I.Introduction

As an increased number of consumers engage in social networking sites (i.e., Facebook, Instagram, and Twitter) on a daily basis (Yuksel, Milne, & Miller, 2016), social media has become an important part of the marketing strategy for firms (Godey et al., 2016). In particular, a recent consumer survey revealed that 72.7 million millennials in the U.S. used social media, these young consumers spent 70% of their social media time on Facebook (Statista, 2016a; Statista, 2016b). Millennials (birth years 1983 to 1999) have received a great deal of attention from marketers because this generation will soon become the largest generation in the U.S. (Goodman, 2015). Their annual spending is estimated to be over $200 billion in 2017 (Solomon, 2015).

Due to rapidly growing young consumers’ interaction with companies through social media in the U.S., most marketers (96%) engage in social media marketing as part of their digital marketing strategy (Stelzner, 2015). Almost 70% of U.S. marketers increased their budget for social media in 2017 (eMarketer, 2016). Social advertising and commerce were reported to be the most important strategies for the digital marketing industry over the last 12 months (eMarketer, 2016). Social commerce refers to the merging of consumer shopping and social media experiences. Social commerce differs from online or mobile commerce because it allows individuals to form virtual communities and participate in various activities such as marketing, selling, buying, and sharing of products and services (Hajli, Sims, Zadeh, & Richard, 2017; Menon, Sigurdsson, Larsen, Fagerstrøm, & Foxall, 2016). U.S. social commerce sales increased nearly 470% between 2012 (three billion) and 2015 (14 billion) and almost 85% of orders from social media platforms came from Facebook (Saleh, 2014).

A recent study shows that social media advertising and commerce are popular among fashion brands to increase consumer website visits and sales (Touchette, Schanski, & Lee, 2015). Among diverse social networking sites, Facebook and Twitter are the most popular with luxury fashion brands (Kim & Ko, 2012) and apparel brands (Touchette et al., 2015). For instance, Burberry, an English luxury fashion brand, offers a free sample of its fragrance to Facebook fans who sign up for the sample on their Facebook page. Express, the U.S. casual fashion brand, allows consumers to browse products and make efficient transactions on Facebook without moving to their online shopping site (Touchette et al., 2015).

Prior research has investigated consumers’ attitudes and intentions toward online and mobile commerce, whereas little research has examined fashion shopping behaviors in social commerce. A few research studies investigated consumer motivations for shopping through Facebook pages (Anderson, Knight, Pookulangara, & Josiam, 2014), the effects of social media marketing efforts of luxury fashion brands on customer equity and purchase intention (Kim & Ko, 2012), factors affecting continuous use intention of social network service marketing (Ko & Kim, 2014), and brand equity and consumer response (Godey et al., 2016). Despite the popularity of social media platforms among millennials (Bolton et al., 2013), empirical findings are limited to reveal factors associated with young consumers’ intentions to use social commerce in fashion shopping. Given the rapid growth of millennials’ social media use as discussed above, further investigation is needed to understand those factors.

The technology acceptance model (TAM; Davis, 1989; Davis 1993; Davis, Bagozzi, & Warshaw, 1992) was adopted to develop a theoretical framework for the present study. The TAM explains that an individual’s information technology acceptance behaviors are determined by five variables including perceived ease of use, usefulness, and enjoyment; as well as attitudes and intentions to use information technology. Drawing on the TAM, the purpose of this study was to examine the effects of perceived ease of use, perceived usefulness, and perceived enjoyment of social commerce on young consumers’ attitudes toward social commerce and their intentions to use social commerce in fashion shopping. This research contributes to expanding the body of research on digital retailing for fashion products.

