27 Mar Customer Engagement
Customer engagement and online reviews
This study aims at understanding the role of customer engagement in writing online reviews by shoppers with speciﬁc focus on mobile devices for shopping. Mobile devices are becoming ﬁrst screen for the customers and are being used by marketers to have interactive communication making it more suitable for building customer engagement. The research in this space however is in a very nascent stage. Current study is one of the ﬁrst few empirical studies exploring the role of customer engagement in writing online reviews. The study explores mediating role of customer engagement in satisfaction – online review intention and trust – online review intention relationships. Further moderating role of trust and satisfaction levels in customer engagement – online review intention is explored. This study contributes to marketing literature in the space of customer engagement, online reviews and mobile shopping behaviour. Further, this study provides a framework to managers for mo- tivating the customers in writing online reviews. Also recommendations for retailers in exploiting customer engagement on mobile platforms are provided to address merchants and advertisers for better management of a new technology.
Online reviews are becoming increasingly important sources of in- formation for shoppers impacting as much as 20–50% online purchase decisions (Mathwick and Mosteller, 2016). Potential shoppers often check online reviews posted by other customers who bought and used those products. Such reviews give them a glimpse of purchase and usage experience of other users. These reviews are considered to be more credible by shoppers as compared to seller promoted advertisements (Bickart and Schindler, 2001; Godes and Mayzlin, 2004). Re- searchers have found that shoppers speciﬁcally log on to online web- sites like Amozon.com to check reviews as part of their product evaluation journey (Bughin et al., 2010; Simonson and Rosen 2014).
Online reviewers serve the role of information service provider for the potential shoppers by ‘‘shaping how customers serve themselves before, during, and after purchase’’ (Ostrom et al., 2015). Such reviews play crucial role in inﬂuencing shoppers’ choice of products, services (e.g, camera, airline, hotel) as well as retailer (e.g., Amazon.com, Make-mytrip.com). Online reviews, therefore impact the business of several ﬁrms in a multi-sided platform like online marketplaces. Researchers recognize that by posting online reviews, customers derive great social value within the community (Balasubramanian and Mahajan, 2001). The act of reviewing is, therefore, being considered as one of the most inﬂuential expressions of customer engagement (Mathwick and Mosteller, 2016).
Customer Engagement is a psychological state of mind that leads to frequent interaction with the focal object (e.g., a brand or a medium). Customer engagement is a long term relationship that arrives out of emotional as well as utilitarian motivational drivers. Emerging literature in the domain indicates that customer engagement may lead to several favourable outcomes for brands and ﬁrms beyond repurchase intentions. Among other things, these outcomes may include posting likes and reviews on online and social media. Customer engagement is a signiﬁcant construct in online and social business environments (Brodie et al., 2013; Hollebeek et al., 2014) and may oﬀer valuable insights in customers’ propensity to post online reviews. The act may be perceived by customers as making them inﬂuential in their relationships with the brands, and the online organisations.
With online reviews becoming key inﬂuencers customer purchase decisions, researchers are exploring drivers of online reviews in various settings. There is a strong stream of literature indicating that high level of satisfaction leads to customer loyalty exhibited in the form of re- purchase, referrals and favourable word of mouth (Anaza and Zhao, 2013; Anderson and Sullivan, 1993; Shankar et al., 2003). Within the speciﬁc context of online media, satisfaction has been found to have a signiﬁcant impact on customer propensity to post online reviews (Maxham III and Netemeyer, 2002; Ranaweera and Prabhu, 2003). Another construct closely associated with satisfaction and widely ex- plored, as an inﬂuencer in posting online reviews is the trust in the brand or the ﬁrm. Trust has been found to have signiﬁcant impact on customers’ propensity to stay with the ﬁrm and provide favourable reviews (Harris and Goode, 2004; Kim et al., 2009; Ranaweera and Prabhu, 2003). While there is established literature on role of satisfaction and trust in motivating people to post online reviews (Anderson and Sullivan, 1993; Gvili and Levy, 2016; Kim et al., 2009; Oliver, 1980; Shankar et al., 2003), there has been increasing interest in under-explored role of customer engagement in online reviews (Kim et al., 2013; Rossmann et al., 2016). Engaged customers are likely to visit retailers website frequently for reasons beyond immediate purchase need. Such engaged customers are likely to be more emotionally invested with the retailer and therefore more likely to respond to re- quests for writing reviews about their purchases.
Customer interaction with retailers’ websites for shopping and other related activities however are increasingly shifting to mobile devices.
With rise in penetration of smart phones, and changing trends in society, mobile phones have become personal companion for the users. Increasingly, customers are using mobile devices for accessing online content more frequently that through personal computers. Especially in emerging economies like India, where access to computers is limited, and mobile phone penetration is high, mobile devices may play signiﬁcant role in generating engagement and online reviews. Further, people in emerging economies spend considerable time in traveling using public transport where mobile phones act both as a productivity enhancer as well as a source of entertainment. Shopping, browsing and writing reviews on the go using mobile devices, therefore, may be a source of instant gratiﬁcation. The behaviour of these customers in the said context is likely to be of interest for online retailers, brands as well as merchants selling through online retailers. Such usage patterns are likely to enable ﬁrms to interact with customers using their mobile devices more frequently and eﬀectively as compared to other modes of communication. Mobile devices are therefore likely to be more eﬀective in building customer engagement with the ﬁrm. With that perspective, the study looks at exploring the role of satisfaction, trust, and customer engagement through mobile phones in posting online reviews. The investigation takes a speciﬁc instance of mobile shopping applications for fashion and lifestyle products to achieve the research objective. This research oﬀers a contribution to academia in the form of an addition to the body of knowledge in three emerging streams in marketing literature – customer engagement, mobile shopping and online reviews.
