Keywords

1 Introduction

With the rapid development and widespread popularity of smart phone devices, mobile applications play increasingly important role in marketing campaigns, especially location-based social networking service (LBSNS). The emerging LBSNS creates an era of GeoLife2.0 where people can share life experiences and connect to each other with their location histories (Zheng et al. 2009). Famous mobile applications using LBSNS include Yelp, Foursquare, and Dianping in China. These services provide an access to a valuable source of data on the geographic location of users as well as the online social connections among them. As a consequence, it offers a great opportunity to better understand individuals’ behaviors by taking into account both spatial and social factors.

LBSNS provides merchants an efficient and effective approach to advertise and promote their products to individual consumers, and also to engage the consumers in product information dissemination. Such applications take advantage of location and social network information to enhance the interaction between buyers and sellers and to strengthen customer loyalty on brands. Previous academic studies have also realized the crucial role of location and social network on mobile marketing. However, the effects of these two factors are never considered simultaneously. In practice, many companies have been making use of mobile social media to leverage on location-based and social-based communications in their commercial pursuits. For example, the famous mobile marketing application, Foursquare, allows users to notify others about their location by ‘checking-in’ at a certain place. A company could spontaneously decide to launch a sales promotion that is only valid for 1-2 h and broadcast this information to all mobile devices within a certain range. This kind of location-sensitive and time-sensitive marketing practices will result in a unique route of propagation via social connections (Scellato et al. 2011). Combining this information with sophisticated data-mining techniques will give traditional merchants the ability to make efficient customized sales promotion (Kaplan 2012).

In our study, mobile marketing is exemplified as a time-limited, group-based “two for the price of one when you get a friend to buy together” promotional campaign advertised via mobile devices. Field experiment was conducted to explore the influence of location and social network factors on consumers’ decision to accept mobile promotions. In the designed field experiment situation, consumers face a dilemma of whether to invite a stranger nearby or to insist on waiting for their friends to enjoy the deal. We are interested in how consumers make the decision when their purchase is constrained with both location and social relationship. First, whether the location of customers plays significant role on the consumption of promoted goods? Second, whether the strength of social ties influence the consumption of promoted goods? Third, what is the priority of customers’ choice, spatial-dependent or social-dependent? We then conducted follow-up surveys that revealed users’ personality features and psychological states as supporting materials to explain our field experiment observations.

The reminder of the paper is organized as follows. Section 2 outlines previous findings on the effects of location and social network on consumers’ acceptance of mobile marketing. Section 3 describes the design of field experiment and the supporting survey study. Section 4 illustrates data analysis results based on field experiment observations and survey findings. Section 5 makes discussion about the empirical results and draws conclusions.

2 Literature Review

Although a number of academic studies have been conducted on consumers’ acceptance of mobile marketing and related issues, many of them have been conceptual work using structural equation modeling (Leppaniemi & Karjaluoto 2005; Merisavo et al. 2007). Technology acceptance model (TAM) suggests that perceived usefulness and perceived ease-of-use are two fundamental factors to determine consumers’ adoption behavior, which is the basic theoretical framework applied in empirical studies. These empirical studies explored the effectiveness of various types of mobile marketing practices, such as mobile emailing and messaging (Barwise and Strong 2002), and mobile couponing (Wehmeyer and Müller-Lankenau 2005). Shankar et al. (2010) reviewed some key factors that enable or hinder the acceptance of mobile marketing in the retail environment. The enablers included networking, prices of mobile device and applications, the utility of mobile application, and trust with the application. The inhibitors are such as consumption inertia, economic barriers, limited knowledge, and distrust of marketing.

To better understand how mobile marketing works and to make it more effective, we focus on two advanced mobile services, that is, location-sensitive function and social network management. Prior studies suggested that location-relevant information and social network connections may increase consumers’ willingness to accept mobile marketing practices. Location-based service is positively associated with consumers’ perceived usefulness of mobile advertising, and social networking service enables the social-context awareness of mobile marketing. The influence of location and social network on consumers’ acceptance of mobile marketing are examined independently in the extant literature.

The increasing utility of mobile applications benefit from location-based services to a great extent. According to Bauer et al. (2005), delivering a store’s information via SMS advertising to consumers are effective and desirable only when they are within the physical vicinity of that store. Pura, M. (2005) examined the fundamental value of location-based mobile services from six dimensions: monetary, convenience, social, emotional, conditional, and epistemic value. Results showed that both commitment and behavioral intentions to use location-based services are strongly influenced by conditional value of getting customized information based on users’ location. Mobile marketing is defined by Scharl, Dickinger, and Murphy (2005) as using a wireless medium to provide consumers with time- and location-sensitive, personalized information that promotes goods, services and ideas. They addressed the underlying importance of location to personalized mobile marketing. The study of Ho (2012) provided empirical evidence of the effectiveness of location personalization in mobile marketing and explored factors that not only attract individuals to mobile services but also keep them engaged in the services in the long term.