Ⅱ.Literature Review

1.Theoretical framework: Technology acceptance model

The TAM model (Davis, 1989) originated from Fishbein and Ajzen’s (1975) theory of reasoned action (TRA), which posits that an individual’s intention to perform social behavior is likely to be influenced by the individual’s beliefs regarding social pressure and his/her attitude toward that behavior. Individuals are willing to perform particular behaviors when they perceive expectations from other social members who perform that behavior (Fishbein & Ajzen, 1975). Based on the causal link between beliefs, attitudes, and intentions to perform a behavior, Davis (1993) proposed that an individual’s information technology acceptance behaviors are determined by perceived usefulness, perceived ease of use, attitudes toward using information technology, and intentions to use information technology. Perceived ease of use is defined as, “the degree to which a person believes that using a particular system would be free of effort” (Davis, 1989, p.320). Perceived usefulness refers to “the degree to which a person believes that using a particular system would enhance his or her job performance” (Davis, 1989, p.320). Individuals who perceive a high level of usefulness for information technology are more likely to use the technology (Davis, 1989). According to Davis (1993), perceived ease of use has a direct impact on perceived usefulness. When individuals perceived that using information technology would be easy, they believed the system would be useful and perform their job as they expected. Attitudes toward using information technology refer to an individual’s positive or negative response toward using information technology (Davis, 1993). Intentions to use information technology refers to an individual’s willingness to use information technology (Davis, 1993). Davis’ later research (Davis et al., 1992) added perceived enjoyment to the TAM model and suggested potential influence of perceived ease of use on perceived usefulness and enjoyment. Perceived enjoyment is defined as an individual’s perception that using information technology is enjoyable (Davis et al., 1992).

The TAM has been applied to numerous studies that investigate purchase behaviors in the contexts of online commerce (e.g., Ha & Stoel, 2009; Tong, 2010; Yu & Park, 2014) and mobile commerce (e.g., Kim, Ma, & Park, 2009; Ko, Kim, & Lee, 2009; Nysveen, Pedersen, & Thorbjørnsen, 2005). Research on online fashion commerce yields mixed results. For instance, Ha and Stoel (2009) found that perceptions of usefulness and enjoyment positively influenced attitudes toward online commerce, whereas perceptions of ease of use did not influence attitudes toward online commerce. Yu and Park (2014) indicated that perceived enjoyment had a positive, direct impact on attitudes, but not the other two variables.

Research on mobile fashion commerce consistently provide empirical supports for the relationships between variables in the TAM model. For instance, Kim et al. (2009) reported that perceived ease of use of mobile phones had a significant impact on perceived usefulness and perceived enjoyment, which influenced on attitudes toward mobile commerce. Ko et al.’s (2009) mobile commerce research found that Korean consumers who perceive mobile commerce as useful, easy to use, and enjoyable perceived value in using mobile commerce and showed positive intentions to adopt mobile commerce in fashion shopping. Nysveen et al. (2005) found that ease of use, usefulness, and enjoyment significantly affected attitudes toward using mobile services, which influenced the intentions to use mobile services.

2.Effect of perceived ease of use on usefulness and enjoyment of social commerce

Perceived ease of use influenced both perceived usefulness and perceived enjoyment (Davis et al., 1992). Empirical supports have been found for the relationship between perceived ease of use, perceived usefulness, and perceived enjoyment (e.g., Cheema, Rizwan, Jalal, Durrani, & Sohail, 2013; Ha & Stoel, 2009; Kim et al., 2009; Kwon & Chidambaram, 2000; Nysveen et al., 2005; Tong, 2010). For instance, two studies (Cheema et al., 2013; Ha & Stoel, 2009) consistently found that perceived ease of use significantly influenced perceived usefulness of online commerce. Tong’s (2010) findings confirmed the positive effect of perceived ease of use on perceived usefulness of online commerce not only for the U.S., but also for Chinese consumers. Similarly, Nysveen et al. (2005) found the positive influence of perceived ease of use on perceived usefulness for adolescents using mobile services (e.g., payment and text messaging). Researchers (Kim et al., 2009; Kwon & Chidambaram, 2000) have found empirical support that perceived ease of use was a significant determinate of perceived enjoyment. Individuals who felt online and mobile commerce was easy to use were more likely to perceive online and mobile commerce as useful and enjoyable. Based on the positive relationships between the three previously studied variables from previous studies on online and mobile commerce, the following hypotheses were proposed:

3.Effect of perceived ease of use on attitudes toward social commerce

Examination of the relationship between perceived ease of use and attitudes toward online commerce yielded mixed results. Empirical support for the positive relationship between perceived ease of use and attitudes toward online commerce was found in previous studies (Chen, Gillenson, & Sherrell, 2002; Chen & Tan, 2004; Lee, Fiore, & Kim, 2006; O’Cass & Fenech, 2003; Vijayasarathy, 2004). For example, Lee et al.’s (2006) findings revealed a significant, positive impact of perceived ease of use on attitudes toward image interactivity technology of retail websites. However, researchers (Cho & Fiorito, 2009; Ha & Stoel, 2009; Liu, Tucker, Koh, & Kappelman, 2003) found insignificant effects of perceived ease of use on attitudes toward online commerce. Likewise, perceived ease of use of co-designing shoes had no direct effect on attitudes toward co-designed shoes; however perceived ease of use indirectly influenced attitudes toward co-designed shoes through product performance risk (Yu & Park, 2014). Perceived ease of use was regarded as a utilitarian motivation in online commerce (Childers, Carr, Peck, & Carson, 2001; Lee et al., 2006) which has been found to have a significant impact on attitudes toward online commerce (Childers et al., 2001). In accordance with these findings, the present study proposed that consumers who perceive social commerce as easy to use may be likely to exhibit a positive attitudes toward social commerce. Thus, the following hypothesis was proposed:

4.Perceived usefulness of social commerce and attitudes toward social commerce

Previous research demonstrated perceived usefulness as a strong determinant of attitudes toward online and mobile commerce (e.g., Ha & Stoel, 2009; Kim, 2012; Lee et al., 2006). Kim (2012) found that perceived usefulness positively influenced attitudes toward online commerce. Two studies (Ha & Stoel, 2009; Lee et al., 2006) found that the effect of perceived usefulness was much stronger than that of perceived ease of use on attitudes toward online commerce. Likewise, research on mobile commerce found perceived usefulness to be a significant indicator of attitudes toward adopting mobile commerce (Yang, 2012). In mobile technology literature, consumers who perceived mobile communication as useful had positive attitudes toward mobile communication (Nysveen et al., 2005). Based on the positive relationship between perceived usefulness and attitudes toward online and mobile commerce, it is logical to suggest a positive relationship for social commerce. Thus, the following hypothesis was proposed:

5.Perceived enjoyment of social commerce and attitudes toward social commerce

Researchers (Kim et al., 2009; Lee et al., 2006; Yang, 2012; Yu & Park, 2014) found that perceived enjoyment significantly influenced attitudes toward online and mobile commerce. For example, Yu and Park’s (2014) findings showed that perceived enjoyment of co-designing had a strong and positive influence on attitudes toward online co-design practice. Compared to the impact of perceived ease of use and usefulness on attitudes, perceived enjoyment had the strongest direct impact on attitudes toward adopting online commerce (Lee et al., 2006) and mobile commerce (Kim et al., 2009). Yang (2012) found that perceived enjoyment positively influenced attitudes toward adoption of mobile commerce. Based on the positive relationship between perceived enjoyment and attitudes toward online and mobile commerce, it was plausible to suggest the positive influence of perceived enjoyment on attitudes toward social commerce. Thus, the following hypothesis was proposed:

6.Attitudes toward social commerce and intentions to use social commerce

According to Fishbein & Ajzen (1975), an individual’s attitude toward a behavior positively influences the specific behavior. Prior studies have found that positive attitudes lead to intentions to use online and mobile commerce (e.g., Lee et al., 2006; Park & Kim, 2007; Yang, 2012; Yoh, Damhorst, Sapp, & Laczniak, 2003). Researchers have confirmed that consumers’ intentions to use online commerce were determined by positive attitudes toward online commerce (Lee et al., 2006; Park & Kim, 2007; Yoh et al., 2003). Similarly, consumers’ intentions to use mobile commerce were influenced by positive attitudes toward mobile commerce (Kim et al., 2009; Yang, 2012). A study on mobile services (Nysveen et al., 2005) also found the positive relationship between attitudes and intentions to use mobile services. Since the relationship between attitudes and intentions to use online and mobile commerce has been found to be significantly related, it is logical to suggest a positive relationship between attitudes toward social commerce and intentions to use social commerce. Thus, the following hypothesis was proposed:

Ⅲ.Method

1.Sample and data collection

A convenience sample of college students over 18 years of age was recruited from a major mid-South university in the U.S. to conduct an online survey. The college student sample was appropriate because millennials are a highly educated generation who often obtain college degrees (Goodman, 2015). Most college students are active social media users (eMarketer, 2015). After obtaining approval from the Institutional Review Board (IRB, a self-administered online survey was conducted for data collection. As an incentive for participation, students received extra credit points added to their course grade (i.e., 5 points out of 1,000). A total of 768 email invitations were sent to the class list by the course instructors. Participants were given seven days to log onto the Website and complete the survey.