In the following section, relevant literature on customer engage- ment, online review, satisfaction and trust is provided to derive the proposed model and testable hypothesis. This is followed by detailed research methodology, data analysis, results, discussion and implica- tions. This study demonstrates that customer engagement mediates the relationships satisfaction → online review intention and trust → online review intention. Further, trust moderates the relationship between customer engagement and online review intention. Implications for the practice include insights for retailers, merchants, and brands.
2-Theory and Hypothesis
2.1-Deﬁning customer engagement and online reviews
Customer engagement is a state of mind of being emotionally in- vested with the focal object (brand or medium), which leads to customers’ frequent interactions with the focal object. As the literature on customer engagement is still in nascent stage, the operationalization of the same is still evolving and is yet to converge. While there are several diﬀerent conceptualizations, researchers’ views agree that customer engagement (CE) is a psychological state that leads to frequent inter- actions with the focal object (brand or medium) that go beyond transactional motive of merely a purchase. Also, researchers have proposed that engagement leads to several outcomes beyond repurchase, including posting likes, reviews and participation in co-creation of products and services (Brodie et al., 2011a,b; Calder et al., 2009; van Doorn et al., 2010). Scholars have conceptualized engagement as a multidimensional construct (Bowden, 2009b; Calder et al., 2009; Hollebeek, 2011; Mollen and Wilson, 2010) covering cognitive, emotional and behavioral dimensions. Table 1 gives a snapshot of recent literature on the conceptualization of customer engagement by scholars.
There is a strong body of conceptual literature emerging in the ﬁeld of customer engagement construct, however the empirical studies where scales have been developed and tested are relatively few (Calder et al., 2013; Hollebeek et al., 2014; Sprott et al., 2009; Vivek et al., 2014; Zheng et al., 2015). Amid the emerging empirical literature, Calder et al. (2013) looked at customer engagement with a focal medium, Hollebeek et al. (2014), and Sprott et al. (2009) investigated engagement with the brand while Zheng et al. (2015) investigated engagement in brand community and social networking sites. A scale developed by Vivek et al. (2014) took a more broad base view at engagement with any focal object ranging from a brand to an organization to a medium where they looked at a three-dimensional view of CE, including conscious attention, enthused participation, and social connection. As this study deals with mobile phones as a medium for shopping, conceptualization around experiencing a medium as suggested by Calder et al. (2013) was deemed to be most appropriate for this study.
Calder et al. (2009) proposed that engagement comes from experiencing a medium in a certain way. They deﬁned an experience as a consumer’s beliefs about how a medium ﬁts into his/her life. Customer experiences could be driven by customer motivations for interactions with the focal object (medium, service/ brand). In the language of measurement models, experiences are ﬁrst-order constructs while engagement is a second-order construct. The model developed by Calder et al. (2009), has been used in subsequent studies on social media, print media, live concerts, mobile media, online retail etc. Mobile shopping sites/ applications provide customers a convenient and compatible medium to shop from their chosen retailer. Also, the focus of this is to investigate role of mobile phones as a medium for building engagement and subsequently generate online reviews. Therefore the model of measuring customer engagement as a higher order construct with underlying customer experiences arising out of usage of mobile app as lower level constructs was deemed to be suitable for this study. Based on the relevant literature and objective of this study, Customer Engagement is conceptualized as A psychological state that leads to frequent interactions with the focal object (mobile shopping apps in this case) that goes beyond transactional motive of immediate purchase intention. The motives for interactions with the focal object may be utilitarian (e.g., looking for new product launch, promotional oﬀers, deals etc. in a speciﬁc category) with the objective of information for potential purchase in future or hedonic (e.g., looking for entertainment in new market trends, scenic pictures, etc.) with the objective of keeping oneself abreast of environment.
The Internet and information technology provide a new opportunity for consumers to share their product evaluations online. Online retailers like Amazon, Flipkart etc. often request shoppers to share post-purchase reviews on their respective websites. The consumer reviews include customers’ experiences with product quality, as well as services of the online service provider. There is strong evidence suggesting the growing importance of consumer reviews in consumer purchase decisions and product sales (Chen and Xie, 2008). Consumer-created in- formation in the form of online reviews are considered to be more credible than seller-created information due to the trustworthiness of the information source (Bickart and Schindler, 2001). Online consumer reviews, as consumer-created product information, can therefore be viewed as a special type of word-of-mouth communication (Godes and Mayzlin, 2004).
Traditionally, the word-of-mouth communication (WOM), has been shown to have a signiﬁcant impact on consumer choice (Katz and Lazarfeld, 1966; Engel et al., 1969; Arndt, 1967), as well as post-purchase product perceptions (Bone, 1995). Word-of-mouth has also been shown to be more eﬀective and credible than the conventional marketing tools of personal selling and various types of advertising (Katz and Lazarfeld, 1966; Engel et al., 1969). Similar to word-of-mouth communication, research has shown that online reviews have higher credibility, empathy and relevance to customers than marketer created sources of information (Bickart and Schindler, 2001; Godes and Mayzlin, 2004). Diﬀerent from the traditional WOM however, the in- ﬂuence of which is typically limited to a local social network, the im- pact of online consumer reviews can reach far beyond the local com- munity via the Internet. Further, technology enables sellers to eﬀectively initiate and broadcast consumer online reviews via its own website (Chen and Xie, 2008).
