A Case Study on Trust Cultivation on Indonesian Facebook Online Craft Stores

Advances in communications technology have allowed new forms of trade, i.e. through online means between sellers and customers. The use of Facebook social networking platform as a tool for online trade has brought a question on the factors influencing the creation of trust in among its users who are engaged in a community of craft makers (sellers and customers). This research hypothesised that three independent variables influence trust, namely intensity of interaction, similarity of interest and endorsement. The research was done through questionnaires distributed to 151 samples. Data analysis was performed through SPSS, using Spearman's correlation to note the strength of relationship between variables. Results of the questionnaire show a medium strength correlation between the independent and dependent variables. menggunakan korelasi Spearman untuk mencatat kekuatan hubungan antar variabel. Hasil kuesioner menunjukkan korelasi kekuatan sedang antara variabel independen dan dependen.


Introduction
Traditionally, market activities are done on a face to face basis, in brick and mortar shops or markets in which sellers and customers meet with each other and perform transaction at the location of the seller. However, communications technology developments have allowed business activities involving customers and sellers separated by distance, such as through mail-order catalogues and television. The rise of the World Wide Web and the Internet in the 1990s has also resulted in an increasing ease of business activities. The Internet has developed from a tool of communications, into an economic or business tool (Armstrong & Kotler, 2005). With increasing developments of the Internet, and a greater reach, it can compete directly in penetration compared to other, conventional media.
Internet users use their electronic networks to find information, chat with others, develop relationships, read and reply to e-mails for their correspondence, and develop networks through social media. In Indonesia, according to statistics from the Internet World Stat (December 2011), 55 million out of almost 250 million Indonesians are users of the Internet. The value of the Internet as a marketing tool is not only characterised by its penetration, but also increased by the rise of Web 2.0, which is characterised by collaboration and data sharing between individual users (Straubhaar, LaRose & Davenport, 2010), notably in social network sites such as Facebook. Facebook allows its users to upload photographs and tag them on their friends, update their status, associate themselves with certain causes by joining groups (political affiliation, hobbies and other causes). In general, Facebook is considered to be an appropriate tool for reaching other users in an effective manner (Treadaway & Smith, 2010).
According to Socialbakers (2012) The technological development of social networks have allowed sellers and customers to interact in almost real time, and conversations do not involve mere exchanges of information directly related to transactions, but also sharing of sensitive personal information, such as addresses, Blackberry PIN numbers, bank account numbers, and even telephone numbers.
It is generally known that many social network users tend to be rather liberal in providing private information on their pages. This is in fact facilitated by the social network providers which provides the opportunity for users to include personal information such as their addresses and phone numbers in šZ ]OE ‰ P •U v š} Z•Z}µš }µš[ ÁZ š À OE they have in mind or are doing, by posting statuses, photographs and videos, thus creating an atmosphere of openness.
Yet, the Internet is not all about transparency of information. According to Bhattacherjee (2002), there is a potential of issues raised by online transactions due to lack of trust from customers to online sellers. Such lack of trust is understandable, as the online business mechanism does not allow potential customers to see directly the physical form of the business. Hoffman et al. (1999) confirms the above by stating that in general customers do not have an adequate amount of trust toward online sellers in relationship exchanges, especially related to money and personal information.
When trust cannot be built due to inappropriate application of relationship exchanges, the sustainability of the online business is questioned (Bhattacherjee, 2002). Thus, trust is an important component in developing online exchange relationships, which is characterised by uncertainty, anonymity, o l }( }všOE}o (OE}u šZ µ•š}u OE[• ‰ OEšU and potential opportunism (Hoffman et al, 1999).
The importance of trust in online transactions is shown in the large body of studies on the interactions of trust and other variables such as familiarity and purchase decisions (Slyke et al., 2006;Nath, 2007).Thus, one can see here that there are two divergent characteristics of the Internet, in the first, there seems to be a greater willingness to share personal information as shown in interactions in social media; while on the other hand there is also a reluctance of using online shopping methods due to concerns about privacy.
Apparently, in the case of online shops located in social networks, they could overcome the trust barrier by developing certain methods to build trust between the parties, despite the limitations of security technology.

