Determinants of Young Sri Lankan Individuals’ Intention to Engage in Viral Marketing

Despite the prolific growth of viral marketing, factors that determine successful viral marketing campaigns remain largely unknown. Since the extant literature postulates that individual who pass along messages play a vital role in viral marketing, this paper examines the factors that determine the individuals’ message pass along intention in light of the behavioural theory proposed by Ajzen, namely, the Theory of Planned Behavior (TPB), and the interpersonal relations theory proposed by Schutz, namely, Fundamental Interpersonal Relations Orientation (FIRO). The data collected from 391 young undergraduates in Sri Lanka were analyzed using the Structural Equation Modeling method. The results revealed that perceived behavioural control, attitude towards viral marketing, subjective norms, and need for affection and need for control are influential in determining the intention among young individuals to engage in viral marketing. Thus, the findings reveal that firms may need to develop different strategies to manipulate those critical factors that affect one’s intention to engage in viral marketing so as to enhance the effectiveness of viral marketing campaigns.


Introduction
The proliferation of information and communication technology (ICT) opened many opportunities as well as challenges for businesses around the globe. Marketing communication has expanded widely with the support of ICT, whereby it is no longer restricted to the conventional one-way firm-to-consumer or to the more recent two-way or bidirectional communication. Communication flows in diverse directions, exploring the links or relations that individuals have with others through information and communication technology (Camarero & San Jose 2011). Since customers face great exposure to the world it is imperative that organizations work on tapping them by using every possible method by which they gather exposure. Among the many new marketing communication tools that have evolved, viral marketing has become a very important tool (De Bruyn & Lilien as cited in José-Cabezudo & Camarero-Izquierdo 2013) because of its inherent characteristics that strongly support marketing innovations.
Viral marketing can be defined as electronic word of mouth whereby some form of marketing message related to a company, brand or product is transmitted in an exponentially growing way, often through the use of social media applications (Kaplan & Haenlein 2011). The terms "viral marketing" and "eWOM" are often used interchangeably (Camarero & San Jose 2011;Ho & Dempsey 2010;Vilpponen, Winter & Sundqvist 2006) and this study also uses does likewise. Previous studies have shown that eWOM is more influential than traditional marketer-created information because it is perceived to be more impactful, convenient and credible (Cheung & Thadani 2012, Henning-Thurau, Gwinner, Walsh & Gremler 2004, Bickart & Schindler 2001. Moreover, industry research reports have also substantiated the fact that when making purchase decisions, internet users trust online reviews posted by unknown consumers more than they trust the traditional media (eMarketer, Inc 2013).
Despite the propagation of eWOM over the years, it is only very recently that eWOM received serious managerial and academic attention (Cheung & Lee 2012). The extensive studies done on viral marketing have been mainly about the effectiveness of eWOM communication where the focus was on purchase intention, purchasing decision and repurchase intention (Cheung & Thadani 2012;Chan & Ngai 2011) with only limited academic attention given to the critical factors affecting successful viral marketing campaigns (Cheung & Lee 2012, Chan & Ngai 2011, Henning-Thurau, Gwinner, Walsh & Gremler 2004. According to Chan and Ngai (2011), among the studies conducted with regard to eWOM, only 20% focused on the factors critical to efficacious viral marketing campaigns. They point out that the success of eWOM campaigns is largely determined by two factors, namely, the message and the sender. Moreover, even with the rapid growth of Social Networking Sites (SNS), much of viral marketing research has been more focused on website interactions (Chan & Ngai 2011) and less on the intention to engage in viral marketing through SNS.
Further, most of the research on eWOM has been done in developed countries in the western world but little in the Asian context and that also confined to developed countries like Korea and Hong Kong (Chan & Ngai 2011). Therefore, the findings of such research may not be applicable to developing countries because of the socio-cultural and economic differences between developed and developing nations. In viral marketing in Sri Lanka, many of the Sri Lankan brands have still not begun to reap the maximum benefits of using eWoM through SNS (Thajudeen 2012). An exception to this has been recorded in the Sri Lankan political arena where eWOM was used effectively as a promotional tool. According to Abbas & Singh (2014), Prime Ministerial candidate Narendra Modi realized the potential of social media and used it as a key tool in his campaign. In the Sri Lankan political arena there was an unprecedented use of the social media by both presidential candidates as well as by the politically charged but yet unaffiliated youth (Gunawardena 2015). With this in mind, this paper aims to investigate the factors affecting Sri Lankan youths' intention to engage in viral marketing, particularly in political events via social networking sites (SNS).
The remainder of the paper is organized as follows: section two presents the theoretical background of the study; section three outlines the research method; section four presents the results and findings of the study; and section five provides the discussion and conclusions of the study followed by research implications, limitations and directions for future research.

