Become a Brand Ambassador, measured with
Attachment Styles and Employee Engagement
2
Become a Brand Ambassador, measured with
Attachment Styles and Employee Engagement
16 January 2017 Renkema, Martin R.W. S2774445 Oldenoert 108 9351 KS Leek mrwrenkema@gmail.com m.r.w.renkema@student.rug.nl 06-43475031
Master Thesis Marketing Intelligence University of Groningen
Faculty of Economics & Business Department of Marketing
PO Box 800, 9700 AV Groningen (NL)
First supervisor: Dr. ir M.J. Gijsenberg (m.j.gijsenberg@rug.nl) Second supervisor: Dr. J.T. Bouma (j.t.bouma@rug.nl)
External supervisors: Dr. L.H. Teunter (l.teunter@metrixlab.com),
3 Management summary
This paper explains why some employees recommend their firm to someone else and other employees do not (positive/negative word-of-mouth). Our findings are based on attachment styles, employee engagement and employee empowerment. We conducted are research with the online panel of MetrixLab, a Macromill group company.
Worldwide 70% of the employees are disengaged at work (WeForum, 2016), while engaged employees can lead to a boost of twenty percent in productivity and profitability (Gallup, 2016). Kumar and Pansari (2016) found that employee engagement has a positive relation to the firm performance. But how to increase the level of employee engagement, and the positive tone-of-voice of an employee? Therefore we did a study in the working environment of The Netherlands. Our study is conducted in four sectors; B2B & Service, B2B & Manufacturing, B2C & Service and B2C & Manufacturing. Overall we see that our research is applicable in each sector.
We developed new specific attachment styles. These attachment styles are job-specific and based on the business-specific attachment styles of Mende, Bolton & Bitner (2013). Our new job-specific attachment styles are not an extension of the general attachment styles (Bowlby, 1973). Our job-specific attachment styles anxiety and avoidance had a significant negative effect on employee engagement. This suggests that employees who have a low level of anxiety and/or avoidance will also be less engaged in their work. The level of anxiety and avoidance can be influenced by a trust person at work, trainings options and good supervisor (Parker and Smith, 2011).
4 To know how to change/remain your employee as a brand ambassador of the employer, several activities can be done. To deal with the level of anxiety and avoidance a supervisor or trust person can help. To change the level of employee engagement very different activities can be done, like team-activities, trainings, rewards/benefits, etc. Overall our model is useful for (Human resource) managers and owners of firms because our model has the methodology to measure the level of positive tone-of-voice and we formulated the implications how to increase this level.
5 Preface
I started my life as a student at the University of Applied Sciences, and finished successful my Bachelor Marketing. After one year of working, I made the best decision I ever made and went back to the college bench and I started with my pre-master Marketing. In that year I made the choice to do the MSc Marketing Intellingence track. In this master year I met some great inspiring co-students and professors. And this report is the final piece of my master and student life.
For accomplishing this final version of my thesis I would like to thank some lovely people who helped me realizing my thesis. First, I would like to thank two people of the University of Groningen. I would like to thank my great supervisor from the University, dr. ig. Maarten Gijsenberg. Dr. Ig. Maarten Gijsenberg gave direct, active and useful feedback by mail and Skype calls. The second person is Dr. Jelle Bouma for the opportunity and arranging my graduating internship at MetrixLab, a Macromill group company, and for the feedback moments in the beginning of the process.
At MetrixLab I would like to thank some great specialists in their work. First, I have to thank dr. Linda Teunter, a women with great knowledge about analytics and a lovely person which gave us as colleagues always a smile on our face. I would also like to thank Gustave van Brederode and Edwin van der Harst for the critical feedback and useful insights about the business and survey knowledge. Also I would like to thank the Customer Value department for being of a few projects, and the daily lunch with the “(Teunter) tosti’s”.
Finally, I would like to thank, my parents and brother for their support during my whole period as a student. And especially I would like to thank my girlfriend Charlotte, for her unlimited positivity and support during the whole project.
6
Table of contents Page
1. Introduction 8
2. Theoretical framework 11
2.1 Employee engagement 12
2.2 Brand ambassadorship 13
2.3 Employee engagement and brand ambassadorship 15
2.4 Attachment styles 16
2.4.1 General attachment styles 16
2.4.2 Employee specific attachment styles 17
2.5 Attachment styles and employee engagement 19
2.6 Attachment styles and brand ambassadorship 20
2.7 Influence of Employee empowerment 21
2.8 Sector dependent 22 3. Methodology 24 3.1 Procedure 24 3.2 Subjects 25 3.3 Measurements 25 3.4 Plan of analysis 27 4. Results 31 4.1 Data 31 4.1.1 Data cleaning 31 4.1.2Reliability 31 4.1.3 Descriptives 34
4.2 Attachment styles on brand ambassadorship 38
4.2.1 General attachment styles on brand ambassadorship 38 4.2.2 Job-specific attachment styles on brand ambassadorship 39 4.3 Attachment styles on brand ambassadorship, mediated by employee 41
engagement
4.4 Moderation effect of employee empowerment 43
4.5 Validation 44
5. Discussion and conclusion 46
7 7. Limitations and suggestions for further research 53
References Appendices
-Appendix A Questionnaire
8 1. Introduction
Every year many firms start, some successfully and some unsuccesfully. Every owner wants to be successful and wants to increase their profit. An owner can do this by focusing on selling more products or services or be more efficient internal. But there are more options to be successful as employer. Many studies already proved that customer engagement has a positive affect for the company (Thakur, 2016; Apeness Solem, 2016). Although, the internal side with employee engagement is getting more important in scientific research. AbuKhalifed and Som (2013) & Al Merhzi and Singh (2016) already suggested that employee engagement is going to be an important topic in research in the common years. Kumar and Pansari (2016) confirmed this and found that not only customer engagement, but also employee engagement has a positive influence on firm performance. Employee engagement can lead to a boost of twenty percent in productivity and profitability (Gallup, 2016). The strange thing is that Gallup (2016) found that only thirteen percent of the employees globally is engaged in their job. So the question which arises is, how to increase employee engagement? This is important, because employees who are highly engaged think that they can positively affect customer service (Crim and Seijts, 2006).
Results of studies (Bakker et al., 2012; Cristina de Melloe Souza Wildermuth. & Pauken., 2008; De Clercq et al., 2010; Karatepe & Olugbade, 2009; in AbuKhalifeh and Som, 2013) show that employees who are highly engaged will also be more positive about their employer and stay longer at the firm. For human resource managers it is important to increase the level of employee engagement to increase the level of performance, increase the level of loyalty and save recruitment costs. An additional outcome of employee engagement is that engaged employees create more passion with the employer and their work activities (Brown et al., 2005; Tuskey et al., 2013 in Sicilia, Delgado-Ballester & Palazon, 2016). A positive employee will also have a more positive tone of voice in a conversation during and outside work activities. From a branding perspective it can be said that there will be a positive atmosphere created around the firms’ name. But how can an employee give branding values to the company?
