Negative eWOM and perceived credibility : a potent mix in consumer relationships

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY-NC-ND 4.0 https://creativecommons.org/licenses/by-nc-nd/4.0/ Negative eWOM and perceived credibility : a potent mix in consumer relationships © 2022, Emerald Publishing Limited Accepted version (Final draft) Izogo, Ernest Emeka; Jayawardhena, Chanaka; Karjaluoto, Heikki Izogo, E. E., Jayawardhena, C., & Karjaluoto, H. (2023). Negative eWOM and perceived credibility : a potent mix in consumer relationships. International Journal of Retail and Distribution Management, 51(2), 149-169. https://doi.org/10.1108/ijrdm-01-2022-0039 2023 1 Negative eWOM and perceived credibility: a potent mix in consumer relationships Ernest Emeka Izogoa* Lecturer in Marketing & Research Associate Chanaka Jayawardhenab Professor of Marketing & Heikki Karjaluotoc Professor of Marketing aEbonyi State University, PMB, 053 Abakaliki, Ebonyi State, Nigeria & University of Johannesburg, Kingsway Campus, South Africa; Email: [email protected], Tel: +234 (0)7036297276 bUniversity of Surrey, Guildford, Surrey GU2 7XH, United Kingdom, Email: c.jayawar[email protected]; Tel: +44 (0)1483 683981 cUniversity of Jyväskylä, FIN-40014 University of Jyväskylä, Email: [email protected], Tel: +358 40 576 7814 2 Abstract Purpose – Based on the foundations of the schema theory, the elaboration likelihood model, and customer experience literature, this research examines how the interplay between a consumer’s previous shopping experience(s) and perceived credibility of negative online word-of-mouth leads to improved consumer-firm relationship quality. Design/methodology/approach – We utilised series of scenario-based experiments (N = 918) to test our research hypotheses. Findings – We show that a focal customer’s previous shopping experiences attenuate the perceived credibility of negative word-of-mouth on social media by other customers, which in turn weakens consumer-firm relationship quality. We also show that positive and negative perceptual experiences are asymmetric. Research limitations/implications – First, the online shopping experiences described in the experimental scenarios were generic and did not refer to any particular product/service. Thus, calibrating products and services into categories, and studying how product type differences impact online shopping experiences warrant further research. Practical implications – From a practical perspective, we demonstrate that not only does enhancing consumer-firm relationship quality demand meticulous integration of consumers' website and social media experiences, in positive vs. negative perception scenarios, relationship quality wane as review frequency increases. Originality/value – We contribute significant insight to the existing literature by specifically adopting the premise that consumers’ previous online shopping experience(s) will influence how credibly they will perceive negative online WOM posted on social media. Keywords eWOM, perceived credibility, online shopping experience, relationship quality, elaboration likelihood model 3 Introduction Most people who shop online seek out experiences of other shoppers about products/services they are about to buy on social networks. Such advice/reviews are considered to be electronic word-of-mouth (eWOM) and are defined as “all informal communications directed at consumers through Internet-based technology related to the usage or characteristics of particular goods and services, or their sellers” (Litvin et al., 2008, p. 461). Such shared customer experiences have become very influential (Mukerjee, 2020; Irshad et al., 2020) because the faceless online reviewers are increasingly becoming opinion leaders of online communities (Dalman et al., 2020; Litvin et al., 2008). This is especially so when eWOM is perceived to be credible – the extent to which a piece of information is believable or true. Research shows that 95% of shoppers read a review prior to making a purchase while 82% specifically seek out negative reviews (McCabe, 2018). But somewhat problematic is that “roughly half (51%) of those who read online reviews say they generally give an accurate picture of the true quality of the product, but a similar share (48%) believes it is often hard to tell if online reviews are truthful and unbiased” (Smith and Anderson, 2016). Likewise, more than half of the product reviews shared about certain products on Amazon are potentially deceptive or questionable (Pyle et al., 2021). Given that the credibility of reviews plays an important role in online consumers purchase decisions and consumers appear to seek out negative reviews in particular, we pose two important managerial questions. First, how does consumers’ previous shopping experience(s) influence their perceived credibility of negative online word-of-mouth (PCNWOM)? Second, in light of the influence of reviews on purchase decisions, how should retailers manage relationships with shoppers with different levels of online shopping experience? These questions are vital because previous customer experience influence consumer attitude (Park et al., 2021; Zheng and Bensebaa, 2022; Lao et al., 2021) and experience has been touted as a significant 4 differentiator of how consumers process information (Dagger and O’Brien, 2010; Lao et al., 2021; Talwar et al., 2021). Despite Verhoef, Kannan, and Inman (2015) stating that the changing retail landscape brought about by digitalisation have affected the business models of many retailers, our understanding of the interactions of customer experience and WOM in general remains poor (Talwar et al., 2021). Similarly, there have been calls to investigate how subsequent customer experience is influenced by customers’ perceptions of a retail brand and the potential asymmetric effects of positive and negative perceptions (see Verhoef et al., 2009). We contextualise our inquiry within an emerging market context where retailing dynamics are different, and thereby advance extant understanding of consumption experience from a developing market perspective. We contend that our efforts are the first to investigate the potential asymmetric outcome of positive and negative perceptions and how such polarised effect interact with shoppers’ experience level to influence PCNWOM. We draw on the