Theoretical Background
Scholars widely perceive online content forwarding as a unique behaviour (Lee, Ham & Kim 2013, Camarero & San Jose 2011, Yang 2013, Yang, Liu & Zhou 2012, Hsu & Lin 2008) and behavioural intention as a more influential predictor of behaviour (Pavlou & Fygenson 2006). Intentions are presumed to capture the motivational factors which influence behaviour and they are indications of how hard people are willing to try or how much of an effort they plan to exert to perform a particular behaviour (Ajzen 1991). In this regard, the Theory of Planned Behaviour (TPB), an extension of the Theory of Reasoned Action (TRA), provides a useful framework for dealing with the complexities of human social behaviour (Ajzen 1991, Ajzen & Madden 1986. TPB is considered to be a well-accepted model that has been used to predict behaviours ranging from classroom participation to dishonest behaviours such as cheating in tests, shoplifting, lying and avoiding assignments. It permitted more accurate predictions of intentions (Ajzen & Madden 1986). According to TPB, the individual's intention to engage in a particular behaviour depends on his attitude to such behaviour -the degree to which a person makes a favourable or unfavourable evaluation or appraisal of the behaviour in question, subjective norms -perceived social pressure to perform or not to perform the behaviour and perceived behavioral control (PBC) -the perceived ease or difficulty of performing the behaviour (Ajzen 1991, Ajzen & Madden 1986. PBC is a construct which differentiates TPB from TRA where PBC was incorporated owing to the limitations of the original TRA model in dealing with behaviour over which people have incomplete volitional control (Ajzen 1991).
Since intention is presumed to capture the motivational factors which influence behavior, in a study of viral marketing it is pivotal to look at interpersonal communication motives which drive individuals to engage in viral marketing. Related to this, the theory of Fundamental Interpersonal Relations Orientation (FIRO) (Schutz 1966) emphasizes that people engage in interpersonal relations as they are motivated to express one or more of three interpersonal needs, namely, inclusion, affection and control. As engaging in viral marketing is a form of maintaining interpersonal relations, motives that drive the indiviual's intention to engage in viral marketing can be explained in light of the FIRO theory (Schutz 1966). According to FIRO, the need for inclusion refers to the individual's desire to be a part of a group; need for affection refers to an individual's desire to love or to be loved by others; and need for control refers to the need to exert power over others or give power over the self to others.
When analyzing the need for inclusion in depth it is seen that while belonging to a group, the individual desires to be identifiable or different from other people (Schutz 1966). This has led to claiming that the need for inclusion is an integration of need to belong and need to be different or individuation. While an individual wants to be a part of a group, he wants to obtain recognition or unique identification among the group members. It has been hypothesized that the individual's intention to engage in viral marketing depends on three factors proposed by Ajzen (1991) and four factors proposed by Schutz (1966), as explained below.

Attitude
Much of the prior studies related to techonology adpotion and knowledge sharing supported the individual's intention to engage in a particular behaviour being positively related to his/ her attitudes (Mishra, Akman & Mishra 2014, Taylor & Todd 1995, Pavlou & Fygenson 2006, Bock, Zmud , Kim & Lee 2005, Ranasinghe & Dharmadasa 2013. This has been further substantiated in the context of viral marketing, where Lee, Ham and Kim (2013) came up with the finding that attitudes positively affect an inidvidual's intention to pass along online video advertisements. Further, in an effort to determine mobile viral marketing behaviour among young Chinese customers, it was revealed that intention to forward useful electronic messages and entertaining messages was determined by the attitudes to viral messages of senders (Yang, Liu & Zhou 2012). Yang (2013) asserted that young Chinese consumers' frequency of product recommendations on social networking websites is influenced by their viral marketing attitudes. Based on the above, it can be assumed that the more favourable the attitude towards a particular behaviour, the stronger the intention to engage in it. Thus, we hypothesise that H 1 : Attitude towards viral marketing positively influences intention to engage in viral marketing of political events through SNS.