Another information source closely related to online consumer review is professional reviews from third parties (e.g., Carwale.com, Tripadvisor, PC Magazine, PC World). Professional reviews are pro- vided by experts to build up the product reputation, oﬀer product in- formation, and serve as indirect advertisements (Zhou and Duan, 2016). Empirical studies have demonstrated a signiﬁcant relationship between professional reviews and user decisions (Basuroy et al., 2003; Chen et al., 2012; Chen and Xie, 2008; Lee and Tan, 2013). Professional re- views tend to focus on product attribute information (e.g., performance, features, and reliability) and their review ratings are likely to be correlated with the performance of these attributes. Diﬀerent from professional reviews, online consumer reviews are posted by users based on their personal experiences and focus on whether and how a product matches a speciﬁc individual’s preference and usage condition (Zhou and Duan, 2016). Online consumer reviews are likely to be more re- levant to consumers as it often describes product attributes in terms of usage situations and measures product performance from a user’s perspective. Consumer-reviews, therefore, help less-sophisticated consumers (i.e., novices) in ﬁnding their best-matched products (Bickart and Schindler, 2001). Further, as unpaid, voluntary sources of in- formation, consumer reviews are considered more credible.
Given the widespread impact of consumer reviews, ﬁrms are adjusting their marketing communication strategy to respond to this emerging source of WOM information (Chen and Xie, 2008). While both professional reviews, as well as consumer reviews, have their own importance in inﬂuencing the potential customer’s choice, this study investigates the factors inﬂuencing consumer reviews. There is a growing body of literature on psychological motivation including altruistic (helping) motives with a desire to help other consumers make informed buying decisions or due to egoistic motives, with a desire for self-reputational enhancement (Bendapudi et al., 1996; Hennig-Thurau et al., 2002). Researchers also recognize that by posting online reviews, customers derive great social value (Balasubramanian and Mahajan, 2001) and become empowered consumers (Labrecque et al., 2013). With customers using mobile phones as a key mode to access online content, read reviews and shop online, the channel might be playing an important role in generating online reviews. Responding to a review request promptly soon after receiving it using a mobile device may give the customer a sense of instant gratiﬁcation rather than thinking and posting it at a later time when using a PC device. However, consumer usage of mobile phones for writing reviews for products bought in the past is underexplored.
2.2-Conceptual model and hypothesis development
Researchers have found that higher levels of satisfaction is likely to lead to higher levels of loyalty among customers (Anaza and Zhao, 2013; Anderson and Sullivan, 1993; Shankar et al., 2003). Satisﬁed customers are likely to exhibit loyalty through repurchase intentions and writing favourable reviews (Maxham III and Netemeyer, 2002; Oliver, 1980; Ranaweera and Prabhu, 2003). The likelihood of custo- mers writing online reviews will depend on a) the extent to which the product or service performance exceeds the customer’s expectations to share their positive experience, or b) the extent that the customer’s expectations are not fulﬁlled, motivating them to engage in negative reviews warning others, and/or seeking retaliation (De Matos and Rossi, 2008). Hirschman (1970), in his seminal work on customer loyalty, suggested that customers with a strong attachment to the ﬁrm actively look for mechanisms to make themselves inﬂuential regarding the products of those ﬁrms. Researchers have speciﬁcally explored the role of satisfaction in building loyalty in online retailing environment thereby popularising the terms e-satisfaction and e-loyalty (Anaza and Zhao, 2013; Sahadev and Purani, 2008). Writing positive reviews on retailers’ portal or third-party portals are common practices among satisﬁed customers that inﬂuences potential shoppers (Gvili and Levy, 2016). This study therefore proposes
H1. Customer satisfaction will have a positive eﬀect on intention to write online reviews.
Trust refers to “a willingness to rely on an exchange partner in whom one has conﬁdence” (Moorman et al., 1993, p. 82). Trust has an
important eﬀect on customer’s propensity to leave or stay with the same service provider (Garbarino and Johnson, 1999; Morgan and Hunt, 1994; Singh and Sirdeshmukh 2000). Empirical ﬁndings have shown that higher levels of trust are associated with a greater tendency to oﬀer favourable reviews (Garbarino and Johnson, 1999; Gremler et al., 2001; Ranaweera and Prabhu, 2003). This is based on the rationale that customers mostly provide recommendations to other individuals of their reference group, such as a friend or a relative, and, thus, a customer will be more likely to endorse a provider that he or she has previous experience with and conﬁdence in (Gremler et al., 2001). But even when customers are oﬀering advice to others, no matter who the receiver is, there is a risk of being wrong and a reviewer would not like to be wrong (Mazzarol et al., 2007). Trust creates beneﬁts for customers such as lower anxiety, uncertainty, and vulnerability about the trans- action. These beneﬁts inﬂuence satisfaction, which in turn aﬀects re- views, especially in a service context that is relatively more complex (Garbarino and Johnson, 1999; Hennig-Thurau et al., 2002). This study therefore proposes
H2. Trust will have a positive eﬀect on intention to write online reviews.
2.3-Customer engagement and online reviews
Customer engagement is being explored as a construct to facilitate favourable behaviour among existing customers including commit- ment, loyalty and online word of mouth (Brodie et al., 2011a, b; Calder et al., 2009; Pham and Avnet, 2009). Researchers have proposed that higher frequency of interaction with the focal object is an indicator of higher engagement. Further, scholars are also investigating various antecedents of customer engagement including trust and satisfaction (R. J. Brodie et al., 2011a, b; Jaakkola and Alexander, 2014a, b; So et al., 2014). Table 2 provides a snapshot of recent literature on antecedents and consequences of customer engagement.