2.Survey instruments

Reliable and valid scale items were adapted from previous studies and modified to relate to the topic of social commerce. The questionnaire consisted of four sections: (a) perceived ease of use of social commerce, perceived usefulness of social commerce, and perceived enjoyment of social commerce, (b) attitudes to ward social commerce, (c) intentions to use social commerce, and (d) demographic characteristics including age, gender, ethnicity, and education attainment. All scale items except demographic information were measured using five-point Likert-type scales ranging from strongly disagree (1) to strongly agree (5). At the beginning of the survey, participants were asked to indicate their past and current experience of social media use.

Twelve items from two studies (Davis et al., 1992; Kim et al., 2009) were adopted to assess perceived usefulness of social commerce, perceived ease of use of social commerce, and perceived enjoyment of social commerce. The Cronbach’s alpha value for a five-item perceived ease of use was .92, a five-item perceived usefulness scale was .92, and A six-item perceived enjoyment scale was .95 (Kim et al., 2009). Spears and Singh’s (2004) seven-item attitude scale was adopted in this study. The Cronbach’s alpha value for the scale was .95 (Spears & Singh, 2004). A five-item intention scale developed by Engel, Blackwell, and Miniard (1995) and Wakefield and Baker (1998) was adopted from Lee et al. (2006). The reliability of the scale was .97 (Lee et al., 2006).

3.Data analyses

Data was collected from the online survey and analyzed using SPSS. First, descriptive statistics (e.g. means, variances, and standard deviations) were used to summarize data from demographic variables Second, exploratory factor analysis (EFA) was performed as a data reduction technique for the five variables (i.e., perceived ease of use of social commerce, perceived usefulness of social commerce, perceived enjoyment of social commerce, attitudes toward social commerce, and intentions to use of social commerce). Third, a Cronbach’s alpha coefficient for each of the five measures was calculated to assess reliability as well as test discriminant validity of the constructs. Fourth, a Pearson correlation coefficient was calculated to determine the direction and magnitude of the relationship between variables. Fifth, confirmatory factor analysis (CFA) tested measurement model and convergent validity. Finally, the hypothesized relationships proposed in the present study were tested through structural equation modeling (SEM).

Ⅳ.Results

1.Sample characteristics

The sample was comprised of college students over 18 years of age at a major Mid-Southern university. There were 524 valid and complete responses from 768 online surveys distributed through email invitation, for a 70% response rate. The majority of the sample was female (71%) students between the ages of 18 to 24 years old; the other 29% were male. The students were from diverse majors; including Agricultural, Food, and Life Sciences, Arts and Sciences, Business, Education and Health Professions, and Engineering. The majority of participants were Caucasian or European (84.4%) and almost half of the participants reported more than $100,000 for annual household income. Participant characteristics of the sample are presented in <Table 1>. The majority of participants reported being active users of social media. Instagram was the most preferred and frequently used social networking site followed by Facebook (20.1%) and Twitter (10.5%). More than half of the participants indicated that they have been visiting social media networking sites five or more times a day (64%). Slightly less than 50% of participants indicated that they have purchased fashion products through social commerce. Of those participants who have made a purchase via social commerce, nearly 28% reported having purchased one to ten items over the last year. Interestingly, half of the participants indicated that they prefer to shop in a traditional brick and mortar store. Online commerce was the second preferred channel (36%) followed by social commerce (9.2%), and mobile commerce (5.6%). While young consumers were found to shop across multiple channels, they still preferred to shop at brick and mortar stores for fashion products.

2.EFA

EFA was conducted with varimax rotation to extract one factor each for items assessing perceived ease of use, perceived usefulness, perceived enjoyment, attitudes toward social commerce and intentions to use social commerce. An eigenvalue measuring greater than 1.0 determined the number of factors extracted for each construct. As shown in <Table 2>, items with factor loadings of .50 or higher on one factor and factor loadings of .30 or lower on the other factor were retained on one factor. Findings showed that each variable had a single factor dimension with high factor loadings (.67~.92). Internal consistency of each measure was tested with a Cronbach’s alpha value and composite reliability (CR) greater than .70.

3.Correlations between the variables and discriminant validity

As predicted, the results of Pearson correlation coefficient showed strong relationships among the variables. Perceived ease of use was highly correlated with perceived usefulness (.65), perceived enjoyment (.66), attitudes toward social commerce (.62) and intentions to use social commerce (.69). Perceived usefulness was highly correlated to perceived enjoyment (.73), attitudes (.67) and intentions to use social commerce (.69). Perceived enjoyment was strongly correlated to attitudes toward social commerce (.79), which may be a result of using positive words to describe attitudes. Attitudes toward social commerce was highly correlated to intentions to use social commerce (.69). Correlation values less than .85 (Kline, 1998) confirmed discriminant validity of the constructs <Table 3>.

4.CFA: Measurement model analysis and convergent validity

CFA examined the fit of a five-factor measurement model using a maximum-likelihood estimation procedure in Mplus 7.0. The goodness-of-fit of the measurement model was determined by examining the Chi-square value, comparative fit index (CFI≥.95), root mean square error of approximation (RMSEA≤ .08), and standardized root mean square residual (SRMR≤.08) (Hair, Black, Babin, Anderson, & Tatham, 2006; Hu & Bentler, 1999). The results indicated that the five-factor measurement model yields a satisfactory fit to the data: [χ²=185.60 (df=80), p≤ .000], CFI=.99, RMSEA=.05, and SRMR=.02. All confirmatory factor loadings were significant and measured higher than .85 with highly significant t-values ranging from 29.28 to 129.86. In addition, the average variance extracted (AVE) of each construct was equal to or above .50, which confirmed convertgent validity for the constructs in the measurement model (Fornell & Larcker, 1981).

5.Structure model analysis and hypotheses testing

A structural model was estimated using the maximum- likelihood estimation procedure in Mplus 7.0. (Muthén & Muthén, 2012). The structural model was found to fit to the data well: [χ²=369.47 (df=84), p≤ .001], CFI=.96, RMSEA=.08, SRMR=.07. As shown in <Fig. 1>, five proposed structural paths were statistically significant in the predicted direction (p≤.001), except one path from perceived ease of use of social commerce to attitudes toward social commerce. As predicted, perceived ease of use of social commerce positively affects perceived usefulness of social commerce, supporting H1 (b=.79, t=34.58, p≤.001). Perceived ease of use of social commerce positively affects perceived enjoyment of social commerce, supporting H2 (β=.79, t=35.94, p≤.001). There was no significant impact of perceived ease of use of social commerce on attitudes toward social commerce (β= .13, t=1.99). Thus, H3 is not supported. Perceived usefulness of social commerce positively affects attitudes toward social commerce, supporting H4 (b=.18, t= 3.39, p≤.05). Perceived enjoyment of social commerce positively affects attitudes toward social commerce, supporting H5 (b=.62, t=12.61, p≤.001). Finally, attitudes toward social commerce positively affects intentions to use social commerce, supporting H6 (b= .75, t=33.97, p≤.001).

6.Further structural model analysis

In the structural model, perceived ease of use of social commerce did not have a direct significant impact on attitudes toward social commerce as previous studies found inconsistent results (Chen & Tan, 2004; Hwang, Chung, & Sanders, 2016; Lee et al., 2006). The relationship between perceived usefulness and perceived enjoyment of social commerce was not hypothesized in the structural equation model because of a lack of empirical evidence from previous studies. The significant relationship between perceived ease of use and enjoyment yielded a potential relationship between perceived usefulness and perceived enjoyment of social commerce based on modification indices. Based on modification indices in the structural model, the insignificant path was removed between ease of use and attitudes toward social commerce and a potential path was added between perceived usefulness and enjoyment of social commerce. As a result, the structural model showed improved fit to the data: [χ²=283.37 (df=84), p≤.001], CFI=.98, RMSEA=.07, SRMR=.06. The included path (b=.51, t=10.95, p≤.001) significantly improved the model fit and six structural paths were statistically significant in the adjusted model <Fig. 2>.

V.Discussion and Implications

Social media has become an integral part of consumers’ daily lives. Individuals connect with one another on social networking sites to like, share, and post about information and experiences. As noted earlier, millennials frequently interact with companies through social media. Growth of social media users has led marketers to incorporate social media as part of their digital marketing strategy. Accordingly, an increased number of fashion retailers has been using social media as an advertising tool and a retail channel. Despite the popularity of social media among millennials, empirical findings are limited to reveal factors associated with young consumers’ intentions to use social commerce in fashion shopping. There fore, the need to conduct new research is identified in the context of social commerce.

Results of the study confirmed the TAM is suitable in explaining millennials’ adoption of social commerce in shopping for fashion products. This study found that perceived ease of use of social commerce influences perceived usefulness and enjoyment of social commerce in agreement with findings from online shopping research (Cheema et al., 2013; Ha & Stoel, 2009; Kim et al., 2009; Tong, 2010). The findings that usefulness and enjoyment of social commerce had a positive impact on millennials’ attitudes and intentions to use social commerce in fashion shopping supported previous studies on online shopping (e.g., Ha & Stoel, 2009; Kim, 2012; Yu & Park, 2014). Although the direct path coefficient between perceived ease of use and attitudes was insignificant, this study suggests an indirect effect of perceived ease of use on attitudes and intentions based on the significance of all other path coefficients in the model. Thus, the results of this study reinforced the importance of informative and experiential elements on social media platforms to enhance behavioral intentions among young consumers (Yuksel et al., 2016).

Further structural model analysis revealed a direct, positive influence of perceived usefulness of social commerce on perceived enjoyment of social commerce, which has not been explored in prior studies. The impact of perceived usefulness was stronger than the impact of ease of use on perceived enjoyment. The findings imply that utilitarian and hedonic benefits are important to augment young consumers’ positive attitudes and intentions to purchase fashion on social media. Aligned with Kim et al.’s (2009) study, enjoyment had a much stronger impact on attitudes compared to usefulness. The findings suggest that utilitarian benefits from an efficient, time saving experience may engender positive feelings (e.g., enjoyable, fun, and exciting), which are crucial to increase millennial shoppers’ fashion purchasing on social media.

The findings from this study also suggest managerial implications for social commerce retailers and marketers. The impact of utilitarian benefits greatly influences hedonic benefits, which in turn leads to consumers’ positive attitudes and intentions to use social commerce when shopping for fashion products. Thus, marketers should design social commerce experiences to be easy to use and useful for consumers to access fashion product information or experience product features. For example, brands such as Kate Spade and Tory Burch have adopted shoppable photo tags on Instagram that offer consumers in-app details about a specific product’s price and description and features a “Shop Now” button. This tool assists consumers with instant product information and enhances their shopping experience by allowing them to view product pages all within the app without ever leaving their Instagram feed (Constine, 2016).

While this study contributes to extending the body of social media research, several limitations should be recognized. The sample was limited to college students at a Mid-Southern university; thus the results may not represent the U.S population in general. The majority of the sample consisted of primarily Caucasians or European women located in the Mid-South; therefore the results may not characterize individuals from diverse backgrounds and other specific regions of the country. Extending the study to other regions of the U.S would contribute greatly to understanding consumer intentions to use social commerce while shopping for fashion products. Moreover, majority of the participants (95%) in the study reported being active social media users and approximately 84% of the participants indicated visiting social media networking sites over five times a day. Thus, the lifestyle of the participants and their previous experiences with social media may affect the findings of this study. Finally, this study examined the relationships among perceived ease of use, usefulness, enjoyment, attitudes, and intentions to use social commerce in shopping for fashion products. The focus on shopping for fashion products affects the generalizability of the results to other product categories. To increase the validity of the model, future research should test the proposed model for other product categories. Further investigation is needed to explore other factors (e.g., perceived risks and perceived value) that affect product purchase through social commerce.