- Email: [email protected]
Service quality of frontline employees: A profile deviation analysis
Journal of Business Research 66 (2013) 1338–1344
Contents lists available at SciVerse ScienceDirect
Journal of Business Research
Service quality of frontline employees: A profile deviation analysis☆ Neeru Malhotra a,⁎, Felix Mavondo b, 1, Avinandan Mukherjee c, 2, Graham Hooley d, 3 a
Aston Business School, Aston University, Aston Triangle, Birmingham B4 7ET, UK Monash University, Clayton Building 11E, Room 272, Clayton Campus, Welington Rd., Clayton, 3168, Australia c Department of Marketing, School of Business, Montclair State University, Montclair, New Jersey, USA d Marketing Department, Aston University b
a r t i c l e
i n f o
Article history: Received 1 April 2011 Received in revised form 1 September 2011 Accepted 1 November 2011 Available online 3 March 2012 Keywords: Profile deviation analysis Configuration theory Bank branches Call centers Service Quality Commitment
a b s t r a c t Using a configuration theory approach, this paper conducts a comparative study between frontline employees in phone and face-to-face service encounters for a retail bank. The study compares the top performers in service quality in relation to three components of organizational commitment and their demographics by applying a profile deviation analysis. The results show that the profile deviation for face-to-face employees is significantly negative, while for call center employees nonsignificant. Although the study finds no significant differences in the three components of commitment, significant differences exist in the total experience and age of the best performers. Also, affective commitment dominates the profile of high performers, while poor service providers seem to exhibit a higher level of continuance commitment. This study demonstrates the utility of profile deviation approaches in designing internal marketing strategies. © 2012 Elsevier Inc. All rights reserved.
1. Introduction Organizations operating in today's highly competitive business environment need to differentiate on service quality as a means of achieving a competitive advantage, and frontline employees are central in determining this quality (Gustaffson, 2009). The frontline is the touch point of the company; therefore, the service that frontline employees provide is critical in developing customer relationships, gathering customer information, and in creating customer satisfaction, loyalty, and brand commitment (Burmann & Konig, 2011; Fang, Palmatier, & Grewal, 2011). Previous research indicates that the organizational commitment of frontline employees exerts a strong, positive influence on their service quality (e.g., Malhotra & Mukherjee, 2004; Vandenberghe et al., 2007). However, the actual commitment-profile differences between high and low quality service performers have not been substantively researched.
☆ The authors acknowledge the research support of Kyungwon Lee. ⁎ Corresponding author. Tel.: + 44 121 2043151; fax: + 44 121 3334917. E-mail addresses: [email protected] (N. Malhotra), [email protected] (F. Mavondo), [email protected] (A. Mukherjee), [email protected] (G. Hooley). 1 Tel.: + 61 3 990 59249. 2 Tel.: + 1 973 655 5126: fax: + 1 973 655 7673. 3 Tel.: + 44 (0) 121 204 4643. 0148-2963/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.jbusres.2012.02.034
Using a configuration theory approach, this study compares employees in the two most difficult and important types of frontline service situations; face-to-face and telephone encounters. First, this study identifies the profile of the top performers in each of the encounters. Second, this study makes comparisons between the ideal and non-ideal profiles to investigate whether deviations from the ideal result in a decrease in service quality in each context. Any difference this study finds among the profiles of ideal performers in the two contexts is useful to explore, especially for companies employing multi-channel delivery, in order to recruit and manage frontline employees effectively through appropriately designed internal marketing strategies (see Lings, 2004; Wieseke, Ahearne, Lam, & VanDick, 2009). This study offers four contributions to the services marketing literature. First, this study applies a profile deviation analysis with a basis in configurational theory, which is a methodological innovation because profile deviation has rarely been used in marketing. Profile deviation approaches provide significant advantages over traditional approaches such as regression analysis, slope analysis, and subgroup analysis, particularly with individuals as the unit of analysis, in assessing fit in a way that is consistent with the multidimensional and holistic perspective of services marketing (Vorhies & Morgan, 2003). Previous research mainly applied the configuration theory and the profile deviation approach to organizations as units of analysis to assess organizational performance (Chen, Huang, Sung, & Huang, 2009; Kabadayi, Eyuboglu, & Thomas, 2007). But, this study uses these concepts to predict individual behavior.