As can be seen in Table 2, the literature on antecedents and con- sequences of customer engagement is largely conceptual or exploratory with small sample size. Further, there is a lack of convergence among the propositions of scholars that is common where the central construct is still emerging. With reference to the role of customer engagement in online reviews, Vivek et al. (2012) in their conceptual paper proposed word of mouth to be a potential consequence of customer engagement. Further, Hollebeek (2011) in his conceptual paper and Dwivedi (2015) and Thakur (2016) in their respective empirical paper found a signiﬁcant role of customer engagement in customer loyalty. There is also evidence in marketing literature on a signiﬁcant relationship between loyalty and referrals / online reviews (Cheung and Lee, 2012; Harris and Goode, 2004; Srinivasan et al., 2002). There is a whole new stream of research that is emerging in customer engagement in online reviews (Mathwick and Mosteller, 2016). Customer engagement construct has a strong aﬀective component that repeatedly drives customers back to the focal object, i.e., mobile shopping application. Such psychological state is likely to motivate customers to act favorably towards the shopping app and the retailer that may include responding positively to the re- quests for writing online reviews. Therefore, this study proposes,
H3. Customer Engagement will have a positive eﬀect on intention to write online reviews.
2.4-Satisfaction, customer engagement and online reviews
Customer engagement may be driven by satisfaction, i.e., a satisﬁed customer is likely to engage more with the focal construct as also observed by van Doorn et al. (2010). In the case of mobile and online retail, a satisﬁed customer is likely to interact with the medium (i.e., website/ mobile application) to check back on new oﬀerings, trends, promotions etc. oﬀered by the retailer demonstrating engagement. Such engagement is likely to inﬂuence customers’ behaviour beyond transactional purchase relationship and is likely to provide a favourable response to a request for writing reviews. Conversely, a dissatisﬁed customer is less likely to engage further with the service provider and subsequently demonstrate facvourable behaviour. In their paper on engagement and satisfaction, Calder et al. (2013) explored predictive power of satisfaction vs engagement in the adoption of diﬀerent media. There is an emerging stream of literature which posits role of engagement in long-term non-transactional relationships (Brodie et al., 2011a, b; Pansari and Kumar, 2017) and writing online reviews as a mechanism to express the same is likely. Further, there are some re- searchers who have proposed/ validated the relationship between customer engagement and loyalty and the likelihood of writing online reviews (Dwivedi, 2015; Hollebeek, 2011; Thakur, 2016; Vivek et al., 2012; Zheng et al., 2015). Based on existing literature and the current analysis, it is likely that satisfaction will inﬂuence customer engagement, which in turn is likely to inﬂuence intention to write online re- views. This study, therefore proposes a mediating role of customer engagement in satisfaction – online review intention relationship.
H4. Customer Engagement will mediate the relationship between satisfaction and online review intention.
2.5-Trust, customer engagement and online reviews
There seems to be a lack of convergence in existing literature regarding the relationship between customer engagement and trust. Some scholars have proposed trust to be an antecedent for customer engagement (Bowden, 2009b; Brodie et al., 2011a; Jaakkola and Alexander, 2014a, b; van Doorn et al., 2010). Other researchers, on the contrary, have proposed trust to be a consequence of customer engagement (Harwood and Garry, 2015; Hollebeek, 2011; So et al., 2014; Vivek et al., 2012). Trust is a long term relationship that builds over time as it refers to customers’ willingness to rely on the service provider with conﬁdence (Moorman et al., 1993, 1992). Such long term relationship built on trust is likely to build eﬃciency and eﬀectiveness (Morgan and Hunt, 1994). The customers’ engagement with a retailer’s mobile site/ application in the form of visiting/ downloading the same on her device is likely to be inﬂuenced by customer’s trust in the retailer. In other words, the act of engagement with retailers’ mobile portal is unlikely by the customers with lack of trust in the relationship. This study, therefore, would like to consider trust as an antecedent to customer engagement. Further, trust is likely to have a strong role in customer writing online reviews (Garbarino and Johnson, 1999; Gremler et al., 2001). Considering the likelihood of trust inﬂuencing customer engagement which in-turn is likely to inﬂuence propensity to write online reviews, this study proposes mediating role of customer engagement in trust-online review relationship.
H5. Customer Engagement will mediate the relationship between trust and online review intention.
2.6-Moderating role of satisfaction level
Customer engagement is a psychological state whereby the desired outcome of writing online reviews is driven by customer’s positive feeling with the focal object. Satisfaction, being a function of a positive experience with the retailer, is likely to amplify the eﬀect of customer engagement on online review intention. A satisfactory experience will motivate an engaged customer to respond favorably to a retailer’s re- quest for providing online reviews while using the shopping app. Conversely, when a customer is not satisﬁed with the retailer, the positive eﬀect of customer engagement on intention to write online re- views is weakened due to the discrepancy in aﬀective component of usage experience of shopping app and cognitive (usage outcome) experience with the retailer.
Therefore, it is proposed that customer engagement will have a stronger (weaker) inﬂuence on online review intention for customers with higher (lower) satisfaction level. Based on the current analysis, this study proposes,
H6. Satisfaction level will moderate the relationship between customer engagement and online review intention such that higher the satisfaction level higher is the eﬀect of customer engagement on intention to write online reviews.
