Increasing Awareness and Reputation of Merck S.A. Portugal through Employee Advocacy

Merck KGaA is a global science and technology company, focusing on the healthcare industry, life science, and performance materials. Merck in Portugal is a relatively small subsidiary with a midsized office employing approximately 114 employees. Merck S.A. Portugal focuses mainly on the healthcare market. As the company went through a major transformation in 2016, it created a need to communicate on a low-cost basis and, in an efficient way, using digital as its main channel. The current paper shows the implementation of an employee advocacy strategic plan as a solution to digital communication. Employee advocacy is a new communications concept derived from employee social media usage and employee word of mouth. It has been noted that employee networks are wider than company’s own and employee word of mouth is more trustworthy. Before the tool could be implemented, a research on what motivates employees to commit to brand citizenship behavior was performed. Findings revealed that in the case of Merck S.A. Portugal employees, individual internal drive and brand knowledge are factors affecting willingness to share on social media. Based on the results, a series of internal strategies were applied on- and offline.

The goal of the current in-company project was to incorporate a digital, a social media centered presence of Merck S.A. Portugal online, which involved employees as the main advocates of the company.
An Employee Advocacy program was decided to be launched. The employee advocacy tool allows employees to share company content on social media without fear of endangering compliance. However, before an employee advocacy tool could be implemented the readiness of the organization to commit to Brand Citizenship Behavior was to be measured. Based on the result of BCB the author made conclusions relative to employee advocacy and the early results obtained by using such a program for the purpose of communication. uploading employee generated text/photos and contribute to an already written article online (Rokka, Karlsson, Tienari, 2014) or sharing their knowledge within their social network offline with friends and family, for example (Morokane, 2014).
From the organization's point-of-view, employees are middle men between external stakeholders and the company. Employees, deliver the message of the company to enhance reputation and elevate brand awareness among stakeholders (B2B clients, B2C customers, friends, family, colleague's etcetera). (Rokka, Karlsson, Tienari, 2014) From the stakeholder's perspective, employees are the ones who provide brand reality for the listener (Piehler, King, Xiong 2013). Therefore, employee communication needs to be aligned with the company's communication regarding, brand/ -s, identity, and other messages intended to be communicated, that the reality perceived by the stakeholders is aligned with the intended message. When the messages are aligned, an increase in awareness and reputation may be obtained, not to mention a possibility to shape the image of the company (Morokane, 2014).
Therefore, the company is in charge of giving the right guidance and tools for the employee, and encourage them to speak on behalf of the company. On the other hand, it is not sufficient to have only the company engaged, the employees need to be willing to collaborate as well (Grace, King 2012 According to Piehler et al. (2016) and Burmann & Zeplin (2005) brand citizenship behavior reflects the behavior of an employee which is congruent with the brand promise and -identity and the outcome is to strengthen the brand. Grace and King (2012) define brand citizenship behavior as extracurricular and voluntary employee activities on behalf of the brand/company for example, positive word of mouth. Therefore, it can be widely understood that brand citizenship behavior is extra-role behavior conducted by the employee voluntarily and is in favor of the brand. Burmann and Zeplin (2005) originally defined brand citizenship behavior as a spur from a previously established concept known as OCB (organizational citizenship behavior). As a criteria for organizational citizenship behavior, an employee must ´live the brand´, indicating that employees exercise voluntary behaviors which are outside of role-expectations in their day-to-day work. However, brand citizenship behavior takes the concept of OCB further, claiming it to bring the brand to life. Therefore, brand citizenship behavior refers to additional, employee conducted voluntary pro-brand behavior which takes place externally.
