Keywords

1 Introduction

The rise of social media has provided a revolutionary form of communication media for narrowing the distance between merchants and consumers. However, the understanding of such issue remains limited. Explicit engagement in social media to communicate product information to consumers potentially increases sales [1,2,3]. In practice, how to integrate social media into commercial activities is unclear. Mangold and Faulds [4] echo similar concerns as they claim that the degree to which the media is effectively utilized determines its value realization. We aim to address this concern by contextualizing the research in group-buying context where group-buying websites use microblogging for deal communication and sales generation.

Anchoring on Construal Level Theory (CLT), we propose that effective use of social media depends on the degree to which consumers perceive the distance of deal communication based on who communicates it. CLT indicates that psychological distance is linked to the level of mental construal, such that psychologically distant (proximal) objects will be construed at a higher (lower), abstract (detailed) level; and high (low) level construal will bring to mind distant (proximal) objects [5]. Psychological distance is determined by the spatiotemporal distance of the event or object from the self, here and now. Spatiotemporal distance of social media communication perceived by consumers can affect their construal level, which in turn, guides how they evaluate the communication [6]. Deal communication in social media varies in their spatial distances. For instance, consumers perceive the postings at the city-level social media account of merchants as spatially proximal compared with the postings at the account of the headquarters (i.e., national-level). Psychologically distant and proximal postings could be evaluated differently. The relationship between the construal level and evaluation (purchasing willingness) has been investigated in existing CLT literature. However, the results differ. Trope et al. [6] suggest that consumer willingness to purchase a product is higher for distant objects because abstract thinking focuses more on the desirability of the object (e.g., how interesting the product is) and less on the feasibility of the object (e.g., how expensive or useful the product is). By contrast, Williams et al. [7] argue that the purchasing willingness of a consumer is higher for proximal objects because distance promotes emotional detachment. The present study empirically tests these conflicting arguments.

This study contributes to researchers by addressing the means through which social media can be best integrated into daily commercial activities from the merchant perspective. This study adds to the existing social media literature that mainly focuses on the effects of consumer generated contents, such as consumer reviews and product ratings, on sales performance [8, 9].

2 Literature Review

The popularity of social media motivates researchers to examine how business and economic values can be realized through social media communication, which is an important indicator of communication effectiveness for merchants. Most studies focus on word-of-mouth (WOM) communication through social media. For instance, several studies investigate the effects of WOM on the box office sales of movies. Liu [9] discovers that the volume, instead of valence (percentage of positive/negative messages), of WOM messages offers significant explanatory power for both aggregate and weekly box office revenue. Duan et al. [8] examine the awareness effect of online user reviews on daily box office performances of movies, and reveal that the volume of online posting significantly influences box office sales. Chen and Xie [10] consider online user reviews as endogenous factors that influence movie sales. In addition, the volume of online reviews, which indicate the importance of the awareness effect, significantly influences box office sales. However, the rating of online user reviews has no significant impact on the box office revenues of movies, which indicate little persuasive effect on consumer purchase decisions. Moon et al. [11] find that high advertising spending on movies supported by high ratings maximizes movie revenues. In other contexts, Ye et al. [12] reveal that user ratings positively affect hotel room sales, which imply that ratings as a new form of communication on social media have business values for retailers. Besides the volume of WOM communication, Chevalier and Mayzlin [13] determine a significant relationship between the length of reviews and online book sales. Clemons et al. [14] reveal that the variance of ratings and the strength of the most positive quartile of reviews play a significant role in determining which new products grow fastest in the market. The effectiveness of reviews is also affected by other factors. Forman et al. [15] suggest that consumers use reviewer disclosure of identity-descriptive information to supplement or to replace product information when making purchase decisions and evaluating the helpfulness of online reviews in the online community. Consumers can rate reviews that contain identity-descriptive information more positively, and the prevalence of reviewer disclosure of identity information is associated with increases in subsequent online product sales. User-generated social media content conveys more information compared with conventional online platforms. Luo et al. [16] argue that social media-based metrics, such as consumer ratings and blogs, are significant indicators of firm equity value compared with web traffic and Google searches.

The abovementioned studies focus on investigating social media communication generated by consumers. Social media affords communication among consumers, which generate economic values for merchants. Although the findings are valuable, the fact that merchants have limited or no control over what users post in social media indicates that improvements still need to be introduced in terms of how social media communication can be utilized by merchants to increase sales. Thus, a few studies have started to explore how to communicate promotional messages to consumers on social media sites. For instance, Goh et al. [2] show that the information richness of social media communication generated by merchants on Facebook positively affects sales. Our study aims to enrich the understanding of effective social media communication by merchants.

3 Construal Level Theory (CLT)

We propose that social media communication generated by merchants are processed by consumers, and the manner how consumers construe communication affects its effectiveness. This process can be explained by CLT. CLT is an account of how psychological distances between individuals and their targets influence individual construal levels of those targets. The theory suggests that psychological distance is an important determinant of whether the primary or the secondary peripheral characteristics are used as the basis of individual evaluation [5, 6]. CLT assumes that people mentally construe objects that are psychologically near in terms of low-level, detailed, and contextualized features, while, at a distance, they construe the same objects or events in terms of high-level, abstract, and stable characteristics [6]. Psychological distance is egocentric, and its reference point is the self, here and now. Differences in time, space, social distance, and hypotheticality constitute different distance dimensions [17]. Various empirical studies have investigated the effects of psychological distance on construal level. Liberman et al. [18] examine temporal differences in construal and report that people think about a set of objects in more superordinate abstract terms in the distant future condition as compared with in the near future condition. Fujita et al. [19] show that the events in a film are described by more abstract language for people who believe such events were located in a spatially distant location. Social distance is influenced by the similarity or power difference among individuals [6]. Liviatan et al. [20] report that students prefer subordinate and superordinate action identifications to describe the activities of students of similar (similar targets) and different (dissimilar targets) classes, respectively. Aside from temporal, spatial, and social distances, hypotheticality affects psychological distance. An improbable event would seem more distant than a probable event [21]. Different dimensions of psychological distance affect mental construal, and these construals guide prediction, evaluation, and behavior [6]. Prediction, evaluation, and behavior are based on high-level construal aspects as psychological distance increases. Nussbaum et al. [22] study the manner by which temporal distance influences the confidence of predicting future outcomes. They find that temporal distance enhances confidence when deriving from a high-level construct. Concerning evaluation, Liberman and Trope [23] examine desirability concerns, which high-level construals of an activity should emphasize, and feasibility concerns, which low-level construals of an activity should emphasize. As temporal distance increases, desirability effect increases, while feasibility effect decreases. Applying desirability and feasibility concerns in consumer purchasing behavior, abstract mental construal increases consumer willingness to purchase a product because abstract thinking focuses more on the desirability of the object (e.g., how interesting the product is) and less on the feasibility of the object (e.g., how expensive or useful the product is) [5]. Thus, psychological distance could promote the purchasing willingness of consumers. Similar arguments can be found in several consumer behavior studies. For instance, consumers who have higher construal level place more importance on perceived fit in evaluating brand extensions [24]. Based on the above arguments, effectiveness of social media communication depends on whether consumers perceive communication messages distant (spatially).

However, recent studies question this main proposition of CLT and re-examine the relationship between psychological distance and construal level. Williams et al. [7] argue that psychological distance and construal level are actually two separate concepts. They suggest that distance promotes emotional detachment, whereas abstract thinking promotes a focus on positivity. Other studies have also shown that consumers have weaker emotional responses to distant stimuli, e.g., [25]. van Boven et al. [26] suggest that emotional intensity reduces perceived psychological distance. Given that psychological distance reduces the intensity of emotional reactions during judgment and choice, they find that consumer willingness to purchase highly desirable products decreases [7]. The influence of emotional reactions on consumer behavior and advertising communication has been widely studied in consumer research literature, e.g., [27]. The effect of emotional reactions on purchasing behavior has been studied in the online context. For instance, Hausman and Siekpe [28] find that emotional reactions toward a website positively affect online purchasing intentions. Park and Kim [29] show that consumer emotional reactions to products are key influential factors for consumer commitment, which results in higher purchasing intentions. According to this stream of studies, for social media communication, which is distant to consumers and causes emotional detachment [7], the effectiveness decreases. The present study examines how the distance between consumers and social media communication affects the effectiveness of communication. Hypotheses are developed based on these two opposite views.

4 Hypotheses Development

According to CLT, spatial distant object increases psychological distance and is construed at a high level [5]. By construing the object in a high level, the abstract and stable characteristics of the object would be presented [6]. Liberman and Trope [23] find that desirability effect increases and feasibility effect decreases when psychological distance increases. High-level construals of the objects emphasize desirability concerns, whereas low-level construals of the object emphasize feasibility concerns. Therefore, spatially distant object induces abstract thinking from consumers who will focus more on the desirability of the object and less on the feasibility of the object [5]. Increases in desirability effect would enhance consumer willingness to purchase. Conversely, feasibility concerns should be more prominent in decisions regarding the spatially proximal object. Ironically, although feasibility aspects are often more multifaceted compared with desirability aspects, increasing reliance on low-level construal for spatially proximal objects often reduces consumer confidence and decisiveness [5]. Therefore, the increase of feasibility effect would diminish consumer willingness to purchase.

One company can have multiple accounts in the same social media site based on the level of its business operations. A company can have one account as its national (headquarter and country-level) account and other accounts for each of its branches (city-level), which are mainly used to deliver information from the local branches. From the perspective of local consumers, the national account of a company is spatially distant, whereas its local account is spatially proximal. Accordingly, local consumers will have high mental construals of communication sent through the national account of a company. The prevalence of desirability (high mental construal) over feasibility (low mental construal) should increase consumer confidence and decisiveness, and attract better sales performance of the promoted product. Thus, we hypothesize:

  • H1a: Compared to promotional messages on local accounts, promotional messages on national accounts result in higher sale performance.

The above counterintuitive view of CLT is still debatable. The gist of CLT is that psychological distance has a linear relationship with abstract thinking. These two variables are often treated as identical and interchangeable. However, a recent study has contested this relationship and has found evidence suggesting that these variables should not be treated as identical because they are not interchangeable. Specifically, distance promotes emotional detachment [7], and reduces emotional concern [25, 26]. Abstract thinking promotes a focus on positivity in which the objects are perceived more pleasant [7]. Therefore, being psychologically distant from an object does not necessarily mean that consumers will construe the object at a high level or in an abstract manner. Given that psychological distance promotes emotional detachment, a psychologically distant object, which results in lower emotional reactions, diminishes consumer willingness to purchase, e.g., [28]. Therefore, compared with promotional messages communicated through the local account of a company, those communicated through its national account should have worse sales performance of the promoted product. Thus, we hypothesize:

  • H1b: Compared to promotional messages on local accounts, promotional messages on national accounts result in lower sales performance.

5 Research Methodology

Two sources of data were collected, namely, merchant communications on social media site and the sale performance of the merchant, which represents the effectiveness of the communications. We selected a daily-deal group-buying website, Meituan, one of the most popular Groupon-like daily-deal group-buying websites in China. SinaWeibo, which is a Twitter-like micro-blogging site in China, is the most widely used social media site in major cities in China and hosts a verified company account of almost every business in the country. Thus, we selected SinaWeibo as the social media site in our study and crawled the messages posted by Meituan. In this study, we only collected deals featured in the Beijing and Shanghai websites, the two biggest cities in China where the website first started its business. Data were collected from December 2011 to February 2012. The number of coupons sold for the deals in the Meituan Beijing and Shanghai sites was crawled hourly. At the same time, all deal-related data (i.e., face value of the deal, price, title, start time, end time, tipping point, availability, and expiration time of the coupon) were also captured. After excluding lucky draw deals, for which the price is 0, and the deals without expiration date, 1131 deals were collected (583 deals from Meituan Beijing branch and 548 deals from Meituan Shanghai branch). Among these 1131 deals, 184 deals were sold out before the end of feature time online. Thus, we only considered 947 deals for the analysis on total number of coupons sold. To control the product types, the deals were categorized into grocery/dining, leisure/entertainment, and others. This categorization is widely applied in Chinese daily-deal websites. The feature duration of the deal refers to the number of days the deal is featured on the group-buying website. The expiration time of the coupon refers to the number of days the coupon is valid for redemption after the end time of the deal.

In SinaWeibo, we crawled the number of followers from the headquarter account of Meituan and its two local accounts in Beijing and Shanghai hourly. All tweets, including contents and posting times on these three accounts were also crawled and saved. In total, we obtained 1128 tweets from headquarter account, 2219 tweets from local account in Beijing, and 2490 tweets local account in Shanghai. These tweets were classified into three categories based on their contents, namely, promotional tweets (information on the promoted product), business-related tweets other than promotions (e.g., company news and lucky draws), and non-business-related tweets (e.g. jokes and weather conditions). We used three count variables for each deal to determine how many times they were promoted in the Meituan headquarter/branch accounts. We control the frequency and types of tweets on the national and local accounts of the company. The number of tweets for each type and from each account is calculated for the duration when a particular deal is featured on the group-buying website. We also control the number of followers on each account, which is calculated as the average number of users for the duration when a particular deal was featured on the group-buying website.

6 Data Analysis and Results

According to H1a and H1b, promotional messages on the national account had higher or lower sales performance as compared with promotional messages on the local account. We used the negative binomial regression to test this hypothesis, given that our dependent variable, i.e., the total number of coupons sold for each deal, was counted with over-dispersion (μ = 1395, σ = 3600). The frequency of promotion of each deal on the national account (both explicit and implicit) and the respective local account (both explicit and implicit) were the independent variables. Control variables were obtained from two sources. First, deal-related variables from group-buying retailers, such as the category of the deals, tipping point, availability, price, and expiration time of the coupon, were included in the analysis. Three deal-related variables, namely, the city where the deal is featured, face value of the deal, and duration of the feature, were dropped because of a multi-collinearity problem in the analysis. Second, social media communication-related variables, such as the number of followers on the national and the respective local accounts as well as the number of tweets for each account level and category were included in the analysis. Due to the multi-collinearity problem of the control variables, we ran four models with different sets of control variables. For each set, we used control variables that did not have a multi-collinearity problem.

Table 1 indicates that the number of sold coupons could increase up to 5.12 (10^0.709) times by posting a promotional message on the headquarter account, whereas the number of sold coupons could increase up to 1.36 (10^0.135) times by posting a promotional message on the local account. The national account was a more effective social media communication tool. Thus, H1a is supported, while H1b is not supported. This finding supports the CLT.

Table 1. Field data results
Full size table

7 Discussion

Inspired by the question on how to integrate social media into daily commercial activities, this study investigates how psychological distances, reflected by the spatial distance, between social media communication and consumers affect the effectiveness of the communications. Based on CLT, spatially distant social media communication should increase the effectiveness of communication because of the high mental construal of consumers, which increases the desirability and decreases the feasibility concern. However, there is an opposing view on this stand where psychological distance could promote emotional detachment. According to this view, spatially distant social media communication could decrease the effectiveness of communication because of consumer emotional detachment from the message. One way to increase its effectiveness is by making the message more abstract. To address this opposing view, we analyze the sales performance of online group-buying website and their respective microblogs. We find that promotional messages that are posted on the company national account, which is spatially distant from consumers, generate better sales performance than message posted on the local account. This finding is consistent with CLT.

This study offers several implications. First, we introduce and empirically examine an opposing view to CLT, which argues that psychological distance and construal level are two separate constructs, and that the former is linked to emotional detachment, e.g., [28, 29]. Second, this study enriches social media literature. Although social media is attracting more attention from the industry, empirical studies, especially theoretical frameworks anchored studies, still lack. This study relies on the concept of psychological distance and on the two opposing views of construal level and emotional reactions to explain social media phenomenon. Third, this study contributes to the group-buying literature. The daily-deal group-buying phenomenon has attracted researchers from different fields, such as information systems [30] and marketing [31]. The optimization of the sales performance, which is measured by the number of sold coupons, is a central topic in this area. We confirm in this study that the herding behavior/observational learning effect exist in group-buying websites. We have also extended group-buying studies to social media platforms. Existing studies on group-buying mostly focus on data derived from group-buying websites, such as the number of sold coupons, to determine consumer purchasing behavior on such websites [30], and characteristics of deals, websites, and merchants [31]. By integrating social media and group-buying data, studies on group-buying can attract a broader audience and offer more valuable implications.