9 customer you can create a certain attachment with the brand. To check what kind of connection there is we investigate the influence of different attachment styles. The general attachment theory of Bowbly (1973) provides a framework for understanding why people form close emotional bonds with each other. Mende & Bolton (2011) and Mende, Bolton & Bitner (2013) developed an extension of the general attachment styles, the business-specific attachment theory, which is about the bond between a customer and a firm. We will adapt the theory of Mende & Bolton (2013) in a new extension of the general attachment styles. We will investigate the relation between an employee and the employer with the job-specific attachment style.
In this study it is important to include relationships that influence the employee-firm relationship. The main relationships in this research are the links between employee engagement and brand ambassadorship and the relationship between attachment styles and employee engagement. These relationships form an important step in the process of what explains who is an ambassador of the brand.
The aim of this study is to provide an extension of current research knowledge about the attachment styles that drives employee engagement and the direct link with brand ambassadorship. Business-specific attachment styles are already provided (Mende and Bolton, 2011; Mende, Bolton and Bitner, 2013) but in our study we develop job-specific attachment styles, an extension of current knowledge. Additionally this study provides insight about the impact of the attachment styles and employee engagement on ambassadorship across different industries in The Netherlands. Finally, we analyze if the relation on brand ambassadorship is affected by employee empowerment (Kumar and Pansari, 2016).
10 attachment styles can have a direct or indirect effect on brand ambassadorship. For human resource it is useful to know if the employee is anxious or avoidant in general and in their relation with the firm. A firm can conduct our research to find out what the level of anxiety and avoidance is. A change in the level of anxiety and avoidance of an employee influences their level of engagement which has an effect on the level of being a promoter for your employer. To come to these new insights, we give answers to the next research questions:
1. What is the impact of employee engagement on brand ambassadorship? 2. What is the impact of attachment styles on employee engagement?
3. To what extent is there a direct effect from attachment styles on brand ambassadorship? 4. Has employee empowerment an influence on the outcome employee engagement?
5. To what extent is there a difference in influence from attachment styles and employee engagement on brand ambassadorship across different industries?
11 2. Theoretical framework
In the theoretical framework we will explain the five research questions (introduction, p.6). We will do this by explaining the different variables and relationships between these variables. Figure 1 is the conceptual model and gives a visual overview of our study. In our research employee engagement has a central role, and we try to find relations with this variable. We know from previous research some affects on employee engagement (Fernandez and Moldogaziev, 2013; Philo, 2016). And some research is done about the effect of employee engagement. An important example found by Kumar and Pansari (2016) is that employee engagement positively influences the firm performance.
In this study we use a different outcome of employee engagement, namely the brand ambassadorship. Brand ambassadorship can be explained as an employee who is very positive and recommends his/her own employer. To see the relation, a positive word-of-mouth is an outcome of a higher engagement level (Brown et al., 2005; Tuskej et al., 2013 in Sicilia, Delgado-Ballester & Palazon, 2016). We also investigate the relation between the different attachment styles directly on brand ambassadorship. After that we include and investigate the influence of this direct relationship by including the mediator employee engagement. Additionally, Kumar and Pansari (2016) found that employee empowerment has an effect on the outcome employee engagement. So, we include the moderator employee empowerment in the relation between the mediator employee engagement and brand ambassadorship. Finally, employees work in different sectors, with different specifications and humans, so to check if there are different outcomes across industries we split our research into four sectors.
12 We will explain the different variables of the conceptual model (figure 1) in this chapter. In section 2.1 we will discuss employee engagement, followed by brand ambassadorship in section 2.2 and the relation between these both variables in section 2.3. Followed by the attachment styles (section 2.4) and the influence of attachment styles on employee engagement (section 2.5). After section 2.5 the direct relation of attachment styles on brand ambassadorship will be explained (section 2.6). In the last sections the influence of the employee empowerment (2.7) and the sector dependency (2.8) will be explained.
2.1 Employee engagement
Employee engagement is a key business driver for organizational success (Lockwood, 2007). But what is exactly employee engagement? Some researchers called it job engagement (Rich et al., 2010 in Saks and Gruman, 2014) or work engagement (Schaufeli & Salanova, 2011 in Saks and Gruman, 2014). Saks and Gruman (2014) conclude that there continues to be lack of consensus on the meaning of employee engagement as well as concerns about the validity of the most popular measurement of employee engagement. Bhuvanaiah and Raya (2014) did a meta-analysis about definitions for employee engagement. We describe employee engagement in our research as ‘’the extent to which employees are motivated to contribute to organizational success and are willing to apply discretionary effort to accomplishing tasks important to the achievement of organizational goals” (Kenexa work trend report, 2012 in Bhuvanaiah and Raya, 2014, p.65). This definition fits in our study because it is about the employee and his/her level of engagement which contributes to organizational perspective (Bhuvanaiah and Raya, 2014) and goals. An increase in the level of employee engagement positively influences the firm performance (Kumar and Pansari, 2016).
13 the firms’ culture (Kumar and Pansari, 2015) to identify them with the firm. Team-activities and creating new ideas can improve the identification. Therefore human research can for example, exploit mentorship programs. Emotional bonded is also an important part to be committed to the company. If employees are not committed, some will look for new job opportunities outside the firm. To increase commitment companies can change their work environment, give performance incentives, create reward and/or benefit structures (Kumar and Pansari, 2015). This should be in line with the policy of the firm. Fourth, is the level of loyalty. For example, nonloyal employees can give away firm secrets, documents or are not friendly to colleagues. To increase loyalty, development programs can be set up, so the employee can update/increase knowledge and skills (Kumar and Pansari, 2015). Another option is to do something for the employees’ family, for example; providing a car, reimburse fuel, giving discount on insurances. Finally, the level of performance indicates how productive an employee is, for example a high performance level has a positive impact on behavior and the firm (Kumar and Pansari, 2015). If the performance level is high, it confirms the policy or human research otherwise and changes have to be evaluated and made (Kumar and Pansari, 2015). Human research can also develop orientation programs, incentives and/or feedback session to increase individual performance.
The five elements of employee engagement together form the employee engagement level. The higher employees rate on the different elements, the higher the level of employee engagement is. Firms would like that the employees rate high on each elements because more engaged employees will have higher propensity to engage in positive word-of-mouth (Brown et al., 2005; Tuskej et al., 2013 in Sicilia, Delgado-Ballester & Palazon, 2016). Positive word-of-mouth can be explained by being an ambassador for your employer, which we will discuss in the next section.