N. Malhotra et al. / Journal of Business Research 66 (2013) 1338–1344
Second, this study conducts an empirical test to see if deviations from the ideal profile result in a decrease in service quality. In this respect, an important question is how different the top performers in call centers are from those in bank branches in face-to-face and telephone types of encounters. Third, this study explores whether the form of commitment matters in the commitment-service quality relation. Previously, a majority of the studies in marketing adopted one-dimensional approach to organizational commitment while the implications of a three-component model of commitment that comprises affective, normative, and continuance commitment, (Allen & Meyer, 1990) have remained largely unexplored (Culpepper, Gamble, & Blubaugh, 2004; Malhotra, Budhwar, & Prowse, 2007). Affective commitment (AC) is the extent of an employee's emotional attachment to, identification with, and involvement in the organization; normative commitment (NC) denotes an employee's feelings of obligation to stay with the organization; and continuance commitment (CC) is the commitment based on the costs that the employee associates with leaving the organization. From the literature, AC seems to have more of a relation to service quality than NC and CC (Meyer, Stanley, Herscovitch, & Topolnytsky, 2002; Vandenberghe et al., 2007). Because not all forms of commitment necessarily have an association with high job performance, the profile of top service quality performers needs to be understood with respect to the three components of commitment. Fourth, this study also investigates whether the form of service delivery matters for performance fit by comparing face-to-face and phone services. This study integrates the service-quality literature with the profile-deviation analysis literature to provide valuable new insights into the theory and practice of service excellence, and is the first to attempt to understand the role of configuration theory in designing internal marketing strategies. 2. Configuration theory Configuration theory has been used by management over the last two decades to assess complex, multidimensional phenomena implied in fit or congruence relations in ways that are more consistent with the holistic framing of strategic management and marketing strategy than traditional approaches like interactions or contingency theory. Traditional approaches lack correspondence between verbal and statistical approaches when testing the theory. This lack of correspondence means that a weak link exists between theory building and theory testing that leads to inconsistent research findings. Configurational theories posit higher effectiveness for employees that resemble the ideal type that theory defines. The increased effectiveness comes from an internal consistency or fit among the relevant structural and strategic factors (Doty, Glick, & Huber, 1993). Researchers who treat configurations as categories rather than the ideal types fail to test the core thesis of the theory. Many researchers have developed categories of employees by using cluster analysis and then comparing the means of the categories across effectiveness measures (Smith, Guthrie, & Chen, 1989). The appropriateness of this approach is questionable. When treating configurations as categories, they predict that marginal members of the category are as effective as the central members. On the other hand, when treating configurations as ideal types, they predict employees that marginally resemble the ideal configuration as less effective than those that closely resemble the ideal (Doty et al., 1993). Modeling the ideal-type employee should begin by recognizing that an ideal type is a theoretical construct, that is, a singular and discrete phenomenon rather than a nominal category. Hence, any empirical test should involve a rich multivariate approach to define the ideal type. Employees do not have to be classified into nominal groups because the crux of the matter is the deviation of the employee from the ideal type. The deviation measure (a Euclidian distance) can then be used to predict employee effectiveness in a way consistent with theory.