Subjective Norms
Subjective norms reflect consumer perceptions of whether particular behaviours are accepted, encouraged and adopted by the consumer's circle of influence (Pavlou & Fygenson 2006, Hsu & Lin 2008, Palka, Pousttchi & Wiedemann 2009). The impact of subjective norms on the intention to engage in a given behaviour has been studied along diverse lines of research, the findings of which reveal positive relationships. Substantiating the findings of TPB (Ajzen 1991), empirical studies have shown that subjective norms influence intention to adopt technology (Taylor & Todd 1995, Pavlou & Fygenson 2006, Hsu & Lu 2004and Mishra, Akman & Mishra 2014. Moreover, prior research underlines that subjective norms of knowledge-sharing positively affect the intention to share knowledge (Bock , Zmud , Kim & Lee 2005, Ranasinghe & Dharmadasa 2013. In addition, a significant positive impact of subjective norms on intention to engage in viral marketing has been shown in different contexts (Yang 2013, Yang, Liu & Zhou 2012, Lee, Ham & Kim 2013. Since the extant literature supports the thesis that subjective norms positively impact on intention to engage in behavior, the following hypothesis is proposed: H 2 : Subjective norms in respect of viral marketing positively influence intention to engage in viral marketing of political events through SNS.

Perceived Behavioural Control
Perceived behavioural control (PBC) is the person's belief as to how easy or difficult his/ her behaviour is likely to determine his/her intention (Ajzen 1991). An individual's intention to engage in a particular behaviour will be shaped by how he/she percieves his/her ability to perform. The positive impact of PBC on the individual's intention has been recognized in the existing literature on technology adoption (Han, Hsu & Sheu 2010, Taylor & Todd 1995, Grunert & Ramus 2005, Askew, et al., 2014, Ross, et al., 2009, Cheung & Vogel 2013. In regard to knowledge-sharing behaviour, Lin and Lee (2004) found that percieved ease of sharing knowledge with others positively impacts on the individual's intention to engage in knowledge-sharing. Since PBC is shown to generate a significant positive effect on intention to engage in techonology adoption as well as in knowledge-sharing, this study hypothesises that: H 3 : PBC positively influences intention to engage in viral marketing of political events through SNS.

Need to Belong
Every individual has a strong desire to be a part of a particular community and to maintain interpersonal relationships. Schutz (1966) recognizes that the need to belong involves including significant others in a circle of acquaintances and remaining "in the loop". By nature, people are in need of love and social acceptance, which is referred to as the "need to belong" (Leary, Kelly, Cottrell & Schreindorfer 2013). In their study exploring the determinants of word of mouth in online consumer platforms, it was found that the feeling of being emotionally involved with the group had a significant impact on consumers' eWOM intentions (Cheung & Lee 2012). Further, Phelps, Lewis, Mobilio, Perry and Raman (2004) asserted that consumers' motivations to pass along emails are backed by the desire to connect and share with others. Thus, the need to belong could be considered to have a positive influence on the intention to engage in viral marketing of political events in Sri Lanka. Thus, it is hypothesized that H 4 : The need to belong positively influences intention to engage in viral marketing of political events through SNS.