2.2 Brand Ambassadorship
14 possible more positive because the conversation partner received positive information about the firm, from the employee. This is what we call in literature, positive word-of-mouth. Word-of-mouth can lead to a snowball effect, which means that more people will be talking positive about the firm. Word-of-mouth is a cheap and powerful online/offline marketing tool (Chen and Berger, 2016; Ya, Vadakkepatt & Joshi, 2015). This positive tone of voice comes from engaged people because those people have a higher propensity to engage in positive word-of-mouth (Brown et al., 2005; Tuskej et al., 2013 in Sicilia, Delgado-Ballester & Palazon, 2016) and disregard negative brand information (Swaminathan et al., 2007 in Sicilia, Delgador-Ballester & Palazon, 2016).
Brand ambassadors are those employees who carry the brand’s promise and message into the world (McEwen and Robison, 2007). Unfortunately, many firms fail in recognizing that employees need to be supported to become and remain brand ambassador (McEwen and Robison, 2007). Therefore human resource should focus on the connection from employee ‘’to the brand individually, it’s easier to build an unified and inspired team a through it a profitable corporate brand (Ikalafen, 2008, p.67)’’. If an employee is spreading positive information about the firm, you could say someone is engaged (Brown et al., 2005; Tuskej et al., 2013 in Sicilia, Delgado-Ballester & Palazon, 2016).
Traditional recommendation is measured with the Net Promotor Score (NPS) developed by Reichheld (2003). The developed question and scale of Reichheld (2003) is normally used for a customer-firm relation. In this relation, a customer can recommend a brand to someone else. Our study has a different setting, namely employee-firm relation, where an employee can recommend his/her employer to someone else. But recommendation in a employee-firm relation can be interpreted differently than in the customer-employee-firm relation. Therefore we choose to make two recommendation questions. The first recommendation question is about as an employee you can recommend people to join the firm and become a colleague. The second recommendation is as an employee you can recommend the firm as a supplier to customers, friends and/or family.
15 measurement is also on a positive scale. The NPS is based on an 11-points scale (0-10). The respondents can be placed in one of the three categories ‘:detractors’ (score 0-6), ‘passives’ (score 7-8) or ‘promoters’ (score 9-10). The NPS is calculated by taking the percentage promoters minus the percentage detractors.
2.3 Employee engagement and Brand Ambassadorship
In section 2.1 and 2.2 we discussed the elements employee engagement and brand ambassadorship separately but we try to find a significant relation between the two variables. Employees who are highly engaged score overall higher on the elements satisfied, committed, loyalty, identification with the firm and have greater performance (Kumar and Pansari, 2015). When the employee engagement is high, this will have an effect on the emotional and personal connection with the firm. ‘’Passionate ambassadors are the result of a sustained effort. Brand's best interest to be backed by engaged employees who are crystal clear on what's expected of them (McEwen and Robison, 2007, p.7)’’. Tsang, Lee & Li (2011) found that the elements employee commitment, -satisfaction and –identification have a positive relation with the brand image and –awareness. Loyal employees stay longer at the firm, which can mean that they are positive about the firm otherwise they probable are looking for another job. Higher performance should also lead to higher engagement because it asks for more effort in the work activities. Every firm should see the importance of engagement, because ‘’every employee should be viewed as a brand ambassador”, says Sam Stern in B-to-B Marketer (Schwartz, 2016, p.12). If the employee is not a brand ambassador, this can be a result of a lower level of engagement. Which means that employees rate overall lower, so on average they are less satisfied, less committed, less loyal, less identified with the firm and/or have lower performance. We expect this will result in being less positive about the firm, and that they are less an ambassador for the brand than highly engaged employees. The following hypothesis can be made:
16 2.4 Attachment styles
Attachment styles are normally used in psychological studies to investigate the emotional relation between people (Bowbly, 1973). Two of those attachment styles are anxiety or avoidance. We will use the general attachment styles anxiety and avoidance and we will use the specific attachment styles (Mende and Bolton, 2011) which will be adapted into an employee-firm relation. The general and specific attachment styles can have an influence on the level of employee engagement and ambassadorship. First, the general attachment styles will be explained and secondly the employee-specific attachment styles.
2.4.1 General attachment styles
Bowlby (1973) developed a framework for the attachment theory to understand why people have close emotional bonds with others. Based on Bowlby’s (1973) study, Mikulincer and Shaver (2007) described an attachment style as the systematic pattern of relational expectations, needs, emotions and social behaviors that results from the internalization of a particular history of attachment experiences. The dimensional models of attachment generally have two primary dimensions: attachment-related anxiety and attachment-related avoidance. Anxiety is the extent to which a person worries that relationship partners might not be available in times of need, has an excessive need for approval and fears rejection and abandonment (Mende and Bolton, 2013). Attachment style avoidance is the extent to which a person has excessive need for self-reliance, fears depending on others, distrusts relationship partners’ goodwill and strives for emotional and cognitive distance from partners (Mende and Bolton, 2013).
17 2.4.2 Employee-specific attachment styles
The employee-specific attachment styles are an extension of the general attachment styles. Mende and Bolton (2013) developed a scale in the customer-firm specific setting. In their research the business-specific attachment styles confirmed that “specific attachment styles predict customers’ preferences for closeness better than established marketing variables do (p.125).’’ These attachment relations presumably also can be proven in an employee-employer relation. Probably there will be similarities but also differences compared with the customer-firm relationship. To find the similarities and differences between the firm- and employee-specific relations we discuss the attachment styles anxiety and avoidance separately.
First we will discuss the similarities of the specific attachment styles anxiety. As a customer and employee you can both create anxiety. The level of anxiety can increase when someone does not receive the attention which the person needs. As a customer this feeling of anxiety can be created if the firm does not give you as much attention as you need, for example if someone is a loyal customer and never receives a thank you. It creates the feeling that the customer is not important enough. This feeling of anxiety can also develop in an employee-firm relation. As an employee, someone wants the intention to be important and get the feeling appreciated. For example, an employee received a Christmas present from the employer but a colleague did not receive one.
Then the differences between the two relations for the attachment style anxious. If a customer is anxious about a brand, the customer will not be loyal and will search for other products or services. The same applies for if an employee who is anxious about the employer. It can result in being less loyal to the firm. This can have an effect on his/her performance and eventually the employer will look for another job. These outcomes can have very different financial effects. If a firm loses one customer it can be substitute on short term, but losing an employer can have a negative contribution on human resource and work productivity of other colleagues besides the negative influence of the unemployed person.
18 distance because the firm is pushing (commercial) too much. This can also count for an employee, sometimes the employer asks too much from the employee. For the employee this can be a moment to create some distance, for example the employee does not want to have a burnout. Another reason is that the firm is changing strategy which is different than what the employee wants to achieve with the company. The level of avoidance is depending on the fear or distrust somebody creates with the firm. Less avoidant individuals are resilient and experience better outcomes regardless of the level of partner support they receive(Girme, Overall, Simpson & Fletcher, 2015).
The context of avoidance is also different in both relations. When a customer has a high level of avoidance, the preference for a brand is less and the propensity that the customer goes to a substitute brand is high. For a customer it is much easier to be less loyal to a brand. For an employee it is different if the avoidance is high. As employee you distrust your colleagues and have less attention for the firm activities. This can have an influence on the level of engagement. Less engaged employees are less productive and create less value for the firm.