The employees who commit to BCB are generally referred to as brand ambassadors or employee advocates. An antecedent or requirement of BCB is brand commitment (Burmann, Zeplin. 2005). Brand commitment is defined as an affective, psychological attachment to the brand which drives employees to feel responsible of communicating it to others, in order to reach the brand goals. Burmann and Zeplin (2005) also define the antecedents of brand commitment as; compliance (consistent behavior with code and conduct), identification (belonging to the group) and internalization (pairing brand values with self-concept). These three antecedents are not interdependent, and therefore fulfilling one of the three alternatives is sufficient for BCB. However, before BCB can take place, employees need a sufficient know-how and resources to be consistent. Helm, Renk, Mishra (2014) found in their research that brand pride is a factor which affects BCB (brand citizenship behavior). They also found that brand self-congruity (brand internalization, according to Burmann and Zeplin (2005) affect BCB, and therefore is an antecedent in the creation of brand ambassadors and employee advocates. Self-congruity means pairing the employees own values with the company´s to find a consistent ground (Burmann and Zeplin 2005). Self-congruity can be directed by the company through organizational socialization, previously presented by Grace and King (2012). Burmann and Zeplin (2005) also highlighted the importance of identification with the brand, by claiming that when an employee feels it belongs to a group, success and failures are sensed on a personal level. According to Piehler et al. (2016) when an individual identifies himself with a brand he automatically thinks and acts for the benefit of the brand, follows code and conduct as well as, commits to behavior which is beyond role requirements hence to, incorporating the brand to the self-concept. Also other authors referred to in this paper also agreed that, selfidentification with the brand leads to brand citizenship behavior. (Rokka, Karlsson, Tienari 2014), (Dreher 2014), (Xiong, King, Piehler 2013), (DuBois Gelb, Rangarajan 2014).
When considering a corporate brand, Burmann and Zeplin (2005) draw a parallel between brand commitment and organizational commitment, referring to it as the same thing.
Therefore, being committed to the corporate brand is also considered as to being committed to the employing organization.
2.2 eWord of Mouth Grace and King (2012) investigate in what drives employees to feel brand commitment and what drives them to perform brand citizenship behavior. A vertinent of brand commitment is referred to as a positive emotion an employee has towards an organization (Grace and King 2012) or quoted from " Burmann and Zeplin (2005:284) /…/ extent of psychological attachment of employees to the brand, which influences their willingness to exert extra effort towards reaching the brand goals /…/" The higher the brand commitment, the higher the probability that the employee commits to brand citizenship behavior. In the core, the factor affecting employee's emotions (BC) and motivation to act in order to go over and beyond (BCB) is a balanced relationship. The organization needs to provide the employee with information and knowledge related to the work and the brand itself (organizational socialization). Moreover, the environment has to be supportive and give space for relationship exchanges (relationship orientation). On the other hand, the employee has to be open to dialogue, have a desire to understand and be engaged to the work (employee receptiveness).
Employee receptiveness is a character by the employee himself and therefore is difficult to be affected by the company. (Grace, King 2012) Thus, Grace and King (2012) suggest the following research model; Employee receptiveness has a strong effect on brand commitment (emotion) and brand citizenship behavior (action). Thus, a receptive employee may be a valuable asset for the company, due to willingness to perform brand citizenship behavior and feeling commitment to the brand. H3 and H4 were both validated. Furthermore, providing receptive employees with the possibility and knowledge to take part in communication may result in elevated job satisfaction. (Grace, King 2012) Applying BCB in the social media context may be a solution for marketers who in today's world are challenged to get through the clutter of noise. On the internet, the key for being listened to is to be authentic. Therefore; viral and buzz marketing are not perceived as authentic according to Morokane (2014), but merely a weak attempt to make an impression on customers. As earlier discussed, brand ambassadors are employees who commit positive word of mouth of their company. The act of WOM can take place both, offline (Morokane 2014) or online (Rokka, Karlsson, Tienari 2014) and it involves the employee's personal network and contacts. A potential customer looking to obtain more knowledge about a product or service relies on the word surging from personal networks. Thus, employees have power to influence and reassure people in their networks and therefore are important assets for the company (Morokane, 2014) (Dreher, 2014).
As of word of mouth, it has been studied for decades as the most authentic manner to communicate with customers (Harquail, 2009). Therefore, electronic word of mouth (eWOM on social media) is considered authentic and trustable. Furthermore, it is vital to point out that, there needs to be a reason, or rationale embedded in the employee's words and a motivation from the employee. A reason can be, educating close friends, being the professional, specialist or solely pride. The employee needs to be driven by internal knowledge and engagement to speak in favor of the company (Morokane, 2014).