Need for Individuation
While belonging to a particular group, individuals seek to stand out in their group and to make themselves different from others. It has been accepted that social behaviour often occurs with reference to other people and is also evaluated in those terms. Some behavioural options are chosen for the purpose of creating a particular social image whereas others are selected for the function of self-expression (Maslach, Stapp & Santee 1985). Anything that enhances an individual's self-image, self-expression and self-determination drives the individual to engage in that behaviour. For example, if an individual perceives that spreading eWOM will enhance his/her self-image, he/she will engage in eWOM. In an effort to reveal the underlying motives for consumers engaging in WoM communication, it has been revealed that the need for self-enhancement is a key motivator that drives individuals to engage in positive WOM (Sundaram, Mitra & Webster 1998, Henning-Thurau, Gwinner, Walsh & Gremler 2004. Individuals expect to show their expertise, to enhance their status and seek appreciation for their enhanced self-image through passing along messages. Chung and Darke (2006) found that individuals were more likely to engage in WOM for products that tell others about who you are and which make you unique and which are closely aligned to one's self-image. In their work, Ho and Dempsey (2010) found that individuation or the need to stand out in a crowd was a key motive for an individual to engage in eWOM. Thus, the need for individuation could be considered to have a positive influence on the intention to engage in viral marketing related to political events. Thus, it is hypothesized that:

Need for Affection
Love and care are among the foremost needs of an individual. Affection has been defined as the need to establish and maintain a relationship of mutual warmth, care and love with others (Schutz 1966). Individuals desire to express and receive love and care in any relationship they maintain in a physical or virtual context. To have such mutual warmth, the individual should be willing to help others in a selfless manner. The act of doing something for others without anticipating any reward in return is altruism (Sundaram, Mitra & Webster 1998)

Need for Control
The need for control is the desire to exert power over others or to give power over one's self to others (Rubin, Perse & Barbato 1988). Schutz (1966) emphasized that the need for control is an imperative motivation for individual interpersonal relationships including interpersonal communication. Substantiating the finding that the need for control is an interpersonal need (Schutz 1966), the Interpersonal Communication Motives (ICM) scale developed by Rubin, Perse, & Barbato (1988) has also identified the need for control as a motive for individuals to engage in interpersonal communication. Smith (2012) suggested that achieving superiority and influencing power is a main motivation for passing eWOM communication in websites and blogs. In addition, Sun, Youn, Wu, and Kuntaraporn (2006) found that online opinion leadership is positively related to online forwarding in the context of music-related communication. Opinion leaders are presumed to be influential members of their social network with a desire to control others (Chaney 2001). Hence, the need for control could be considered to have a positive impact on the intention to engage in viral marketing relating to political events in Sri Lanka. Thus, it is hypothesized that: H 7 : The need for control positively influences intention to engage in viral marketing of political events through SNS.
The conceptual framework of the study based on the above discussion is given in Figure 1.

Data Collection and Sample
Data for the study was collected from 400 undergraduates registered for different degree programmes at four higher educational institutions in Colombo district, aged between 18 and 24 and active in at least one social media network. Convenience sampling was used and data collected through a questionnaire. Much of the previous research has proposed that young adults are the most appropriate segment for conducting studies on viral marketing as they form the demographic group with the highest rates of technology adaptation and viral marketing penetration (Lee, Ham & Kim 2013, Camarero & San Jose 2011, Ho & Dempsey 2010. The age group 18-24 years was considered because, according to the UN, youth is defined as those persons between the ages of 15 and 24 years (UNESCO 2014) and to qualify as a voter, an individual must have attained the age of 18 years by the date of the relevant register -June 1st. In order to increase the response rate, anonymity was assured and as an incentive a summary of the research findings was offered to the participants. Further, several follow-up procedures were used such as reminder notices to the participants and reminder messages to SNS accounts.
The questionnaire used in this study consisted of three sections: screening questions to ensure respondents were within the age group and have an SNS account; predictor constructs and intention measures; demographic information. To measure the constructs, already established measurement scales were adapted to the Sri Lankan setting. Altogether, there were forty two (42) questions, all of the Likert style rating scales which enable respondents to indicate how strongly they agree or disagree with the statements (Saunders, Lewis & Thornhill 2011). A five-point Likert scale was used from 1 = strongly disagree and 5 = strongly agree (Refer Appendix 01).

Method of Analysis
Prior to hypotheses testing, the data was examined, and with the support of SPSS, outliers were identified, missing values were treated and multivariate assumptions for the data set tested. The hypotheses of the study were then tested using Structural Equation Modelling (SEM) through AMOS. Following Anderson and Gerbing's (1988) two-step approach, SEM was conducted to analyze the data. The measurement model was first constructed and tested with confirmatory factor analysis (CFA). Secondly, the structural model was specified in order to examine the causal relationships among variables or predictors and criteria.

Results and Findings
With the support of SPSS, the data was purified and multivariate assumptions tested. It has been proven that all multivariate assumptions been met. In addition before proceeding into inferential statistics researcher assessed the unidimensionality, validity and the reliability of data. validity and the reliability of the data. Hence, the study was conducted with scaled items and Cronbach's Alpha was used to measure internal consistency (i.e., reliability). Hair, Black, Babin, & Anderson (2009) recommended that Cronbach's alpha values ranging from 0.6 to 0.7 can be treated as the lower limit of acceptability. Accordingly, Cronbach's alpha for each item was calculated and several items had to be removed from the scales in order to exceed the cut off value of Cronbach's alpha. The NTB construct was identified as a significantly poor variable in terms of internal consistency as the alpha value was 0.484.
With the removal of two items, namely, NTB3_R and NTB7_R, the Cronbach's alpha value was indicated as 0.642 after the purification. All the other variables were considered to be internally consistent as the corresponding alpha values were greater than 0.7. Convergent validity measures the extent to which a scale has a positive correlation with other measures of the same latent construct (Sekaran & Bougie 2014). Assessing factor loadings is a commonly used method to determine convergent validity where higher factor loadings represent higher convergent validity while all factor loadings should be statistically significant. (2009) determined that the threshold value for higher factor loading is the standardized loading estimates which should be 0.5 or higher, and ideally 0.7.

Hair, Black, Babin, & Anderson
The Average Variance Extracted (AVE) which calculates the mean variance extracted for the items loading on a construct is another indicator of convergence. It is accepted that AVE of .5 or higher is a good rule of thumb to ensure adequate convergence (Bagozzi & Yi 1988).
Further, Construct Reliability (CR) was also introduced as an indicator of convergent validity and the values greater than.6 were desirable for CR (Bagozzi & Yi 1988).
Discriminant validity emphasizes the extent to which a construct is truly distinct from other constructs in a model (Hair, Black, Babin & Anderson 2009). This indicates that the measured variables have more in common with the construct they are associated with than they do with the other constructs. A more rigorous way to determine discriminant validity can be the comparison of the square root of the AVE and the correlation coefficients. If the square root of AVE is larger than the correlation coefficients, discriminant validity is established (Malhotra & Dash 2011).
Accordingly, a measurement model was used to assess convergent and discriminant validity. Table 1 shows AVE, CR, Correlations, Square Roots of AVE values for each variable. In respect of the measures, all factor loadings were statistically significant and the standardized loading estimates met the minimum threshold of 0.5, which provided a reasonable estimation of convergent validity among item measures. Table 1 shows that following Bagozzi and Yi's (1988) proposition, the CR of the constructs was confirmed with all constructs demonstrating a value which exceeds 0.6, which is another indicator of convergent validity. However, AVE for all constructs exceeded the threshold of 0.5 except for the constructs NTB and NTI. According to Malhotra and Dash (2011), AVE is considered to be a more conservative measure than CR. Thus, on the basis of CR alone, the researcher has the discretion to conclude that the convergent validity of the construct is adequate. Therefore despite the low AVE values for constructs NTB and NTI, convergent validity was confirmed with sufficient backing from factor loadings and CR. Discriminant validity for all the constructs except PBC and INT was established with the square root of AVEs being larger than the corresponding correlation coefficients.

Structural Model
The causal relationships among the latent constructs in the model and hypotheses were  The statistical results given in Table 3 show that out of the seven hypotheses proposed, five were supported while two were unsupported. Among the interpersonal communication motives, only two hypotheses were supported, suggesting that the need for belongingness and need for individuation may not be possible motives of a young individual's intention to engage in viral marketing of political events through SNS.

Discussion
The findings of the study confirmed that perceived behavioural control is the most pivotal motivation that governs a young individual's intention to engage in viral marketing of political events in Sri Lanka. This finding was consistent with several prior researches related to the use of information technology and studies related to antecedents of specific behaviors (Han, Hsu & Sheu 2010, Taylor & Todd 1995, Grunert & Ramus 2005, Askew, et al., 2014, Ross, et al., 2009, Lin & Lee 2004).
However, it was interesting to note that the path coefficient (β) established in the study was relatively high compared to the findings of the aforementioned studies. The difference in the degree of the relationship could be mainly due to the differences in the context of the research.
In this study, the concern was with intention to engage in viral marketing specifically of political events. This is different from the use of internet for buying purposes or the use of computer resource centres. Unlike other generations' youth are investigative, freely expressive, and the need to authenticate (Smith 2011) thus they perceive lesser obstacles from the political system to express their independent political viewpoints. Further, nowadays young individuals with a wired life style may be technologically capable of navigating and expressing themselves in SNS very effectively. Thus, young individuals believe they possess more resources and opportunities and face fewer obstacles or obstructions, resulting in greater perceived behavioural control (Ajzen 1991), which, in turn, form their intentions resulting in a high path coefficient.
With a positive and significant standardized beta coefficient the study confirms that subjective norms have a positive impact on intention to engage in viral marketing related to political events in Sri Lanka. This finding is also in line with many of the previous research findings  (2013) and Yang, Liu & Zhou (2012) . However, the established path coefficient is smaller compared to that shown by Hsu & Lin (2008), Pavlou & Fygenson (2006) and Lee, Ham & Kim (2013). The difference in the degree of relationship could be due to differences in the contexts of the research. The majority of prior studies on viral marketing intentions were conducted either in Western countries or in south east Asian countries.
The findings of the study are in line with the original TPB model which postulates that the individual's intention to engage in a particular behaviour depends on the attitude towards such behaviour, subjective norms and perceived behavioural control (Ajzen 1991, Ajzen & Madden 1986, Ajzen 2002. Another interesting finding is that percieved behavioural control, which is a major determinant that differentiates TPB from TRA, is recognized as the most vital factor affecting one's intention to engage in viral marketing in a given context. Thus, the research is in line with the same conclusion reached by Taylor & Todd (1995) and determines TPB as a viable model that provides a broader explanation of technolgy adoption and engagement in viral marketing.
Among the four interpersonal communication motives which affect intention to engage in viral marketing based on Schutz (1966), only two, the need for affection and the need for  (2004), which identified concern for other consumers as a key determinant in engaging in viral marketing. Considering the participants of the study, the signficance of affiliation orientation can be related to Sri Lankan culture as well.
The study found support for the hypothesized relationship between need for control and intention to engage in viral marketing of political events with a positive and significant path coefficient. This was also confirmed in the study done by Sundaram, Mitra and Webster (1998) as achievement obtained by using a product and by being an intelligent shopper motivates them to share information with each other with the need to change and influence the behaviour of others in the network. Further Sun, Youn, Wu & Kuntaraporn (2006) found that online opinion leadership with a desire to control others is positively related to online forwarding in the context of music-related communication. In addition, the study established a path coefficient of 0.23, which is somewhat similar to the relationship established in the study of Sun, Youn, Wu & Kuntaraporn (2006). There are other similar researches that did not establish such a relationship (Ho & Dempsey 2010, Sun, Youn, Wu & Kuntaraporn 2006. The insignificant relationship can be attributed to the purpose served by SNS. According to studies conducted by Ross, et al (2009) and Sheldon (2008), most 'Facebook Friends' were individuals known from the offline world; in other words, the Facebook was not used as a tool to meet new people online or to belong to new groups and it was just used as a tool to maintain contact with the groups to which they already belong.
Further, based on substantial amounts of literature, researchers have hypothesized a relationship between one's need for individuation and intention to engage in viral marketing (Sundaram, Mitra & Webster 1998, Ho & Dempsey 2010, Daugherty, Eastin & Bright 2008Henning-Thurau, Gwinner, Walsh & Gremler 2004). However, the present study did not establish such a relationship as hypothesized. The findings of the study are much aligned with a study conducted by Yap, Soetarto & Sweeney (2013) and Ghane, Shokrizadeh, Omidvar & Kamyab (2014). It is not suggested that viral marketing is used as a means of showcasing one's uniqueness because the results of this study may have been influenced by the special personality traits of Sri Lankans. As Lee (1994) says, by nature Asians have compliant and obedient personalities and thus may not be willing to stand out from the crowd.

Conclusion
This study addressed a gap in the extant literature on viral marketing by determining factors affecting towards individuals intention to engage in viral marketing through SNS. Researcher examined motives which drives an individual to engage in viral marketing of political events through SNS in light of well established theories TPB (Ajzen 1991) and FIRO (Schutz 1966).
The findings of the study have implications for both marketing practice and theory, and it enlighten about possible further research that can be undertaken on intention to engage in viral marketing. will enable readers to show their appreciation for the reviews received. This may satisfy the contributors' need for control by presuming that they were able to influence the decisions of others.

Theoretical and Practical Implications
Further, the findings substantiated that individuals tend to engage in writing and spreading messages when they feel that their friends and relatives accept and engage in viral marketing on SNS. Platform operators can work on sending personalized updates to individuals about those friends who engage in diffusing messages. Also, posting a public note of appreciation to contributors for spreading the message will enable the individuals in his friends' list to know that he engages in viral marketing. This may give a boost to individuals to get involved in diffusing messages as a norm among their friends.
The findings also reveal the pivotal impact of perceived behavioural control on the individual's intention to engage in viral marketing. To enhance the viral nature of the message, marketers should always work on enhancing the customers' level of confidence in the ease of performing the task with few obstacles and resource constraints. For example, most young Sri Lankans find it difficult to diffuse highly sophisticated rich content due to Internet bandwidth issues in the country. Thus, they may avoid diffusing messages of such calibre which may freeze the buzz. In order for practitioners to foster eWOM engagement activities on SNS, user-friendliness needs to be considered when providing service features such as searching, listening, and viewing, purchasing and sharing. Perceived behavioural control can be enhanced by platform operators through usability evaluation, service process reengineering, or user-centered design. Digital marketing practitioners can improve the system by designing a more controllable user interface, eliminating unnecessary plug-ins, enabling easier communication among peers and organizing the information architecture to enable users to more easily navigate its services in their fan pages and applications.

Limitations and Further Research
While this study's findings have considerable potential implications, there are several limitations. Firstly, the researcher has drawn her sample from undergraduates as they are considered appropriate participants because of their tech savvy and substantial use of SNS.
However, to get a broader picture of viral marketing behaviour among young individuals future research needs to pay attention to wider groups from diverse settings. This will enhance the generalizability of findings which is a foremost objective of conducting a quantitative study.
Secondly, the study relates to the context of viral marketing of political events in Sri Lanka. The use of viral marketing for political events is considered viable since many of the Sri Lankan brands are still not able to reach maximum benefits by utilizing eWoM through SNS (Thajudeen 2012) and the exception for this is an unprecedented use of social media in political campaigns (Gunawardena 2015). However, in general people do not have much of a good perception about politics in Sri Lanka and it may have impacted the responses. Thus, it is necessary to project the derived model to other settings such as viral marketing of fast moving consumer goods, fashion, etc. in order to get a broader view.
Thirdly, though the study was limited to the determinants of intention derived from FIRO theory and TPB model, there are many interpersonal psychological and behavioural theories that can be used to explain the determinants of eWOM intention. Examples of such theories are social exchange theory, equity theory, balance theory and theory of reasoned action (Chan and Ngai 2011). Hence, further research can be enriched by using more motivational and behavioural theories and exploring interactive behaviour, which has not been sufficiently explored (Liu 2007). Further, the need for belongingness and individuation might not have a direct impact on intention to engage in eWOM. Instead, it might have an indirect effect through mediation, which can be tested in future research. Finally, it is also possible that consumers' pass-along intentions may also depend on the type of content of the message (Lee, Ham, & Kim 2013) where people perceive some messages more viral than others.
Therefore, future research needs to test how eWOM motives impact through different types of online messages.