Cross-cultural electronic word-of-mouth: a systematic literature review

Step 2: locate and create an extensive list of prospect relevant research papers

According to Verma and Yadav (2021), Scopus and Web of Science are the most reputed bibliometric databases. However, as Scopus is the largest abstract-based bibliometric database, consisting of more than 20,000 peer-reviewed journals from various publishers, its coverage is broader than Web of Science. Thus, we conducted keyword searches for journal articles in the Scopus database.

We limited the sample of empirical studies to those in which both eWOM and cultural dimensions were significant themes. Thus, we used two sets of keywords representing eWOM and cultural differences. The keywords relevant to eWOM were adopted from a previous eWOM literature review. The following are primary and synonym keywords used for eWOM:

“e-WOM,” “eWOM,” “electronic word of mouth,” “online word of mouth,” “online customer review,” “online consumer review,” “online reviewer,” “online rating,” “online review,” and “user generated content.”

Below are the primary and synonym keywords representing cultural differences:

“Hofstede,” “cultural dimensions,” “cultural differences,” “cross-culture,” “cross-cultural study,” “individualism,” “collectivism,” “power distance,” “masculinity,” “uncertainty avoidance,” “long-term orientation” and “indulgence.”

To conduct an extensive and systematic search, for each search, these two sets of keywords were exhaustively combined using “AND” operators, resulting in a total of 120 search terms. Specifically, the following sets of keywords were used: “e-WOM” AND “Hofstede,” “e-WOM” AND “cultural dimensions,” “e-WOM” AND “cultural differences,” …, “user generated content” AND “uncertainty avoidance,” “user generated content” AND “long-term orientation,” “user generated content” AND “indulgence.” We used a Boolean operator “OR” in the article title, abstract and keywords tab of the Scopus database to search for the research papers. The initial search resulted in 141 articles. To avoid missing any relevant articles, we cross-validated this search with the Web of Science and SAGE publication databases. The results indicated that SCOPUS had already incorporated all the papers from the Web of Science and SAGE publication databases.

Step 3: choose and analyze relevant studies according to predetermined inclusion and exclusion criteria

Inclusion and exclusion criteria were used to delimit the search results to extract only the most relevant articles for the literature review. Journals are believed to represent the highest level of research through which academics and practitioners acquire information and disseminate new findings (Nord and Nord, 1995). Therefore, the search results were based solely on academic journals. Book reviews, conference papers, research notes and editor prefaces were disregarded because of their limited contribution to knowledge development. After the first inclusion and exclusion criteria had been applied, the number of articles was reduced to 124.

The second exclusion criterion of language was then applied to the search results. Only articles published in English were included in the set of articles for further review. After the second inclusion and exclusion criteria of the English language had been applied, the number of results remained the same at 124 research articles.

The final inclusion and exclusion criteria were based on the quality of the paper and relevancy to the cross-cultural eWOM domain. Only papers that had a defined sample and an empirical methodology were included. The abstract and full text of each article were read in detail to assess their suitability.

In total, 61 journal articles from 38 journals, published from 2006 to the beginning of October 2021, were examined. Thus, the recent trends in eWOM and culture research were captured based on studies published over the past 15 years.

Step 4: analyze and synthesize the relevant literature

To organize the main findings of each paper, we followed Cheung and Thadani (2012) in adopting a social communication framework. By identifying who says what to whom and with what effect, this framework incorporates four major elements: the communicator (source) is the person who transmits the communication, the stimulus (content) is the message transmitted by the communicator, the receiver (audience) is the individual who responds to the communication and the response (main effect) is the receiver’s response to the message.

In our study, we first outline the current knowledge from the literature. Subsequently, we identify knowledge gaps that can potentially be used for future research.

Figure 1 shows an overview of the research methodology.

Descriptive results

Among the 61 articles on cross-cultural eWOM studies, 54% were published in business, management and accounting journals; 15% were published in social sciences journals; 15% were published in computer science journals and the rest of the papers were published in journals incorporating a mixture of disciplines, such as economics, econometrics and finance, arts and humanities.

The number of published papers shows that the Journal of Business Research has published the greatest number of articles (5), followed by the International Journal of Contemporary Hospitality Management (4), International Marketing Review (4), International Journal of Hospitality Management (3), Journal of Global Marketing (3) and Tourism Management (3). Other journals that have contributed to the topic are: Electronic Commerce Research and Applications, Journal of Electronic Commerce Research, Journal of Interactive Marketing, Journal of Intercultural Communication Research, Journal of International Marketing, Journal of Internet Commerce and Journal of Marketing Communications. The articles from these journals combined account for 59.02% of the literature on cross-cultural eWOM. The variety of journal disciplines indicates the widespread and growing interest in cross-cultural eWOM research across various academic fields, such as business, marketing, tourism and management information systems.