1339
Management researchers have conceptualized the configuration model as an interaction, selection, or systems approach to fit (Drazin & Van de Ven, 1985). The interaction approach is the basis for many contingency theories that define fit as the statistical interaction of two variables (Schoonhoven, 1981), although population ecologists adopt the selection approach to develop taxonomies. Neither of these approaches is consistent with the complex fit assertions in configurational theories (Venkatraman & Prescott, 1990). Thus, this study conceptualizes fit through a systems approach, which is the most appropriate. Drazin and Van de Ven describe this approach as the most complex and promising approach for future research. The systems approach includes profile deviation and gestalt and defines fit in terms of consistency among multiple dimensions (see Kabadayi et al., 2007). Fit is high to the extent that an employee is similar to an ideal type along many dimensions, and effectiveness is highest in the ideal types because the fit among factors is at a maximum in those configurations. When applying fit to employees, an opportunity exists to provide a more holistic profile that leads to superior decisions in selecting and rewarding outstanding employees. When considering fit among multiple elements simultaneously and examining the effects on outcomes, as this study does, configuration should be conceptualized and measured via profile deviation (PD) analysis (Doty et al., 1993; Venkatraman, 1990; Vorhies & Morgan, 2003). PD analysis views fit as the degree to which a particular case (a customer contact employee, in this study) matches an ideal profile (an optimal standing within a dataset) (Hult, Boyer, & Ketchen, 2007; Venkatraman, 1990; Zajac, Kraatz, & Bresser, 2000). The PD analysis in this study (Fig. 1) assesses the fit between the three components of commitment and service quality as the degree to which the commitment and demographic characteristics of a frontline employee differ from those of an ideal profile in achieving service quality (Vorhies & Morgan, 2003; Zajac et al., 2000). Employees whose profiles the innermost circle (A) represents are the ideal type of employee who delivers the highest service quality to customers. Employees that inner circle (B) represents deliver lower quality than those in A but higher than those in C; and those in C, in turn, deliver better service quality than those in circle D (Fig. 2). The prediction of PD is that, as the Euclidian distance increases from the ideal, service quality deteriorates. Thus, the more a frontline employee is like the ideal profile (definition from AC, NC, and CC components and the total experience and age demographics), the more superior the service quality is that the employee delivers. Hence, the key proposition relating the profiles of service employees and the outcome of interest (i.e., service quality) is: Proposition 1. The ideal type of employees will deliver significantly superior service quality than non-ideal employees. With the growth in multi-channel strategies among services, the two types of encounters involving frontline employees, face-to-face and telephone are becoming increasingly crucial to manage. However, several differences exist in customer service delivery between face-to-face and telephone encounters that have a bearing on the type of frontline employees suited for these encounters. In face-to-face encounters, both verbal and non-verbal behaviors (e.g., employee physical appearance and dress) are important determinants of service quality; the customer plays a role in creating quality service through his/her behavior during the interaction, and people can create quality perceptions relating to the environment where the service takes place (Burgers, Ruyter, Keen, & Streukens, 2000). On the other hand, in phone encounters, the customer has less influence on service quality. The service environment and tangibles are not part of the quality perceptions; their judgement of service and quality comes purely from intrinsic dimensions like reliability, responsiveness, assurance, and empathy (Boshoff & Tait, 1996; Burgers et al., 2000). Further, performance in branches involves less scripting and standardization as compared to call centers, which provides an opportunity to the more able employees
1340
N. Malhotra et al. / Journal of Business Research 66 (2013) 1338–1344
Organizational Commitment • Affective • Normative • Continuance
Employee profile fit with ideal profile Deviation from an ideal employee profile leads to inferior service quality
Service Quality
Demographics • Total experience • Age
Fig. 1. A model of profile deviation for customer-contact employees.
to use their initiative, experience, and commitment to stand out. Keeping this in mind, this research proposes: Proposition 2. The ideal type of employees will differ significantly with respect to their commitment profile and demographics in the two types of service encounters.
3. The methodology 3.1. Sample This research comes from telephone call centers and branches of a major retail bank in the United Kingdom. Because the perception is that most financial products are high-involvement, complex, and low on differentiation, the frontline employee, and not the service itself, provides a source of differentiation and creates competitive advantage (Burgers et al., 2000). Hence, this study measures service quality from the viewpoint of the frontline employee. Researchers mailed self-administered anonymous surveys to the heads of customer services and the regional branch managers who then arranged for further distribution to their respective frontline employees. The researchers used stratified sampling to ensure that no selection bias exists in the sample. They also provided selfaddressed pre-paid envelopes along with the questionnaires, and the employees returned the completed questionnaires directly to one of the researchers. Six hundred and forty employees in the call centers and 300 employees in the branches received the questionnaires. This distribution
D C B A
A: The ideal type of employees who delivers highest service quality to customers. B: Employees who deliver lower service quality than A. C: Employees who deliver better service quality than D, but lower service quality than B. D: Employees who deliver lower service quality than C. Fig. 2. Profile deviation as concentric circles. A: The ideal type of employees who delivers highest service quality to customers. B: Employees who deliver lower service quality than A. C: Employees who deliver better service quality than D, but lower service quality than B. D: Employees who deliver lower service quality than C.