Employees as Brand Ambassadors and Advocates
As discussed in the previous section, Brand Citizenship Behavior is a positive character of an employee and therefore also found relevant to be nurtured. As previously hinted, in the past decade social media has pushed its way into the world of organizations. Social media is accessible around the clock, on different devices and therefore employee social media use is inevitable (Dreher, 2014). Although social media has become part of people's everyday lives (2013) present three psychological states related to attitude and behavior; 1. Employee perceived brand knowledge, 2. Employee perceived brand importance, 3. Employee perceived brand role relevance. It was found that 2. Employee perceived brand importance and 3.
Employee perceived brand role relevance affected employee brand commitment, whilst 1.
Employee perceived brand knowledge had an impact on employee brand equity (EBE). The employee commitment to the brand (employee brand commitment) is related to EBE.
Employee Brand Equity directly affects brand endorsement, brand allegiance and brand consistent behavior, which are the factors generating brand ambassadors. The findings indicate that when employees believe that the brand is important and relevant, the employees are more committed to the brand. Whereas, knowledge of the brand itself leads directly to employee brand equity. Thus, when employees personally feel as part of the whole organization and have sufficient brand knowledge they are motivated to emit their perspective of the company to the clients. Without the identification and symbiosis with the organization and the brand, the employee does not feel necessary to conduct favorable additional company endorsement. The knowledge provided to employees comes from inside the company, indicating good brand communication and training of job position and relevance to the whole organization. The same results were found in the research made by Morokane (2014) where the employee willingness to commit positive WOM of an organization was studied.
However, Grace and King (2012) proved in their research that solely activeness on behalf of the company is not enough for employees to conduct brand citizenship behavior. Employees need a personal motivation to participate as brand ambassadors.
Rokka, Karlsson, Tienari (2014)   important since clients and other stakeholders of the company have a bond of trust to the employee and therefore also have more faith in his word. However, a negative incidence can lead employees spreading content online which, on the other hand damages the reputation and the brand. Therefore, it is considered beneficial to train staff about the message that is aimed to be delivered, online and offline.
A parallel can be drawn from Gelb DuBois and Rangarajan (2014) (2014), apart from Figure 4 below).
Perceived external prestige was proven to not have an impact on employee endorsement, and irrelevant to the discussion above. Therefore, it is left outside of the discussion. The findings of the research indicate that internal marketing does affect employee word of mouth (employee endorsement). Moreover, the findings made by Grace and King (2012) agree with Morokane's (2014) findings that employees who identify themselves with the organization are more likely to commit positive WOM. It was also found that internal engagement positively affects employee endorsement. Another relevant finding revealed is that stakeholder's view of the organization affects the employee's view of the organization, indicating that the overall image of an organization is imposed from both sides, employees and stakeholders.
Taking a strategic aspect in the physical incorporation of employee advocacy into an organization Dreher (2014)  Thus, 'research' indicates a company-wide social media audit from the perspective of the organization and the employee digital footprint. The major factor of launching an employee advocacy program is to provide 'access' on social media at work and enhance its usage during working hours. Employees can be asked to update profile information during working hours.
On the other hand, the management has to be committed to the new culture. Therefore, 'commitment' from the management is an extremely important aspect on launching employee advocacy. In order to have a team to follow up and take care of the steps in implementation as the back office of the project a 'social media team' is necessary, according to Dreher (2014).
The participating employees need to be communicated clearly the rules of the game to avoid process. An employee advocate is the same as a brand ambassador, however with a minor difference. When a company uses an employee advocacy program it is accessible for the whole organization in general. Thus, Terpening, Li and Littleton (2016) claim in their study that companies engaged earlier in brand ambassador programs which resulted in employees being brand ambassadors of the company. Thus, the difference lies in the access. Brand ambassadors are claimed to be selected and trained employees whereas an employee advocate can be anyone. Nevertheless, the structure of implementing a program and the variables needed before implementation are same for both. The separation between these two terms, employee advocate and brand ambassador was made by Altimeter in their report entitled "Social Media Employee Advocacy: Tapping into the power of an engaged workforce" (Terpening, Li, Littleton 2016). It is vital to point out that the difference between employee advocate and brand ambassador has not been made in previously discussed literature.
Therefore, the author continues to examine brand ambassadors and employee advocates as the same.
Terpening, Li and Littleton (2016)  On behalf of reputation, several aspects of improvement were identified. Emotional Reputation (recommend, trust, say something positive, buy products, invest, work) and Rational Reputation (products & services, innovation, workplace, governance, citizenship, leadership and performance) both scored average in the study. The gravest aspect needed to be improved, was awareness among generic public. It needs to be pointed out that generic public includes many important stakeholders for the company: old acquaintances, colleagues from current and former businesses, family members, friends, colleague friends, possible future employees, opinion leaders, students, un-recognized patients, caregivers, etc. These profiles were regarded as having an importance to the company, but had not been actively targeted in means of communication before. Therefore, evidence from the positioning and reputation study indicated that a change needed to be made. Furthermore, pointing out that Merck conducted the brand and image change in 2015-2016, a more proactive stance was to be taken regarding communication.
Therefore, an employee advocacy program was decided to be implemented to Merck in Portugal. The employee advocacy program allows employees to participate in the company's communication on social media through sharing company content from a platform called Smarp. The employee advocacy program was believed to be lucrative due to the possibility of speaking to stakeholders in an authentic manner and obtaining a wider audience than through traditional advertising, and for a lower cost. (Reputation study Merck S.A. Portugal, 2015)