Social media applications including microblogs, blogs, virtual communities, and content communities, etc. have already become part of the standard communication repertoire in marketing campaigns. The combination of social media applications and mobile devices amplifies individual’s urge to share opinions with friends at anytime and anywhere. Mobile social media has two forms communication, company-to-consumer communication and user-generated content (UGC) (Kaplan 2012). Due to homophily and contagion in social networks, advertisements and promotions will be well spread through social ties once companies accurately target potential customers (Rana et al. 2009). According to (Hill et al. 2006), marketing campaigns based on social network will significantly increase customers’ trust, and then contribute to adoption by word-of-mouth.

The conceptual work in previous literature revealed that location and social network are both important drivers of consumer acceptance of mobile marketing. However, existing studies regard location and social network as two independent factors that may affect consumers’ acceptance of mobile marketing. Seldom or never have research explored the interaction effect of these two factors on consumers’ decision. Furthermore, papers focusing on adoption intention and satisfaction based on mobile applications and services are evidently insufficient compared with that based on social networking sites. In terms of research methodology, studies in this field are mainly conceptual work combined with survey investigations. In our paper, we conduct field experiment to observe the real reaction of consumer in designed experiment situation in which location and social network may simultaneously affect consumer’s decision.

3 Methodology

We designed a time-limited, group-based “two for the price of one when you bring a friend to buy together” sales promotion situation for our field experiment. We posted an APP testing invitation to the BBS forum of a large public university in China and uploaded our mobile marketing application for free downloading. 99 subjects successfully installed our mobile promotion application and performed registration. We did not reveal the experiment objective to the subjects, and we collected real usage data in a natural, unobtrusive environment.

The mobile promotion application was developed specifically for the purpose of this study. It enables location-based service (LBS) as well as social network management. The main interface of the application includes six modules: my friends’ list, import of mobile phone contact, discovery of someone nearby, historical records, and latest promotions. The friends list of each subject was obtained from their self-reported social connections. Subjects could get the list of people who is currently around him within a distance of 1.5 miles by clicking ‘someone nearby’. Friends and strangers are distinguished in the nearby list by background color as it is shown in Fig. 1. The order of users appearing in the nearby list is randomly arranged. The ranking would be updated once user refresh the list. The latest promotion details would be pushed to inform all the subjects once a promotion was launched. As the promotion plan in our experiment was group-based, the subjects were required to invite a friend to redeem the promotional coupon together. The mobile application provided a function for subjects to make such invitations. The communication involved two actors, inviter and invitee. As an inviter, he should decide whether to invite a friend or someone nearby (the second interface in Fig. 2). Once the invitation is sent out, the invitee will receive a notification. The invitee should decide whether to accept this invitation (the third interface in Fig. 2). The specific invitation processes are demonstrated in Fig. 2.

Fig. 1.
figure 1

The interface of “someone nearby” module (Grey background indicates a friends nearby and white background indicates a stranger nearby) (Color figure online).

Full size image
Fig. 2.
figure 2

Invitation processes

Full size image

We cooperated with one of the most popular bakery shop outside the university campus. Customers may be sensitive to the promotion price of the product. In order to avoid this unexpected effect, we manipulated one control factor, the price of promoted goods, in our field experiment. We made a within-subject design across two dessert packages, providing 10 RMB set and 30 RMB set. The sales promotions of these two packages were launched in two separate weekends with the same group of subjects, and subjects could spontaneously decide whether to participate in the promotions. Within the limited available time of the promotion, subjects had to make a decision about whether to invite a friend who may not be able to show up at that time or someone nearby who may not be an acquaintance. The invitation behaviors of subjects were recorded by our mobile applications. Subjects could make invitations several times to different people, but the deal could be redeemed only when at least one invitee accepted the invitation. Throughout the experiment, only the cashiers of the bakery shop (but not the authors of this paper) could interact with the inviter and invitee who came to buy and redeem the deal, therefore their responses should reflect their true decisions.

Regardless of whether subjects successfully invite someone to buy and redeem the deal, a follow-up survey was posted to elicit some necessary information. The questionnaire consists of four parts. First part was the personality test about the propensity to engage in social relationship exploratory and exploitative behavior (Mom et al. 2007). Second part measured the perceived time pressure of subjects during the two experiments. Third part inquired the subjects about the daily usage of location-based services and social network services in mobile applications, the spending habits of the goods promoted in our experiment, and the attitudes to group buying and mobile promotion. The last part gathered the demographical information of the subjects, such as, age, sex, and major. The complete questionnaire of the follow-up survey is shown in Appendix A. The results of follow-up survey serve to supplement the field experiment observations.

4 Results

Table 1 describes all the variables and their transformation form used in the data analysis below.

Table 1. Description of field experiment variables and transformations
Full size table

The social network graph of 99 subjects and the directed graph of invitation relationship in the two field experiments are shown in Figs. 3, 4, and 5. In the field experiment of low-price product, 51 invitation relationships are overlapped with the social connections among the total 138 invitation trials involving friends as well as strangers. The overlapping rate of the two networks is 37.0 %. In the field experiment of high-price product, 40 invitation relationships are overlapped with the social connections among the total 103 invitations. The overlapping rate of the two networks is 38.8 %. The overlapping connections of social ties and invitation relationships are colored red in the graphs shown in Figs. 4 and 5.

Fig. 3.
figure 3

Harel-Koren fast multiscale graph of social network among 99 subjects

Full size image
Fig. 4.
figure 4

Harel-Koren fast multiscale graph of invitation relationship in low-price product promotion experiment (red lines indicates overlapping with social ties) (Color figure online).

Full size image
Fig. 5.
figure 5

Harel-Koren fast multiscale graph of invitation relationship in high-price product promotion experiment (red lines indicates overlapping with social ties) (Color figure online).

Full size image

We applied two sample t-test and two-way ANOVA statistics to examine the effects of location and social network. At first, we ignore the price difference between the cases of two experiments and conduct two sample t-test based on the aggregated data. Two sample t-test on populations of Deal = 0 and Deal = 1 indicates the distance between inviter and invitee in successful invitation group is significantly different from that in failing invitation group. T-test on populations of Relation = 0 and Relation = 1 indicates that the successful rate of social relationship based invitation is significantly larger than invitation between strangers. Then we conduct two-way ANOVA in which price is taken as one factor that may decide whether the invitation will be accepted (Deal) and whether the inviter will choose a friend or a stranger (Relation). As the influence of price factor is controlled in two-way ANOVA, distance range between inviter and invitee (Distance_norm) and social network embededness (Relation) demonstrate significant influence on the response variable (Deal), that is, whether the invitation will be accepted.

In order to figure out factors that affect subject’s decision on whether to accept the promotion invitation, we conducted logistic regression on the dependent variable “Deal”. Table 2 and 3 depict the coefficient estimation results based on untransformed and transformed data sets. In low-price promotion setting, social relationship take significant effect but the influence of distance is not significant. When the value of promotion product increases, distance is regarded as a significant factor that may influence subject’s decision as well as social relationship.

Table 2. Logistic regression on deal based on untransformed variables
Full size table
Table 3. Logistic regression on deal based on transformed variables
Full size table

In addition to the acceptance of mobile marketing, we are also interested in the decision of inviters who should determine whether to invite a friend or a stranger. Thus logistic regression on variable Relation was made. Except for centrality properties such as degree and betweenness, the distance between invitee and shop and the distance between inviter and shop are found to be significant factors to influence the choice of friend or stranger. Furthermore, we find that the personality measurement scale in follow-up survey has significant association with subject’s choice of friend or stranger. Social relationship exploitative scores are positively related to the intention to choose strong social ties while social relationship exploratory scores are negatively related to strong tie dependency (Table 4 and 5).

Table 4. Logistic regression on relation based on experiment dataset
Full size table
Table 5. Logistic regression on relation based on personality assessment score
Full size table

5 Disscussion and Conclusion

The exploratory field experiment examines the significant influence of geographic location and social network on individual’s acceptance of mobile marketing. Based on experiment data analysis and follow-up survey investigation, we obtain the following interesting findings. In terms of the acceptance of promotion invitation, we find that social relationship always plays significant role to affect the acceptance of invitation regardless of the price of promoted product. Logistic regression on Deal indicates that the effect of geographic location becomes significant when the price of promoted product is high. With regard to the choice of strong social tie or weak social tie, we find that geographic location has significant influence on inviter’s choice of friend or stranger only when the price of promoted product is high. Social network centrality of the inviter is positively associated with the intention to invite a friend, no matter how much the promoted price is. The personality test of subjects indicated that the intention to invite friend or non-friend was influence by the social types of inviters as well. Social relationship exploratory individuals tend to invite non-friends while social relationship exploitative individuals are inclined to invite friends.

Managerial implications of the field experiment should be highlighted as well. For mobile advertising application developers, they should recognize users’ demand for group-based mobile promotions, and their willingness to mobilize social resources, both of exploitative (existing friends) and explorative (strangers) nature, to obtain group-based deals. LBSNS’ social network function can aid in this aspect to enhance mobile promotion effects, but the convenience of communication between individuals who are proximate in location but having weak social tie needs be improved for such social resources to be mobilized. As for merchants, location of target consumers should be the priority consideration when launching high-price promotions. Merchants should take advantage of word-of-mouth effect through social ties when designing their sales promotion scheme.