included all employees who satisfied the initial profile screening. This distribution yielded 342 useable questionnaires from the call centers and 170 useable questionnaires from the branches, generating net response rates of 53.5% and 57% respectively. The employee sample in branches comprises 16% males and 84% females, with a mean age of around 36 years, and an average organizational tenure of around 12.5 years. The profile of the respondents is representative of the typical employee profile in branches. Similarly, the sample in call centers comprises 36% males and 64% females. The mean age of the employees is around 30 years. The average organizational tenure is around 3.5 years. The similarity between population and sample profiles ensures that no significant non-response bias exists. 3.2. Measuring instruments In this study, frontline employees of the bank evaluated their own performance in terms of service quality on a shortened and adapted version of SERVQUAL (Parasuraman, Zeithaml, & Berry, 1988). Many studies have effectively used employees’ perceptions of service delivery (Babakus, Yavas, Karatepe, & Avci, 2003; Boshoff & Tait, 1996; Singh, 2000) in measuring performance. After extensive discussions with managers of call centers and branches, researchers selected the appropriate items from the dimensions of SERVQUAL that were applicable in measuring service quality in both encounter scenarios. In both cases, they selected only those items that pertained specifically to employee-related aspects of service quality (see Boshoff & Tait, 1996). A three-component scale (Meyer, Allen, & Smith, 1993) measures the three components of organizational commitment. The scale (18 items) has been extensively used by several researchers (Malhotra & Mukherjee, 2003, 2004) and has been well accepted for reliability and validity. A five point, Likert-type scale measures all items, ranging from strongly disagree to strongly agree. Researchers also collected data on the demographics of the respondents in terms of their age and total work experience. 4. Results This study standardizes the data (mean-centered with a mean of zero and a standard deviation of one) to remove the effects of different measurement units and potential multicollinearity (Jaccard & Turrisi, 2003). All the Chronbach's alphas are greater than 0.7 and therefore are acceptable (Nunnally, 1978). An exploratory factor analysis follows this standardization. The eight items that measure service quality yield one factor, but the three components of commitment emerge as three distinct factors. The study runs a confirmatory factor analysis (CFA), and the measurement model exhibits strong psychometric properties and acceptable fits (see Table 1). The study also tests common method variance (CMV). A correlation matrix of
N. Malhotra et al. / Journal of Business Research 66 (2013) 1338–1344
the constructs in this study shows that the smallest correlation is between AC and CC (0.002). This correlation becomes the conservative estimate for CMV (Lindell & Whitney, 2001). Adjusting all corrections for this estimated value of CMV shows that all the correlations that were previously significant remain significant (Malhotra, Kim, & Patil, 2006). Thus, CMV is not a problem in this data. The extrapolation procedure that Armstrong and Overton (1977) suggest assesses potential non-response bias. When assessing the first quartile versus the last quartile of the respondents in the sample group, no significant difference exists on any of the four summated measures in the survey (i.e., service quality, AC, NC, and CC), indicating that the data are free from systematic difference or non-response bias. 4.1. Testing for measure equivalence Measure equivalence of service quality is important to check because respondents from the two settings could possibly interpret service quality differently. First, the study establishes the psychometric properties of service quality in each sample separately. These properties are acceptable (see Table 1). Using common service quality items, the researchers then perform multiple group comparison in AMOS version 6. The first step is to test for configural equivalence that establishes that both samples map the same measure, that is, they map the same indicators and latent variables. Configural equivalence can not be rejected (χ 2 = 184.2, df = 40, Cmin/df = 4.6, RMSEA = 0.08; IFI = 0.91, CFI = 0.91). Testing for metric equivalence, that is, constraining the factor loadings to equality across the samples, shows that measure equivalence at this level again can not be rejected because the chi-square difference test is not significant (Δχ 2 = 196.9, Δdf = 47, p > 0.05). Imposing additional constraints of equality on intercepts leads to an acceptable, strong factorial equivalence (Δχ 2 = 189.76, Δdf = 48, p > 0.05). Support for the strong factorial
1341
equivalence indicates that any approach that specifies the model will not alter this finding (Vandenberg & Lance, 2000). Overall, these results suggest that support exists for measure equivalence that allows all comparisons and the interpretation of the findings to be valid (Byrne, 2004; Van Herk, Poortinga, & Verhallen, 2005). 