Trust Conceptualization in Online Businesses
One of the reasons customers prefer not to shop online is the lack of trust between online businesses and customers. During online transactions, there is not just an exchange of goods/services and money, but also of sensitive information such as credit card numbers. There is often a temporal gap between the payment and the delivery of the items bought.
Moreover, according to Winand and Pohl (2000) in Grabner-Kraeuter (2002), during computer mediated transactions, elements that normally occur during interpersonal communications, such as facial expressions and body gestures, often disappear or become inapplicable. According to Einwiller et al. (2000) and Einwiller and Will (2001), online transactions between sellers and customers can result in a number of potential risks.
During online transactions, customers cannot directly see, touch or feel the product, check the quality of the object in person, and also risk loss of sensitive information (Lee & Turban, 2001). Thus, trust is necessary between the seller and the customer as a mental shortcut to resolve the risks of online transactions. According to Luhmann (1989) in Grabner-Kraeuter (2002), customers try to minimise risks in online transactions by developing trust, which • OEÀ • • ^u Z v]•u š} OE µ šZ complexity of human conduct in situations where people have to cope Á]šZ µv OEš ]všÇX_ According to Lewicki and Bunker (1995), trust is defined as how the two parties expect each other to behave during transactions. Another definition of trust is that it is a confidence in an expectation of the trµ•š [• P}} Á]oo (Baier, 1994;and Gambetta, 1988;in Koehn, 2003). Thus, trust is thus a longterm issue that needs to be fostered and developed between parties in online transactions (Grabner-Kraeuter, 2002). Mayer et al. (1995) proposed an integrative d (]v]š]}v }( šOEµ•š • ^šZ willingness of a party (trustor) to be vulnerable to the actions of another party (trustee) based on the expectation that the other (trustee) will perform a particular action important to the trustor, irrespective of the ability to monitor or control that the other party ~šOEµ•š •_X µ•š}u OE• OE šZ šOEµ•š}OE•U and the SNOS are the trustees, as the customers are expected to provide their sensitive information and make payments before receiving the goods or services, hence making them vulnerable to misbehaviour of the SNOS.

Causes of trust in online shops
There are various studies about trust development in online shopping, as can be seen in the following table:

Functions of Facebook as Community/Network
The social media is often considered to be synonymous with social networks such as Facebook (Giannini, 2009;Brennan & Schafer, 2010). A social network is defined as a community resulting from a number of users in one Web site designed to allow interaction, discussion and sharing of information according to a common interest (Giannini, 2009). A social network is also defined by Dictionary. }u • ^ Á •]š where one connects with those sharing personal or professional interests, place of origin, education at a particular • Z}}oU š _X Social networks such as Facebook is an online social community combining current news and entertainment, complementing existing functions of Web sites such as chat rooms, mailing lists, instant messengers and news groups (Straubhaar, LaRose & Davenport, 2010). As an online social community, Facebook is often utilised as an effective tool to influence opinions or behaviour of users in the community. This is facilitated by the flexibility with which users share their opinion or interests by posting on walls in various communities within Facebook (Darren & Julie, 2009).

Facebook as Online Shopping Tool
While there have been a number of studies on Facebook as social networks Argan, 2012;Darvell, Walsh, & White, 2011;Kim & Lee, 2011;Kalpidou, Costin, & Morris, 2011), there is a lack of information about Facebook as a medium for online transactions. According to Belch and Belch (2012), the selection of social networks as a promotion medium is of interest to sellers due to its ability to disseminate information at low cost, with a high level of engagement with other users and specific targeting of potential customers. Through Facebook, for example, sellers can find out personal information of customers, such as age, address and interests, so sellers can direct customers to appropriate information, based on their preferences (Lane, King, & Reichert, 2011). In other words, sellers will be able to bring their stores to the right customers (Brennan & Schafer, 2010). According to Trusov, Bodapati, and Bucklin (2010), sellers use Facebook to promote products and services, and improve brand image.
A social network is also beneficial in its ability of disseminating word of mouth in a viral manner (Belch & Belch 2012). More interestingly, a social network allows the creation of friendship relations (online communities) that becomes an added value in the creation and development of trust between sellers and customers.
According to the ForeSee study, Facebook has the highest level of use among online shoppers. In general, the majority of online shoppers (69%) are also users of social media sites, and the breakdown of these social media sites show that Facebook is dominant, as shown in the following table.  (Brennan & Schafer, 2010).