2.7-Moderating role of trust level
Customer engagement has a strong hedonic component that creates positive association with the retailer. Trust, ‘willingness to rely on an ex- change partner’ (Moorman et al., 1993) is developed over a period of association based on several positive interactions. Lack of trust on the contrary may be a function of unpleasant experiences or perceived op- portunistic behaviour of the exchange partner in the past. Trust in the retailer, a positive construct, therefore, is likely to strengthen the eﬀect of customer engagement (a positive association) on online review intention. Contrariwise, lack of trust the retailer is likely to weaken the positive eﬀect of customer engagement on intention to write online reviews.
An engaged customer who also has developed strong trust (dis-trust) in the retailer trough multiple interactions is more (less) likely to take out time and respond favorably to retailer’s request for writing online review for past purchase. Based on the current analysis, this study proposes,
H7. Trust level will moderate the relationship between customer engagement and online review intention such that higher the trust level higher is the eﬀect of customer engagement on intention to write online reviews.
Based on the relevant literature, a conceptual framework is pro- posed (Fig. 1).
3-Method and analysis
The research methodology in this investigation was designed to test the proposed model for the role of customer engagement (Fig. 1) in inﬂuencing customers to write online reviews using mobile devices through survey method with a large sample size. The survey data facilitated validation of the psychometric properties of experience scales through conﬁrmatory factor analysis and then a second-order factor model for engagement. The ﬁnal step was to test the research hypothesis around motivators for writing online reviews with emphasis on mediating role of customer engagement.
The research instrument for the study was designed with items from validated scales for measuring customer engagement, satisfaction, trust and online review intentions. This study was focused on shopping of fashion and lifestyle products using mobile shopping applications/ site. To capture the speciﬁc context, verbatim of the existing scales were slightly modiﬁed for this study. Three scholars working in similar re- search area examined the face validity of the modiﬁed measures. Following their advice, some measures were reﬁned to improve their validity. The ﬁnal items of the research instrument and the source are provided in the Appendix A. Three diﬀerent versions of questionnaire employing a distinct, randomly assigned sequence of the data collection items with seven-point Likert scales anchored in ‘strongly disagree’ (1) through to ‘strongly agree’ (7) were designed. This was done to reduce the occurrence of primacy and recency eﬀects.
The respondents were asked to select a fashion or lifestyle mobile shopping application (like Myntra, Jabong, Fashion & You or lifestyle e- retailers like Flipkart and Amazon) they had made purchased from, and to complete the entire questionnaire for that mobile shopping application/ mobile shopping site only. The choice of fashion and lifestyle retailers was appropriate as the study explored both utilitarian and hedonic aspects while dealing with customer engagement. A pure utilitarian category like grocery may not be appropriate for exploring such constructs and hence the choice of a category that has both utilitarian and hedonic purchase motivations.
The study was conducted in Mumbai, the ﬁnancial capital of India. Mumbai is a large metropolitan city with a mix of the population from diﬀerent parts of the country. The research instrument developed for the study was administered using survey method to a set of respondents who had made more than one purchases in past six months using mo- bile devices from a fashion/ lifestyle retailer. For data collection, 1500 questionnaires were distributed. A total of 439 responses were received indicating around 30% response rate. Of this, 18 responses were found incomplete or un-usable due to extreme patterns and were removed from further analysis. 421 responses were used for ﬁnal analysis. This provided a sample size that exceeds the recommended minimum (Bentler and Chou, 1987; Hair et al., 1998).
The qualiﬁed respondents were current users of mobile devices for shopping (31% female; 66% below 30 years of age; 82% with professional experience of over 5 years). Young professionals are appropriate sample as they are comfortable with the usage of mobile devices for shopping and spend a considerable amount of time on mobile phones. Further, they have suﬃcient disposable income to spend on both utilitarian as well as discretionary needs. While the gender mix may not represent the user base of mobile telephony, it does reﬂect the population mix of working population in the country.
Before proceeding with further analysis, common method variance was tested using Harman’s single method test (Podsakoﬀ and Organ, 1986). The factor analysis did not produce a single factor or one general factor that accounted for the majority of the variance. Each factor ac- counted for more than the 5% cut-oﬀ thereby establishing that common method variance was not a problem.
The data collected in the study was analysed with structural equation modeling (SEM) using SPSS AMOS 20. Following a two-step analytical procedure (Hair et al., 2006), the measurement model was ﬁrst evaluated for validity reliability and statistical ﬁt. This was followed by an assess- ment of the structural model, path analysis and test for mediation.
4.1-Construct operationalization and conﬁrmatory factor analysis
Customer engagement was conceptualized as second order construct with six dimensions – social-facilitation, self-connect, intrinsic enjoy- ment, time-ﬁller, utilitarian experience and monetary evaluation ex- periences as per existing literature (Calder et al., 2009; Thakur, 2016). Scholars in marketing and information systems have a perpetual debate on the suitability of reﬂective vs. formative indicators for second order constructs. Some of the decision rules that favor selection of re- ﬂective model include direction of causality from construct to indicator items, indicators as manifestations of the construct, indicators being interchangeable, dropping an indicator should not alter the conceptual domain of the construct and indicators are expected to covary with each other (Jarvis et al., 2003). Also, reﬂective indicator models yield more meaningful measures of reliability and tests for construct validity en- abling genralisation (Bagozzi and Yi, 2012). Speciﬁc to this research, consensus deﬁnition of engagement as a psychological state arising from context dependent experiences (Brodie et al., 2011a, b) indicates that varied experiences are manifestations of engagement, experiences in usage of a focal medium (mobile device in this case) are inter- changeable, dropping one experience from the measurement model is unlikely to alter the construct and these experiences are likely to vary with each other (Calder et al., 2009). Further, researchers in the domain have extensively conceptualized customer engagement as a multi-dimensional construct (Brodie et al., 2013; Calder et al., 2009; Hollebeek et al., 2014; Vivek et al., 2014) that is best measured using second order reﬂective model. This study, therefore, conceptualized customer engagement as a second order reﬂective construct measuring engagement with customer experiences arising out of usage of mobile shopping applications as ﬁrst order constructs.