4.2. Testing the predictions of profile deviation The study follows the literature on configuration theory and PD analysis to test the predictions. The first step in the PD analysis is to identify ideal employee-based profiles that can be used as benchmarks against which their fit can be examined (Doty et al., 1993; Vorhies & Morgan, 2003). To identify these profiles, the study examines the frequencies of the outcome variable (service quality) coupled with the general guidelines in PD studies of selecting about 10% of the cases to be included in the ideal profile (Venkatraman & Prescott, 1990). The cutoff point becomes the top 10% of the employees where a significant drop-off in service quality is apparent (resulting each ideal profile comprising a range of 17 to 34 cases). One concern is that these tests might be self-fulfilling prophecies because one dataset provides both the ideal profiles and the rest of the data. To be clear, the development of the profiles comes from employees’ perceptions of their own service quality. The hypotheses in essence examine the expectation that the closer the perceptions of commitment of a firm's employees, in general, align with those held by the firm's best employees; the stronger the perception of their own service quality. To test for PD, the development of an empirically derived calibration sample comes from the top 10% of employees rank-ordered by service quality (Venkatraman & Prescott, 1990; Vorhies & Morgan, 2003). The calibration sample uses the top 10% of employees showing the highest service quality to identify the ideal profile in terms of the
Table 1 Instrument items and psychometric properties. BRANCHES ITEM Affective Commitment I would be happy to spend the rest of my career with this organization. I really feel as if this organization's problems are my own.* I do not feel a strong sense of 'belonging' to my organization (R). I do not feel 'emotionally attached' to this organization (R). I do not feel like 'part of the family' at my organization (R). This organization has a great deal of personal meaning for me. Normative Commitment I don't feel any obligation to remain with my current employer (R). Even if were to my advantage, I don't feel right to leave my organization now. I would feel guilty if I left my organization now. This organization deserves my loyalty. I would not leave my organization right now because I have a sense of obligation to the people in it. I owe a great deal to my company. Continuance Commitment Right now, staying with my organization is a matter of necessity as much as desire. It would be very hard for me to leave my organization right now, even if I wanted to. Too much in my life would be disrupted if I decided I wanted to leave my organization now. I feel that I have too few options to consider leaving this organization. If I had not already put so much of myself into this organization, I might consider working elsewhere. One of the few negative consequences of leaving this organization would be the scarcity of available alternatives. Service Quality When I promise a customer that I will do something by a certain time, I do so. I perform the service right the first time. When problems occur, I give them all my attention in an effort to solve them speedily. I am never too busy to respond to the requests of my customers. I treat all customers courteously. I have the knowledge and ability to answer customers' questions. When a customer has a problem, I provide him/her with individual attention. My behavior instils confidence in my customers.
Factor Loading
CALL CENTERS
α
C.R.
AVE
0.83
0.83
0.52
Factor Loading
0.54
0.68
0.79 0.81 0.80 0.59
0.68 0.76 0.79 0.58 0.89
0.88
0.55
0.65 0.88 0.87 0.78 0.58 0.63
C.R.
AVE
0.82
0.83
0.50
0.79 0.81 0.80 0.59 0.78
0.77
0.51
0.86
0.85
0.50
0.89
0.89
0.50
0.51 0.81 0.80 0.62 0.40 0.45 0.84
0.83
0.48
0.72 0.55 0.64 0.79 0.72 0.69
0.61 0.64 0.74 0.73 0.82 0.70 0.88
0.72 0.62 0.80 0.58 0.74 0.67 0.78 0.71
α
0.88
0.50 0.69 0.64 0.67 0.67 0.71 0.65 0.76 0.82
CFA Fit Statistics: IFI = 0.91; CFI = 0.91, TLI = 0.90; RMSEA = 0.04; CMIN/DF = 1.95 (R) = reverse-coded item; *deleted from further analysis due to poor factor loading; α: Cronbach's Alpha; CR: Construct Reliability; AVE: Average Variance Extracted
1342
N. Malhotra et al. / Journal of Business Research 66 (2013) 1338–1344
three commitment components and the demographics. Thus, the benchmark for the comparisons of the rest of the sample is the calibration sample. The calculation of the Euclidian distance is according to the formula (Venkatraman & Prescott, 1990): 2 MISALIGN ¼ ∑ bj X sj −X cj ;
ð1Þ
j¼1
where MISALIGN = Degree of misfit, X are aspects of commitment and demographics, and X-bar is the mean of the X for the calibration sample (or the mean of the non-ideal 10% random sample). A regression model tests for PD. Several advantages exist from using PD like the ability to add more variables without leading to increased demand on sample size; thus, yielding a more holistic description of the ideal service employee. Developing a more holistic profile can lead to superior employee selection and can be used for rewarding outstanding employees on some objective measure. Alternative non-ideal models, where the average employees are the benchmark for PD scores, also examine whether the ideal profiles provide strong explanatory power. Thus, to further test the robustness of the model, this study uses a random sample of 10% of the employees as another calibration sample (non-ideal profile). This method tests whether deviation from a random sample of employees has performance implications. The resulting PD score substitutes for the one from the top 10% of performers in order to re-estimate the model. If the PD (random) is not significant, then this result provides further evidence of the robustness of the test and provides more support for the research proposition in this paper. The results support the concept that calibrating ideal employee profiles produces stronger PD coefficients and greater explanatory power than the random benchmarks (see Chow, 1960; Cohen, Cohen, West, & Aiken, 2003). The results in Table 2 indicate that PD for the face-to-face employees is significant and negative (β = − 0.3, p b 0.001). This finding supports the hypothesis that employees who deviate in their profile from the best performing employees deliver