Schaufeli (2014) notes that work engagement and employee engagement are used interchangeably; however, work engagement refers to the relationship between the employees and their work, while employee engagement refers to the relationship between employees and the organisation. In the context of this study, the focus is on work engagement.

According to Schaufeli, Bakker and Salanova (2006), work engagement is a positive, fulfilling work-related state of mind that is characterised by vigour, dedication and absorption. Consequently, work engagement may be assessed by examining vigour, dedication and absorption (Schaufeli et al. 2006). Vigour is defined as high energy levels, tenacity in difficult assignments and the desire to master certain skills and tasks. Dedication is the feeling of inspiration, accompanied by satisfaction, and of work being important and challenging to the individual. Absorption is being engrossed in one’s work, accompanied by happiness and deep levels of attentiveness. Employees, who experience a state of work engagement, have an effectual vital association with their work and feel that they meet the requirements of their role (Schaufeli et al. 2006).

Work engagement is becoming more important for business leaders and managers, due to its relevance to service-related outcomes; this is mainly due to work engagement being predictive in relation to performance (Barnes et al. 2014). Not only does work engagement predict performance accurately, but it has also been associated with finding meaning in work (Hoole & Bonnema 2015). Additionally, researchers are now suggesting that complete work engagement could also lead to meaningfulness at the work place (Hager 2018). Meaningfulness is influenced by job enrichment, work-role fit and co-worker relations (May et al. 2004). Job enrichment has an influence on the degree of meaningfulness experienced through work, while meaningfulness has an even larger impact when the individual has sound working relationships with co-workers (May et al. 2004). In terms of the current research, literature shows that when employees experience meaningfulness at the workplace through job design, it may influence employees’ interaction with clients and consumers (Kahn 1990). Finding meaning in one’s work has also been associated with improved job performance (Rodell 2013). Consequently, if employees are not fully engaged, they do not focus on customers’ needs. There is potential benefit in each interaction of employees with customers, which may have an impact on the customer’s perception of the organisation and brand image; if employees are disengaged it would probably have a negative impact on organisational performance (Kumar & Pansari 2016).

Work engagement in a South African environment was found to be associated with psychological meaningfulness (Geldenhuys et al. 2014; Van Zyl, Deacon & Rothmann 2010). According to Geldenhuys et al. (2014), meaningful work is especially important in the South African environment, as this results in employees being committed to their organisations. Previous South African research found that work engagement and meaningfulness were associated with satisfaction in life (Williamson & Geldenhuys 2014). Although the relationship between work engagement and meaningfulness is well-documented from a South African perspective, the indirect effect meaning might have on the relationship between work engagement and perceived customer value in a South African environment remains uncertain.

Further, there seems to be a research opportunity for a more thorough examination of the relationship between work engagement and performance (Christian & Slaughter 2007). In the case of this study, the performance variable is perceived customer value. Perceived customer value has been associated with customer satisfaction (Han & Xie 2019) and sales performance (Chakrabarty, Brown & Widing 2010). From a South African perspective, perceived customer value has also been associated with consumer satisfaction (Mbango & Toerien 2019).

The measuring instrument involved two steps. In the first step, biographical information from respondents was collected, while views on work engagement, meaningfulness through work and on perceptions of customer value were collected in the second step.

The holding company from the manufacturing sector was requested to participate in the research by contacting their executive management and human-resource departments. An introductory meeting was held with representatives from the various human-resource departments to explain the rationale behind and focus of the study. A covering letter, as well as a questionnaire, were distributed to sampled employees, via email, through their human-resource departments. Respondents were given 30 days to complete the questionnaire and were reminded once a week. In the covering section of the questionnaire, ethical considerations were addressed by assuring participants of confidentiality and non-prejudice. All the participants were free to voluntarily participate and were not coerced in any way. The holding company from the manufacturing sector granted permission to conduct the research.

Data was cleaned, captured and analysed with the Statistical Package for Social Sciences (SPSS) and Analysis of a Moment Structure (AMOS), version 26 (IBM Corp 2019). Descriptive statistics (means and standard deviations) were calculated in order to determine central tendencies for work engagement, meaningfulness at work and perceived customer value. Internal consistency was determined by calculating Cronbach’s alphas (α) and inter-item correlations (r). Cronbach’s alpha threshold was set at a moderate level (α = 0.60; Robinson, Shaver & Wrightsman 1991) while inter-item correlation was considered sufficient (r = 0.20–0.40; Pallant 2007).