To establish reliability and validity of customer engagement scale, ﬁrst order CFA followed by second order CFA was conducted. One indicator item of the monetary evaluation was dropped due to low factor loading (Hair et al., 2010). Further analysis was carried out with 18 indicator items. The ﬁrst order CFA results for the six-factor, 18-item CE scale indicated the model provided acceptable ﬁt to the data: χ2 (32) = 116.699; χ2/df = 3.647; GFI = 0.956; CFI = 0.981 and RMSEA =0.069. The six factors measured with 18 indicator items had high correlation and were validated for second order reﬂective construct model for customer engagement. These six factors converged into the second order construct CE explaining 81 per cent of variance explained by the six constructs. The ﬁt indices (χ2 (44) = 470.71, χ2/df = 3.22, GFI.0.86, RMSEA.0.09, NFI.0.79, CFI.0.9) suggest that the proposed model represents a good ﬁt to the data.
This was followed by reliability and validity test of the complete measurement model, using conﬁrmatory factor analysis (CFA). The ﬁt indices (χ2 (288) = 999.11, χ2/df = 3.47, GFI = 0.85, RMSEA =0.07, NFI = 0.86, CFI = 0.9) suggest that the model with the nine latent variables represents a good ﬁt to the data (Tables 3, 4). The instrument demonstrates evidence of both convergent (signiﬁcant critical ratios, average variance extracted > 0.50 in all occasions) and discriminant validity (“square root” of AVE of each latent variable is greater than the correlations among the latent variables) (Fornell and Larcker, 1981).
4.2-Structural model – path analysis, mediation and moderation
The next step in the analysis involved testing of the structural model and corresponding proposed relationships. Structural equation mod- eling and path analysis are standard tools for estimating the strength of relationships between multiple constructs especially while dealing with latent constructs (Hair et al., 2006; Kline, 2010). The model (Fig. 2) was investigated to evaluate the variance explained by the model in predicting customer propensity to write online reviews and to establish the mediating role of the customer engagement.
The overall ﬁt measures (χ2 (312) = 1226.67, χ2/df = 3.93, GFI= 0.817, RMSEA = 0.035, NFI = 0.96, CFI = 0.87) indicate that the hypothesized model (Fig. 2) is a reasonable representation of the structures underlying the observed data (Fornell and Larcker, 1981; Hair et al., 2006). The results of the analysis produced a satisfactory picture regarding the signiﬁcance of estimated coeﬃcients. The model explained 72% variance in the dependent variable i.e., customer pro- pensity to write online reviews. The next step was to examine the hy- pothesized explanatory paths (Hair et al., 2006). As proposed, satisfaction (b = 0.374, C.R. = 5.128, p < 0.01), trust (b = 0.156, C.R.= 2.172, p = 0.03) and customer engagement (b = 0.416, C.R. = 5.239, p < 0.01) had statistically signiﬁcant impacts on customer pro- pensity to write online reviews. Hypothesis H1, H2, and H3 are there- fore retained.
Towards examining the mediating role of customer engagement, i.e., indirect eﬀects of satisfaction and trust on online reviews via customer engagement, a path analysis was performed using the boot- strapping method in AMOS (Byrne, 1998). First, the direct eﬀects of satisfaction (b = 0.707, p = 0.002), and trust (b = 0.343, p = 0.002) on online review intention are all signiﬁcant. Having established these direct eﬀects, the indirect eﬀects were then tested (Table 5).
The indirect eﬀect of satisfaction on online reviews via CE was found to be signiﬁcant (b = 0.255, p = 0.002). However, the direct eﬀect of satisfaction on online reviews was still signiﬁcant (b = 0.446, p = 0.005) suggesting that CE partially mediates the eﬀect of satisfaction on online reviews, in partial support of H4a. Next, the indirect eﬀect of trust on online reviews via CE was signiﬁcant (b = 0.175, p = 0.002), at the same time direct eﬀect was not signiﬁcant (b = 0.180, p= 0.092), indicating CE fully mediates the relationship between trust and online review intention, supporting H5a.
Once support for the main eﬀects had been found, the next step was to include the suggested moderator variables into the model in order to gain further insights. As the proposed moderators were latent variables, it was important to convert the data into categorical variables. Two step process was followed for that – derivation of factor scores for satisfaction and trust (through imputation), followed by median split. For testing moderation through invariance analysis, new categorical vari- ables were created – Staisfaction_Level and Trust_Level and grouping was done by a median-split on their the respective scores. As satisfaction and trust had a full spectrum of values, the variables were divided into three levels – low, medium and high. The High_Satisfaction group consisted of 106 subjects, and the Low_Satisfaction group consisted of 108 subjects. Similarly, the High_Trust group consisted of 108 subjects, and the Low_Trust group also consisted of 108 subjects. The subjects in the middle groups were excluded for more accurate analyses. Mean comparison test was conducted between the groups (High_Satisfaction, Low_Satisfaction and High_Trust, Low_Trust respectively) before pro- ceeding with further analysis. The diﬀerence in both the cases was found to be statistically signiﬁcant (Yi and La, 2004).
