Informational quality

The brand page is principally presented as an application of multimedia technology. In order to achieve its purpose, it has to provide some functional value for visitors. Brand pages are quite different in how they present information which may be a function of the product, service, brand philosophy or several other factors. However, without attempting to understand the underlying philosophy or objectives of the brand, we attempt to look at a variable denoting the subjective assessment of the brand page. Wu [27] attempted to codify the information available on a brand page using several parameters. This includes the accuracy of the information, the validity of data, the objectivity amongst others. This is also referred to formally as an informational quality which can be conceptualized as the perceived overall value and excellence of information posted on the brand page. A brand page high on this overall measure will provide a great deal of value for visitors [28] . It can thus be posited that the level of customer engagement will increase with higher levels of information quality. Therefore, we propose:

H1: Information quality of the brand page positively influences customer engagement with the brand page.

Social networking sites were initially mooted as information sharing platforms and have slowly morphed into platforms for social interaction [29] . Interaction on brand pages takes the form of exchanging opinions, reviews and experiences which are related to the brands in question [25] . The provision for customer interaction has become so imperative that today most brand pages elicit opinions from visitors and encourage interaction. This was in line with the argument made by Liang, Ho, Li, & Turban [30] support of users’ social interaction is important for brand pages since this, in turn, increases the level of confidence in the brand. An additional advantage of such encouragement is the strengthening of visitors’ social identities with the brands. Therefore, we hypothesize that increasing interactivity between customers should increase social identity and ultimately customer engagement.

H2: The quality of social interaction in the brand page positively influences the customer engagement with the brand page.

The Self, as an abstract concept, is partially created by an individual dynamically through consumption [31] [32] . Thompson [33] describes the self as “a symbolic project, which the individual must actively construct out of available symbolic materials, materials which the individual weaves into a coherent account of who he or she is, a narrative of self-identity”. Self-concept has been defined as “the relatively stable schemata of oneself that are generalized to the extent that they refer to an individual’s view of him- or herself across different situations” ( [34] , p. 95). Social media are popular platforms for self-expression, communication and self-promotion. The emergence of what Castells [35] has termed “mass self-communication” as a global system of networked interaction—a system within which platforms like Facebook, LinkedIn, Google Plus¹ and Twitter offer crucial tools to galvanize the expression of online identity.

The personal image or the product user image is “the image formed of a generalized user of the product through advertising” [36] . Brands are consumed not only for their utilitarian purpose but also for the symbolic associations of the brands projected in their image [37] . Prior research suggests that customer’s consumption is directly related to their perception of similarity of the product-user image to their own self-image [36] [38] [39] .

The relationship of a person with a brand in terms of the strength and depth has been a subject of study for quite a while now [36] . Several researchers have reported how consumers tend to feel that possessions are an extension of themselves and help define their sense of self [40] [41] . While describing brands, consumers also go so far as to project their own personality characteristics onto them [42] .

Goffman [43] argued that individuals tend to sculpt their personalities using the mode of intentional presentation. Sometimes, several of these displays are not intentional but yet become a part of self-expression. In the context of SNSs, while conscious self-promotion is the overt objective of a user, there is also unconscious self-expression. These are designed by the owners of these platforms using sophisticated interface technologies. Attachment to a brand is thus likely to increase in cases where customers project their self on brands. Therefore, we hypothesize that:

H3: Self-expressive behaviour of the respondent has a positive impact on customer engagement with the brand page.

Brand pages on the SNSs are typically very heterogenous in their appearance and content. This is usually a reflection of the brand imagery and the objectives of the brand for the page. Brand pages also vary by the type of media used on the page. This can be differentially measured using the variable of media richness. The media richness theory [44] defines contextual cues in a communication medium as its “richness”. “Richness” in the form of visual cues generally tend to increase the effectiveness of the medium. Daft and Lengel [44] have attempted to study media richness and find that it is a function of several characteristics of the medium. This could the variety of language, the number of cues and the capacity for prompt feedback amongst others. SNS pages, as used by brands, offer different media types which, in turn, affects media richness of these pages. Brand pages and posts therein can be bare in the form of text or could include audio-visual media. Hence, one could conceptualise a continuum of message richness for these posts on brand pages. At one extreme one could have text-only posts which are leaner media-types and at the other extreme one could have posts containing embedded photos, videos, or URLs. The effectiveness of the medium will also increase with this gradation. Hence, we anticipate that the use of richer media, such as photos, videos, or URLs, is associated with higher customer engagement. Therefore, we argue that the inclusion of richer media types such as photos, videos, or URLs, is associated with higher customer engagement and it is expected that more media rich brand page will show stronger relation with the antecedents of customer engagement.

Studying the moderating effects of media richness on the relationships between the independent and dependant variables will be of theoretical interest [45] [46] . In customer relationships, moderation takes place when the variable M (moderating variable) modifies the relationship between the variables X (independent variable) and Y (dependent variable) [47] by either increasing or decreasing the strength of the relationship. For the purposes of this study, media richness was manipulated at two levels: less and morerich. Based on the above discussion, three further hypotheses can thus be formulated.

H4: Media richness will strengthen the positive relationship between information quality of the brand page and customer engagement with the brand page.

H5: Media richness will strengthen the positive relationship between interaction quality of the brand page and customer engagement with the brand page.

H6: Media richness will strengthen the positive relationship between self-expressive behaviour elicited on the brand page and customer engagement with the brand page.

All the hypotheses discussed above are visually presented in Figure 1.

Data for the analysis was collected using a questionnaire survey. Most of the scales used to measure the constructs in this study are adapted from previous studies in order to ensure content validity. Items measuring customer engagement—vigour, dedication, and absorption were adopted from Cheung et al., [18] . The self-expressive behaviour was measured using a scale developed by Higie and Feick (1989). To study interpersonal interaction and information quality, items from previous studies were adopted [48] . Media richness was measured by using consumer perception about the richness of the brand page in terms of inclusion of media types such as audio-visuals, photos and other links.

The survey was conducted in three stages for the present study: qualitative field interviews followed by a pilot study and the final study through a questionnaire. 20 exploratory interviews were conducted (8 academicians, 4 industry experts and 8 end consumers) with the intent to clarify the constructs and structure of the study framework. The academicians were selected on the basis of their expertise in the areas of digital marketing and word of mouth marketing. The industry experts were practitioners who were handling roles pertaining to digital marketing. The experts were drawn from the acquaintance pool of the authors. The end-consumer respondents were selected using convenience sampling. The interviews lasted between 40 - 60 minutes. This step was necessary for development of the instrument. The instrument was then checked for its face validity and content validity issues with the help of academic and industry experts. Later, a pilot study was conducted using the instrument with a sample of 154 respondents. The objective of the pilot study was to test the reliability of the scales used to measure different constructs in the instrument. After examining the results from these preliminary respondents, the questionnaire was finally distributed amongst the final sample for the research among students/executives/alumni of an urban university. The sample was diverse with respect to key demographics including computer knowledge. Each item was measured using a six-point Likert agreement scale (with 6 denoting strongly agree and 1 denoting strongly disagree)

Nearly 1500 respondents were contacted and administered a screening questionnaire ascertaining whether they were active on social media and were exposed to brand pages on social media. Around 955 respondents were found eligible and the final questionnaire was distributed amongst them. Only 667 completely filled questionnaires were returned. However final usable sample after the removal of suspicious pattern data, outliers etc. was 479. There were 23 items in the model and thus the sample size of requirements of both 5:1 [49] and 10:1 [50] were met. Table 1 provides certain descriptive statistics of the sample. Nearly 60% of the respondents were males while 40% were females. A large majority of the respondents were in the age group of 20 - 30 years and a significant majority of respondents (46.56%) had encountered brand pages on Facebook.

Customer engagement emerged as a second-order construct. In the Confirmatory Factor Analysis (CFA), customer engagement emerged as a three-factor second-order reflective construct. The second order was considered as the CFA showed high correlations between the four constructs. The results clearly support that customer engagement is explained through the three dimensions viz., vigour, dedication, and absorption. The fit indices (χ²/df = 3.65, GFI = 0.89, RMSEA = 0.074, NFI = 0.884, CFI = 0.82) suggest that the model with the three latent variables represents a good fit to the data. Thus, customer engagement was a holistic construct encompassing the four sub-constructs. The other three constructs were first-order reflective constructs.

Test of convergent validity was conducted to assess the extent to which the scale

items were theoretically related. Convergent validity assessment involved three criteria, 1) at least 0.70 composite reliability, 2) at least 0.50 average variance extracted [52] , and 3) greater than 0.50 item loadings on their expected factor and less than 0.40 on other factors [53] .

Table 2 shows that all convergent validity criteria were met, i.e., CR values ranged from 0.79 to 0.87, average variance extracted values ranged from 0.52 to 0.57, and item loadings were higher than 0.60.

For the SEM model, we used AMOS20 to test whether the empirical data conformed to the proposed model. The model included 23 items describing 6 latent constructs: interpersonal interaction, information quality, self-expressive behaviour, customer engagement and its three dimensions, viz., vigour, dedication, and absorption. We examined the model fit of our research model, as shown in Table 3. The criteria used in the SEM were as suggested by Hair et al. [51] . Although the value of GFI (0.87) was slightly less than the recommended value (0.90), all other fit indices were acceptable. Thus, the results indicated adequate model fit between our research model and the empirical data.

To test the significance of each hypothesized path in the research model, AMOS reports raw and standardized estimates for all specified paths, as well as standard errors and test statistics for each path. The result of the structural equation model is shown in Figure 2. All the relationships were supported in the model. 71% variance in customer engagement was explained by the model. The results indicate that information quality has a significant positive relationship with customer engagement. Interpersonal interaction has a significant and positive relationship with customer engagement and self-expressive behaviour also has a positive relationship with customer engagement.

Media richness was then included as part of the model to test for moderation. The sample was split into two to compare the coefficients of each pair respectively. The significance of the moderating effects was found by calculating T-value for cross-multiple of the moderator and dependent variables [54] [55] . Table 4 has results for the comparison of path coefficients. The path coefficients from interpersonal interaction and self-expressive behaviour to customer engagement are significantly higher for more rich media brand pages as compared

** for p ≤ 0.01 and * for p ≤ 0.05; S: Supported, PS: Partially Supported, NS: Not Supported.

to less rich media. However, media richness showed no impact on the relationship between information quality and customer engagement. This indicated that if the brand page is less media rich then interaction quality has insignificant impact on customer engagement. Thus, more media rich brand pages are important to engage customers. Firms need to create media rich interface to engage customers. Similarly, customers exhibiting self-expressive behaviour are more inclined to brand pages that are media rich.