19 2.5 Attachment styles and employee engagement
In this section we discuss the relationship between attachment styles and employee engagement. We split the attachment styles because as the study of Mende, Bolton and Bitner (2013) shows, the specific attachment styles are a better indicator than the general attachment styles. First, Brennan, Clark & Shaver (1998) found that a low level of general attachment styles anxiety and avoidance leads to a stronger relationship with others. Which in our context means that both, low level of anxiety and avoidance lead to a stronger relation between employee and employer. In the same study Brennan, Clark & Shaver (1998) found that a high level of general attachment styles anxiety and avoidance results in a negative impact on satisfaction, which is in our study an element of employee engagement. Kivlighan, Lo Coco & Gullo (2012) confirmed the outcomes of Brennan, Clark & Shaver (1998) in an aggregated perception of attachment anxiety and avoidance. They found that anxiety and avoidance were positively related to group conflict and that aggregated perception of attachment anxiety and avoidance were negatively related to group engagement.
H2a General attachment style anxiety has a negative influence on employee engagement. H2b General attachment style avoidance has a negative influence on employee engagement.
Secondly, the employee specific attachment style anxiety and avoidance. Parker and Smith (2011) did an interesting study about anxiety on the business floor. Too little anxiety creates complacency, boredom and stagnation of learning and produces insufficient energy for optimum learning (Parker & Smith, 2011). On the other hand too much anxiety fosters resistance, aggression and withdrawal which produces chaotic energy that inhibits learning (Parker & Smith, 2011). For every individual it is different, but there should be just enough anxiety which fosters creativity, heightens engagement and drives growth. Finally Sadatsafavi, Walewski & Shepley (2015) found that a lower level of job-related anxiety is related to a higher level of job satisfaction, which is an element of employee engagement.
20 predicts less engagement coping suggestions (orienting toward stress) and heightened disengagement coping suggestions (avoiding or denying stress) now and over time. In line with the study of Abaied and Rudolph (2010) we suggest that a higher level of avoidance results in a lower level of engagement.
H3a Job related anxiety has a negative influence on employee engagement. H3b Job related avoidance has a negative influence on employee engagement.
We expect, based on the study of Mende, Bolton and Bitner (2013) that the employee-specific attachment styles are a better indicator than the general attachment styles. That is why we formulate the following hypothesis:
H4 Job specific attachment styles are a stronger indicator of employee engagement than general attachment styles.
2.6 Attachment styles and Brand Ambassadorship
When accounting the indirect effect we still think there could be a direct effect from the attachment styles on brand ambassadorship. Increase in the level of anxiety can result in that customer overreact to critical incidents (Mende, Bolton & Bitner, 2013), this can lead to negative word-of-mouth and thus a negative effect of ambassadorship. Anxious employees are often involved in unpleasant relationships (Simpson, 1990, p.972), which can result in a lower NPS score. A lower NPS score that an employee less propagates his/her employers’ brand . As mentioned in section 2.5 the level of anxiety should be just enough. Of course this differs per person, but in this study we try to look for a figure per sector. Just enough anxiety drives growth and generates productive energy that promotes (Parker & Smith, 2011) the brand/employer.
21 attachment styles it is expected that an increase in the level of the attachment styles has a negative influence on brand ambassadorship. Therefore the following hypotheses are formulated:
H5a When accounting for the indirect effect of general attachment style anxiety through employee engagement, attachment styles have a negative influence on brand ambassadorship.
H5b When accounting for the indirect effect of general attachment style avoidance through employee engagement, attachment styles have a negative influence on brand ambassadorship.
H6a When accounting for the indirect effect of job-specific attachment style anxiety through employee engagement, attachment styles have a negative influence on brand ambassadorship.
H6b When accounting for the indirect effect of job-specific attachment style avoidance through employee engagement, attachment styles have a negative influence on brand ambassadorship.
We also checked based on the study of Mende, Bolton and Bitner (2013), if the specific attachment styles are a better indicator than general attachment styles. We did this because our theory about employee-specific attachment is new and therefore we compared this to the proven general attachment styles.
H7 Employee-specific attachment styles are a stronger indicator of employee engagement than general attachment styles.
2.7 Influences of employee empowerment
22 Empowerment enhances the relationship between employees and customers because an empowered employee is in a better position to engage and deliver firm value (Kumar and Pansari, 2016). In this study employee empowerment enhances the relationship between employees and the firm. Employee empowered can affect the outcome of employee engagement (Kumar and Pansari, 2016). And as mentioned a higher employee engagement positively influences the ambassadorship. Jain and Jain (2014) also conclude that more research has to be done about the relationship between empowerment and other organizational or individual variables. That is why in this study the moderation effect of employee empowerment on the relation employee engagement on ambassadorship will be analyzed. Kumar and Pansari (2016) applied and proved the measurement tool for employee empowerment of Menon (2001). Therefore this method will be applied in this study.
H8 Employee empowerment has a significant positive effect on the outcome of employee engagement on brand ambassadorship.
2.8 Sector dependency
Employee engagement scores different across countries (Farndale & Murrer, 2015) and across industries (Kumar and Pansari, 2016). Kumar and Pansari (2016) already found that business-to-business companies tended to score higher than business-to-consumer companies on the influence from employee engagement on firm performance. Secondly, Kumar and Pansari (2016) found that service-based companies scored higher than manufacturers from the influence of employee engagement on the firm performances.
23
TABLE 1 MATRIX SECTOR DEPENDENT
B2B B2C
Service-based ++* +*
Manufacturing-based +/-*** -**
*Strong(er) influence, **Weaker influence, ***Not stronger, not weaker
We have chosen these four different sectors because they explain most of the business markets in The Netherlands (except charities). In a B2B environment, each employee is responsible for taking care of his own customers. The number of customers is limited in B2B, and the service provider’s focus is ensuring a good relationship with every customer because each customer has a higher transaction value than in a B2C firm. Additional, in a B2B environment, each employee is responsible for taking care of a set of customers (Kumar and Pansari, 2016, p.503). In a B2C firm, the employee cannot be the reason for the relationship between the firm and the customer; the product or service offered by the firm creates this relationship. The pool of customers is large, and the employee-customer interaction is low, compared to a B2B company.
Engaged employees focus on ensuring that their customers have positive interactions (especially in service firms) with the firm by addressing customer concerns in the best possible manner (Kumar and Pansari, 2016, p.503). Although employees are important in ensuring a smooth workflow in the manufacturing sector, it is the machines and the technology that are the competitive advantage of manufacturing firms (Kumar and Pansari, 2016, p.503)
H9 There is a significant difference from the influence from employee engagement on brand ambassadorship across industries.
H9a The effect of employee engagement on brand ambassadorship is stronger in the B2B sector than in the B2C sector.