An exploratory factor analysis (EFA) was conducted for individual variables in order to assess the internal structure. Principal component analysis was used as the extraction method. Examinations included the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Bartlett’s test of sphericity. According to Field (2009), KMO values should be as close as possible to 1 and Bartlett’s test of sphericity should be significant, while eigenvalues larger than 1.00 were used as criteria for factor selection. The minimum threshold set for factor loadings was considered adequate at 0.40 (Hinkin 1998).

Once reliability and the internal structure were confirmed, the work engagement, meaningfulness through work and perceived customer value variables were subjected to a confirmatory factor analysis (CFA) by making use of AMOS. We considered the Root Mean Square Error of Approximation (RMSEA) and relative/normed chi-square (χ²/df) as absolute fit-indices, while incremental fit-indices included the Tucker-Lewis Index (TLI) and Comparative Fit Index (CFI). The cut-off criteria for adequate CFA were considered to be: RMSEA < 0.08 (Van de Schoot, Lugtig & Hox 2012), χ²/df < 2.0 (Tabachnick & Fidell 2007) CFI and TLI > 0.90 (Wang & Wang 2012).

Spearman’s correlations were calculated in order to determine the type and nature of the relationships between the variables under investigation. Correlations were firstly analysed for statistical significance (p < 0.05), and additionally assessed for practical significance. Practical significance was considered at r ≥ 0.10, medium effect at r ≥ 0.30, and large effect at r ≥ 0.50 (Cohen 1988). Additionally, the final model was tested for multicollinearity. Thresholds for multicollinearity was set at 4.00 for variance inflation factors (VIF), and not less than 0.20 for tolerance (Hair et al. 2010). Direct and indirect effects were assessed by computing a simple mediation model with PROCESS (Hayes 2017), which is installed in SPSS. Direct effects were considered significant at p < 0.05, while indirect effects were considered significant if the lower-level confidence interval (LLCI) and upper-level confidence interval (ULCI) excluded zero (Zhao, Lynch & Chen 2010).

Table 1 shows the results obtained from the descriptive, reliability and correlation analyses. According to Table 1, all variables, except psychological safety met the set of minimum criteria for internal consistency. However, Pallant (2007) indicates that scales with less than 10 items may present as problematic, so low Cronbach’s alphas and, therefore, inter-item correlations should also be considered. Psychological safety was assessed against only three scale items. The inter-item correlation for psychological safety (r = 0.25) was considered adequate and the variable was deemed reliable.

Closer examination of the mean and standard deviation scores revealed that all variables were within each scale’s ‘positive’ range. Therefore, we are able to confirm that, according to the respondents, there was a conducive level of work engagement; they experienced meaningfulness at the workplace and had a positive perspective on providing customer value.

With regard to the relationships between the variables under investigation, the results from the Spearman’s rank order correlations demonstrated that work engagement, meaningfulness and perceived customer value had a significant positive relationship. A strong positive relationship (r = 0.65; p < 0.05; large effect) existed between work engagement and meaningfulness. In reality, this result translates into a situation where respondents, who reported positive levels of work engagement, would also be inclined to report high levels of meaningfulness. In terms of the relationship between meaningfulness and perceived customer value, the result was a moderate positive relationship (r = 0.35; p = 0.05; medium effect). This result suggests that positive levels of meaningfulness will also be associated with positive levels of perceived customer value. The correlation calculation between work engagement and perceived customer value demonstrated a positive result (r = 0.29; p < 0.05; small effect) as well. Although interpreted as a small effect, this result implies that respondents, who reported conducive levels of work engagement, were also likely to report conducive levels of perceived customer value.

Concerns with regard to potential multicollinearity were dismissed, due to results obtained from VIF and tolerance computations. Scores for VIF ranged from 2.74 to 1.56, and for tolerance from 0.37 to 0.63. Consequently, work engagement, meaningfulness and perceived customer value may be applied independently with confidence in the simple mediation model.

