KOREASCHOLAR

Introduction
The social Web is becoming more and more visual. While Facebook remains the leading social networking site (SNS), visual-based platforms such as Instagram and Pinterest have showed a significant increase in the proportion of online users worldwide (Greenwood, Perrin, & Duggan, 2016; Statista, 2017; Pew Research Center, 2017). The trend toward visual communication on SNSs is influenced by the emergence of advanced mobile technologies and the shifting habits of how social media users consume content today. While images displayed on Instagram are past-oriented sharing of memories, or present-oriented displays of creativity, Pinterest presents opportunities for future-oriented inspiration to evolve (Jin, Lin, Gilbreath, & Lee, 2017; Phillips, Miller, & McQuarrie, 2014). Pinterest also serves as a searchable archive of visual content, allowing users to seek and include brand-relevant pins and boards into their own aspirations and inspirations, thus facilitating the sharing of branded content (Phillips et al., 2014). Extant industry reports also suggest that Pinterest users are more likely to image-search for products, interact with sponsored pins, and show favorable attitudes toward advertising, compared with Facebook or Twitter users (McMullen, 2015; Ruscillo, 2017; Williamson, 2015). Considering the benefits of social media for global marketing and the role that imagery plays for developing international brands (Mikhailitchenko, Javalgi, Mikhailitchenko, & Laroche, 2009), this research focuses on the visual aspects of consumer engagement with branded content on Pinterest and investigates how such engagement leads to consumer-brand relationship building and Pinterest-inspired purchase intention.
Theoretical Frameworks
The extant consumer research has documented the superiority effect of visual content on facilitating persuasion (Childers & Houston, 1984; Miniard, Bhatla, Lord, Dickson, & Rao, 1991). Recent research on social media marketing also suggests that branded content with visual elements help generate favorable consumer responses (De Vries, Gensler, & Leeflang, 2012; Trefzger, Baccarella, & Voigt, 2016). The presentation of branded posts is directly related to the flow construct and trust (Bart, Shankar, Sultan, & Urban, 2005; Hoffman & Novak, 1996). Along this logic, consumer engagement, the degree to which consumers consume, contribute, and create social media content in relation to brands (Schivinski, Christodoulides, & Dabrowski, 2016), is likely to be influenced by brand-related visual cues on Pinterest. While consumption refers to consumers’ acquisition of both brand-created and user-generated content (Muntinga, Moorman, & Smit, 2011), contribution refers to consumers’ inputs to content created by brands or by other consumers (Shao, 2009). Further, creation is the highest level of engagement, which refers to consumers’ generation of brand-related content that could stimulate further consumption and contribution (Muntinga et al., 2011). Therefore, the visually dominant and appealing presentation style, such as appearance, layout, and images, on Pinterest would provide users with high information content, and is likely to increase both active and passive consumer engagement with brands (Bart et al., 2005). Consumer engagement is then expected to encourage social media users’ willingness to actively promote and endorse brand-related content, including organic and sponsored pins. Such volunteer display of their preferred brand-related content is likely to help engender eWOM communication and make such content become more influential in the social Web (Chu & Kim, 2011). As a result, users’ brand-related Pinterest activities are likely to enhance their relationships with brands (Rapp, Beitelspacher, Grewal, & Hughes, 2013) and their intention to purchase from the brands featured in the pins and boards (King, Racherla, & Bush, 2014). All hypothesized relationships are therefore formulated and displayed in Figure 1.