The earliest article found was published in 2006, which is of no surprise since the major online review websites were launched during the early 2000s. For ease of interpretation, the timeframe was divided into three equal intervals. During the first subperiod (between 2006 and 2011), only four articles (6.56%) were published. In the second subperiod (between 2012 and 2016), the discipline’s development accelerated slowly, with 11 research articles (18.03%) published. Since 2017, the growth of cross-cultural eWOM research publications has been phenomenal, with 46 articles (75.41%) having been published by the beginning of October 2021. Overall, the findings show an upward trend starting around the early 2010s in the number of publications on cross-cultural eWOM studies, with a relatively steeper increase toward 2020. This rapid growth depicts the increasingly strong research attention on cross-cultural eWOM.

Among the 61 cross-cultural eWOM studies, Hofstede’s cultural dimensions theory was the most used theoretical foundation. Only 15 of the 61 studies based their research on other cultural theories.

Synthesized cross-cultural eWOM findings

In this study, we reviewed the 61 cross-cultural eWOM studies and identified variables related to the four key elements (communicators, stimuli, receivers and responses) of social communication. Extending the framework proposed by Cheung and Thadani (2012), Figure 2 depicts the interrelationships among the key elements. Factors associated with each element of the framework are as follows:

Communicators:

• Source credibility – message source’s perceived ability (expertise) or motivation to provide accurate information.

• Social tie – the extent to which two individuals are socially related.

• Homophily – the degree to which individuals are similar in age, gender, education and social status.

• Information giving – the tendency for an individual to give an opinion about products or services.

Stimuli:

• Argument quality – the persuasive strength of arguments embedded in an informational message.

• Comprehensiveness – the completeness of a message.

• Review rate – the rating given by communicators on a product.

• Valence – whether the message is positive or negative.

• Review emotion – the extent to which communicators express their emotions in the message.

• Volume – number of reviews.

• Recommendation sidedness – two-sided information contains both positive and negative statements, one-sided information contains only either positive or negative statements.

• Recommendation consistency – the degree to which a particular eWOM recommendation is consistent with existing recommendations.

• Recommendation rating – the overall rating given by other readers on an eWOM recommendation.

Receivers:

• Information seeking – the tendency that an individual seeks information about products or services from others.

• Prior knowledge – prior knowledge of the review topic and the platform.

• Confirmation of prior belief – the level of confirmation/disconfirmation between the received information and their prior beliefs about the products or services.

• Homophily - the degree to which individuals are similar in age, gender, education, and social status.

Responses:

• Attitude – reviewer's overall evaluation of a person, messages, products or services.

• Perceived usefulness – the perception of an eWOM message being useful.

• Trust - the perception of the degree of truthfulness of the message.

• Risk perception – the uncertainty a consumer has when making purchase-related decisions.

• Information adoption – a process in which people purposefully engage in using information.

• Perceived product quality – perceived level of product quality.

• Purchase intention – the willingness to purchase a product in the future.

• Willingness to recommend – the willingness to recommend products in the future.

• Time spent – the amount of time spent on searching and considering product choice.

• Sale – the number of products sold.

• Market share – the portion of a market dominated by a product.

Tables 1 and 2 summarize the direct effects of cultural values on these factors. It is worth nothing that Tables 1 and 2 only include relationships with a theoretical base, specifically, those relationships hypothesized by the authors. This section first outlines the current knowledge from the literature in each element. Subsequently, knowledge gaps that can potentially be used for future research are identified.

Communicators

Most research studies were in agreement that individualistic consumers engage in more information giving than collectivistic consumers (Fong and Burton, 2008; Cheong and Mohammed-Baksh, 2020; Kitirattarkarn et al., 2021) because individualistic societies appreciate the expression of different opinions, whereas collectivistic cultures focus on the preservation of harmony, respect for hierarchy and the saving of face within the group. Therefore, collectivistic cultures tend to express opinions less to avoid challenging other opinions within the group. These findings were consistent for both horizontal and vertical dimensions of individualism (Choi and Kim, 2019). However, the results were inconsistent for the horizontal and vertical dimensions of collectivism (Chu and Choi, 2011; Lee et al., 2018; Choi and Kim, 2019).

Consumers tend to share information with their closely related peers. As collectivism is positively associated with the degree of homophily and social ties (Chu and Choi, 2011; Pezzuti and Leonhardt, 2021), the positive effect of social ties on information giving is more salient in collectivistic cultures (Chu et al., 2020). In addition, Litvin (2019) found that low uncertainty avoidance travelers plan their travels more extensively; therefore, their reviews tend to appear to be from the perspective of high expertise reviewers.