SMARP
Smarp is a social media sharing tool which allows employees to share content on various social media channels. Each employee uses his own social media account for sharing, in other words, the content shared is personal and the choice to share is individual.
Smarp was created in 2011 by two Finnish students. Initially the focus was on professional branding and creating a role for employees on social media. The idea behind Smarp was the potential in employee's networks, which tend to be 10-15 times larger than corporate social media accounts. Nowadays the company is focusing on mutual benefit for the employing organization and the employee, following the mantra "humanizing brands". It was also discovered that the employee word of mouth is far more trustworthy than the communication conducted by a company. Shortly after launching the platform the company started to grow its business. Nowadays Smarp is a highly successful and acknowledged Employee Advocacy tool. The company is located in Finland, Sweden, Singapore, Dubai, England and New York.

Hypothesis Formulation and Conceptual Model for Merck Case
According to Grace and King (2012) an employee needs to be exposed to internal  (2013), where the authors found that if an employee does not identify himself and create a sense of unity with the organization, he will not conduct brand citizenship behavior.
The author chooses to follow Burmann and Zeplins (2005) example (previously discussed) by defining corporate brand as the organization itself to avoid confusion, since the organization holds several brands under a common name.
Applying the above mentioned findings to the case of Merck, the following hypothesis (2) was formulated: Hypotheses 2: Self-identification with the organization positively affects brand citizenship behavior, in the case of Merk, S.A.
Dreher (2014) Applying the above mentioned findings to the case of Merck, the following hypothesis (4) was formulated: Hypotheses 4: Internal culture positively affects brand citizenship behavior in the case of Merk, S.A. (2014) found out in their study that employee advocacy needs to be voluntary. Employees forced to contribute brand citizenship behavior will result in revolt and most likely adverse communication. Also Morokane (2014) proved that employees need to be receptive. It means to say, the employee must have an internal will to contribute to BCB. Employee receptiveness is an individual character which cannot be affected by the company. Also Grace and King (2012) suggest that employees need a certain level of engagement towards the work to contribute in BCB, which signifies an internal drive.