Methods
To test the proposed conceptual model, an online survey was conducted with participants recruited from a consumer panel administrated by Qualtrics. A total of 467 Pinterest users (58.2% female, Mage = 28.9, SDage = 5.6) comprised the sample, including Caucasian (76%), African American (7.5%), Hispanic (7.4%), Asian (6.9%), multiracial (1.3%), and others (.9%). The participants spent an average of 9.5 hours on Pinterest per week. Smartphones (54.8%) emerged as the most dominant devices for their Pinterest usage, followed by computers (34.9%) and tablets (10.3%). They had an average of 33.7 boards (Range = 1,233), 1,384.1 Pins (Range = 69,000), 930.1 likes (Range = 100,000), 319.0 followers (Range = 44,231), 65.0 following topics (Range = 3,442), 173.0 following people (including brands, Range = 7,000), and 102.0 following boards (Range = 6,000). To note, 78.3% of the participants reported that they made Pinterest-inspired purchases frequently (including 24.8% always, 24.4% usually, and 29.1% often). In addition, the participants were asked to answer questions pertaining to their perceptions of visual cues related to a self-selected brand on Pinterest, engagement with brand-related content, brand endorsing behavior, consumer-brand relationships, and Pinterest-inspired purchase intention (see Table 1).
Results
To test the proposed hypotheses, a two-step modeling approach following Anderson and Gerbing’s (1988) procedures was employed. The results of confirmatory factor analysis (CFA) showed that the measurement model achieved acceptable fit for the data (χ2 = 1178.30, df = 393, χ2/df = 2.99, RMSEA = .065, SRMR = .054, CFI = .930, TLI = .922). The constructs had good composite reliability (>.70). Factor loadings of the construct indicators were all above .70 with significant t values, which indicated good convergent validity for each of the construct items (Nunnally & Bernstein, 1994). Following Fornell and Larcker’s (1981) criteria, the results achieved satisfactory discriminant validity because the average variance extracted (AVE) for each factor was above .50 and greater than the squared correlations between each pair of constructs. The factor scores were then used to perform a path analysis for hypothesis testing which showed satisfactory fit of the conceptual model (see Figure 2). The results indicated that visual cues positively influenced consumer engagement with brand-related content (β = .68, p < .001), which positively led to brand endorsing behavior (β = .86, p < .001). Consequently, brand endorsing behavior positively enhanced the development of consumer-brand relationships (β = .97, p < .001) and increased Pinterest-inspired purchase intention (β = .48, p < .001). In turn, all hypotheses were supported.
Discussion and Conclusion
In response to the thriving academic interests in understanding how social media users connect and interact with international brands across SNSs (Hennig-Thurau, et al., 2010), this research sheds light on factors that determine social media users’ active discovery, creation, and distribution of visual branded content and the effects of consumer engagement on social media marketing outcomes. By focusing on Pinterest, the research findings showed that users, both men and women, actively seek inspiration from Pinterest across brand-relevant categories. They are highly engaged, open to brand activities, and ready to take Pinterest-inspired actions in the real world. The visual presentation of branded content plays an important role in generating consumer engagement, which is likely to act as a form of perceived relationship investment and subsequently stimulates brand endorsing and advocacy behavior (Simon & Tossan, 2018). Engaged Pinterest users search and select brand-relevant visual content, including sponsored pins, as part of their self-presentation to others (Phillips et al., 2014). Through such interactions with brands and other like-minded users on the platform, their relationships with brands and purchasing behavior are likely to be further enhanced. This research integrates several important literature streams in the conceptual model and employs a representative adult sample to provide theory-driven and generalizable insights into how visually-focused social media marketing works. Marketing managers face opportunities and challenges that social media and the interconnected international business environment present. Therefore, it is crucial to understand how social media users interact with brands on different platforms in order to tailoring brand strategies accordingly for fostering consumer engagement. Specifically, as Pinterest’s popular visual discovery engine continues to attract myriads of global users, managers should harness the power of visual content marketing and engage current as well as potential consumers on the platform, which could have a significant impact on consumer-brand relationship building and business performance in domestic and foreign markets. Building on the findings of this research, future studies should further delve into the psychological underpinnings and motivations that drive consumer engagement with visual branded content on SNSs across cultures and discover how paid, owned and earned brand communication on visual-based SNSs can help strengthen consumer-brand relationships.

• Jhih-Syuan (Elaine) Lin(National Chengchi University, Taiwan (ROC))
• Kuan-Ju Chen(National Central University, Taiwan (ROC))