Though the empirical evidence in the literature offered fruitful insights, a few research questions remain open for future research. First, most existing research suggests that collectivistic consumers tend to post less compared with individualistic consumers because consumers from collectivistic cultures want to avoid challenging other opinions within the group. However, none of these studies measured whether this conjecture truly explains this phenomenon. Future research may address this gap by comparing the number of reviews by collectivistic consumers between products with strong conflicting reviews and products with strong congruency among reviews. If the said mechanism is correct, we expect that the number of reviews from collectivistic consumers would be significantly higher in products with strong congruency among reviews. Specifically, future research may examine the interaction between collectivism and congruency among reviews. The main negative effect of collectivism and the number of reviews must be significant, and this negative effect should be attenuated for products with high review congruency. Further, it could even become positive because collectivistic consumers may prefer sharing reviews that are consistent with other reviews to connect with their in-group peers (Lee et al., 2019).

Second, self-enhancement was found to be one of the factors determining eWOM sending behavior (De Angelis et al., 2012). Despite existing arguments on the possibility that the presence of self-enhancement is much weaker in collectivist-dominant Asian culture (Kitayama et al., 1997), no research has yet examined the cross-cultural effect of self-enhancement on the generation of eWOM. Future research may attempt to address this gap.

Third, self-efficacy was found to be an antecedent of eWOM (Huang et al., 2009). It would be interesting to examine this relationship across cultures because self-efficacy appears to differ among cultures (Oettingen, 1995).

Finally, previous research suggests that individuation (the willingness to stand out or be different from others) positively affects eWOM forwarding (Ho and Dempsey, 2010). However, as consumers from collectivistic cultures value interpersonal harmony above personal preference, they are less willing to stand out from the group. Therefore, we suspect that the impact of individuation on eWOM forwarding is attenuated in consumers from collectivistic cultures. Future research may attempt to prove this conjecture.

Stimulus

Most studies focused on observing the impact of culture on the review rate. Though some studies documented opposing findings (Kim et al., 2018; Stamolampros et al., 2019), the majority found that review rate was related to collectivism (Chiu et al., 2019; Mariani et al., 2019; Mariani and Predvoditeleva, 2019; Stamolampros et al., 2019) and low power distance (Gao et al., 2018; Mariani et al., 2019; Mariani and Predvoditeleva, 2019). This is because customers from countries with a high score on individualism and/or power distance dimensions have higher expectations. Therefore, they have a greater tendency to complain about disconfirmations in the perceived service quality. On the other hand, Stamolampros et al. (2019) argued that because individuals from high power distance cultures accept differences in social status, they are more likely to tolerate failures from the more powerful service providers. This argument is in fact very similar to that of studies proposing power distance to be negatively related to review rate. The difference is whether they see the service providers as holding a higher status or lower status. Further, Kim et al. (2018) posited that individualistic cultures prioritize enjoyment of life and fun, resulting in a positive attitude and optimistic outlook, which eventually leads to a more positive review rate.

Masculinity is found to be negatively associated with the review rate (Mariani and Predvoditeleva, 2019; Stamolampros et al., 2019). This is because masculine consumers are less tolerant of service failures and perceive themselves to have the power to confront service providers over the unsatisfactory experience (Torres et al., 2014). Additionally, long-term orientated consumers were found to be less likely to give a negative review rate (Mariani et al., 2019) because they are not willing to take the risk of compromising their long-term relationships with the service provider. The relationship between indulgence and review rate was found to be positive (Mariani et al., 2019; Stamolampros et al., 2019) because individuals from indulgent societies have a more positive attitude and are therefore more optimistic and more likely to remember positive emotions. However, individuals from restrained societies are less happy, less likely to remember positive emotions and more pessimistic. Further, uncertainty avoidance was found to negatively affect the review rate (Litvin, 2019; Mariani et al., 2019; Mariani and Predvoditeleva, 2019; Stamolampros et al., 2019). This is because customers with higher uncertainty avoidance are more risk-averse; therefore, they tend to search for the attributes of the product and service, leading to higher expectations. The cultural distance was also found to be related to a negative review rate (Mattson, 2017; Stamolampros et al., 2019). Because when the cultural distance between customers and sellers is higher, there is a higher chance of cultural gaps and differences, and these differences may translate into consumers’ dissatisfaction. Similar to the review rate, review valence was found to be associated with indulgence (Wen et al., 2018), high power distance (Wen et al., 2018) and low uncertainty avoidance (Fang et al., 2013), with mixed findings in individualism dimension.

In this review, it was noted that several studies also paid much attention to volume and recommendation consistency. The findings showed that the volume of eWOM messages was significantly more in individualistic cultures (Fang et al., 2013; Chiu et al., 2019). This is in line with the previous notion that individualistic consumers tend to engage in information giving more than collectivistic consumers. Also, higher volume was found in low power distance societies than in high distance power societies because individuals from high power distance are recognized as introverts and more reluctant to express their feelings (Fang et al., 2013). With online movie review data, Chiu et al. (2019) found that indulgence positively influences eWOM volume.

For recommendation consistency, various cultural theories were applied across the studies reviewed to explain the difference between East Asians and Westerners. Using Hofstede’s theory, Chiu et al. (2019) posited that members of individualistic Western societies are more likely to value freedom of expression, while members of collectivistic East Asian societies are more likely to seek group consensus rather than directly expressing their opinions. On the other hand, Wang et al. (2021) proposed that because consumers from individualistic cultures are more likely to engage in voice behaviors and are more involved, they are more susceptible to the anchoring effect of prior eWOM. Drawing on thinking style theory, Kim et al. (2018) posited that East Asians (holistic thinking style) are more likely to be influenced by their surroundings and peripheral cues, such as prior eWOM, while Western customers (analytical thinking style) tend to see an object as independent of its context and are therefore likely to evaluate a service or product without external influences. These studies are inconsistent both in their findings as well as theoretical background. Future research may attempt to clarify these inconsistencies.

Receivers

The findings of the studies included in this review were mostly consistent with the notion that collectivistic consumers are more likely to seek eWOM than individualistic consumers (Fong and Burton, 2008; Cheong and Mohammed-Baksh, 2020). This is because collectivistic cultures perceive that the act of using the information gained from a referent group enhances their relationships. On the other hand, individualistic societies are independent and rely less on others when they seek information. Other research delved deeper into the HVIC. The results showed that both horizontal (Chu and Choi, 2011; Lee et al., 2018) and vertical (Lee et al., 2018; Choi and Kim, 2019) collectivism were positively related to information-seeking behavior. However, Choi and Kim (2019) argued that some individualistic consumers also seek information from eWOM. Specifically, vertical individualistic consumers seek eWOM information because they believe that winning or doing better than others is crucial; therefore, they use all information available to them to be more competent.

Responses

Based on our framework, the response is a function of stimuli, communicators and receivers, and these relationships are moderated by the cultural backgrounds of the receivers of eWOM. Tables 1 and 2 shows that only three factors are directly affected by the cultural values of the receivers: perceived usefulness, trust and purchase intention.

Applying Hofstede’s cultural theory, Noh et al. (2013) found significant effects of collectivism on perceived usefulness. Further, Kim et al. (2018) found that cultural distance negatively affects perceived usefulness. This is because consumers who read reviews written by reviewers from similar cultural backgrounds assume that the reviewers are also similar in terms of preferences and attitudes, leading to an increased perceived usefulness of reviews. In terms of the trust, because uncertainty avoidance is related to the risk averseness of individuals, people from high uncertainty avoidance cultures tend to be more sensitive to information from unknown sources, which is generally the case with eWOM. In line with this argument, Furner et al. (2014) documented that consumers with high uncertainty avoidance exhibit lower trust. Regarding purchasing intention, compared with consumers from restraint societies, consumers from indulgent societies are usually more optimistic, more positive and happier; therefore, they have a higher tendency to engage in impulsive buying behaviors, making more spontaneous purchase decisions regarding products with positive reviews (Ruiz-Equihua et al., 2020).

Despite the various contributions offered in these studies, our synthesis suggests that additional research efforts may be undertaken to further develop the cross-cultural eWOM literature. Though the extant literature has investigated the moderating roles of culture on various relationships, it is fragmented, with one study for each relationship. Future research should attempt to conduct a replicated study for these relationships.

Second, research findings showed that eWOM significantly affects levels of trust and loyalty (Awad and Ragowsky, 2008). However, the extant literature suggests that there exists a cross-cultural difference in customer loyalty (De Silva Kanakaratne et al., 2020). Therefore, future research may attempt to discover differences in the relationship between eWOM and consumer loyalty across cultures. The findings may assist practitioners in leveraging their target customers’ cultural orientation when designing their loyalty reward programs.

Another venue of research concerns the anonymous nature of eWOM. This characteristic results in an important aspect unique to eWOM, fake online reviews, and this aspect may actually be its own biggest drawback. It is well known that product sellers manipulate product reviews to increase sales (Chen et al., 2011). In different regions, the propagators possibly adopt different strategies to disseminate fake reviews. Thus, future research may explore cultural differences in the posting of and responses to fake reviews.

Fourth, research suggests that consumers seek out both the most negative reviews and the most positive reviews to get a range of customer feedback. However, negative reviews are given more importance by consumers (Park and Lee, 2009). Future research may examine this negativity bias across cultures. It is likely that different cultures are predisposed to a different spectrum of eWOM valence.

Finally, visual eWOM is an emerging area of research that has received no attention in the cross-cultural context. Visual eWOM can be either a product review or the increasingly popular product “unboxing” videos posted on Instagram, YouTube, TikTok and other social media platforms. Using information processing theory, future research should investigate how consumers decide which visual eWOM to trust and which visual eWOM to reject.