Rokka, Karlsson and Tienari
Applying the above mentioned findings to the case of Merck, the following hypothesis (5) was formulated: Hypotheses 5: Employee pro-activeness positively affects brand citizenship behavior in the case of Merk, S.A.
Based on the above mentioned hypothesizes the author presents a conceptual model. The

established hypotheses aim to define which independent variables have an impact on Brand
Citizenship Behavior in the case of Merck. In this case, the author wishes to define which factors affect employee willingness to conduct BCB in the form of WOM of the company.
Based on the following five hypotheses the conceptual model was drawn accordingly:

Merck Corporate Engagement Survey
The table below demonstrates questions which were considered important in the realization of the thesis. The questions were also used in the primary data collection executed by the author.
One of the questions in the questionnaire presented to employees spoke about As One for

Primary Data
The primary data of the thesis was collected through a quantitative survey made on Google Typeforms. The author seeks to measure the correlation between five established hypotheses and Brand Citizenship Behavior in the case of Merk. The questions are retrieved from previously tested scales and Merck Corporate Engagement Survey 2015 and used to measure the same purpose as the original authors did. The answers are measured on a 1-7 Likert Scale, which is the same measure used in the previously used and validated scales.

Questionnaire and Sample Design
The quantitative data was collected via a questionnaire. The questions are divided into seven sections.
1) Demographics to measure gender and age. Williams & Anderson (1991); Ganesan & Weiz (1996). The answers were measured on a 1-7 Likert scale, strongly disagree to strongly agree. Appendix 2   5 (2015) and Meyer and Allen (1997). The answers were measured on a 1-7 Likert scale, strongly disagree to strongly agree. Appendix 2 6) Internal Culture. Employees of Merck S.A. Portugal were asked to answer four questions about their perception on internal culture. The questions for internal culture were fully adapted from the Merck Engagement Survey (2015). The answers were measured on a 1-7 Likert scale, strongly disagree to strongly agree. Appendix 2 7) Employee Pro-Activeness. The employees of Merck S.A. Portugal were asked to answer five questions related to their own will to contribute voluntarily to brand endorsement. The questions used in the questionnaire were adapted from validated scales by Bettencourt (1997) and Diamantopoulos and Lonhdorf (2014) as well as, Merck Engagement Survey (2015). The questions used from Diamantopoulos and Lohndorf (2014) were modified by the author to be suitable for the questionnaire used in the in company project. The answers were measured on a 1-7 Likert scale, strongly disagree to strongly agree. Appendix 2 The population analyzed in the study comprises the employees of Merck S.A. Portugal. It needs to be pointed out that the employees chosen are working in the physical office as well as, the sales representatives working outside of the office. Other employees, such as Patient Support Nurses etc., were left out of the survey due to vague attachment to the brand/organization. Some of the questions used in the questionnaire were modified to be better suitable due to the specific pool of respondents, in order to obtain answers close to real life. On the other hand, the questions were gathered from authors who used validated scaled as well as, the Merck Engagement Survey (2015).
The questionnaire was sent out to all employees on the general emailing list MEPO_Portugal and additional employees not updated to the current emailing list, the number of recipients was 149. The amount of responses to the final survey was 103. The response rate was 69%, which was considered sufficient. The timing of the questionnaire was three weeks, between 20.2.2017 and 11.3.2017. The employees were sent two reminder emails to fulfill the questionnaire before the deadline to make sure the response rate was sufficient.

Demographics
The respondents of the survey came from different age groups, gender and they had spent varying time working for Merck in Portugal. In the questionnaire the employees were also asked to indicate what type of content they like to share most and in which social media channels.
The age distribution is presented in the table below. The biggest age group in the survey were employees between 40-50 years old, with 45 respondents corresponding to 43,7 % of the sample (103 respondents). The smallest group were the youngest employees, 20-30 years old with 7 respondents representing 6,8 % of the sample.