schema theory (Bartlett 1932 in Brewer and Nakamura, 1984), the elaboration likelihood model (ELM) (Petty and Cacioppo, 1986), and customer experience literature (notably Verhoef et al., 2015; Verhoef et al., 2009) to build a unique experienceperception-relationship quality model. By so doing, we make several theoretical and practical contributions. From a theoretical perspective, our study advances the literature based on the foundations of schema theory and the elaboration likelihood model (ELM) by specifically adopting the premise that consumers’ previous online shopping experience(s) will influence how credibly they will perceive negative online WOM posted on social media. From a practical perspective, we demonstrate that positive and negative perceptual experiences are asymmetric. Thus, not only does enhancing consumer-firm relationship quality demand meticulous integration of consumers' website and social media experiences, in positive vs. negative perception scenarios, relationship quality wane as review frequency increases. 5 The next section presents the study’s conceptual framework and develops the hypotheses. Thereafter, we explain the experimental method, and present the results of data analysis. Finally, we discuss the research implications and point out future research directions. Conceptual framework Verhoef et al. (2009, p. 32) posit that “customer experience construct is holistic in nature and involves the customer’s cognitive, affective, emotional, social and physical responses to the retailer”. Specifically, online customer experience (hereafter referred to as OCE) has been described as the overall psychological effect that a customer feels and which arises from customer interactions with numerous virtual touchpoints (Rose et al., 2012). Customer experience has gained traction because firms have realised that success is inextricably tied to the delivery of seamless shopping experiences to customers; thus, organisations are making significant financial investments to create positive and memorable experiences that trigger greater sales (Rose et al., 2012). Positive customer experiences attract greater emotional attachment to their brands, enhanced customer satisfaction and loyalty (Anshu et al., 2022). Full comprehension of the domain of OCE must start with a deeper understanding of every direct and indirect customer-firm interactions (Frasquet et al., 2015). OCE has been shown to be a subjective, holistic, and multidimensional concept that originates from customers’ interactions with the online environment (Adhikari, 2015; Rose et al., 2012; Gentile et al., 2007). For instance, Klaus (2013) found that OCE comprises the functional and psychological dimensions. While the functional dimension in Klaus’ (2013) framework comprises usability, interactivity, communication, product presence and social presence, the psychological dimension constitutes value for money, trust, and context familiarity. Pandey and Chawla (2018) adopted four dimensions of OCE including 6 informativeness, interactivity, ease of navigation, and visual engagement as the functional elements of OCE, while the six psychological dimensions of OCE include e-negative beliefs, e-distrust, e-logistic ease, e-self efficacy, e-enjoyment, and e-convenience. We adopt the view that customer experience includes sensory, affective, cognitive, behavioural, and relational dimensions (Schmitt and Zarantonello, 2013) because it is encompassing and reduces the redundancy that might be inherent in adopting two-dimensional model with a variety of several sub-dimensions. Despite the growing body of extant works on customer experience, Nguyen et al. (2022) and Pandey and Chawla (2018) argue that the consequences of OCE are not completely understood. Drawing upon the foregoing, our conceptual model (figure 1) captures customers’ experiences across two touchpoints – all channels through which consumers experience a firm or its products and service – and its consequent effect on consumer-firm relationship quality (RQ hereafter). It postulates that consumers’ previous online shopping experience(s) will influence how credibly they will perceive negative eWOM posted on the social media. Review source credibility reflects the confidence and assurance of the message’s source (Belanche et al., 2021). The model further postulates that perceptions from exposure to negative eWOM posted by other consumers will influence the consumer-firm RQ. RQ is a higher-order construct consisting of three different but interrelated constructs: trust, satisfaction and commitment (Dagger and O’Brien, 2010; Crosby et al., 1990). While satisfaction is the affective state that arise from consumers’ overall assessment of the service experience (Dagger and O’Brien, 2010), we define trust as the confidence in an exchange partner’s reliability and integrity. Morgan and Hunt (1994) defined relationship commitment as the enduring desire to maintain a valued relationship. These components of RQ are built based on 7 consumer-firm encounters or interactions through customer facing employees, marketing communication, or eWOM. Figure 1 Conceptual model Hypotheses development Schema theory According to Aronson et al. (2010, p. 85) “Schemas [are] mental structures [that] people use to organise their knowledge about the social world around themes or subjects, and influence the information that people notice, think about, and remember.” The schema theory assumes that when exposed to a phenomenon, the individual abstracts generic cognitive representation in his/her mind; thus, individuals enact meanings when new information interacts with old information represented in the schema (Bartlett 1932 in Brewer and Nakamura, 1984). For Sensory experience Emotional experience Relational experience Behavioral experience Experience type x shopper type Cognitive experience Perceived credibility of negative eWOM Relationship trust Relationship satisfaction Relationship commitment Source credibility x review frequency Experience Perception Attitude H1a H1b H1c H1d H1e H5a H5b H5c H6-H8 Experience Perception Relationship quality (RQ) 8 instance, a shopper exposed to eWOM will judge its credibility by abstracting schemas of previous experiences with the firm in question or its products/services. Customers evaluate market offerings based on their own