An exploratory factor analysis (EFA) was conducted on each individual variable in order to examine the structure. The examinations of eigenvalues and scree plots confirmed the individual factors for each variable. For vigour, the first factor extracted, explained 73.72% of the variance with an eigenvalue of 2.21. Factor loadings ranged from 0.75 to 0.77 for vigour (KMO = 0.72; Bartlett’s test of sphericity = p < 0.01). With regard to dedication, the first factor extracted, explained 74.33% of the variance with an eigenvalue of 2.23 and factor loadings ranging from 0.67 to 0.90 (KMO = 0.70, Bartlett’s test of sphericity = p < 0.01). For absorption, the first factor extracted, explained 63.57% of the variance, with factor loadings ranging from 0.67 to 0.72 (KMO = 0.70; Bartlett’s test of sphericity = p < 0.01). For perceived customer value, the single factor extracted, explained 60.03% of the variance with an eigenvalue of 2.40. Factor loadings for perceived customer value ranged from 0.56 to 0.92 (KMO = 0.72; Bartlett’s test of sphericity = p < 0.01).

For psychological meaningfulness, the first factor extracted during the EFA explained 70.90% of the variance with an eigenvalue of 4.26. Factor loadings ranged from 0.69 to 0.89 for psychological meaningfulness (KMO = 0.85; Bartlett’s test of sphericity = p < 0.01). The single factor extracted for psychological safety explained 51.00% of the variance with an eigenvalue of 1.52. Factor loadings ranged from 0.40 to 0.62 for psychological safety (KMO = 0.60; Bartlett’s test of sphericity = p < 0.01). In terms of psychological availability, the single factor extracted, explained 58.81 of the variance with an eigenvalue of 2.94. Factor loadings for psychological availability ranged from 0.63 to 0.86 (KMO = 0.81; Bartlett’s test of sphericity = p < 0.01).

A latent variable model was constructed in AMOS in order to conduct the CFA. Work engagement was considered by assessing vigour, dedication and absorption, while perceived customer value was considered as a one-dimensional variable. Meaningfulness at work was considered by assessing psychological meaningfulness, psychological safety and psychological availability. The model demonstrated adequate fit (χ² = 448; df = 289; p < 0.001; χ²/df = 1690; TLI = 0.90; CFI = 0.91; RMSEA = 0.068). Based on the results obtained from the CFA, validity for the present study was deemed acceptable.

The results of the simple mediation model are illustrated in Table 2. Work engagement was considered the independent variable, perceived customer value the dependent variable and meaningfulness the proposed mediator.

According to the results of the standardised regression coefficients, the first regression (1) in which work engagement was considered the independent variable and the proposed mediator, meaningfulness at work, the dependent variable, yielded a significant result (b = 0.36, t (150) = 10.41, p < 0.01). The second regression (2) in which the proposed mediator, meaningfulness, was considered the independent variable and perceived customer value the dependent variable, also yielded a significant result (b = 0.57, t (149) = 2.83, p < 0.01). The final regression (3) in which work engagement was considered the independent variable and perceived customer value the dependent variable, did not yield a significant result (b = 0.12, t (149) = 1.05, p = 0.29). The indirect effect of meaningfulness at the workplace on the relationship between work engagement and perceived customer value was also significant (b = 0.20, SE = 0.08, LLCI = 0.06 and ULCI = 0.36). The indirect effect from the bootstrap analysis with a 95% confidence interval did not include zero; therefore, (a x b) is significant (Zhao et al. 2010). Consequently, meaningfulness may be considered a mediator for the relationship between work engagement and perceived customer value. The results for the simple mediation model are shown in Figure 2 below.

Literature findings indicate that perceived customer value is related to improved customer satisfaction (Han & Xie 2019; Mbango & Toerien 2019), including improved sales performance (Chakrabarty et al. 2010). The findings of this study indicate that meaningfulness at the workplace predicted improved perceived customer value. The findings further specify that having engaged employees is not enough, but that the manufacturing organisation also relies on creating meaning at work for employees in order to improve their perceptions of customer value. South African organisations wishing to improve indications of meaningfulness at work should direct their attention towards person-job fit, self-concept-job fit (Scroggins 2008), calling orientation, job design and co-worker relations (Fouché, Rothmann & Van der Vyver 2017). The improvement of work engagement would not only be applicable to the employees’ finding meaning at the workplace, but will also be extended to the customer with enhanced perceptions of customer value, which in return will improve employee productivity and organisational performance.