24 3. Methodology
In this chapter we start with the operationalization part of our research. First, we will explain the procedure, followed by the subjects. Thirdly, the different measurement tools as regression analysis, mediator and moderator mediation analysis will be explained. Finally, we look further with the plan of analysis.
3.1 Procedure
For our study we developed a questionnaire by using questions and statements from proven research. The questionnaire started after the welcome page with questions about demographic background and in which sector the respondents were working. On this page we checked if respondents were working otherwise the respondents were screened out the questionnaire. On the next page the general attachment styles were asked. After the general attachments questions we started with the employee-firm relation question. This place in the survey is chosen because we wanted to prevent bias with the general attachment style in a employee-firm relation. This started with the recommendation questions (NPS). The NPS questions explain how much the employee recommended his/her employer to a friend, family or customer and as a supplier to someone else. Than the job specific statements were introduced and asked. On the next two pages the employee engagement questions were asked. On the last page the employee empowerment questions were asked. In every section the statements were randomized to prevent bias (Mitchell, 1974).
25 (The Netherlands) because the level of employee engagement can differ over countries (Farndale & Murrer, 2015).
3.2 Subjects
We tried to get a representative sample of the Dutch (working) population. The total research sample consists out of 841 working people. The respondents were in the age of 16-72 years (M=47.95; SD=11.128). 54,6% (n=459) of the sample was male and 45.4% (n=382) was female. When looking at the sector dependency, we found that 24.5% (n=206) of the respondents worked in the B2B & Service industry and 17.2% (n=145) in the B2B & Manufacturing industry. The most employees worked in the B2C & Service industry (47.8%; n=402) and the smallest group worked in the B2C & Manufacturing industry (10.5%; n=88). The distribution is not perfect, but the size of every group was big enough to give validated results.
3.3 Measurements
First, all the variables have to be measured to test the relationships between the variables. In appendix A1 shows the list of questions which were asked, in appendix A2 is the Dutch translated version.
26 Ambassadorship is measured with the Net Promotor Score (NPS). The NPS is a Customer Feedback Metrics (CFM) which explains how positive or negative the employer is about the firm. Reichheld (2003) proposed and proved the NPS in his publication ‘’The One Number You Need to Grow’’. Van Doorn, Leeflang & Rijs (2013) did a study about an alternative research method than NPS, namely: Alternative NPS (Percentage of respondents rating 8, 9 or 10 minus the percentage of respondents rating 0-5) or other loyalty intentions (adapted from Zeithaml, Berry and Parasuraman, 1996) and found that NPS is neither superior nor inferior to other metrics. The metric NPS is scientific proven and managerial most used, that is why we also apply this method to measure how much an employee recommends his/her employer to someone else. Ambassadorship is measured in two separate ways, with an colleague and supplier recommendation question. Normally an NPS question is asked in an customer-firm relation. In our study the relation is different, namely employee-firm. Because we did not find this relation on an NPS question in previous studies, we divided the NPS into two different recommendation questions. The first one is an NPS question of how would you recommend your employer to family or friends to become a colleague. The second NPS question is, how would you recommend your employer as an supplier to someone else.
The general attachment styles were developed by Bowlby (1973). Brennan, Clark and Shaver (1998) decreased the amount of statements and provided a smaller measurement instrument. Fraley, Waller and Brennan (2000) made a revised version of the Experiences in Close Relationships (ECR), attachment styles. Totally the ECR and ECR-R was measured with 36-item statements of attachment styles. Verbeke, Bagozzi and Van den Berg (2013) made a smaller measurement consisting eleven items, six items to measure anxiety and five items to measure avoidance. To keep the questionnaire as parsimonious as possible, we used the short and proven version of Verbeke, Bagozzi and Van den Berg (2013).
27 theory of Mende, Bolton and Bitner (2013), from a customer-firm relation to an employee-firm relation. We had to do this because the relation employee-employee-firm or customer-employee-firm is different. It is essential in our research that we do it correctly, because these questions are not available in previous scientific research. We changed as little as possible from the proven statements of Mende and Bolton (2011). The statements were still divided in two groups, anxiety and avoidance. We will show you one example, Mende and Bolton (2011) developed the statement ‘’I worry that (name of firm) doesn’t really like me as a customer’’ which can be answered on a five-point Likert scale. In our study we adapted this statement in an job-specific statement: ‘’I worry that my employer doesn’t really like me as an employee’’. We applied the statements based on translation and pre-testing, by people with different backgrounds.
Then, employee empowerment is measured by the questions of Menon (2001). Kumar and Pansari (2016) applied and proved this measurement. Employee empowerment exists out of three sub elements (perceived control, perceived competence and goal internalization) with each three statements on a five-point Likert scale, from 1 (strongly disagree) to 5 (strongly agree). By combining the statements we use a Cronbach’s Alpha reliability test.
Finally, the industry dependency is asked with two closed questions. The first question is about what is the primer direct customer of the employer, B2B or B2C. The second question was about what does someone’s employer deliver, a service or a product. To test the relationship between the different variables we made a plan of analysis.
3.4 Plan of analysis
28 consistency of the variables. A minimum value of 0.6 has to be measured to qualify if the internal consistency is sufficient (Malhotra, 2009).
The second step is to investigate the relationship between the different variables. We start by analyzing our key relationship, employee engagement on brand ambassadorship. After that we analyze the relationship from attachment styles on brand ambassadorship, these two relations will be analyzed with a regression analysis. By applying a regression analysis, we have to take several assumptions into account (Leeflang, Wieringa, Bijmolt & Pauwels, 2015, p.100). These assumptions are the disturbance term, homoscedasticity, normality and assumption that the mean of the disturbance term is zero. Additional we investigate which explanatory variable has the greatest influence on brand ambassadorship (Leeflang et al., 2015). We will do this by comparing the adjusted R-squares.
29 FIGURE 3.1 MEDIATION ANALYSES
Yi = Brand ambassadorship X = Anxiety of Avoidance Z = Employee engagement
Fourth, we check the influence of employee empowerment between mediator employee engagement on brand ambassadorship. Kumar and Pansari (2016) used employee empowerment as moderator of the outcome of employee engagement on customer engagement, which had a significant positive effect. We test if the moderator employee empowerment will have an influence on the outcome of employee engagement on brand ambassadorship. Therefore we will apply the moderation mediation analysis (figure 3.2, Hayes, Process model 14). For assessing the moderation mediation effect we use the same bootstrap method of Hayes and Preacher (2008) as with the mediation analysis.
Figure 3.2 Moderation mediation analysis, model 14 of Hayes
Y= Brand ambassadorship X= Anxiety or Avoidance Mi = employee engagement V= employee empowerment
30 is low or not there we can suggest that the job-specific attachment styles are not an extension of the general attachment styles, and we have to interpret attachment styles separately.
31 4. Results
4.1 Data
Before starting to analyze the relations between the variables, we will clean the data and then look at the reliability of the variables.
4.1.1 Data cleaning
After the data was collected we analyzed the data on outliers or respondents which should not be in the target group. “Data cleaning includes consistency checks and treatment of missing responses” (Malthotra, 2009, p.461). Because of the screen-out questions, everybody in the sample was in our target group. No extreme cases were found, and all the participants answered all the questions.
4.1.2 Reliability
After the data cleaning we were able to start the analyses. The first step, is that we had to recode statements. The statements which we had to recode were; general anxiety, general avoidance and job specific attachment style avoidance. All the statements were on a five-point Likert were ‘1’ was the highest score. We had to recode this that ‘5’ was the highest score, which means a high level of anxiety or avoidance.
Second, after we recoded the variables we conducted the Cronbach’s Alpha reliability test, “to check the internal consistency reliability of the different items per variable” (Malhotra, 2009, p.319). We can qualify the internal consistency as sufficient if the results measured a minimum value of 0.6. This coefficient can vary from 0 to 1. The results of the Cronbach’s Alpha reliability test for the general attachment styles anxiety and avoidance are in table 4.1. All the results are above 0.6, so we computed a new variable for anxiety and avoidance by measuring the average of all the different elements per sector.
Table 4.1 Cronbach’s Alpha Reliability test, general attachment styles per sector B2B & Service B2B & Manufacturing B2B & Service B2C & Manufacturing
Anxiety 0.874 0.883 0.848 0.892
32 We also applied the reliability analysis for the job-specific attachment styles. The results of the Cronbach’s Alpha reliability test for the job-specific attachment styles are in table 4.2. The values are all above 0.6, so we computed a new variable for job-specific attachment styles anxiety and avoidance by measuring the average of all the different elements per sector. If we look at table 4.1 and table 4.2 we can see that the reliability of anxiety is in both types of attachment styles higher than avoidance. This indicates that the internal consistency is higher for anxiety than avoidance.
Table 4.2 Cronbach’s Alpha Reliability test, specific attachment styles per sector B2B & Service B2B & Manufacturing B2B & Service B2C & Manufacturing
Anxiety 0.821 0.814 0.848 0.840
Avoidance 0.749 0.718 0.738 0.742
33 attachment styles are not an extension we had to interpreted the general and specific attachment styles separately.
In the fourth step, we tested the reliability of employee engagement. Employee engagement exists out of five elements (satisfaction, identification, commitment, loyalty and performance). First, we had to check if these elements score per sector a higher Cronbach’s Alpha of 0.6 (Malhotra, 2009). The results, table 4.4, show that all the Cronbach’s Alpha were above 0.6. Except for the element performance in the sector B2B & Service (.553), B2C & Service(.575) and B2C& Manufacturing(.512). We continued with the Cronbach’s Alpha reliability test to conduct the total employee engagement variable. All the five elements were included, and the results show that all the Cronbach’s Alpha were above 0.6. For total employee engagement the Cronbach’s Alpha in the B2B & Service sector was .862, in the B2B & Manufacturing was .870, in the B2C & Service was .849 and in the B2C & Manufacturing it was .869. So, we continued with the new computed variable total employee engagement for each sector.
Table 4.4 Cronbach’s Alpha Reliability test Employee engagement
B2B & Service B2B & Manufacturing B2C & Service B2C & Manufacturing
Satisfaction .733 .701 .691 .684 Identification .861 .865 .842 .908 Commitment .799 .782 .768 .786 Loyalty .808 .744 .783 .810 Performance .553 .691 .575 .512 Total employee engagement .862 .870 .849 .869
34 perceived competence and goal internalization. After that we conducted a Cronbach’s Alpha reliability test of the new computed variables, the results show that the minimum Cronbach’s Alpha value was .618 across the sectors (appendix B, table 4.5). So, we computed a new variable, total employee empowerment for each sector.
4.1.3 Descriptives
Totally 841 respondents completed the questionnaire, which was conducted on the online research panel of MetrixLab. In table 4.6 is an overview of the sample distribution for each sector. The distribution across the sectors is not perfectly, but because the number of respondents per sector is big enough (lowest n=88 in B2C & Manufacturing) our results are validate enough.
Table 4.6 Respondents per segment
B2B B2C Total Service-based 24.5% (N=206) 47.8% (N=402) 72.3%(N=608) Manufacturing-based 17.2% (N=145) 10.5% (N=88) 27.7% (N=233) Total 41.7% (N=351) 58.3% (N=490) 100% (N=841)
Secondly, the descriptive of the attachment styles. Figure 4.1 shows the average results of the general attachment styles anxiety and avoidance per sector. The average for general attachment style anxiety is 2.46 on a five-point Likert scale. Which means that the respondents are general anxious in relationships. The average of the general attachment styles avoidance is 2.81 on a five-point Likert scale. This suggests that being avoidant, like do not share emotions has a bigger influence, than being anxious, like somebody will forget you. 2,53 2,84 2,3 2,78 2,47 2,82 2,51 2,78 0 1 2 3 4 5
B2B & Service B2B & Manufacturing B2C & Service B2C & Manufacturing General anxious
General avoidance
35 After the general attachment styles we analyzed the job-specific attachment styles anxiety and avoidance (figure 4.2). The average for job-specific attachment style anxiety is 2.53 on a five-point Likert scale and the average for job-specific attachment avoidance is 3.02. The same as with the general attachment styles, respondents are more avoidant than anxious in a relation. This suggests that employees are more sensitive of avoidance behavior (like, sharing emotions) than being anxious (start new connections) in a relationship on the work floor.
Figure 4.2 Average specific attachment styles per segment
We conducted a paired t-test, to test if the respondents differ in their attitude towards general versus specific attachment style. In the B2B & Service sector, we found a medium significant difference in the scores for general attachment anxiety (M=2.53, SD=,861) and job-specific attachment anxiety (M=2.65, SD=,789) conditions; t(205)=-1.691, p=.092. These results suggest that the employees in this sector are more anxious at work than at home. In the B2B & Manufacturing sector we did not found a significant difference in the scores for the general attachment anxiety (M=2.30, SD=.859) and job-specific attachment anxiety (M=2.42, SD=.808) conditions; t(144)=-1.406, p=.162.
In the B2C & Service sector we did not find a significant difference in the scores between the general attachment styles anxiety (M=2.47, SD=.808) and job-specific attachment anxiety (M=2.50, SD=.829), conditions t(401)=-.672, p=.502. Finally in the B2C & Manufacturing we also did not find a significant difference in the scores for the general attachment anxiety (M=2.51, SD,858) and the job-specific attachment style anxiety (M=2.57, SD=.837) conditions, t(87)=-.590, p=.557). We found only one medium significant relation between anxiety in a relation with home or being anxious at work. This suggests that only employees in the B2B & Service sector behave differently at home and at work. It can be
2,65 3,05 2,42 2,9 2,5 3,07 2,57 2,92 0 1 2 3 4 5
B2B & Service B2B & Manufacturing B2C & Service B2C & Manufacturing Specific anxious
36 that employees in the B2B & Service sector have less connection with their colleagues because there is more focus on performance.
We also conducted a paired t-test if the respondents differed in their attitude towards the general attachment style avoidance and the job-specific attachment style avoidance .In the B2B & Service we found a significant difference in the scores for the general attachment style avoidant (M=2.84, SD=,697) and job-specific attachment style avoidant (M=3.05, SD=.687), conditions t(205)=-3.226, p=.001. In the B2B & Manufacturing we did not find a significant difference in the scores for the general attachment style avoidant (M=2.78, SD=,774) and job-specific attachment style avoidant (M=2.90, SD=.662), conditions t(144)=-1.614, p=.109. In the B2C & Service we found a significant difference in the scores for the general attachment style avoidant (M=2.82, SD=,650) and job-specific attachment style avoidant (M=3.07, SD=.667), conditions t(401)=-5.708, p=.000. In the B2B & Manufacturing we did not find a significant difference in the scores for the general attachment style avoidant (M=2.78, SD,718) and job-specific attachment style avoidant (M=2.92, SD=.671), conditions t(87)=-1.218, p=.227. These results suggest that by employees in the B2B & Service, their personality trait of avoidance behavior on work is significant bigger than at home. This could mean that employees in the service industry are more avoidant to ask for help or share their emotion at work than at home. Overall, the results suggest that in the B2B & Service sector the employees are less anxious and avoidant at home then at work.
37 Fourth and finally we measured the NPS for brand ambassadorship per segment. As mentioned in the literature review we divided the ambassadorship in two recommendation questions. The first one is to recommend your friends, family or clients to work for your employer. The second recommendation is that an employee would recommend the firm as a supplier to someone else. The overall NPS for the colleague recommendation is -17. If we analyzed for the separate sectors, we saw that in the sector B2B & Manufactory the NPS is the highest (-4, figure 4.3). Which suggested that these employees recommend other people to join their employer more than other sectors do. It can suggest that these people are more satisfied with their work circumstances and their employer. The lowest NPS is in the B2B & Service sector (-24), which can suggest that people are afraid of new colleagues (competition) or are less satisfied about the primary and secondary work circumstances.
Figure 4.3 NPS on the recommendation to others to work at your organization
The second NPS, about how much does someone recommend the employer as a supplier to others (firms, customers and/or family). The overall NPS is -10. Which suggest that employees are not very satisfied about the product or service they deliver to business and customers. Employees in the B2B & Manufacturing are the most promoters (+9, figure 4.4). This suggests that employees were very satisfied about the produced products they deliver to their customers (businesses). The lowest score was measured in the B2C & Service (-16). Which suggests that employees are not very satisfied about the ,quality, of the service they deliver to consumers. 13 18 13 13 50 60 57 52 37 22 30 35 0% 20% 40% 60% 80% 100% B2B & Service B2B & Manufacturing B2C & Service B2C & Manufacturing
Promotors Passives Detractors
-24
-4 -17
38 Figure 4.4 NPS on the recommendation to others that your firm as supplier
The NPS of the colleague recommendation is overall lower than the NPS of the supplier recommendation. This suggests that employees are satisfied or proud of what they deliver (service or product) but are not satisfied or proud of their own work circumstances. In the B2B & Manufacturing the employees are the most brand ambassadors, both the highest on colleague- and supplier recommendation.
We tested the correlation and the Cronbach’s Alpha of the two recommendation. We found that we were able to combine the two recommendations, because all the Cronbach’s Alpha were above 0.6 across the sectors (appendix B, table 4.7) and the lowest correlation found was .535, and significant. Because of our theoretical suggested difference between the supplier and colleague recommendation we continued with the two recommendations separately.
4.2 Attachment styles on Brand Ambassadorship
To analyze the relation between attachment styles on brand ambassadorship we conducted a regression analysis. With a regression analysis we measure the effect of one or more independent variables on a metric dependent variable. To compare the effect we used the unstandardized beta coefficient.
4.2.1 General attachment styles on Brand Ambassadorship
By conducting a regression analysis from the general attachment styles anxiety and avoidance on brand ambassadorship we found for the supplier recommendation none significant results (appendix B, table 4.8). By conducting the regression on colleague
20 28 14 16 46 53 56 60 35 19 30 24 0% 20% 40% 60% 80% 100% B2B & Service B2B & Manufacturing B2C & Service B2C & Manufacturing
Promotors Passives Detractors
39 recommendation (appendix B, table 4.9) we found one significant result (p=.027) and one medium significant (p=.066) result. In the B2B & Service the result was significant (p=.027) with a small adjusted R2 of 0.025. General avoidance was in this relation negatively significant (B=-.625, p=.008) to colleague recommendation. General avoidance was positive medium significant (B=0.311, p=.099) related. Which is a surprisingly phenomenon, because this means that the more avoidant someone is in the relation with the firm, the more someone would recommend to join the organization.
The second medium significant (p=.066) result we found was in the B2B & Manufacturing sector, with an adjusted R2 of .024. General avoidant had no significant relation on colleague recommendation. General anxiety had a medium significant negative relation on colleague recommendation (B=-0.468, p=.057). All the other relations were not significant. If we looked at the adjusted R2, we conclude that the general attachment styles anxiety and avoidance had no or very little influence on brand ambassadorship. So, there is little or no significant influence from the general attachment styles on brand ambassadorship and a low or no significant correlation between the two attachment styles. We overall conclude that the general attachment styles did not have a (significant) influence in our conceptual model. For further analysis we excluded the general attachment styles.
4.2.2 Job-specific attachment styles on Brand Ambassadorship
40 unstandardized beta value of -1.376 (p=.000), which indicated that if the employee rate 1 point higher on avoidant the NPS will decrease with -1.376. Table 4.10 is an overview of the results of the job-specific attachment styles anxiety and avoidance on colleague recommendation. The results indicates that employees which are more anxious and avoidant will give a more negative colleague recommendation. Across the sectors we found that avoidant behavior, like do not ask for help if necessary, has (general 2 times) more negative effect on the colleague recommendation than anxiety had.
Table 4.10 Regression, job-specific attachment styles on colleague recommendation
Sector Adj. R2 F(sig.) Independent variables Unstandardized Beta
B2B & service .171 22,126 (p=,000) Anxiety -.482***
Avoidance -.948***
B2B & manufacturing .241 23.850 (p=,000) Anxiety -.599*** Avoidance -.1.254*** B2C & service ,250 67.814 (p=0,000) Anxiety -.483***
Avoidance -1.14***
B2C & manufacturing .358 25.288 (p=.000) Anxiety -.809*** Avoidance -1.376*** *<0.10, **<0.05,***<0.01
41 Table 4.11 Regression, job-specific attachment styles on supplier recommendation.
Sector Adj. R2 F(sig.) Independent variables Unstandardized Beta
B2B & service .132 16.600(p=,000) Anxiety -.171 Avoidance -1.192*** B2B & manufacturing .096 8.651 (p=,000) Anxiety -.292
Avoidance -.861***
B2C & service .107 24.904 (p=,000) Anxiety -.068 Avoidance -1.106*** B2C & manufacturing .193 11.407(p=,000) Anxiety -.646***
Avoidance -.819***
*<0.10, **<0.05,***<0.01
4.3 Attachment styles on Brand Ambassadorship, mediated by employee engagement After the direct effect of the attachment style on brand ambassadorship we added a mediator, employee engagement, in the analysis. We analyzed the mediation of employee engagement for job-specific attachment styles on colleague and supplier recommendation in each sector separately. By conducting a mediation analysis (figure 3.1), we applied three regression analyses on each variable to propose the mediation model. We used the Process of Hayes, which allowed us to include one independent variable. But we had two independent variables in our study, job-specific anxiety and avoidance. So, we included anxiety in the independent variable and placed avoidance as covariate, and vice versa.
42 bootstrap estimation approach with 5000 resamples (Hayes, 2008). The results indicated that the indirect coefficient was significant b=-0.2407, SE=.0872, 95% CI= -0.4414, -0.0977. Being more anxious was associated with around -0.24 point less colleague recommendation as mediated by employee engagement.
Secondly, we looked at the supplier recommendation per sector (table 4.12). In all the sectors we saw that there are negative significant results (appendix B, table 4.14). Looking at the B2B & Service sector the results indicated that anxiety was a significant predictor for employee engagement, b=-3.4121, SE=.864, p=<.01, and that employee engagement was a significant predictor of colleague recommendation, b=.0756, SE=.0152, p=<.01. These results support the mediation analysis. Anxiety was no longer a significant predictor of colleague recommendation after controlling for the mediator, employee engagement, b=.0870, SE=.1942, not significant, which mean there is a full mediation. The adjusted R-square is .2230, which indicates that 22.30% of the variance in colleague recommendation was explained by the predictors. The indirect effects were tested by using a bootstrap estimation approach with 5000 samples (Hayes, 2008). The results indicate that the indirect coefficient was significant b=-0.2593, SE=.0962, 95% CI= -0.4935, -0.1082. Being more anxious was associated with around -0.26 point less colleague recommendation with a mediator employee engagement. In appendix B, table 4.13 are the results of the mediation analyses on colleague recommendation across the sectors. And in appendix B, table 4.14 are the results of the mediation analyses on supplier recommendation across the sectors.
43 Table 4.12 Mediation analyses: specific attachment styles on ambassadorship, mediated by employee engagement Independent variable Mediator Dependent variable B2B & Service B2B & Manufacturi ng B2C & Service B2C & Manufacturin g Anxiety Employee engageme nt Ambassador-ship (Colleague)
Full Partial Partial Partial
Avoidance Employee engageme nt
Ambassador-ship (Colleague)
Full Full Full Full
Anxiety Employee engageme nt Ambassador-ship (supplier) Inconsistent mediation Inconsistent mediation Inconsistent mediation Partial Avoidance Employee engageme nt Ambassador-ship(supplier)
Full Full Full Full
4.4 Moderation effect of employee empowerment
44 relation between employee engagement and colleague recommendation (appendix B, table 4.15)
Secondly we also adapted the analysis in the supplier recommendation stage. We found that only in the B2C & Service we found a little significant effect (B=-.0279, p=.0190, 95% CI=-.0512, -.0046) with an adjusted R-square of .2291 (p=.000). In the other sectors we did not find a significant influence (appendix B, table 4.16). From both analyzes we suggest that employee empowerment did not have a moderation effect on the relation (V, figure 3.2) between the mediator employee engagement and brand ambassadorship, except in the B2C & Manufacturing sector (table 4.17). This gives a surprising result, because it indicates that the more empowerment employees receive, like in a bakery getting the opportunity to decide to bake a special bread, the less someone recommends other people to join the organization. But the adjusted R-square is almost ten percentage higher, this indicates that the NPS is better explained by including employee empowerment.
Table 4.17 Comparison mediation analysis with and without the moderation effect of employee empowerment
Sector Dependent Variable Adjusted R-square
Without empowerment With empowerment B2C & Service Colleague
recommendation .3839*** .4039*** B2C & Manufacturing Colleague recommendation .4730*** .5751***
B2C & Service Supplier
recommendation
.2052*** .2291***
4.5 Validation
45 We checked the two models on the square, p-value and the adjusted R-square. “The chi-square statistic is used to test the statistical significance of the observed association in a cross-tabulation” (Malhotra, 2009). The p-value explains if the model is significant. The adjusted R-square explains the percentage of model that is explained by the independent variables. Table 4.18 shows us the results. The p-value is over all the models significant. But looking at the adjusted R2 we see differences. Model 1, colleague recommendation, scored in each sector higher than model 2, supplier recommendation. Based on that we conclude that our model fits best by the colleague recommendation.
Table 4.18 Adjusted R2 and p-values from mediation model on colleague- and supplier recommendation.
Model 1 (DV) Chi-square (F)
Adj. R2 Model 2 (DV) Chi-square (F)
Adj. R2
B2B & Service Colleague recommendation 26.8667*** .2746 Supplier recommendation 20.6130*** .2230 B2B & Manufacturing Colleague recommendation 27.3487*** .3544 Supplier recommendation 19.5257*** .2785
46 5. Discussion and conclusion
In this chapter we discuss and conclude the results of our study. First we looked at only the NPS for colleague and supplier recommendation, and we found only one positive figure. We expected there would be more positive figures. But Van Doorn et al. (2008) already suggested that “an 8 is already a high grade in the Netherlands (p.316).” Van Doorn et al. (2008) conducted also alternative NPS with detractors (0-5) and promoters (8-10), but they found that there is no big difference. De Haan, Verhoef & Wiesel (2015) found in their research also a low NPS value of 6.59 (SD=2.13, n=1375, on a scale 0 untill 10) in the Netherlands. So, the negative NPS is not strange in the Netherlands. Only the supplier recommendation +9 in the B2B & Manufacturing was relatively high. When we saw the descriptive of the NPS data in the B2B & Manufacturing sector, we found that 53.2% were passives which can explain the positive NPS. It can also be that the employees are more proud of what they produce compared to other businesses. We stick to our findings that Reichheld (2003) and Van Doorn et al. (2008) proved that NPS is neither superior nor inferior to other metrics and applicable in a research with well formulated methodology, and our figures can be used.
Second, we found differences between the general and the job-specific attachment styles. Mende and Bolton (2011) proved that business-specific attachment styles were an extension the of general attachment styles. Our findings show that there is only one medium correlation, and a few relatively low correlations and in many cases not even significant. Our study suggests that the job-specific attachment styles are not an extension of the general attachment styles.