Figure 10. Age Distribution
The gender distribution between respondents of the survey is 46,6 percent female (48 respondents) and 53,4 percent male (55 respondents). As it can be noted, the slight majority of respondents are male.

Figure 11. Gender
Regarding years worked for Merck in Portugal, the majority of employees have been with the company for more than 10 years, with a percentage equal to 51,5%. The second biggest group, 37,9%, are employees who have worked for Merck between 0 and 5 years. The smallest group is people employed between 5 and 10 years representing 10,7%.

Figure 12. Years employed for Merck in Portugal
When asked what employees like to share on social media, the answer was the following. 52,4% of the employees like to share content which is related to the pharmaceutical industry.
The value is equal to the willingness to share corporate content, 52,4%. The most popular type of content to share on social media is content related to Merck S.A. Portugal, 63,1%.    To have a more specific view of social media channel preferences, the chart above was further divided into female and male. As it can be seen below, the most popular social media channel for women is Facebook with 28,4%, whereas males prefer Linkedin 33%. For both groups, male and female, Twitter was the least popular platform with 1,9% for both.

Hypothesis Testing
After the survey was completed, a descriptive analysis of the scales was performed to check for mean and standard deviation among each item for each scale.  I am well informed about the values represented by the brand I know which attributes of our brand differentiate us from our competitors.

MEAN (Male) MEAN (Female) MEAN
As it can be seen in figure 18 above, the lowest mean value is 5,7 for "I know how Merck S.A. Portugal is different from our competitors." The highest mean value is 6,3 for the question "I am well informed about the values represented by the brand".
For female respondents the lowest mean score is 5,3 for "I know how Merck S.A. Portugal is different from our competitors" and "I know the value Merck S.A. adds to the society." And the highest mean values are 6,0 for the questions "I know which attributes of our brand differentiate us from our competitors" and "I am well informed about the values represented by the brand".
For the category male respondents, the lowest mean values are 6,1 for "I know how Merck S.A. Portugal is different from our competitors" as well as "I know the value Merck S.A. adds to the society." The highest mean value is 6,5 for "I know which attributes of our brand differentiate us from our competitors."  The values that are important to Merck S.A. are also important to me.

SELF IDENTIFICATION
My values are similar to those of this organization.

Figure 20. Affective Commitment Scale Items by Gender
The highest mean value for hypothesis three, "Affective commitment positively affect employee willingness to commit brand citizenship behavior" is for the question "I am enthusiastic about my work" and "I feel positive about my work" with mean values 6,2. The lowest mean value, 5,8 is for the question "I feel like "part of the family" at our brand".
Among female respondents the highest mean value is divided between two questions, "I feel like "part of the family" at our brand" and "This organization inspires me to do my best work every day". The lowest mean value, 5,6 is for "I feel "emotionally attached" to our brand." For male respondents the highest mean value is 6,4 for the questions "This organization inspires me to do my best work every day", "I am enthusiastic about my work" and "I feel I am enthusiastic about my work.
I feel like "part of the family" at our brand.
I feel "emotionally attached" to our brand.

MEAN (Male) MEAN (Female) MEAN
like "part of the family" at our brand". The lowest mean score is for "I feel "emotionally attached" to our brand", mean 5,9.

Figure 21. Internal Culture Scale Items by Gender
The highest mean value for the following hypothesis is 5,8, "I believe the "As one for Patients" cultural initiative will drive positive change through this organization" and "I am familiar with the "As One For Patients" initiative's goals and deliverables". The lowest mean organizations." Among female respondents the highest mean value is 5,7 for "This organizations mission and goals provide meaningful direction to me" and "I believe the "As one for Patients" cultural initiative will drive positive change through this organization". The lowest mean value is 5,3 for "I am familiar with the "As One for Patients" initiative's goals and deliverables." For male respondents the highest mean value is, 6,2 for "This organizations mission and goals provide meaningful direction to me" and the lowest value 5,8 for "I believe the "As one for Patients" cultural initiative will drive positive change through this organization" and "I can clearly explain what makes working here different from other organizations".

