models analyze customers’ evaluations and classify products into two broad categories: hedonic and utilitarian products (Huang, Lurie, & Mitra, (2009). Both product categories are being presented as detailed below.
2.6 Product Type
Research suggests that hedonic products tap into symbolic values for customers and help them fulfil more abstract needs (Keller, 2001). Consumption of hedonic products offers pleasure and excitement to customers and enhances a whole sensational experience (Dhar & Wertenbroch, 2000). People interact with these products in a peripheral manner and try to elicit fun (Lu, Liu, & Fang, 2016). On the other hand, utilitarian products meet more practical needs (Dhar & Wertenbroch, 2000). This product type pinpoints the functional nature of consumption and focuses on product attributes (Keller, 2001). People experience those products in a cognitive way and extract benefits from its technical characteristics (Lu et al., 2016).
Many studies use other terms like experienced and search goods in place of hedonic and utilitarian products respectively. For that reason, this research applies both terms interchangeably in accordance with previous studies to avoid any meaning inconsistencies.
Except for the digital environment that continuously besiege people’s minds and their intentions, offline experience keeps surrounding everyday life in every aspect.
[22]
Therefore, people are constantly exposed to stimuli, which in turn are being amplified through personal contact and tangible experiences. A lot of studies have examined the impact of offline experience on customers’ motivations for further activity by now. To better introduce this impact theoretically, the general scene is being set briefly below.
2.7 Store Experience
There has been a lot of discussion in terms of how multiple atmospherics can create store experience that motivate for activity. Other research’s results support that factors such as store environment, store design and social parameters affect the perceived hedonic experience and hence customer loyalty (Muhammad, Musa, & Ali, 2014). Accordingly, the store ambient influences people’s personal attitudes and behavior towards the brand (Kumar & Kim, 2014). But while new technology possibilities are constantly emerging and customers get accustomed to online searching, are customers willing to abandon physical stores? Findings from previous studies point out that introducing immersive technologies can trigger customers’ motivation to select specific store as well as to improve their relationship with the brand and to boost purchases (Pantano & Laria, 2012). Consequently, the introduction of technological elements in brick-and-mortar environment has a huge potential for both customers and companies. The latter not only can refine the shopping process, but also by stimulating entertainment cues, they can attract more customers (Pantano & Naccarato, 2010). Regardless of how the firm configures a store, other elements contribute to customers’ attitudes as well. For instance, information recall, degree of personal involvement, customers’ attributions or personal predispositions affect their final decisions (Puccinelli et al., 2009). Yet, when does a retailer have to create an exciting store environment? Previous findings state that arousal has different effect on
[23]
pleasantness and subsequently on shopping behavior. This effect depends on whether the prospective customer is task- or recreational-motivation oriented (Kaltcheva & Weitz, 2006).
To sum up, there are quite a lot of studies referring to offline customer experience that mostly talk about how companies can create the necessary conditions that energize people and encourage their purchase intentions. However, offline customer experience is not limited to store atmospherics. Overall, scholars have obtained deep knowledge in relation to face-to-face context over time. But still little attention has been paid to perceived customer experience in a digital context. Additionally, previous studies examined the effect of web service quality on customer satisfaction and attitudes towards the website. Also others concluded that high-quality service leads to customer satisfaction and loyalty. However, the impact of online customer service on customer purchasing intentions offline has not been tested yet. Especially, a comparison to purchase intentions online will demystify customers’ intentions to switch between online and offline channels. Furthermore, this study aims to add one new moderating variable in this relationship and thus extend the conceptual framework beyond the existing literature. In particular, the moderating role of the user-generated content is examined. Formulating the research question, it results in the next statement:
“The impact of online customer service on offline purchase intentions moderated by the user-generated content”
[24]
3. Theoretical Framework and Hypotheses
3.1 Conceptual Framework
In order to test the effect of online customer experience on customers’ purchase intentions offline, the quality of the online customer service is being modified in the experiment setting. The conceptual framework contains different variables that are interrelated. The independent variable is the online customer service, while the dependent ones are the purchase intentions in online and offline context. In this study, the moderating role of the user-generated content between the online customer service and the purchase intentions is being tested. In particular, the product user rating is being examined as possible moderator. The product user rating is chosen as part of the broader customer review context, since reviews made by average customers outweigh expert reviews numerically. Another independent variable that also acts as potential moderator
Online Customer Service Product User Rating Product Type
Control Variables: Age, Gender, Time spent
Purchase Intentions on Specific Site
Purchase Intentions on Different Site
[25]
in this relationship is the product type. Especially, a distinction between hedonic vs utilitarian products takes place to test whether the product category influences the decision making process. Other control variables are being examined as well, like the age, the gender of participants and the time they spend when searching for a product online and they are interested in purchasing it immediately.
3.2 Constructs & Hypotheses
To better understand the concept of the main constructs in the given framework, the relationship between them is described below. Therefore, the theoretical background behind them is defined and the proposed hypotheses are formed.
3.2.1 Online Customer Service & UGC
Everyday more and more people interact with online retailers and build a relationship over time that can be valuable for both the company and the customers. These transactions include pre-purchase research, order transactions and online customer service (Rose, Hair, & Clark, 2011). People’s active participation in service co-creation brings about behavioral and economic outcomes for the firm, like positive or negative purchase intentions or word-of-mouth (Bolton & Saxena-Iyer, 2009). For that reason, companies have to track and analyze the user-generated content not only on their sites but also on theirs competitors. This practice can give a competitive advantage when firms strive to deliver efficient customer services online (He, Zha, & Li, 2013). In most cases, the online reviews are posted after product consumption or post-purchase service experience. Two solid drivers that urge customers to create a negative review are altruism and vengeance. For instance, when customers are possessed by vengeance they tend to share more negative content online (Presi et al., 2014). In general, the positive
[26]
online customer service brings about a sense of trust and satisfaction. This sense is conveyed through word-of-mouth (WOM) and forms a prerequisite for high return rates and positive purchase intentions (Kassim & Abdullah, 2010).
3.2.2 Product Type & UGC
The user-generated content can influence the online sales not only in relation to the word-of-mouth (WOM) source but also due to the product type and the degree of involvement. That is, for high-involvement products the retailer’s hosted (internal) WOM cannot significantly influence the online sales, because customers pay more attention to external WOM vents for that kind of products (Gu, Park, & Konana, 2012). Especially for niche products, online reviews tend to be more influential since customers search online for alternative information sources (Zhu & Zhang, 2010). This fact leads to H1.
H1. When the online customer service is perceived as positive, purchase intentions online are greater when searching for hedonic products.
For experienced products, reviews with extreme star ratings (1/5 or 5/5 stars) are less helpful for customers’ product evaluations and therefore it is implied that these reviews affect potential online sales less than what moderate reviews do for the same product category (Mudambi & Schuff, 2010). In addition, for experienced goods the volume of reviews is more important than search goods, because the product evaluation depends on subjective criteria and personal interaction with the offering (Cui, Lui, & Guo, 2012). This leads to H2.
H2. When the online customer service is perceived as positive, purchase intentions offline are greater when searching for utilitarian products.
[27]
Park & Lee, (2009), show that for experienced goods the effect of negative online reviews is greater than for positive ones compared to search goods. This fact is also justified by the research of (Huang et al. (2009),which points out the bigger effect of reviews on consumer research and purchase intentions for experienced goods. This leads to H3.
H3. When the online customer service is perceived as negative, there is no difference in purchase intentions across the two product categories.
3.2.3 UGC & Purchase Intentions
Online communities have undeniably attracted too much attention due to the rapid diffusion of the user-generated content. The contemporary technological evolution allows customers to create and share materials online with significant effects on other customers’ product evaluations. In particular, customers’ purchase intentions are influenced by perceived usefulness of the online reviews that in turn affects customers’ trust in online retailers (Elwalda, Lü, & Ali, 2016).
Another study examines the impact of online reviews on purchase intentions moderated by the level of consumer’s conformity. Conformity resides in the proposition that the majority of people should express the most valuable point of view (Huang, Zhang, Hui, & Wyer, 2013). Consequently, conformists tend to accommodate their behavior in order to “fit” with the others. Therefore, this research supports that positive reviews have a greater impact on intensions to purchase for conformists. Also, the review volume seems to influence the perceived persuasiveness between conformists and non-conformists. Specifically, positive reviews are deemed as more persuasive for non-conformists (Tsao, Hsieh, Shih, & Lin, 2015).
[28]
Another reason that considerably affects purchase intentions is the degree at which customers consider a prospective online review credible. The more credible they think it is, the more they trust it and hence lead to positive intentions to purchase (Lee, Park, & Han, 2011). This fact combined with customer responses from paragraph 2.4.2 supporting that people who are influenced by negative reviews tend to avoid purchasing the product online and probably turn to brick and mortar solutions, leads to H4.
H4. When the online customer service is perceived as positive, purchase intentions offline are greater when the user-generated content is negative.
In addition, when the online service is perceived as negative and the online reviews are negative, people seem to lower their expectations for this specific product or the corresponding medium (results paragraph 2.4.2) and thus their intention to purchase decreases. Taking into account this fact and the aforementioned theoretical background, H5 is formed.
H5. When the online customer service is perceived as negative and the user-generated content is negative, there is no difference in purchase intentions.
3.2.4 Product Type & Purchase Intensions
Previous research shows that utilitarian and hedonic value-value from the whole interaction with the product before and after purchase-are connected to positive purchase intentions. Yet, this association is moderated by the level of perceived risk (Chiu, Wang, Fang, & Huang, 2014). Although customers spend equal time when searching online for both experienced and search goods, the way they browse and intend to purchase remains different for both product categories. In other words, customers of experienced offerings are inclined to purchase the product from the
[29] H1 : H2 : H3 : H4 : H5: Hed Uti Hed/Uti -UGC -UGC
primary product source, in contrast to buyers of search goods who prefer “free riding”; shopping from other retailers (Huang, Lurie, & Mitra, 2009b). All hypotheses are being presented graphically below and in Table 10 in Appendix.
Hypotheses
Online Customer Service: OCS User-generated content: UGC
Intentions Online: IntOn Hedonic: Hed
Intentions Offline: IntOff Utilitarian: Uti
4. Methodology
4.1 Research Design
In order to examine the impact of online customer service on customers’ purchase intentions and test the given hypotheses, a 2x2x2 between participants experiment is conducted. The experiment has been designed using Qualtrics software and is available
OCS OCS OCS OCS OCS IntOn IntOff IntOn IntOff IntOn IntOff IntOn IntOff IntOn IntOff Figure 6
[30]
in two languages-English and Greek. The relevant link was posted online on different Facebook groups to attract respondents who are familiar with the online environment, implying that the sample fits properly to the requirements of the current research. In addition, a lottery for a 50-euro cash reward was organized to increase the response rate and thus all data were collected within a week. All participants were informed about the objective of the experiment and were asked for their consent before the experiment had started. All data were handled confidentially and all responses remained anonymously.
The experiment consists of eight different conditions and each participant was randomly assigned to one of them to avoid any carry over effects. In each condition, the quality of online customer service and the product user rating are manipulated accordingly for two product types: hedonic and utilitarian. The online customer service and the product user rating can be either positive or negative. All treatments are presented in Table 1 a & b below and on Questionnaire (Appendices) in detail.
Table 1a: Experiment Conditions
Hedonic
Negative Rating C1 C3
Positive Rating C2 C4
Negative Service Positive Service
Table 1b: Experiment Conditions
Utilitarian
Negative Rating C5 C7
Positive Rating C6 C8
Negative Service Positive Service
Firstly, participants had been invited to evaluate the average user rating of a product and then they were provided with either a positive or a negative scenario depending on the condition. Then they were encouraged to imagine that they are interested in
[31]
purchasing the specific product as soon as possible and thus remaining concentrated before answering the upcoming questions was important for the successful process of the experiment. After they had taken into account the product user rating and the relevant scenario they were asked to indicate how likely it was that they would purchase the product online or in a physical store. Finally, they were asked to provide some demographics like gender and age and to declare the average time they usually spend every time they search online for a product they are willing to buy immediately. Leaving their e-mail address was optional. The whole questionnaire is available in Appendices.
4.2 Variable Operationalization
Online Customer Service
According to previous literature, there are three main factors that affect the efficient online service delivery: the perceived risk, the web content and the user’s convenience (Udo et al., 2010). All three have been incorporated in the given scenario and have been adjusted accordingly in order to manipulate the provided online customer service. The latter is either positive or negative.
For instance, one practice that contributes to risk elimination is the prompt correspondence to customers’ requests and the online assistance provided by customer service representatives (Chen et al., 2010). In the current experiment, people contact the service department via online chat, which is available 24h/day and it is guaranteed that customers get a response within 5 minutes. In addition, web content goes hand in hand with the online service. The brief product description and the technical characteristics being visible on photos, make the product comparison to other products simple and the information given, accessible in people’s mind. In the experiment,
[32]
therefore, customers ask for some help in terms of the product details presented on photos and hence the online service makes the web content meaningful in customers’ product evaluation (Tan et al., 2013). Finally, convenience constitutes a key function of perceived online customer service. Providing pre-purchase flexibility and a wide range of payment alternatives makes online customer service a competitive asset when firms aiming to attract and retain new customers (Chen et al., 2010). Therefore, in this experiment, customers ask whether they can book the product over the phone and make the money transfer tomorrow. The payment via bank account is included as alternative payment method.
By using Descriptive statistics, a manipulation check takes place to assure that the online customer service is manipulated as it was intended at the beginning. Manipulation is based on one of the proposed customer satisfaction measurements in the study by Peterson & Wilson (1992). According to this measure, after participants have been exposed to a specific stimulus, they are asked to indicate in a five Likert scale the level of their satisfaction. Then, the majority of responses (in percentages) create the “top box” of responses and the manipulation is approved or not. Thus, to test whether the online customer service is perceived as positive or negative, participants were asked to indicate in a five-Likert scale from 1 “Not at all” to 5 “Very much” how much satisfied they felt with the customer service. The frequency calculations in Table 2a (Appendix A) show that manipulation exists. When people were provided with a positive scenario, 89% declared that they felt “somewhat” and “very much” satisfied with the service and only 11% were undecided or not satisfied. From this percentage, only 2% were not satisfied at all. When participants were provided with a negative scenario instead, 81% indicated that they were not satisfied with the service and only
[33]
2% were satisfied very much. Manipulation results for online customer service are graphically presented at the following charts.
Product User Rating
Participants were asked to evaluate the average product user rating of either a hedonic or a utilitarian product. According to the findings came from the preliminary survey and the customer responses (paragraph 2.4.2), when people are interested in buying a product online they mostly prefer purchasing electronics and apparel. Therefore, for both product categories a headphone set and a pair of shoes were chosen respectively,
2 3 6 34 55 N O T A T A L L N O T R E A L L Y U N D E C I D E D S O M E W H A T V E R Y M U C H PE R CE N T % CUSTOMER SATISFACTION
POSITIVE
44 37 7 10 2 N O T A T A L L N O T R E A L L Y U N D E C I D E D S O M E W H A T V E R Y M U C H PE R CE N T % CUSTOMER SATISFACTIONNEGATIVE
Chart 1: Manipulation check for positive online customer service
[34]
since those products are considered representative of the different product types. So, participants were given a photo of a product, presenting a brief description of it, its price and the average user rating in the form of star rating. Then they were asked to indicate whether the review is positive, negative or neutral.
According to previous studies, extreme star ratings can have different influence on people’s purchase decisions online depending on the product category. In some cases, they may present a milder effect as well (Mudambi & Schuff, 2010). For that reason, and to extend the theoretical framework a bit further, in the current experiment moderate reviews are assigned to the products. For both product types, the photos presented are identical and the number of ratings is constant. For example, for experienced goods the volume of reviews is more important than search goods, because the product evaluation depends on subjective criteria and personal interaction with the offering (Cui et al., 2012). In that way, any potential parameters that differ between photos and could affect participants’ review evaluation are isolated.
For the product user rating, Descriptive statistics applied again and a manipulation check took place. Manipulation is based on review classification in agreement with the study by Pang & Lee (2005). In their study, people evaluated movie star ratings (from 1 to 5 stars) as more or less positive depending on star number. After all data had been collected, they distinguished responses into three classes: positive, mild and negative. Therefore, they could know whether people perceived the movie’s review valence as positive, negative or neutral. Following the same rationale, frequency checking in Table 2b (Appendix A) shows that manipulation exists indeed. When positive product user rating is being presented, 86% of participants indicate that the online review is positive and only 2% consider the review as negative. The rest 12% declare that the average user rating is neutral. On the other hand, when negative product
[35]
user rating is being resented, 80% of participants indicate that the review is negative and only 9% consider the review as positive. The rest 11% declare that the average user rating is neutral. Manipulation results for product user rating are presented at the following pie charts.
Purchase Intentions
In order to measure the effect of online customer service on customers’ purchase intentions, a measure based on study by Taylor, Houlahan, & Gabriel (1975), was used. This measure specifies customers’ intentions to purchase a specific product at different product development stages. In their study, they used five Likert scales to predict the likelihood that a new product would be purchased after customers had experienced it. Accommodating the same measurement to the current research, participants were asked to indicate in a five Likert scale from 1 Strongly Disagree to 5 Strongly Agree, how likely it was that they would purchase the product on that specific site, on a different site and in a physical store. In order a future comparison between customers’ intentions to be feasible, it was compulsory participants to fill out all question fields.
Positive 86% Negative 2% Neutral 12%
POSITIVE
Positive 9% Negative 80% Neutral 11%NEGATIVE
[36]
4.3 Preliminary Analysis
After all data had been collected, the SPSS Statistics program was used to export the data from Qualtrics software and to prepare them for further analysis. As long as manipulation checks are valid, all data related to customer satisfaction were accumulated under the Online Customer Service (OnCuSe) variable and the numbers 1 and 2 were assigned to positive and negative scenarios respectively. Likewise, all data related to customers’ rating evaluations were gathered under the Product User Rating (ProUserRa) variable and the numbers 1 and 2 are assigned to positive and negative user star ratings respectively. To distinguish the product types (ProType), the numbers 1 and 2 were assigned to hedonic and utilitarian products respectively. Purchase intentions online and offline were divided into three different variables; Purchase Intentions on that Specific Site (PurIntSpeciS), Purchase Intentions on a Different Site (PurIntDiffeS) and Purchase Intentions in a Physical Store (PurIntPhySt).
From 426 answered questionnaires received, 360 were totally completed and thus only those were included in the analysis. Descriptive statistics showed that from the 360 filled responses, 140 participants were males (38.9 per cent) and 220 were females (61.1 per cent) (Table 3-Appendix A). The age (in years) of participants ranged from 17 minimum to 54 maximum (M=25.95 & SD=5.91) and the time spent (in minutes) from 5 minimum to 180 maximum (M=49.78 & SD=43.4).
For one, a “missing value” check was done but no missing values were found. For nominal data (online customer service, product user rating, product type and gender), dummy variables were created by recoding into different variables; OnCuSeDum, ProUserRaDum, ProTypeDum and GenderDum. For online customer service, the values of 0 and 1 were assigned to positive and negative scenario respectively. For product user rating, the values of 0 and 1 were assigned to positive
[37]
and negative product review respectively. For product type, the values of 0 and 1 were assigned to hedonic and utilitarian products respectively and finally, for gender, the values of 0 and 1 were assigned to males and females respectively.
For continuous variables (Age, Time spent and Purchase Intentions), normality checks were done and Kolmogorov-Smirnov & Shapiro-Wilk tests were applied. These tests combined with Skewness-Kurtosis output, Histograms and Q-Q plots reached to the violation of normality only for control variables. In particular, for both age and time spent, 0 was outside the bounds of the 95% confidence interval and the significance level was p=.000<.05 (Table 12 & 13-Appendix A). As the results showed an extremely positive skewness, the transformation equation of X*=1/X was applied. Then another check took place to locate any potential extreme values or outliers that after normalization could still affect the data distribution (Table 17 & 18-Appendix A). However, observing that after normalization the 5% Trimmed Mean was slightly different from the Mean and the Q-Q plots depicted an almost normal distribution, this could not affect our results and thus none element was excluded from the analysis (Pallant, 2010). After normalization, the variables of Age and TimeSpent renamed into NormalAge and NormalTimeSpent respectively.
4.4 Correlation Analysis
Before the main analysis, a correlation matrix of all the main constructs of interest was compiled to observe any important relationships among them. The Pearson correlation results show a significant, positive and strong correlation between the online customer service and the purchase intentions on that specific site (r=.71, p<.01). On the contrary, there is a significant, negative and moderate correlation between the online customer service and the purchase intentions on a different site (r=-.47, p<.01). Also, there is a
[38]
significant, negative and small correlation between the online customer service and the purchase intentions in a physical store (r=-.29, p<.01). Additionally, a significant, negative and small correlation found between the product user rating and the purchase intentions on that specific site (r=-.11, p<.05). However, this correlation actually expresses a positive relationship between the product user rating and the purchase intentions on that specific site. In general, the minus sign of the correlation coefficient means that when the product user rating increases, the purchase intentions on that specific site decrease. Yet this increase in product user rating refers to a numerical increase from 1 to 2, showing a transmission from a positive to a negative product review condition. Therefore, when the product user rating becomes negative, the purchase intentions on that specific site become less, expressing a positive relation. When also the product user rating is positive, the purchase intentions on that specific site increase.
Table 4: Means, Standard Deviations, Correlations
Variables M SD 1 2 3 4 5 6
1. Online Customer Service
3,08 1,57 1
2. Product User Rating 1,46 0,49 -0,025 1
3. Product Type 1,51 0,50 -0,005 -0,003 1
4. Purchase Intentions on that Specific Site
3,12 1,22 0,712** -0,116* 0,042 1 5. Purchase Intentions on a Different Site 3,29 0,89 -0,475** -0,075 0,053 -0,455** 1 6. Purchase Intentions In a Physical Store 3,37 0,95 -0,291** -0,021 -0,004 -0,300** 0,400** 1
*. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed).
[39] C1
C2 C3
5. Results
The interesting point in correlation results is that between the online customer service and the product user rating there is almost no correlation (r=.02). This fact alerts for no interaction between them and thus no moderation effect. In addition, almost no correlation was found between the online customer service and the product type (r=.005). In order to test the existence of any moderation effect, Process Model 11 by Andrew F. Hayes (2015) was ran.
Process Model 1
Figure 1
In the dependent variable field, the PurIntSpeciS variable was used, since this variable is the only one, which performs a significant correlation with the product user rating. The moderation analysis (Table 5) indicates that the coefficient C3 corresponding to the interaction effect, is not significant (p>.05) and hence the moderation hypothesis for product user rating is rejected. As a result, the average user rating does not affect the strength of the effect of online customer service on purchase intentions on that specific site.
1 “In Process Model 1, two conditional effects of X on Y can be formally compared with a statistical test. Evidence
of moderation of X’s effect on Y by M leads to the corresponding claim that any two conditional effects of X on Y for different values of M are different from each other”. Andrew F. Hayes, Supplementary PROCESS Documentation (2016), p.520
X M XM
[40]
Table 5: Process Model 1:Product User Rating-Outcome variable: PurIntSpeciS
Coefficient SE t p Intercept i1 3,12 0,05 61,04 p<0,001 OnCuSeDum (X) c1 -0,33 0,10 -3,24 0,001 ProUserRaDum (M) c2 -1,47 0,10 -14,46 0,000 ProUserRaDum*OnCuSeDum (XM) c3 -0,07 0,20 -0,33 0,738 R²=0,378 p<0,001 F(3,356)=76,825
The potential moderation effect was examined for the product type as well. The moderation analysis (Table 6) indicates that the coefficient C3 is not significant (p>.05) and hence the moderation hypothesis for product type is rejected. As a result, the product type does not affect the strength of the effect of online customer service on purchase intentions on that specific site. In the process options, heteroscedasticity consistent SEs was chosen, assuring in that way the heteroscedasticity assumption of the analysis. This assumption requires that the variance of the error terms is constant across all values of the independent variables (Pallant, 2010).
Table 6: Process Model 1:Product Type-Outcome variable: PurIntSpeciS
Coefficient SE t p Intercept i1 3,12 0,05 60,14 p<0,001 OnCuSeDum (X) c1 -1,47 0,10 -14,20 0,000 ProTypeDum (M) c2 0,14 0,10 1,35 0,178 ProTypeDum*OnCuSeDum (XM) c3 -0,07 0,20 -0,32 0,747 R²=0,363 p<0,001 F(3,356)=68,428
Consequently, a Factorial ANOVA is performed. The acronym Factorial ANOVA stands for Factorial Analysis of Variance and is one of the major statistical techniques for quantitative data analysis. It allows studying the main effect of at least two categorical independent variables on one continuous dependent variable. This type of analysis can also test whether there is any possible interaction between the independent
[41]
variables (Pallant, 2010). Therefore, a Factorial ANOVA is applied to investigate the effect of online customer service, product user rating and product type on Purchase Intentions on that Specific Site. The analysis is controlled for Age, Gender and TimeSpent. The same procedure is repeated for Purchase Intentions on a Different Site and Purchase Intentions in a Physical Store.
Effect on Purchase Intentions on that Specific Site
The results show that there is a significant effect of online customer service on purchase intentions on that specific site, F(1,349)=209.04, p<.001, η²=.37. In addition, there is a significant main effect of product user rating on purchase intentions on that specific site F(1,349)=10.06, p<.01, η²=.03. Also, there is not any significant effect of product type on purchase intentions on that specific site F(1,349)=2.03, p=.15, η²=.006. From results, it is reconfirmed that there is no significant interaction effect between online customer service and product user rating on purchase intentions on that specific site F(1,349)=.19, p=.65, η²=0. There is no significant interaction effect between the online customer service and the product type as well F(1,349)=.008, P=.92, η²=0. Finally, there is a significant three-way effect of online customer service, product user rating and product type on purchase intentions on that specific site F(1,349)=5.93, p<.05, η²=.01. Finally, none of the control variables seems to have a significant effect on purchase intentions on that specific site. Analysis’s results are presented in Table 7 a & b.
[42]
Table 7a: Factorial ANOVA- DV: Purchase Intentions on that Specific Site
SS DF MS F η² Sig
Gender 0,09 1 0,09 0,10 0,00 0,75
Age 0,50 1 0,50 0,54 0,00 0,46
Time Spent 0,03 1 0,03 0,03 0,00 0,85
Online Customer Service 194,98 1 194,98 209,04 0,37 0,00***
Product User Rating 9,89 1 9,89 10,60 0,03 0,00**
Product Type 1,89 1 1,89 2,03 0,00 0,15
Online Customer Service*
Product User Rating 0,18 1 0,18 0,19 0,00 0,65
Online Customer Service*
Product Type 0,00 1 0,00 0,00 0,00 0,92
Product User Rating*
Product Type 0,06 1 0,06 0,06 0,00 0,79
Online Customer Service* Product User Rating*
Product Type 5,53 1 5,53 5,93 0,01 0,01*
Error 325,52 349 0,93
Total 4046,00 360
Note. Statistical Significance: *p<.05; **p<.01; ***p<.001
Table 7b: Factorial ANOVA- Estimated Marginal Means
M SE LB UB
Online Customer Service
Positive 3,87 0,07 3,72 4,01
Negative 2,39 0,07 2,25 2,53
Product User Rating
Positive 3,29 0,07 3,16 3,43
Negative 2,96 0,07 2,81 3,11
Covariates appearing in the model are evaluated at the following values: GenderDum = ,61, Age Normalization = ,0403, Time Spent Normalization = ,0350.
The estimated marginal means reveal that the purchase intentions on that specific site show a tendency to be higher in the positive scenario (M=3.87) compared to the negative one (M=2.39). Furthermore, the estimated marginal means reveal that the purchase intentions on that specific site show a tendency to be higher in the positive