and/or other customers’ experiences (Belanche et al., 2021; Irshad et al., 2020) because consumers influence one another through eWOM (Mukerjee, 2020). Consumers can only perceive integrated interactions if they had previously encountered congruent experiences across the firm’s channels (Bèzes, 2021). According to Zheng and Bensebaa (2022), previous perceptual state triggers several attitude and behaviour. Thus, customers’ previous experiences with a company or its products/services can influence the customers’ perception of available information especially across different customer touchpoints. The utility of the schema theory in explaining the mechanics of our conceptual model also lies in the reasoning that retail customer experiences are complex because of the multiplicity of touchpoints that consumers encounter which implies that inference making is inevitable at some point. Previous OCE and PCNWOM Previous experience facilitates the understanding of consumer perceptions (Bèzes, 2021; Zheng and Bensebaa, 2022). Lao et al. (2021) and Bonfanti and Yfantidou (2021) highlighted the influence that the different components of customer shopping experience has in the shopping context. Consumers rely on previous consumption experiences as decision-making heuristics. Thus, past shopping experiences can influence the credibility of eWOM posted in the social media. Consumers’ perceptions are a function of previous experiences that shape socially shared and unquestioned beliefs (Sokolova and Kefi, 2020; Holttinen, 2014). Per McVee et al. (2005), the schema theory provides the basis for linking up the interpretation of current information with previous experience. Given that previous purchase experience is linked to several attitudinal and 15 has consistent valence as hypothesised in H7. However, the interaction effect of review source credibility and reviews’ frequency on relationship quality (RQ) is likely to be attenuated in the face of consistent negative reviews, as individuals evaluate the weight of evidence supporting the weakening effect of message consistency. Specifically, under situation of consistent negative eWOM, the effect of review source credibility and review frequency will be weakened to a negligible extent irrespective of whether review source credibility and review frequency are at the high or low thresholds. We, therefore, hypothesise as follows: H8. The interaction effect of review source credibility and review frequency on a) relationship trust, b) relationship satisfaction, and c) relationship commitment will be weakened when eWOM valence is consistent. Methodology We devised two pilot studies to validate the experimental scenarios and two main studies to test our research hypotheses. While a large cohort of Postgraduate students participated in the pilot studies for a course credit, the main studies involved both staff and students at a university situated in Southeast Nigeria, a typical emerging market. Following current research (e.g., Izogo and Mpinganjira, 2022), participants were recruited through an invitation sent to an email list obtained from the university’s IT department. Response rate was enhanced by incentivising participation in the main studies. Data quality was ensured across the studies through a filter question: previous online shopping experience (yes/no). We discounted participants who had no previous online shopping experience. Thus, valid participants were able to relate themselves to the experimental scenarios. Overall, the experimental procedure across the studies involved a demonstration of a typical shopping scenario, exposure to the experimental scenarios, and measurement of the latent variables and the manipulation checks. The appendix summarises all 16 the constructs examined in this study. Different scale formats and anchors used in the scales is consistent with its utility in attenuating common method variance (Podsakoff et al., 2003). All the measurement scales were in the 7-point Likert-type scale format. The partial least squares structural equation modelling procedure, independent sample t-test, analysis of variance (ANOVA) and multivariate analysis of variance (MANOVA) aided examination of experimental manipulation and test of research hypotheses. Pilot study We conducted a pilot study to pre-test the experimental scenarios and establish measures of OCE. The pilot study combined two experiments. The first pilot experiment utilised a 2 (experience type: positive vs. negative) × 2 (shopper type: experienced shopper vs. novice shopper) betweensubject factorial design. Consistent with extant works (see Izogo and Mpinganjira, 2021; Wan and Wyer, 2015; Mittal et al., 2008), a scenario-based/vignette experiment was conducted with four scenarios (available on request). Scenario 1 described an experienced shopper who had series of positive experiences. The three remaining scenarios are similar to scenario 1 except that scenario 2 described a novice shopper who had a positive experience while scenario 3 described an experienced shopper who had series of negative experiences. Scenario 4 described a novice shopper who had a negative experience. The second pilot experiment employed a 2 (review source credibility: high vs. low) × 2 (negative experience eWOM frequency: high vs. low) between-subject factorial design. Consistent with Ladhari and Michaud (2015) and Izogo and Mpinganjira (2020), the experiment was scenario-based. The design also called for four scenarios in which review source credibility and negative experience eWOM frequency were manipulated. Review source credibility was manipulated using popular Nigerian celebrities for high source credibility and unknown review writer(s) for low source credibility. This manipulation is consistent with the assumptions of the 17 ELM. For the manipulation of eWOM frequency, 3 reviews reflect high while 1 reflect low frequency. The validity of this manipulation is supported in previous research (see Boyer and Hult, 2006). Scenario 1 depicts a high review source credibility and high review frequency condition. The same opening vignette and reviews captured above were used for the three remaining scenarios except that the names of posters and the number of reviews were varied. Procedure and findings A booklet containing the experimental scenarios for the two combined studies