Aspects Influencing Trust Generation in Facebook as Online Shopping Medium
Facebook is an example of social network website, whose main characteristic is the facilitation of the formation of a virtual community (Giannini, 2009). Within such communities, an entitativity is created, in which users share a high level of cohesiveness, transcending the borders of collectivity (Stewart, 2003). Users of the social network interact in their own ways (Stewart, 2003), yet with shared values and unwritten rules.
In Facebook transactions, for example, potential customers ask about items s/he is interested in, sellers explain the price and specification of the items, customers provide address and contact information for delivery, and so on. Consistent interactions between sellers and customers in their transactions are expected to strengthen trust from either side.
Unlike in other online transaction media, interaction between customers and sellers in Facebook is not limited to transactions. There is a friendship-like relations between the two, such as commenting positively on each other, P]À]vP Z>]l •[ š} Z }šZ OE[• •š šµ• •U comments or pictures, and so on, not only on items being transacted between the two. Such relation is expected to develop a greater sense of trust.
Similar to other online transaction media, identities on Facebook are often not the actual names of the sellers (Kim & Lee, 2011). However, on Facebook, šZ OE ]• }(š v Z( [ š} šZ }vo]v •š}OE U such as the storeowner providing a personal photograph or status, giving a personal face to her store. This greatly assists the interaction and communication between the seller and potential customers, helping to create trust. The more intense the interaction occurring, the greater the trust of the customer towards the seller, and vice versa.

H1. Intensity of interaction between sellers and customers has a positive association with trust from customers.
Social networking sites such as Facebook allows people sharing common interests to form groups, through which to share their interests and perform transactions. These groupings do not have to be a formal one (i.e. with strict boundaries, leadership, rules etc.), but can be an agglomeration of people with shared interests and shared connections.
Members of these groups can share different roles (pure seller, pure customer, and seller-customer), as long as they share the same theme: craft, fashion, celebrities etc., and information related to the similar theme, and interact about the said theme. Hamilton and Sherman (1996) in Stewart (2003) speculated that within an entitativity in which interaction occurs, individuals within are perceived to share similarities. These similarities form the boundaries of the group: people who have similar interest will attempt to join, and conversely, within one group, influential members will cause other members to have similar interests.
According to McKnight et al. (2002), a number of people sharing common characteristics tend to positively perceive each other and tend to have greater trust about each other. In other words, during a Facebook transaction, it helps potential customers to trust the seller when s/he knows that the seller is not simply a person who sells the goods s/he is interested in, but also shares her interests. This similarity of interest is expected to strengthen the trust from the customer side.

H2. Similarity of interest between sellers and customers has a positive association with trust from customers.
The concept of word of mouth is eloquently expressed in Facebook interactions between sellers and customers. Facebook allows community members to share information, especially when a customer is satisfied with the goods or services provided by the seller, on her wall or status, or ‰OE}À] š •š]u}vÇ }v šZ • oo OE[• ‰ P X According to Cheskin Research and Studio Archetype/Sapient in Lim et al. (2006), satisfied customer endorsements is a tool to affect customer trust. Other community members who note the endorsement will have a greater trust as potential customers towards the seller.  customer, also client, buyer or purchaser is usually used to refer to a current or potential buyer or user of the products of an individual or organization, mostly oo šZ •µ‰‰o] OE }OE • oo OEX_ o OEP number of Pinklady and Partner customers are also customers of other online craft stores, and thus it is expected that the respondents of this study will also reflect the general characteristics of customers of online craft stores in Facebook.

The Proposed Model
The response rate resulting in a sample size of 151 respondents is caused by a number of factors: inactive Facebook accounts of former customers, not all customers logging into their account on a daily basis, and lack of response from customers to the questionnaire. The resulting sample size has a sampling error of 7.3% at a 95% level of confidence. This is within the standard of a maximum sampling error of 8% at a 95% level of confidence, according to DataStar Inc.
A pre-test is done on 30 respondents before the actual test. According to Flanagan, Mahler and Cohen (1994) Leech, Barrett and Morgan (2005). The pre-test also resulted in a validity value of 0.678-0.933 for different indicators. The same source also states that the minimum expected validity value, using KMO is 0.5-0.7, placing the indicators in the valid category. Thus, based on the pre-test, the research instruments can be utilized in the actual test.
There is a limitation in the research in the form of low response rate of the questionnaires. Despite the high number of the sampling frame, only about 20% of the questioned respondents replied. There are several possible factors, such as Facebook users not using computers and using gadgets or Blackberries instead to log in to their account, which do not allow filling online questionnaires, irregular use of Facebook (not logging in at the time of the survey delivery) or inactive Facebook account, lack of technological skills, and difficulty of following up questionnaires. This has an impact on the relatively low number of samples obtained by the researcher, but does not severely affect the sampling error of the survey results.
Another limitation to the analysis of the research is the difficulty in comparing results. The researcher has previously intended to compare the customers from two different online stores; however, there was a complete lack of response from the customers of the other online store. The researcher intends that in the future this limitation can be removed by more actively engaging craft online shop owners to cooperate in the collection of data.
Thus we have a description of the customers of Facebook online craft shops, which are dominated by females of productive age (20-39 years of age), and have a monthly income level of less than Rp. 5 million, but managing to spend up to Rp. 499 thousand per month for craft shopping, or about 10 percent of their income.