Towards testing the moderating eﬀects of Staisfaction_Level and Trust_Level, multiple-group analysis was conducted with two groups in each of the case (Arbuckle, 2010; Byrne, 2004). This technique was appropriate as customer engagement and online review intention are
both latent constructs. The technique involves a two step procedure – testing overall diﬀerences in the relationship at model level followed by assesment of outcome variable values at diﬀerent levels of the moderating variable. In the ﬁrst step, an overall Chi-square diﬀerence was calculated for each of the proposed moderating variables, i.e., Stais- faction_Level and Trust_Level. As a standard statistical process, a model with equality constraints was compared to a model that allowed the parameters to vary. This test imposed the null hypothesis that the moderator variables do not have any eﬀect on the parameters (Arbuckle, 2010). For both Staisfaction_Level (Δχ2 = 108.212, ΔDF = 26, p < 0.001) and Trust_Level (Δχ2 = 119.702, ΔDF = 26, p < 0.001), groups were found to be diﬀerent at model level (Steenkamp and Baumgartner, 1998). As a next step, hypothesized path (CE → eWoM) was tested for invariance for diﬀerent levels of sa- tisfaction and trust. In the case of moderating role of Staisfaction_Level, the χ2 test showed insigniﬁcant diﬀerence across groups (Δχ2 = 0.317, p = 0.573), basis which hypothesis H6 is rejected. However, a signiﬁcant diﬀerence (Δχ2 = 5.87, p = 0.015) at the 5% conﬁdence level indicated that the hypotheses H7 proposing the moderating eﬀect of Trust_Level is supported by the data in this study. Further the data validated a higher impact of customer engagement in writing online reviews among HighTrust group (βhigh = 0.403, C.R. = 1.9) vs those in LowTrust group (βlow = 0.096, C.R. = 0.737). Based on these results hypothesis H7 was retained while H6 was rejected.
The objective of this research is to elucidate customer online re- views with a speciﬁc focus on mobile shopping. There is an established literature on the role of satisfaction with products and services in building customer loyalty (Anaza and Zhao, 2013; Anderson and Sullivan, 1993; Shankar et al., 2003) which may be exhibited in various forms including re-purchase and writing online reviews. Further, re- searchers have proposed that trust in the retailer or brand purchased in the online environment would play a key inﬂuence in customer decision to write online reviews (Hennig-Thurau et al., 2002; Ranaweera and Prabhu, 2003). This study validated the signiﬁcant role of satisfaction and trust in writing online reviews that are in line with the existing stream of literature in this space (De Matos and Rossi, 2008; Gvili and Levy, 2016).
In past ﬁfteen years or so, there has been a growing body of literature on customer engagement. There are diﬀerent conceptualizations and researchers do not necessarily converge on the operationalization as well as antecedents and consequences (Pansari and Kumar, 2017). However, most of the scholars have conceptualized customer engagement as a multi-dimensional construct. This study validated the multi-dimensional operational structure of customer engagement as per existing literature (Brodie et al., 2011a, b; Calder et al., 2009; Vivek et al., 2012). Further, scholars largely believe that satisfaction is an antecedent to customer engagement (Calder et al., 2013; Pansari and Kumar, 2017; van Doorn et al., 2010). This study validated the same with respect to mobile shopping context. However, the existing body of knowledge does not converge on the relationship between trust and customer engagement. One stream of research considers trust as antecedent while other consider it as a consequence of customer engagement (Bowden, 2009a; Brodie et al., 2011a, b; Jaakkola and Alexander, 2014a, b; Vivek et al., 2012). This study looked at trust as an antecedent and proposed that trust is essential for the customer to download/ visit mobile shopping site and spend time as well as make a purchase. The empirical results validated the same providing evidence to the proposed hypothesis. The study had further proposed that customer engagement is likely to have a signiﬁcant impact on customer propensity to write online reviews that are fast be- coming an important source of information for potential customers. While there has been a lot of literature around the same, most of the studies are conceptual in nature (Vivek et al., 2012). Building on the literature, empirical evidence in this study provided the evidence that customer engagement adds to the predictive model of writing online reviews in addition to customer trust in the retailer and customer satisfaction with the retailer. Moderating role of trust in engaged customers’ propensity to write online reviews is another interesting ﬁnding of this study. The signiﬁcant inﬂuence of customer trust level in the retailer in inﬂuencing CE → OR path demonstrates the higher the level of trust, higher is the likelihood of impact of customer engagement on intention to write online reviews. On the contrary, the insigniﬁcant inﬂuence of satisfaction level on CE → OR path may be a reﬂection of a short-term relationship based on current transaction only, which may not inﬂuence the likelihood of customer engagement on intention to write online re- views. These ﬁndings further emphasize the importance of building long term relationships leading to overall trust, which is beyond satisfaction in a single transaction. Such relationships are likely to generate a positive response to online review request not only directly but also by inﬂuencing customer engagement to online review intention relationship. With these results, the study provided evidence that engagement plays an important mediating role between satisfaction and trust as antecedents for online reviews. Also trust plays a signiﬁcant moderating role in customer engagement – online review intention. Building customer engagement, therefore, is crucial for the retailers to improve the likelihood of customers in responding to online review requests.
The study has multi-fold theoretical contribution in the emerging areas of online reviews, customer engagement, and mobile shopping.
There is increasing interest among scholars on the inﬂuence of on- line reviews in customer decision making (Chevalier and Mayzlin, 2006; De Matos and Rossi, 2008; Srinivasan et al., 2002). This study contributes to this emerging body of literature whereby online reviews are becoming preferred and more credible sources of information by customers as compared to company sponsored advertorials (Bickart and Schindler, 2001). This study validates the role of satisfaction (De Matos and Rossi, 2008) (Gvili and Levy, 2016) and trust in the retailer (Garbarino and Johnson, 1999; Gremler et al., 2001; Ranaweera and Prabhu, 2003) as signiﬁcant antecedents of online reviews by customers speciﬁcally in mobile shopping apps in tandem with the existing literature.
There is an increasing interest among scholars on the role of engagement in the relationship between customers and brands/ retailers (Bowden, 2009b; Brodie et al., 2013; Hollebeek, 2011; van Doorn et al., 2010). However, the role of customer engagement in motivating users to write online reviews remains underexplored (Vivek et al., 2012). This research broadens the established antecedents of online reviews and thus contributes to the emerging stream of research in the role of cus- tomer engagement in online reviews. Trust is a long-term relationship arising out of several interactions whereby the customer is ready to rely on the exchange partner on whom she has conﬁdence (Moorman et al., 1993). However, engagement is important for the customer who trusts the retailer to go back and respond to the request to write online re- views when there is no personal need for the customer. Further, beyond satisfaction with the current transaction, the customer may not have any functional incentive to motivate her for writing reviews – in other words, the customer may feel ‘what’s there for me’ acting as an inhibitor. The signiﬁcant role of customer engagement in this network would come into play when the customer does not have a current purchase requirement but the engagement will motivate her to go back and write reviews when requested. The conceptualization and empirical validation of mediating role of customer engagement in satisfaction and trust as antecedents of online reviews is, therefore, a strong contribution of this study. Further, the moderating role of trust level in inﬂuencing customer engagement – online review intention is another valuable contribution highlighting the importance of building long-term trust in the relationship for favourable actions. Lastly, this study explored the usage of mobile devices for shopping which are becoming mainstream sources for accessing online content (Grewal et al., 2016; Kleijnen et al., 2007; Shankar et al., 2010). Mobile devices enable a higher level of engagement for service providers especially retailers both in pure-play online/ mobile format as well as multi-channel format. Retailers are actively extending their stores through a mobile channel that enable a higher degree of engagement and hence increase the likelihood of online reviews from customers.
The ﬁndings of this study have implications for several entities in a multi-sided market including the retailers oﬀering mobile shopping applications, the merchants selling their products and for the brands that advertise/ sell with these retailers.
Through empirically testing of the key antecedents, this research seeks to provide managers with strategic tools that drive online re- views. Both anecdotal and empirical evidence attests the importance of online reviews in inﬂuencing potential customer decision in their purchase journey. Findings of this study seek to inform managers regarding what factors to focus on to generate higher levels of reviews, thereby helping potential customers to make favourable purchase decisions. Satisfaction with the products (and services) and trust in the retailer are essential for customers to post online reviews. However, to increase the likelihood of customers posting online reviews post the purchase, retailers need to build customer engagement and motivate them to keep coming back to their sites.
Findings of this study empirically validate the importance of customer engagement in engendering online reviews that are inﬂuential in customer purchase decision. Thus, managers in retail organisations are advised to put in place initiatives to increase customer engagement. Customer engagement arises out of experiencing the retail environment repeatedly and mobile shopping applications could be a critical tool in enabling such engagement. Promoting adoption of mobile shopping apps and using the platform to send customized triggers for fostering repeat visits to retailer app may be crucial to building engagement. The study provided empirical evidence of higher likelihood of writing on- line reviews by engaged customers. Usage of mobile apps for shopping, therefore, is, therefore, likely to increase online reviews in addition to making more purchases.
5.4-Limitations and future research avenues
The research has some limitations that may lead to future research avenues. This research had self-reported data from the respondents and have limitations of the study of this nature. Also, the study looked at satisfaction, trust and customer engagement to predict online reviews. The future studies could look at other emotional, economic, and contextual variables. This study focused on lifestyle and fashion pro- ducts that provide a good context for customer engagement. The framework and results may not be directly applicable to some other for- mats like grocery and may need more investigation and modiﬁcations in the model. Further, online service retailers oﬀering travel, vacation, hospitality products (like Makemytrip.com, Yatra.com, Airbnb.com) as well as information (like Tripadvisor.com, Yelp.com) also rely heavily on online reviews for choices of potential customers. Similarly, in the business of food service aggregators like zomato.co, foodpanda.com, customer reviews play a critical role. Future research may look at taking this study and validating the results across other services.
The author would like to thank the Editor, the Area Editor, and anonymous reviewers for their invaluable feedback, which has helped to improve this paper. The author also expresses gratitude for providing invaluable comments during the development of this paper to Professor Ranjan Banerjee, Professor Sajeev George, Professor Suranjan Das, Professor Moutusy Maity, Professor S Raghunath, Professor Elizabeth Rose as well as the participants at the Research Seminar at SPJIMR, Mumbai India, 2017 Annual Conference of Emerging Markets at IIM Lucknow Noida Campus, India and Academy of International Business– India Chapter Paper Development Workshop.