low levels of service quality. The variance in the model is 8.5% for service quality. This percentage might appear rather low but is consistent with comparable studies (Vorhies & Morgan, 2003). The random PD is nonsignificant (p > 0.1) and the regression coefficient is very small (β = −0.01). This finding provides further support for the robustness of the proposed model. An alternative finding might be interpreted as meaning that deviation from any sample of 10% of the employees has an association with poor service quality and is contrary to the theory. However, the results in Table 2 for the call centers show that PD is not significant. Examining the profile of the top employees for the face-to-face encounters in branches shows that they have the following characteristics: moderate to high AC, moderate to high NC, and comparatively low CC. They have a reasonable amount of experience and appear to be middle aged at an average of 40. Females predominantly staff this service delivery mode as noted in the demographics. On the other hand, for the call centers, the top performers have a similar commitment profile with relatively high AC, moderate NC, and low CC. They have little experience relative to those in face-to-face encounters and are significantly younger by comparison. Thus, the
Table 2 Profile deviation: service quality – face-to-face (branches) & telephone (call centers). Variable
Profile Deviation R Sqr. Adj. R Sqr. F-ratio ***p b 0.00.
Face-to-Face ( Branches )
Telephone ( Call Centers )
10%
Random
10%
− 0.31 (− 3.58***) 0.09 0.07 6.44***
− 0.02 (− 0.15) 0 − 0.02 0.03
0.007 (0.13) 0.05 (0.60) 0.01 0.01 0.01 − 0.01 1.17 0.90
Random
results in Table 3 seem to suggest that no significant differences exist in commitment among the best performers in call centers and face-to-face contact. However, significant differences exist in total experience (p b 0.001) and age (p b 0.001). But, significant differences exist in the commitment structures of employees whose service quality perceptions rank among the bottom 10%. Of specific interest is the result that, although AC is the dominant driver of the super achievers, CC seems to work most for the weak providers. In both call centers as well as branches, AC levels drop significantly between the top and bottom 10% of employees, while CC increases. Even though the poor performers on service quality intend to stay with their organizations, service managers need to train, motivate, and assess them regularly so that they can be of some value to their organizations. 4.3. T-test results This study also conducts a t-test to see if significant differences exist between the top, bottom, and random 10% of employees’ AC, NC, and CC across the sample groups. Only the AC of random employees in call centers is significantly less than the AC of random employees in bank branches at p b 0.05. However, others display no significant differences. 5. Discussion In both call centers and branches, similar profiles of top performers exist in terms of the three components of commitment. Affective commitment is the strongest; normative and continuance commitment follow respectively. These results support the arguments presented in the literature (Allen & Meyer, 1990; Malhotra & Mukherjee, 2004; Meyer et al., 2002) that employees who value organizational goals and identify with the organization are likely to perform better than employees who merely stay under an obligation (normative) or a particular need (continuance). The results further confirm that the continuance commitment dominates the profiles in both contexts for the bottom 10% of performers (Table 4). Thus, the findings indicate that the desirable commitment profile should comprise more affective commitment and less continuance commitment, irrespective of the context. The results also show that profile deviation is significant in branches but not in call centers (Table 2). This finding implies that ideal employees deliver significantly superior service quality than non-ideal employees in branches, but not in call centers. Possibly, this is because in call center environments, frontline work is very routine and leads to excessive standardization that, along with intensive training and electronic monitoring, ensures minimum deviation in standards of performance. As compared to call centers, frontline work in branches is less standardized, more satisfying, and social due to the nature of customer participation in these face-to-face encounters. Good service quality delivery in face-to-face encounters places a great onus on employees’ discretionary efforts and demands much more than just formally acquired skills. These discretionary efforts provide an opportunity to the more committed employees to use their initiative, experience, and ability in order to truly stand out as better performers, as explained by the significant PD results from branches in this study. Although no differences exist in the commitment profiles of top performers, significant differences do exist in total experience (p b 0.001) and age (p b 0.001) between top performers in branches and call centers. Also, the average organizational tenure in the case of call centers is only 3.5 years as compared to 12.5 years in the case of frontline employees working in the branches. Possibly, turnover problems continue to hit the call-center industry and thus limit advancement opportunities. So, the call centers generally attract
N. Malhotra et al. / Journal of Business Research 66 (2013) 1338–1344
1343
Table 3 Profile of top 10% employees. Table profile 10% - mean value between telephone and face-to-face encounters. Variable
Telephone (Call Centers) N = 34
Face-to-Face (Branches) N = 17
F-Ratio
Affective commitment Normative commitment Continuance commitment Total experience Age