Correlations
According to Sarwono (2009) This study uses the Spearman's rho correlation, with SPSS software to calculate the relations between the independent variables and the dependent variable. The values of the correlation will be shown as follows.

Correlation between Intensity of Interaction and Trust
The Spearman correlation value of these two variables shows a value of 0.447 with a significance of 0.000. This means a medium level of correlation between the two and a significant relationship at an alpha of 0.05.

Correlation between Similarity of Interest and Trust
The Spearman correlation value of these two variables shows a value of 0.410 with a significance of 0.000. This means a medium level of correlation between the two and a significant relationship at an alpha of 0.05.

Correlation between Endorsement and Trust
The Spearman correlation value of these two variables shows a value of 0.520 with a significance of 0.000. This means a medium level of correlation between the two and a significant relationship at an alpha of 0.05.

Correlation between Combined Independent Variables and Trust
The Spearman correlation value shows a value of 0.548 with a significance of 0.000. This means a medium level of correlation between the two and a significant relationship at an alpha of 0.05.
The correlations show a medium relationship between the independent variables and the dependent variable, and a significant level of relationship. In this case, we can accept the hypothesis that the creation of trust is related to the independent variables of intensity of interaction, similarity of interest and endorsement. However, the medium relationship between the independent and dependent variables show that there are possible alternative explanations about the creation of trust. The following section will discuss these possibilities.

Insights
Despite Facebook's superiority as an online shopping tool due to the high level of engagement between sellers and customers, there is a medium level of relationship between the independent and dependent variables. Analysis of trust in the context of online shopping thus shows itself as a complex issue, in which there are other possible causes of trust creation, such as technological advances, and provision of information as a standard principle of the online shop (such as privacy protection).
The issues of familiarity and pricing apparently play important roles in creating trust from customers. As mentioned by informant D, "I trust and decide to buy when I know the seller, besides, the cheap price is also a factor in my decision to buy" (Interview with D. Tangerang, January 25, 2013). I trust a seller from my experience with a number of purchases, she gave a good service, such as items sent as I requested, the product offered is the same as the one I received, and because the seller often provides information about new products that I may like and I can consult with her before deciding to buy (Interview with M. Tangerang, January 25, 2013).
The statement above shows that seller's service and fulfilment of expectations can also act in the formation of consumer trust. The following table shows reasons mentioned by customers to buy items from Facebook stores (respondents may choose more than one reason). The table shows that practicality is the main reason why customers decide to do online shopping, and the second main reason is to obtain items that are not available elsewhere. Pricing is a distant third reason, and when related to the relatively low amounts of money spent in craft stores, this may explain why some customers would want to make purchases despite relatively unknown trustworthiness of certain sellers, as mentioned by J, "For new online shops from whom I have never bought, I would try to make a bet by making a small purchase and note the reaction of the seller and how she The use of the Facebook platform for online sales may also influence the less than strong relationship between the independent variables and the creation of trust. Facebook levels the playing field, bringing the sellers and customers to equal position, and provides all parties with a system to check each other, such as by comments and statuses on each party's timeline, making it possible for potential customers to assume the trustworthiness of a seller.
Other factors that may influence the creation of trust may be related to more practical aspects of the seller, such as pricing, brand, quality, availability of items etc. There are also unwritten rules of interaction among the 'sisterhood' of craft sellers and customers, such as giving information and updates, and not tolerating non-paying customers (often termed 'hit and run'). This aspect of trust creation requires further research.

Conclusion
Results of the correlation between the independent variables (intensity of interaction, similarity of interest and endorsement) and the dependent variable (trust) show that each of the variables has a medium level of correlation. The correlation strength increases when the independent variables are combined, showing that the variables act upon each other to strengthen the relationship between the independent and dependent variables.
However, the medium strength of the correlation shows that there are other possible explanations of the creation of trust. The researcher argues, based on findings on the position of pricing in the table of reasons to shop online, the amount of money spent online in a monthly basis, and interview results, that the customers and sellers generally have a baseline level of trust that is already high, which is improved further when the interaction, interest and endorsement are added up.
It seems that there is indeed a feeling of sisterhood in general among the community of craft sellers and customers in Facebook that allows a high level of trust, and despite the small possibility of untrustworthiness of the seller, potential customers tend to put their trust even to new sellers who have joined their community as fellow 'sisters' who will not cheat them, despite the lack of prior interaction experience. It would also be interesting to do further research on online stores whose patrons are males, to see whether an analogous 'brotherhood' exists among them.