and the accompanying survey were developed and administered through a pen and paper approach. The measures of the five OCE dimensions contained in the survey were adapted from Hsu and Tsou (2011), Schmitt (1999) and Gentile et al. (2007) while the measures of PCNWOM were adapted from Freeman and Spyridakis (2004). To ensure that the instruments were randomly administered, they were premixed beforehand while also adhering to order counterbalancing procedures to check response bias. A total of 160 subjects simultaneously participated in the two pilot experiments (76.25% were valid for analysis). The manipulation of the variables in the first pilot experiment through three measures of satisfaction adapted from Crosby et al. (1990) which proved high internal consistency (α: 0.949) and was therefore, averaged to form a composite score was successful (F(3, 118)= 134.614, p< 0.001) whilst the experiment was perceived to be realistic by all the four experimental groups (mean: 4.400–5.548 in a scale of 1–7). The resultant measures of all constructs examined through the Cronbach alpha reliability measure proved internal consistency (α: 0.76–0.939). In the second pilot experiment, although measures of the three components of RQ which were adapted from Ndubisi (2007), Morgan and Hunt (1994), and Crosby, Evans, and Cowles (1990) and measures of PCNWOM demonstrated high internal consistency (α: 0.918–0.966), its manipulation which was solely based on measures of cognition adapted from Mittal et al. (2008) proved unsuccessful both in terms of cognitive immersion (F(3, 18 118)= 1.535, p= 0.209) and PCNWOM (F(3, 118)= 2.108, p= 0.103). We estimate that this manipulation error is traceable to the fact that the manipulation of previous experience had positive valence. In study 2, three remedial steps were taken to correct the manipulation error. Overall, the pilot study validated the OCE scales, experimental manipulations’ quality and the timing of study 1 and 2. Study 1 Subjects, design and experimental procedures This study, like the pilot experiment 1 above in all respects, utilised a 2 (experience type: positive vs. negative) × 2 (shopper type: experienced shopper vs. novice shopper) between-subject factorial design. A total of 420 subjects participated and 378 responses were usable. Additionally, a teaching laboratory commonly used for a variety of postgraduate business classes with 15 participants seating capacity was the setting for stimuli presentation and survey completion. The researcher’s computer situated in the front of the teaching laboratory was connected to an LCD projector with 8’ x 8’ screen. As subjects arrived at the laboratory, they were instructed to sit quietly, and the experimental guidelines were explained to them. After the preliminary guidelines, the researcher proceeded to navigate through the Amazon.com website as an illustrative example of online shopping and its features. Subjects were encouraged to ask questions regarding the online shopping navigation as the demonstration proceeded. After demonstrating online shopping in Amazon.com for about 15 minutes, a booklet containing a described shopping scenario and the accompanying survey were shared to the subjects. To eliminate the effects of previous experiences with existing companies, a fictitious company name (Osas.com) was used in the experiment. The booklets were randomised by premixing them beforehand. The booklets were retrieved 20 minutes after they were administered. 19 Of the 378 usable responses, 52.6% were female (Mage= 27.7years, SDage= 6.6years). All the constructs demonstrated internal consistency (α: 0. 826–0.981; composite reliability: 0.919– 0.986). The scale also demonstrated convergent and discriminant validity (AVEs: 0.820–0.945) while the square root of the AVEs were greater than the highest correlation pair (Fornell and Larcker, 1981). Results and discussion We find that OCE manipulation was successful: positive experience treatments were rated more positively and satisfactorily by experienced and novice shoppers (MExperienced shopper= 6.23 vs. Mnovice shopper= 6.22) than negative experience treatments (MExperienced shopper= 3.28 vs. Mnovice shopper= 2.45; F(3, 374)= 323.40; p< 0.001). The mean value of the aggregate measures employed to assess experimental realism is M=5.48 on a scale of 1–7, an indication of realistic experimental treatments. As illustrated in Table 1, emotional experience (β= -0.488; t= 8.103; p< 0.001) and cognitive experience (β= -0.166; t= 3.003; p< 0.01) have a significant negative effect on PCNWOM. Conversely, sensory experience (β= 0.057; t= 1.437; p> 0.05), behavioural experience (β= 0.022; t= 0.015; p> 0.05), and relational experience (β= 0.001; t= 0.015; p> 0.05) do not significantly predict PCNWOM. The predictive accuracy of our model lies between weak and moderate thresholds (R2= 0.350) (Hair et al. 2011). Additionally, the Q2 for the endogenous construct (i.e. PCNWOM) is 0.317. Thus, the model has acceptable predictive relevance (Hair et al. 2014). Therefore, H1a-e is partially supported. We examined H2–H4 using a two-way ANOVA test. The main effect of experience type was significant (F(1, 377)= 252.75, p< 0.001). Shoppers who previously had positive experience(s) perceived negative experience eWOM to be less credible than shoppers who previously had 20 negative experience(s) (MPositive experience= 3.21 vs. MNegative experience= 5.81). Therefore, H2 is supported. Conversely, the main effect of shopper type (F(1, 377)= 0.99, p= 0.32) and the interaction effect of experience type and shopper type (F(1, 377)= 0.01, p= 0.91) on PCNWOM was not significant. Thus, H3 and H4 are not supported. Summarily, results of study 1 reveal that emotional experience is the most prominent negative predictor of PCNWOM posted in the social media followed by cognitive experience. These findings present an empirical reinforcement of the thoughts of business consultants who claim that 85% and 15% of customer experience are respectively emotional and physical (Shaw, 2007) and the intuition-based reasoning that developed markets’ consumers place more emphasis on ‘value for money’ than emerging markets’ consumers (Marceux, 2015). Additionally, it was established that experience type has a significant main effect on PCNWOM. Such evidence supposes that positive and negative perceptions induces asymmetric effects on PCNWOM but fails to indicate how this sequence of effect affect consumer-firm RQ. Study 2 therefore reports the results of an empirical study that investigated the PCNWOM–RQ link. Hypothesized path Path coefficient t-value H1a: Sensory experience → PCNWOM 0.057 1.437 ns H1b: Emotional experience → PCNWOM -0.488 8.103*** H1c: Cognitive experience → PCNWOM -0.166 3.003** H1d: Behavioral experience → PCNWOM 0.022 0.474 ns H1e: Relational experience → PCNWOM 0.001 0.015 ns Notes: ns= Not significant; **p < 0.01; ***p < 0.001; R2 = 0.350; Q2 = 0.317; observed power = 1.0 Table 1 OCE and its Effects 21 Study 2 Subjects, design and experimental procedures In this study we employed a 2 (experience type: positive vs. negative) × 2 (review source credibility: high vs. low) × 2 (frequency of negative experience eWOM: high vs. low) betweensubject factorial design to remedy the manipulation shortfall in the pilot study. Except for distinguishing between prior positive experience and prior negative experience, everything else was as in pilot experiment 2. The experimental procedures were the same as in study 1 except that the maximum number of subjects in each experimental section which lasted for an average of 35 minutes was 20. To eliminate the effects of previous experiences with existing companies, a new company name (Blue Gate) was devised for the experiment. Subjects were provided with differing scenarios (a brief shopping episode and a selection of consumer reviews), and these scenarios were randomised between-subjects. 400 subjects participated resulting in 380 usable responses. 51.8% of the valid responses were male (Mage= 23.79years, SDage= 3.64years). An independent sample t-test applied to the randomly split sample indicate that response bias was not a serious source of concern because apart from PCNWOM (t(224)= 2.278; p< 0.05), other group comparisons across all the study variables were not statistically significant (p> 0.05). All the constructs demonstrated internal consistency (α: 0.958–0.978; composite reliability: 0.973– 0.986). The scale also demonstrated convergent and discriminant validity since the AVEs (0.901–0.958) were above 0.5 while the square root of the AVEs were greater than the highest correlation pair (Fornell and Larcker, 1981). Results The manipulation quality of review source credibility was checked based on the Wan and Wyer’s (2015) approach by comparing the rating given to celebrities with the rating given to non- 22 celebrities on the following question: “To what extent can you say that the people who posted these comments are popular Nigerian celebrities?” The review source credibility manipulation was successful: the independent sample t-test indicates that reviews posted by celebrities were rated more positively/credibly than the reviews posted by non-celebrities (MCelebrities= 5.94 vs. MNon-celebrities= 2.99; t(378)= 16.06; p< 0.05) along a scale from 1 to 7. The review frequency manipulation examined by asking a dichotomous yes/no question was also successful: the Chi- Square statistic indicated a significant difference between subjects who read three reviews and those who read a single review (χ2(1)= 221.693, p< 0.001). The manipulation quality of prior experience type was assessed by examining cognition across the two groups. The use of cognition is strictly based on the reasoning that negative information carries greater weight than extremely positive information (Park and Lee, 2009). Measures of cognition were negatively worded so that high scores reflect low cognition while low scores reflect high cognition. The manipulation was successful: the independent sample t-test indicates that subjects who read prior positive experience were lower in cognition than those who read prior negative experience (MPrior positive experience= 5.31 vs. MPrior negative experience= 4.83; t(368.256)= 3.33; p< 0.05). We find that PCNWOM (see Table 2) have a significant negative effect on relationship trust (β= -0.681; t= 20.997; p< 0.001), relationship satisfaction (β= -0.697; t= 22.762; p< 0.001), and relationship commitment (β= -0.636; t= 19.555; p< 0.001). Thus, we conclude that H5a-c are all supported. 23 Hypothesized path Path coefficient t-value Cross-validated redundancy (Q2) H5a: PCNWOM→ Relationship trust -0.681 20.997*** 0.371 H5b: PCNWOM→ Relationship satisfaction -0.697 22.762*** 0.447 H5c: PCNWOM→ Relationship commitment -0.636 19.555*** 0.409 **p < 0.01; ***p < 0.001; R2 = 0.350 Table 2 PCNWOM and its Effects Following Wan and Wyer's (2015), a three-way independent MANOVA was utilised to examine H6–H8. As shown in Table 4, the results indicate that after reading other customers’ negative experience review(s), participants were generally more likely to exhibit relationship trust (MPositive experience= 5.12 vs. MNegative experience= 2.41; F(1, 379)= 322.09, p< 0.001); relationship satisfaction (MPositive experience= 5.19 vs. MNegative experience= 2.05; F(1, 379)= 361.13, p< 0.001), and relationship commitment (MPositive experience= 4.99 vs. MNegative experience= 2.42; F(1, 379)= 273.91, p< 0.001) when their previous shopping experiences are positive than when their previous shopping experience are negative. We also find that after reading other customers’ negative experience review(s), participants were generally less likely to exhibit relationship trust (MHigh frequency= 3.49 vs. MLow frequency= 4.04; F(1, 379)= 13.18, p< 0.001); relationship satisfaction (MHigh frequency= 3.37 vs. MLow frequency= 3.88; F(1, 379)= 9.47, p< 0.01), and relationship commitment (MHigh frequency= 3.50 vs. MLow frequency= 3.91; F(1, 379) = 6.88, p= 0.009) when exposed to three negative experience reviews than when exposed to one negative experience review. The above results permit us to confirm that H7a-c are supported. Consistent with our expectations, the main effect of review source credibility on relationship trust (MHigh review source credibility= 3.88 vs. MLow review source credibility= 3.65; F(1, 379)= 2.28, p= 0.13); relationship satisfaction (MHigh review source credibility= 3.76 vs. MLow review source credibility= 3.49; F(1, 379)= 2.58, p= 0.11), and relationship commitment (MHigh review source credibility= 3.83 vs. MLow 24 review source credibility= 3.58; F(1, 379)= 2.61, p= 0.11) were not significant. Additionally, the interaction effects of review source credibility and review frequency (F(1, 379), p> 0.05) and the three-way interaction of these variables and experience type (F(1, 379); p> 0.05) on the three components of RQ were all insignificant. With insufficient power and very low effect sizes (see Table 3), there was enough evidence to confirm that no such effect is evident in the population. Thus, H6a-c and H8a-c are supported. Independent variables Wilks’ Lambda F p-value η2 Experience type 0.509 118.773 0.000 0.491 Review source credibility 0.991 1.070 0.362 0.009 Review frequency 0.967 4.242 0.006 0.033 Experience type × review source credibility 0.989 1.331 0.264 0.011 Experience type × review frequency 0.969 3.895 0.009 0.031 Review source credibility × review frequency 1.000 0.055 0.983 0.000 Experience type × review source credibility × review frequency 0.984 2.047 0.107 0.016 Table 3 Wilks’ Lambda, Effect Size and Statistical Power High review frequency Low review frequency High review source credibility Low review source credibility High review source credibility Low review source credibility Relationship trust Positive experience 4.54a (1.83) (4.17–4.90) 4.66a (1.96) (4.2–5.05) 5.71c (1.20) (5.34–6.08) 5.55c (1.33) (5.15–5.95) Negative experience 2.68b (1.47) (2.23–3.13) 2.07ab (0.83) (1.63–2.52) 2.57b (1.20) (2.10–3.04) 2.30b (0.97) (1.84–2.77) Relationship satisfaction Positive experience 4.79a (2.07) (4.40–5.19) 4.62a (2.34) (4.20–5.05) 5.75c (1.25) (5.35 – 6.16) 5.60c (1.49) (5.19–6.03) Negative experience 2.18b (1.32) (1.68–2.68) 1.88b (1.00) (1.39–2.36) 2.30b (1.37) (1.79–2.81) 1.86bc (0.89) (1.35–2.37) Relationship commitment Positive experience 4.79a (1.82) (4.41–5.16) 4.40a (2.00) (4.00–4.80) 5.46c (1.07) (5.08–5.84) 5.31c (1.34) (4.90–5.71) Negative experience 2.47b (1.53) (2.00–2.94) 2.35b (1.18) (1.89–2.81) 2.60b (1.25) (2.12–3.08) 2.26bc (1.15) (1.78–2.74) Note: The range of values in the parenthesis represent 95% confidence interval. The standard deviations are the decimal values in parentheses. The values with subscripts are the means. Means with different subscripts are significantly different at p < 0.05 Table 4 Dimensions of RQ as a Function of Review Source Credibility, Review Frequency, and Experience Type 31 Anshu, K., Gaur, L. and Singh, G. (2022), “Impact of customer experience on attitude and repurchase intention in online grocery retailing: a moderation mechanism of value Cocreation”, Journal of Retailing and Consumer Services, Vol. 64, p. 102798. Aronson, E., Wilson, T.D. and Akert, R.M. (2010), Social Psychology, 7th edition, Upper Saddle River, New Jersey: Pearson Education Inc. Belanche, D., Casalo´, L. V., Flavian, M. and Ibanez-Sanchez, S. (2021), “Building influencers’ credibility on Instagram: Effects on followers’ attitudes and behavioral responses toward the influencer”, Journal of Retailing and Consumer Services, Vol. 61, 102585 Bèzes, C. (2021), “At the source of integrated interactions across channels”, International Journal of Retail & Distribution Management, Vol. 49 No. 7, pp. 899-918. Bonfanti, A. and Yfantidou, G. (2021), “Designing a memorable in-store customer shopping experience: practical evidence from sports equipment retailers”, International Journal of Retail & Distribution Management, Vol. 49 No. 9, pp. 1295-1311. Boyer, K.K. and Hult, T.M. (2006), “Customer behavioural intentions for online purchases: an examination of fulfillment method and customer experience level,” Journal of Operations Management, Vol. 24 No. 2, pp. 124–147. Brewer, W.F. and Nakamura, G.V. (1984), “The nature and functions of schema,” in Wyer, R. S. & Srull, T. K. (eds.), Handbook of Social Cognition. Hillside, NJ: Lawrence Erlbaum, 119– 160. Crosby, L.A, Evans, K.R. and Cowles, D. (1990), “Relationship quality in services selling: An interpersonal influence perspective,” Journal of Marketing, Vol. 54 No. 3, pp. 68–81. Dagger, T.S. and O’Brien, T.K. (2010), “Does experience matter? Differences in relationship benefits, satisfaction, trust, commitment and loyalty for novice and experienced service users,” European Journal of Marketing, Vol. 44 Nos. 9/10, pp. 1528–1552. 32 Dalman, M. D., Chatterjee, S. and Min, J. (2020), “Negative word of mouth for a failed innovation from higher/lower equity brands: Moderating roles of opinion leadership and consumer testimonials,” Journal of Business Research, Vol. 115, pp. 1-13. Fornell, C. and Larcker, D.F. (1981), “Evaluating structural equation models with unobservable variables and measurement error,” Journal of Marketing Research, Vol. 18 No. 1, pp. 39–50. Frasquet, M., Ruiz-Molina, M.-A. and Molla-Descals, A. (2015), “The role of the brand in driving online customer loyalty for multichannel retailers,” The International Review of Retail, Distribution and Consumer Research, Vol. 25 No. 5, pp. 490–502. Freeman, K.S. and Spyridakis, J.H. (2004), “An examination of factors that affect the credibility of online health information,” Technical Communication, Vol. 51 No. 2, pp. 239–263. Gentile, C., Spiller, N. and Noci, G. (2007), “How to sustain the customer experience: An overview of experience components that co-create value with the customer,” European Management Journal, Vol. 25 No. 5, pp. 395–410. Hair, J.F. Jr, Sarstedt, M., Hopkins, L. and Kuppelweiser, V.G. (2014), “Partial least squares structural equation modelling (PLS-SEM): An emerging tool in business research,” European Business Review, Vol. 26 No. 2, pp. 106–121. Hair, J.F. Jr., Ringle, C.M. and Sarstedt, M. (2011), “PLS-SEM: Indeed a silver bullet,” Journal of Marketing Theory and Practice, Vol. 19 No. 2, pp. 139–151. Herhausen, D., Binder, J., Schoegel, M. and Herrmann, A. (2015), “Integrating bricks with clicks: Retailer-level and channel-level outcomes of online–offline channel integration,” Journal of Retailing, Vol. 91 No. 2, pp. 309–325. Holttinen, H. (2014), “How practices inform the materialisation of cultural ideals in mundane consumption,” Consumption Markets & Culture, Vol. 17 No. 6, pp. 573–594. 33 Hsu, H.Y. and Tsou, H.-T. (2011), “Understanding customer experiences in online blog environment,” International Journal of Information Management, Vol. 31 No. 6, pp. 510– 523. Irshad, M., Ahmad, M.S. and Malik, O.F. (2020), “Understanding consumers’ trust in social media marketing environment”, International Journal of Retail & Distribution Management, Vol. 48 No. 11, pp. 1195-1212. Izogo, E.E. and Mpinganjira, M. (2020), “Behavioural consequences of customer inspiration: the role of social media inspirational content and cultural orientation”, Journal of Research in Interactive Marketing, Vol. 14 No 4, pp. 431-459. Izogo, E.E. and Mpinganjira, M. (2022), “Somewhat pushy but effective: The role of value-laden social media digital content marketing (VSM-DCM) for search and experience products”, Journal of Research in Interactive Marketing, Vol. 16 No 3, pp. 365-383. Izogo, E.E., Mpinganjira, M., Karjaluoto, H. and Liu, H. (2021), “Examining the impact of eWOM-triggered customer-to-customer interactions on travellers’ repurchase and social media engagement”, Journal of Travel Research, doi: 10.1177/00472875211050420. Kim, E.E.K. and Lee, C.H. (2015), “How do consumers process online hotel reviews? The effects of eWOM consensus and sequence,” Journal of Hospitality and Tourism Technology, Vol. 6 No. 2, pp. 113–126. Klaus, P. (2013), “The case of Amazon. com: towards a conceptual framework of online customer service experience (OCSE) using the emerging consensus technique (ECT)”, Journal of Services Marketing, Vol. 27 No. 6, pp. 443-457. Kozinets, R.V. (1998), “On netnography: Initial reflections on consumer research investigations of cyberculture,” in Joseph W. Alba, and J. Wesley Hutchinson (eds.), Advances in Consumer Research, 25. Provo, UT: Association for Consumer Research, 366–371. 34 Ladhari, R. and Michaud, M. (2015), “eWOM effects on hotel booking intentions, attitudes, trust, and website perceptions,” International Journal of Hospitality Management, Vol. 46 No. 3, pp. 36–45. Landmark, A.D. and Sjøbakk, B. (2017), “Tracking customer behaviour in fashion retail using RFID”, International Journal of Retail and Distribution Management, Vol. 45 No. 7, pp. 844-858. Lao, A., Vlad, M. and Martin, A. (2021), “Exploring how digital kiosk customer experience enhances shopping value, self-mental imagery and behavioural responses”, International Journal of Retail & Distribution Management, Vol. 49 No. 7, pp. 817-845. Lee, J., Park, D.-H. and Han, I. (2008), “The effect of negative online consumer reviews on product attitude: An information processing view,” Electronic Commerce Research and Applications, Vol. 7 No. 3, pp. 341–352. Lin, K.-M., Chen, N.-S. and Fang, K. (2011), “Understanding e-learning continuance intention: a negative critical incidents perspective,” Behaviour & Information Technology, Vol. 30 No. 1, pp. 77–89. Litvin, S.W., Goldsmith, R.E. and Pan, B. (2008), “Electronic word-of-mouth in hospitality and tourism management,” Tourism Management, Vol. 29 No. 3, pp. 458–468. Maheswaran, D. and Meyers-Levy, J. (1990), “The influence of message framing and issue involvement,” Journal of Marketing Research, Vol. 27 No. 3, pp. 361–367. Marceux, P. (2015), “How Digitalised Millennials Differ in Developed and Emerging Markets,” Euromonitor International. Available at: http://blog.euromonitor.com/2015/12/howdigitalised-millennials-differ-in-developed-and-emerging-markets.html (11/03/2017). 35 Matzler, K., Strobl, A., Thurner, N. and Füller, J. (2015), “Switching experience, customer satisfaction, and switching costs in the ICT industry,” Journal of Service Management, Vol. 26 No. 1, pp. 117–136. McCabe, K. (2018), “50+ Statistics Proving the Power of Customer Reviews,” Available at: https://learn.g2crowd.com/customer-reviews-statistics (accessed 23/02/2019). McVee, M.B., Dunsmore, K. and Gavelek, J.R. (2005), “Schema theory revisited,” Review of Educational Research, Vol. 75 No. 4, pp. 531–566. Mittal, V., Huppertz, J.W. and Khare, A. (2008), “Customer complaining: The role of tie strength and information control,” Journal of Retailing, Vol. 84 No. 2, pp. 195–204. Morgan, R.M. and Hunt, S.D. (1994), “The commitment-trust theory of relationship marketing,” Journal of Marketing, Vol. 58 No. 3, pp. 20–38. Mukerjee, K. (2020), “Impact of self-service technologies in retail banking on cross-buying and word-of-mouth”, International Journal of Retail & Distribution Management, Vol. 48 No. 5, pp. 485-500. Ndubisi, N.N. (2007), “Relationship quality antecedents: The Malaysian retail banking perspective,” International Journal of Quality & Reliability Management, Vol. 24 No. 8, pp. 829–845. Nguyen, A.T.V., McClelland, R. and Thuan, N. H. (2022), “Exploring customer experience during channel switching in omnichannel retailing context: A qualitative assessment”, Journal of Retailing and Consumer Services, Vol. 64, January, pp. 102803. Pan, L.-Y. and Chiou, J.-S. (2011), “How much can you trust online information? Cues for perceived trustworthiness of consumer-generated online information,” Journal of Interactive Marketing, Vol. 25 No. 2, pp. 67–74. 36 Pandey, S. and Chawla, D. (2018), “Online customer experience (OCE) in clothing e-retail: exploring OCE dimensions and their impact on satisfaction and loyalty–does gender matter?”, International Journal of Retail Distribution Management, Vol. 46 No. 3, pp. 323-346. Park, C. and Lee, T.M. (2009), “Information direction, website reputation and eWOM effect: a moderating role of product type,” Journal of Business Research, Vol. 62 No. 1, pp. 61–67. Park, J., Hyun, H. and Thavisay, T. (2021), “A study of antecedents and outcomes of social media WOM towards luxury brand purchase intention”, Journal of Retailing and Consumer Services, Vol. 58, 102272. Petty, R.E. and Cacioppo, J.T. (1986), “The elaboration likelihood model of persuasion,” in L. Berkowitz (ed.) Advances in Experimental Social Psychology, 19. New York: Academic Press, 123–205. Podsakoff, P.M., Mackenzie, S.B., Lee, J.-Y. and Podsakoff, N.P. (2003), “Common method biases in behavioural research: a critical review of the literature and recommended remedies”, Journal of Applied Psychology, Vol. 88 No. 5, pp. 879–903. Pyle, M.A., Smith, A.N. and Chevtchouk, Y. (2021), “In eWOM we trust: Using naïve theories to understand consumer trust in a complex eWOM marketspace”, Journal of Business Research, Vol. 122, pp. 145-158. Rose, S., Clark, M., Samouel, P. and Hair, N. (2012), “Online customer experience in e-retailing: an empirical model of antecedents and outcomes,” Journal of Retailing, Vol. 88 No. 2, pp. 308–322. Sawyer, D.C. (1997), “Customers’ bad experiences your biggest competitive threat?” Competitive Intelligence Review, Vol. 8 No. 3, pp. 87–88. Schmitt, B. (1999), “Experiential marketing,” Journal of Marketing Management, Vol. 15 Nos. 1-3, pp. 53–67. 