Figure 22. Employee Pro-activeness Scale Items by Gender
The highest mean value is 6,2 "I would not hesitate to recommend this organization to a friend seeking employment." The lowest mean is 5,3 "On social media, I often feel the need to speak favorably about Merck S.A." Female respondents had the highest mean value for "In social situations, I often feel the need to speak favorably about Merck S.A." The lowest mean value is 5,7 for "I would not hesitate to recommend this organization to a friend seeking employment." For Male respondents the highest means score is 6,5 "Given the opportunity, I tell others great things about working here." The lowest mean value is 6,0 for two questions, "On social media, I often feel the need to speak favorably about Merck S.A." and "I would not hesitate to recommend this organization to a friend seeking employment.".   Other people would characterize me as an employee who represents our brand favorably in conversations.
Other people would characterize me as an employee who defends our brand if outsiders criticize it.
Other people would characterize me as an employee who tell others our brand is a good place to work.

Reliability Tests
In order to validate the Likert scales for their internal consistency, Cronbach's Alpha was performed. The results obtained proved that all scales have high internal consistency.
Therefore new compound variables were derived from the original items using the mean (Luo Y. 2003).
The complete figures can be found in the Appendix Cronbach´s Alpha.

Pearson's Correlation
The results obtained for Pearson's Correlation shows that there is a positive correlation among all the independent variables and the dependent variable, indicating that a growth in value for the independent variables also increases the value of the dependent variable.
The p-value value for all correlations is below p <0.05 indicating that there are significant correlations between the dependent and independent variables.  The tables for stepwise method can be found in Appendix 6 Regression Analysis using the Stepwise Model.
As it can be seen in TOL > 0,1 assumption holds, the independent variables are not correlated among themselves and VIF < 10 Assumption holds, there is no serious correlation among explanatory variables.
Durbin Watson is 2,250 which is considered good since the value is close to 2.
On the other hand, the Kolmogorov-Smirnov test on standardized residuals had a result of D0,000 < 0,05 indicating that the result is non-normal. The results of the K-S test can be found in Appendix 6. showcasing what a good and interesting place their employing organization is. The employee sharing content which he is familiar with and considers interesting, places the sharer in the shoes of an opinion leader making the employee look like an expert in his field. Therefore, any employee with the need to be visible and heard will share interesting content he is familiar with, which agrees with the findings made in the research.

ACTIONS TAKEN FOR MERCK PORTUGAL
The motivation to conclude the in company project was the need to establish communications in a cost efficient way by using digitalization. As the author already knew the concept of employee advocacy before starting at Merck Portugal, the communications idea was presented short after researching the actual situation regarding communications. During the writing of the thesis the author started implementing strategies (presented as hypothesis) which prepared the company for the launch of SMARP. The emailing list was also updated at the same time by the IT department, allowing internal messages to fluctuate in a more efficient manner.
Via the newly structured internal communication, brand knowledge was passed to employees as internal messages containing information, news, events and activities communicated on the intranet and email. Later on, employees could post their own photos on the intranet and some were also published in the rotating image gallery. Employee Pro-Activeness was intended to be boosted through tone of communication and company activities. The 'tone' of communication regarding any message sent was pride and joy. It means to say, that nothing was written with a 'flat' tone. Messages were written by using Merck specific font, superlatives, encouragement etcetera. Each achievement was communicated as "our" and  When the petition finally reached the needed signatures to be accepted, the internal messaging was written as "our success" and "we made it happen". Moreover, other activities were conducted on a way to increase employee pro-activeness, such activities were for  (2014) and Burmann and Zeplin (2005) who all claimed that employee advocacy is emotion based. A positive feeling towards the employing organization increases willingness to commit word of mouth. The events and activities were organized keeping in mind Selfidentification with the organization as well as, Affective brand commitment (Burmann, Zeplin 2005 Integration & Goal setting and measurement was conducted after the launch of SMARP. Employees were fully aware that the only channel where "risk free" content could be found, accepted by compliance was on SMARP. By only using SMARP for external Merck Portugal communication all company messages (corporate and local) could be unified in one voice, hence to employees sharing the same content in the same time frame. After the launch and successful onboarding of employees on the tool, goal setting and measurement was started.
KPI´s were set for the entire organization for 2017 with the help of a support consultant from SMARP and other objectives were set case wise by each therapeutic area wishing to use SMARP as part of a campaign.

SMARP in usage
The following text aims to show results obtained from Merck Portugal from an institutional perspective and also provide an example of a therapeutic area which chose to use SMARP in a campaign.
As earlier mentioned KPI´s were set for the entire use of SMARP. In the figure below the KPI´s set in April 2017 are demonstrated. The objectives before KPI´s were to increase employee enrollment via connecting personal social media channels to SMARP, to create content on a timely basis (especially content for Portugal, which had a higher "share -rate" on social media) as well as remind employees of new content to share via push-notifications on iPHONE, iPAD and via email. After the first KPI setting for institutional usage, three tactics to increase usage were set; First, to boost employee willingness to share on SMARP each month a list of "Top Merck Influencers" was published by email and as an internal message on SMARP. SMARP had also a weekly leader board on the platform to show the influencer ranking. It was recognized that the leader board would work as an ego-booster for the most sharing employees and as a motivator to share for the employees who did not appear on the table.   To have up to date information the author used also company specific published works and sources from the internet. However, these sources of information were only used when needed to strengthen the claims. The in company project had a specific target group, the employees, to study.
Therefore, questions from the Merck Employee Engagement Survey 2015 were modified to better correspond to Merck in Portugal.

Future Research and Practical Suggestions
As of future research the author suggests to dig deeper into employee pro-activeness and look into the psychological aspects of internal drivers and motivation. As of brand knowledge, it is considered important to have a specific view of what is included in brand knowledge and what topics regarding the brand and company are considered interesting.
It is also thought of interest to research the concept employee advocacy and the employee advocacy tool to note how employee engagement to the platform can be affected by different company activities offline, which are later on shared online to discover the most "interesting" content shared. It needs to be pointed out that the content employees share with pride (example, opinion article by the Managing Director) was proved to gain much attention on the social media channels of the employees.
As mentioned several times in the thesis, employee advocacy is a new term and it has not been duly researched yet. However, the concept is gaining a lot of attention at this moment.
As a reference already mentioned in the thesis, Linkedin Elevate, is a rising star in employee advocacy which gives good analytics for recruiters etcetera. However, the platform is only limited for Linkedin.
From a practical perspective, implementing a tool such as SMARP is an ongoing project.
After implementation the owner(s) of the tool needs to further develop the usage strategy, content mapping and engagement. Employees need to constantly be informed about the existence of the platform, if not it will be forgotten. SMARP can be used for both internal and external communication. In case employees are wanted to be reached with internal (blocked social media sharing) or external content, push notification function provided by the tool is very helpful to reach also sales force, for instance. The push notification is useful also 2) The mean of the residuals component is zero The assumption holds, residual = 0.
3)There is no correlation among the residual terms The value of Durbin-Watson is 2,255. The value is close to 2, therefore the residuals are assumed to be independent.

4)The variance of the random term is constant
The scatterplot shows that the assuption of homoscedasticity is not valid, the scatterplot shows heteroscedasticity, meaning an uneaven spread around zero. As it can be seen, the dots are spread mostly between 0-1 on the right side of the scatterplot. The assumption is not validated.