3.45 3.06 2.97 10.49 31.53
3.23 3.03 2.78 22.29 40.86
0.94 0.02 0.46 19.65*** 12.87***
***p b 0.001.
Table 4 Profile of bottom 10% employees. Table profile 10% - mean value between telephone and face-to-face encounters. Variable
Telephone (Call Centers) N = 34
Face-to-Face (Branches) N = 17
F-Ratio
Affective commitment Normative commitment Continuance commitment Total experience Age
2.25 2.53 3.07 7.14 29.25
2.72 2.69 2.90 12.94 30.20
4.74* 0.39 0.84 33.01*** 61.30***
*p b 0.05. ***p b 0.001.
younger people, which explains the significant differences in the two samples regarding demographics.
6. Managerial implications and future research As multi-channel service delivery strategies become more popular, organizations need to better understand the right profile for frontline employees that is suitable in different types of customercontact encounters. The results of this study have implications for internal marketing strategists on how to use various rewards and to develop programs that encourage the right commitment profile among their employees. Organizations might be more prudent to foster affective commitment in their frontline employees as affective commitment dominates the commitment profiles of the top service quality performers in this study. Hence, internal marketing strategies should incorporate work variables like role clarity, autonomy, participation in decision making, training, feedback, and job satisfaction to encourage employees’ affective commitment (Malhotra et al., 2007; Meyer et al., 2002). Further, management needs to pay special attention to the profile of frontline staff dealing with customers in face-to-face encounters, because greater variations in service quality delivery exist depending on their commitment and demographic profile. Hence, especially in face-to-face encounters where management has little control over the behaviors of employees and where the discretionary effort matters in delivering excellent service quality, the right commitment profile needs to be encouraged and developed among frontline staff. The results also offer some interesting insights for the recruitment strategies of organizations as the profiles of top performers in face-toface and telephone encounters differ significantly in terms of age and experience. Thus, organizations need to establish different selection criteria for these two types of encounters. A worthwhile investment for organizations would be to incorporate and develop suitable selection mechanisms in their recruitment system that can decipher the psychological profiles of the candidates in terms of their commitment before final recruitment. Also, internal marketing strategies could be devised to incorporate employee rotation between the two types of encounters within the same organization according to employeeencounter fit as their ideal profiles demonstrate. However, because this is a single bank study, the generalizability of results should be tested in the future. The study should be replicated in other contexts using other indicators of performance, like profitability and customer satisfaction. A variety of ways exist in which the calibration sample can be drawn; for example, outstanding
service quality performers can be selected and their commitment or other variables computed to identify ideal profiles. Also, the calibration sample can be taken at any level, which could improve the coefficient of determination, R 2. Some researchers prefer the top 5% while others prefer the top 20%. These can be adjusted to fit with the purpose of the exercise. Future research can also empirically compare profile deviation methods with existing statistical techniques. References Allen, N. J., & Meyer, J. P. (1990). The measurement and antecedents of affective, continuance and normative commitment. Journal of Occupational Psychology, 63(1), 1–18. Armstrong, J. S., & Overton, T. S. (1977). Estimating non-response bias in mail surveys. Journal of Marketing Research, 14, 396–402 (August). Babakus, E., Yavas, U., Karatepe, O., & