This study empirically tests and compares the influence of friends’ recommendations on social media and anonymous reviews on shopping websites in the context of online purchase intention. For this purpose, we analyse the impacts of these two platforms based on the components of information adoption model (IAM) which are borrowed as information quality, information credibility, information usefulness and information adoption. We conduct a survey and find anonymous reviews as more influential on consumer’ online purchase intentions than friends’ recommendations on social media. However, as this result was contrary to that expected, we conduct another study through in-depth interviews in order to enlighten our results found in the first study. In Study 2, we find the reasons why consumers prefer anonymous reviews rather than friends’ recommendations. Information quantity, information readiness, detailed information and dedicated information are factors which make shopping websites superior than social media in terms of the impact of electronic word of mouth (eWOM). Academic and managerial implications are discussed.
The influence of electronic word of mouth (eWOM) on purchase intention has long been known (Bickart and Schindler 2001; Chevalier and Mayzlin 2006; Park, Lee, and Han 2007). Previously, the effects of eWOM on discussion forums (Chiou and Cheng 2003; Huang and Chen 2006), consumer review sites (Cheung, Lee, and Rabjohn 2008; Gauri, Bhatnagar, and Rao 2008), blogs (Chu and Kamal 2008; Lin, Lu, and Wu 2012) and shopping websites (Li and Zhan 2011; Park, Lee, and Han 2007) have been studied by researchers. Also, these platforms have been compared in terms of their influence on consumers’ purchase intentions (Lee and Youn 2009). However, due to being relatively new, far less attention was paid to the influence of eWOM in social media (Cheung and Thadani 2012) although there are a few existing studies (See-To and Ho 2014; Wang, Yu, and Wei 2012). Social media has been defined as group of Internet-based applications that allow the creation and exchange of user-generated content (Kaplan and Haenlein 2010). Users can create content through using variety of facilities provided by social media websites such as pictures and videos. This visually enriched content generated by users can be about anything personal; however, it can also be about brands or their products and services. In fact, this socially extensive environment is considered as a great opportunity to share product-related opinions (Canhoto and Clark 2013; Chu and Kim 2011; Dessart, Veloutsou, and MorganThomas 2015; Tsimonis and Dimitriadis 2014); and recent studies show that consumers increasingly apply social media to obtain information about unfamiliar brands (Naylor, Lamberton, and West 2012; Schivinski and Dabrowski 2016). Social media websites, thus, considered as valuable platforms in terms of eWOM. The emergence of social media, on the other hand, has brought a new aspect to eWOM through enabling Internet users to communicate with people who they already know. While the other online platforms (discussion forums, consumer review sites, blogs, shopping websites) allow eWOM to occur between anonymous users (Dellarocas 2003; Sen and Lerman 2007); people are able to exchange their ideas about products or services with their friends and acquaintances on social media (Chu and Kim 2011; Kozinets et al. 2010) as it encourages users to have online accounts with real identities. Although there is a discussion between researchers about the possible effects of this less anonymity, it has not yet been empirically tested since the social media websites are relatively new eWOM platforms. Some researchers consider the anonymity as an advantage for eWOM since it leads consumers to share their opinions more comfortably (Goldsmith and Horowitz 2006). Anonymity is also regarded as one of the important factors, which ensures higher volume of eWOM (Chatterjee 2001). On the other hand, however, some other researchers argue that social media platforms are more advantageous for the eWOM since the less anonymity has potential to make eWOM information more credible (Chu and Choi 2011; Gillin 2007; Wallace et al. 2009). The latter group of researchers anticipate the eWOM on social media to be more influential on consumers’ purchase intentions as it occurs between friends and acquaintances (Chu and Choi 2011; Moran and Muzellec 2014; Park, Lee, and Han 2007). Indeed, this discussion still has not been elucidated yet. The question of whether the eWOM between anonymous people or the eWOM between people who knows each other was more influential on consumers’ purchase intentions still remains uncertain, although the influence of both types of eWOM were separately tested and found influential (Alhidari, Iyer, and Paswan 2015; Elwalda, Lü, and Ali 2016). This study, therefore, proposed the following question: Is eWOM between people who knows each other on social media or eWOM between anonymous people on other online platforms more influential on consumers’ online purchase intentions? The answer of this question is valuable for practitioners as well as researchers since it will reveal which eWOM platforms are more influential on consumers. In this study, we thus empirically test and compare the influence of friends’ recommendations on social media and anonymous reviews on shopping websites; and we expect the eWOM on social media to be more influential as the above-mentioned latter group of researchers. We analyse the influence of these two platforms on consumers’ online purchase intentions based on the components of information adoption model (IAM; Sussman and Siegal 2003) which are applied as information quality, information credibility, information usefulness and information adoption. We conduct this research through survey. However, as we find completely opposite results than expected, we determine to do another study. We conduct in-depth interviews in order to explain our results found in Study 1 and to provide better understanding for the discussion. In the following sections, we begin with a brief literature review about eWOM on social media and shopping websites, and then we introduce Study 1 and Study 2. Finally, we conclude with the discussion of theoretical and managerial aspects of the findings.
EWOM on social media and other platforms
EWOM is considered as one of the most useful information sources by consumers as it consists of peer opinions and experiences instead of company-generated information (Brown, Broderick, and Lee 2007; Mazzarol, Sweeney, and Soutar 2007; Munnukka, Karjaluoto, and Tikkanen 2015). The Internet has facilitated eWOM communication between customers through a variety of platforms (See Table 1) (Cheung and Thadani 2012). However, there is one major difference between social media and other eWOM platforms; social media allows Internet users to communicate with people who they already know while other platforms enable users to communicate anonymously (Kozinets et al. 2010; Moran and Muzellec 2014). Therefore, consumers are increasingly turning to social media websites for knowledge acquisition about unfamiliar brands (Baird and Parasnis 2011; Goodrich and de Mooij 2014; Naylor, Lamberton, and West 2012; Schivinski and Dabrowski 2016); and social media is considered as an appropriate platform for eWOM (Canhoto and Clark 2013; Erkan and Evans 2014; Knoll and Proksch 2015; Toder-Alon, Brunel, and Fournier 2014). Furthermore, eWOM on social media has been found influential on consumers’ purchase intentions by previous researchers (Iyengar, Han, and Gupta 2009; See-To and Ho 2014; Wallace et al. 2009; Wang, Yu, and Wei 2012). On the other side, other eWOM platforms, consumer review websites (Cheung, Lee, and Rabjohn 2008), discussion forums (Chiou and Cheng 2003), blogs (Lin, Lu, and Wu 2012), have also been found influential on consumers’ purchase intentions. In particular, previous studies have focused on the impacts of eWOM on shopping websites through reviews (Chevalier and Mayzlin 2006; Gu, Park, and Konana 2012; Li and Zhan 2011; Park, Lee, and Han 2007); and findings show that although the reviews are anonymous, they play an important role on purchase intention as with other eWOM platforms. Ultimately, research in this field finds both the eWOM between anonymous users and the eWOM between users who already know each other as effective on purchase intention. However, there is a discussion between researchers about the possible further effects of less anonymous eWOM, which arises with social media websites. In this study, thus, we aim to explore whether the eWOM between anonymous people or the eWOM between people who knows each other was more influential on consumers’ purchase intentions. For this purpose, we choose social media and shopping websites. Social media websites encourage people to communicate with their friends and acquaintances, while shopping websites, like Amazon.com, enable users to communicate anonymously through their online reviews sections; and both types of website allow people to exchange opinions about products and services of brands. Shopping websites were selected among the other anonymous eWOM platforms as reaching users of these websites is more convenient when compared to discussion forums, blogs and reviews sites.
In order to compare the influences of eWOM between anonymous people and eWOM between people who knows each other, we first identify the factors of eWOM information which affect consumers’ online purchase intentions and then build the hypotheses of this study based on the identified factors.
Information adoption model
EWOM conversations consist of basic information transfer. The influence of the information, however, may change from person to person; the same content can evoke different notions among receivers (Chaiken and Eagly 1976; Cheung, Lee, and Rabjohn 2008). To understand how people internalise the information they receive, previous studies have focused on the information adoption process (Nonaka 1994). In the information systems literature, researchers have applied dual-process theories to define how people are affected in adopting ideas or information (Bhattacherjee and Sanford 2006; Sussman and Siegal 2003). However, Sussman and Siegal (2003) take it further and narrow its scope by proposing IAM. The IAM specifically explains how people are affected by the information on computer-mediated communication platforms. The IAM is proposed by considering the elaboration likelihood model (ELM) which posits that people can be affected by a message in two routes, which are central and peripheral. The central route refers to the essence of arguments, while the peripheral route refers to the issues which are indirectly related to essence of the message (Cheung, Lee, and Rabjohn 2008; Petty and Cacioppo 1986; Shu and Scott 2014). As this model particularly focuses on the influence of information on computer-mediated communication platforms, it has been considered as applicable to eWOM studies by many researchers (Cheung, Lee, and Rabjohn 2008; Cheung and Thadani 2012; Shu and Scott 2014). In particular, Cheung, Lee, and Rabjohn (2008) apply this model within the online discussion forums context, while Shu and Scott (2014) use it within the social media context. As this paper focuses on eWOM in social media and shopping websites, we also find using the IAM appropriate. We apply its components into this study as information quality, information credibility, information usefulness and information adoption. Hypotheses of the study were then developed based on the mentioned factors.
Information quality & Information credibility
As a result of extensive usage of the Internet, eWOM information can now be created by almost everyone; and thus, quality and credibility of information has become more critical for consumers (Reichelt, Sievert, and Jacob 2014; Yoo, Kim, and Lawrence Sanders 2015). Information quality has been defined as the strength of the meaning embedded in a message (Yeap, Ignatius, and Ramayah 2014). It plays an important role on consumers’ evaluation about products and services (Filieri and McLeay 2014). Also, Park, Lee, and Han (2007) have found that the quality of reviews on shopping websites affects consumers’ purchase intentions. However, since the eWOM occurs in social media is less anonymous, we predict the quality of information on social media has a stronger effect on consumers’ online purchase intentions than the quality of eWOM information on shopping websites. On the other hand, previous research has shown the positive relationship between information credibility and consumers’ purchase intentions (Dou et al. 2012; Hsu and Tsou 2011; Park, Lee, and Han 2007; Prendergast, Ko, and Yuen 2010). In fact, Wathen and Burkell (2002), consider information credibility as an initial factor in the consumers’ persuasion process. Therefore, we regard information credibility as one of the factors which affects consumers’ purchase intentions; and, within the context of this study, we predict that the credibility of eWOM information on social media is more influential on consumers’ online purchase intentions than the credibility of eWOM information on shopping websites.
Information usefulness & information adoption
Information usefulness and information adoption are the two other elements which were proposed in the IAM by Sussman and Siegal (2003). Information usefulness indicates consumers’ perceptions that using information will improve their performance (Cheung, Lee, and Rabjohn 2008). The relationship between information usefulness and consumers’ purchase intentions has been found worthy to study by previous researchers (Chiang and Jang 2007); and later on, the influence of information usefulness on purchase intention has been demonstrated in following research (Liu and Zhang 2010; Xia and Bechwati 2008). Therefore, we include information usefulness to this study and hypothesise H3. Lastly, information adoption is considered as another factor which might affect consumers’ purchase intentions (Cheung and Thadani 2012). Consumers who engage and adopt eWOM information are more likely to have purchase intentions. However, the information adoption process may change in different platforms (Cheung et al. 2009; Fang 2014). Therefore, we predict that the adoption of eWOM information in social media and shopping websites could have a different influence on consumers’ purchase intentions. In fact, as eWOM information has been exchanged among friends on social media instead of anonymous people like in shopping websites, we anticipate that the adoption of eWOM information on social media is more influential on consumers’ online purchase intentions than the adoption of eWOM information on shopping websites. Figure 1 shows the conceptual framework of this study.
To test the hypotheses, a survey was conducted with 384 university students in the UK. University students were considered appropriate for this study because of the latest statistics which present people between the ages of 18–29 as being the larger part of social media users; 89% of this age group use social media websites, as of January 2014 (PRC 2014). Younger age groups are also more familiar with online shopping. According to latest reports, 83% of 16–24 year olds and 90% of 25–34 year olds who live in the UK use online shopping (National Statistics 2014). The sample size of 384 is deemed suitable for studies where the population comprises millions (at 95% confidence level and 5% margin of error) (Krejcie and Morgan 1970; Sekaran 2006). Sample characteristics are demonstrated in Table 2.
The survey was designed using a multi-item approach; a few items were used for measuring each construct in order to enhance validity and reliability. All variables were assessed with a five-point Likert scale ranging from strongly disagree (1) to strongly agree (5). Items were borrowed from previous related literature and specified according to the context of this study. Particularly, information quality was measured with the items developed by Park, Lee, and Han (2007) (α = 0.85, M = 3.37). The items for information credibility were adapted from following scales (Prendergast, Ko, and Yuen 2010; Smith and Vogt 1995; α = 0.89, M = 3.31). Information usefulness was assessed with the items used by following researchers (Bailey and Pearson 1983; Cheung, Lee, and Rabjohn 2008; α = 0.83, M = 3.55). The items for information adoption were adapted from Cheung et al.’s work (2009; α = 0.90, M = 3.47). Lastly, the items for online purchase intention were adopted from Hille, Walsh and Cleveland’ scale (2015; α = 0.90, M = 4.18). Appendix A1 presents all the measures used for this study
Multiple linear regression analysis was conducted to test each hypothesis. Multiple regression analysis allows assessing the relative impacts of independent variables on dependent variable (Cohen et al. 2003). Table 3 shows the results of analyses; all models for four hypotheses were found significant (p < 0.001). However, all hypotheses were rejected. The eWOM information on social media was not found more influential than the eWOM information on shopping websites. Conversely, based on multiple regression results, quality of eWOM information on shopping websites (β = 0.358) was found more influential on consumers’ online purchase intentions than quality of eWOM information on social media (β = 0.000). Likewise, credibility of eWOM information on shopping websites (β = 0.307) was found more effective than credibility of eWOM information on social media (β = –0.036); usefulness of eWOM information on shopping websites (β = 0.412) was found more effective than usefulness of eWOM information on social media (β = –0.124); and lastly, adoption of eWOM information on shopping websites (β = 0.430) was found more effective than adoption of eWOM information on social media (β = –0.096) on consumers’ online purchase intentions.
Discussion of study 1
In this study, we compare the influence of eWOM on social media and shopping websites. Previous studies have tested the effect of eWOM on social media (Iyengar, Han, and Gupta 2009; See-To and Ho 2014; Wallace et al. 2009; Wang, Yu, and Wei 2012) and the effect of eWOM on shopping websites (through reviews; Chevalier and Mayzlin 2006; Gu, Park, and Konana 2012; Li and Zhan 2011; Park, Lee, and Han 2007) separately; and they both have been found influential on consumers’ purchase intentions. However, the effects of these two different platforms had not yet been compared, although there is one major difference between them in terms of eWOM. The online platforms, except social media websites, mostly allow eWOM to occur between anonymous people as it is in shopping websites through reviews. Whereas, social media websites have brought a new perspective for eWOM through enabling people to exchange their opinions and experiences with friends and acquaintances (Kozinets et al. 2010; Moran and Muzellec 2014). For this reason, we expect a significant difference regarding the influence of eWOM on these two platforms and propose four hypotheses. Indeed, the differences between these two platforms have clearly emerged and we find significant results. However, contrary to hypotheses of this study, eWOM on social media was not found more influential than eWOM on shopping websites; all the hypotheses were rejected. On the other hand, contrary to expectation, the multiple regression results also show that eWOM on shopping websites was found more influential on consumers’ online purchase intentions than eWOM on social media. One of the critical parts of the results is the clarity of outcomes. EWOM on shopping websites was found more effective in terms of all components which we tested: information quality, information credibility, information usefulness and information adoption. Although some prior studies anticipate the eWOM between people who already know each other as more effective, as we hypothesised, (Chu and Choi 2011; Chu and Kim 2011; Park, Lee, and Han 2007); these results are in line with the findings of Yeap, Ignatius and Ramayah’s study (2014) which conclude the online reviews as the most preferred eWOM source. Results clearly indicate that there should be some reasons which make eWOM on shopping websites more influential than eWOM in social media. In other words, there should be some reason for consumers to prefer eWOM on shopping websites although they are anonymous. Thus, in order to provide better understanding for this discussion and to enlighten our results found in Study 1, we conduct another study for this research.
In Study 2 we investigate the factors which lead consumers to prefer anonymous reviews rather than friends’ recommendations on social media. For this purpose, we determine related questions and conduct in-depth interviews.
An exploratory approach was used in Study 2. This approach was chosen in order to explore the comprehensive meaning of results found in Study 1 (Willig 2001). Data were collected in the form of 10 semi-structured, in-depth interviews with university students. A purposive, convenience sampling method was chosen as an appropriate method due to the exploratory nature of this study (Esterberg 2002; Turnbull and Wheeler 2014). Ten students (5 male and 5 female) were selected among the participants of Study 1 to achieve interrelated findings. Interviews reached theoretical saturation (Morse 1994) with interview 10; and the data collection process was completed at this point. The data were initially open coded by researchers during the interviews; the emerged themes were noted. Then, after interviews, voice recordings were transcribed to organise the data. The transcripts were examined through thematic analysis to identify key themes within the data (Braun and Clarke 2006). Finally, selective coding was applied to classify possible quotes to present findings (Fielding 2005).
In Study 2, four main themes arose from the data. Consumers prefer online reviews on shopping websites rather than friends’ recommendations on social media because of four main reasons which are as follows: information quantity, information readiness, detailed information and dedicated information
Most interviewees expressed the importance of number of the reviews; and they indicated that they do not frequently see friends’ posts about brands or their products and services on social media. However, shopping websites have lots of product reviews and this gives an opportunity to compare the different comments. Therefore, information quantity is one of the reasons for consumers to prefer online reviews. Reviews on social media, from my friends? I don’t see them often, not often at all. Once in a blue moon, that’s why I wouldn’t consider it really … and even if you see, what’s the chances that you’ll have two or three reviews from friends at one time on the same product? (ID: 1) I prefer reviews, because it is more. For example, on Amazon.com, you can see up to 300 reviews; but from friends on social media … it is just one product and it is just one or two friends. This is not really tangible. So, if I want to buy a product, numbers of the reviews gives me that confidence … Because, you know, if 300 people are saying good things, that means, something must be good about this product. (ID: 10) Interviewees also talked about how information quantity helps them to overcome the anonymity issue on online reviews. We of course don’t know the personality of reviewers on shopping websites; he or she can be a person who doesn’t like anything. However, when I continue to read the other reviews, I understand the overall opinion about the product. If there are 5000 comments and if the majority of people like the product, then how can I consider that specific person’s comment? There might a problem with his personal preferences. (ID: 6)
Always being ready is another reason which makes reviews on shopping websites superior to friends’ recommendations on social media. When consumers need information about products, online reviews on shopping websites are always ready to use. However, social media does not provide that convenience for consumers. Reviews are not always available on social media; you can’t find them always. That’s why I follow the other consumers’ reviews. They are easy to find and ready to use. (ID: 5) When I am looking for information about products, shopping websites like Amazon.com is ready for it, reviews are there … but in social media, it is very difficult to find the information that I need. Even if I scroll down and go back to posts shared within last 2 years, it is still not easy to find. (ID: 4)
Interviewees also described the significance of detailed information. They emphasised that the information shared by friends about brands on social media lack detail, while the reviews on shopping websites provide answers for specific questions. You can reach detailed information about products through online reviews. Plus, you can find the comparison of the different products for the same purpose. Reviewers explain his/her personal experience with the product and sometimes you can see how expert they are. However, on social media, comments of my friends usually don’t have details … Also, they don’t compare different products as it in online reviews. That’s why, reviews on shopping websites naturally more satisfying. (ID: 3) I read quite a lot of reviews on shopping websites … because some people take the time to write, you know, sometimes a paragraph or so … and I mean all the while adding up, this is building an image of the product to me and how it works efficiently. I see some products have faults, so these all contribute. (ID: 1) Furthermore, interviewees mention that they find information about both the positive and negative sides of products through online reviews on shopping websites, whereas the eWOM information on social media refers to either the positive or negative sides of products. On social media, sometimes I see very brief positive or negative information about products and services. Rather than detailed information, I can only see suggestions with very short sentences such as ‘Try it, it’s nice’ and ‘I hate it, please avoid’. … However, I can’t see a great evaluation about products with both positive and negative sides … but it is not like this on shopping websites, for example on Amazon.com. (ID: 7)
This study set out with the aim of comparing the influences of friends’ recommendations on social media and anonymous reviews on shopping websites in the context of online purchase intention. In fact, based on the components of IAM, we hypothesised that eWOM information on social media has a stronger effect on consumers’ online purchase intentions than eWOM information on shopping websites. However, in contrast to our expectations, the results of Study 1 showed that eWOM information on shopping websites is more influential on online purchase intention than eWOM information on social media in terms of information quality, information credibility, information usefulness and information adoption. In order to provide better interpretation for these results, we conduct another study through in-depth interviews. Findings of Study 2 enlightened the survey results and explained why anonymous reviews are more influential than friends’ recommendations. Information quantity, information readiness, detailed information and dedicated information are found as key themes which make shopping websites better than social media in terms of the impact of eWOM.
Theoretical and managerial implications
This study contributes eWOM literature through elucidating an uncertainty. Some previous studies consider the anonymity issue as an advantage for eWOM (Chatterjee 2001; Goldsmith and Horowitz 2006), while many others expect the opposite; second group of studies anticipate the eWOM on social media as more effective since it occurs between people who already know each other (Chu and Choi 2011; Moran and Muzellec 2014; Park, Lee, and Han 2007). This study enlightens the mentioned discussion through its empirical approach. Secondly, we provide new constructs for researchers; findings reached through in-depth interviews (information quantity, information readiness, detailed information, dedicated information) can be tested as a component of new theories and models in the future studies. Finally, we intentionally preferred to examine the consumers’ online purchase intentions rather than purchase intentions in order to contribute to the related literature; because there are less studies focused on online purchase intention (Hille, Walsh, and Cleveland 2015; Wen 2009), while there are many studies about purchase intention (Park, Lee, and Han 2007; Prendergast, Ko, and Yuen 2010; See-To and Ho 2014; Wang, Yu, and Wei 2012). On the other hand, in terms of practicality, results of Study 1 shows consumers’ preferred eWOM platforms, which is very valuable for marketers who want to utilise eWOM marketing. Also, findings of Study 2 highlight the aspects of eWOM information that consumers consider; marketers can develop better eWOM marketing strategies by considering consumers’ expectations.
Limitations and future research
Although the paper provides considerable amount of theoretical and managerial implications, the following limitations should be noted. Firstly, this research was conducted with participants who are university students. Although they constitute the majority of social media website users and are more familiar with online shopping websites, they may not precisely reflect the whole population. In addition, since the study was conducted only in UK, it might be difficult to generalise the findings to other countries. Also, this research regards all social media websites together, rather than focusing on one website such as Facebook or Twitter; likewise, in terms of shopping websites. Future research could focus the eWOM on one specific social media website and one specific shopping website. Future research could also focus the eWOM about one specific product type. Finally, further studies could test our findings by adding them to current models and theories as a part of new components, which can bring new theoretical and managerial insights.
We thank PhD researchers Erhan Aydın and Abdulaziz Elwalda for their valuable comments and insights into this article. Also, thanks to Dr Chris Evans for encouraging this research.