37 Schmitt, B. and Zarantonello, L.L. (2013), “Consumer Experience and Experiential Marketing: A Critical Review” in N. K. Malhotra (ed.), Review of Marketing Research, 10. United Kingdom, UK: Emerald Publishing Ltd, 25 – 61. Shaw, C. (2007), The DNA of Customer Experience: How Emotions Drive Value, New York, NY: Palgrave Macmillan. Smith, A. and Anderson, M. (2016), “Online Shopping and e-Commerce,” Pew Research Center. Available at: http://www.pewinternet.org/2016/12/19/online-reviews/ (accessed 23/02/2019). Sokolova, K. and Kefi, H. (2020), “Instagram and YouTube bloggers promote it, why should I buy? How credibility and parasocial interaction influence purchase intentions”, Journal of Retailing and Consumer Services, Vol. 53, 101742. Talwar, M., Talwar, S., Kaur, P., Najmul Islam, A.K.M. and Dhir, A. (2021), “Positive and negative word of mouth (WOM) are not necessarily opposites: A reappraisal using the dual factor theory”, Journal of Retailing and Consumer Services, Vol. 63, 102396. Verhoef, P.C., Kannan, P.K. and Inman, J.J. (2015), “From multi-channel retailing to omnichannel retailing introduction to the special issue on multi-channel retailing,” Journal of Retailing, Vol. 91 No. 2, pp. 174–181. Verhoef, P.C., Lemon, K.N. Parasuraman, A., Roggeveen, A., Tsiros, M. and Schlesinger, L.A. (2009), “Customer experience creation: determinants, dynamics and management strategies,” Journal of Retailing, Vol. 85 No. 1, pp. 31–41. Vermeulen, I.E. and Seegers, D. (2009), “Tried and tested: the impact of online hotel reviews on consumer consideration,” Tourism Management, Vol. 30 No. 1, pp. 123–127. Wan, L.C., and Wyer, R.S., Jr. (2015), “Consumer reactions to attractive service providers: approach or avoid?” Journal of Consumer Research, Vol. 42 No. 4, pp. 578–595. 38 Yim, C.K.B., Chan, K.W. and Hung, K. (2007), “Multiple Reference Effects in Service Evaluations: Roles of Alternative Attractiveness and Self-image Congruity,” Journal of Retailing, Vol. 83 No. 1, pp. 147–57. Zheng, L. and Bensebaa, F. (2022), “Need for touch and online consumer decision making: the moderating role of emotional states”, International Journal of Retail & Distribution Management, Vol. 50 No. 1, pp. 55-75. Appendix Measurement scales Sensory experience (Hsu and Tsou, 2011; Schmitt, 1999): • How interesting do you think the shopping experience described above was? • How important will quality of pictures of products and the website’s beauty be to you in making that purchase? • How important will the visual quality of videos demonstrating product features be to you in making that purchase? • How important will the clearness and quality of sounds of audio videos demonstrating products features be to you in making that purchase? Emotional experience (Hsu and Tsou, 2011; Schmitt, 1999): • To what extent would you say shopping on [firm] website went in creating positive feelings in you? • To what extent would you say that shopping on [firm] website went in putting you in certain mood (e.g. joyous mood)? • To what extent would you say shopping on [firm] website went in creating fun-like feelings in you? • To what extent can you say that shopping on [firm] website didn’t appealed to your inner feelings in anyway (-) Cognitive experience (Hsu and Tsou, 2011; Schmitt, 1999): 39 • How helpful was the online shopping activity in stimulating your curiosity/interest in online shopping? • How helpful would you say the quality/price relationship was to you in making the purchase on [firm] website? • How helpful do you think the ease with which you can shop around/access products on [firm] website was to you in making the purchase? • Overall, how helpful do you think [firm] website was in enabling you solve your purchase problems? Behavioural experience (Hsu and Tsou, 2011; Schmitt, 1999): • How important do you think shopping on [firm] website was in making you think about your lifestyle? • How important do you think shopping on [firm] website was in reminding you of the things you can do? • How important would you say shopping on [firm] website from the comfort of your home or office was in enabling you change your lifestyle? • Shopping on [firm] website did not make you think about your lifestyle (-) • Shopping on [firm] website did not remind you of what you can do (-) Relational experience (Hsu and Tsou, 2011; Gentile et al., 2007): • How likely are you to share your [firm] shopping experience with other people (e.g. friends, colleagues, family members etc.)? • To what extent do you think shopping on [firm] website makes you feel a sense of belonging to the wider society (i.e. do you think shopping Osas.com will make you feel perceived positively by your peers, family members, colleagues etc?) • To what extent does shopping on [firm] website make you think about relationships with others (e.g. friends, colleagues, family members etc) • To what extent do you think shopping on [firm] website doesn’t make you feel a sense of belonging to the wider society (-) Perceived credibility of negative eWOM (Freeman and Spyridakis, 2004): • I will perceive that information as credible/reliable 40 • I will perceive the information as trustworthy • I will perceive the information as accurate • I will perceive the information as biased (-) Relationship trust (Crosby et al., 1990; and Ndubisi, 2007): • I can still rely upon blue gate to keep to their promises • [Firm] is a trustworthy company • I still believe that [firm] puts the customers’ interests before its own interest • I still feel safe to transact businesses with [firm] Relationship satisfaction (Crosby et al., 1990): • Overall, how satisfactory would you say your shopping experience with [firm] was? • Overall, how favourable would you say your shopping experience with [firm] was? • Overall, how pleased would you say your shopping experience with [firm] was? Relationship commitment (Morgan and Hunt, 1994): • My relationship with [firm] is something that I’m still committed to • I’m still willing to maintain my relationship with [firm] indefinitely • I’m still willing to put maximum effort to ensure that I maintain my relationship with [firm]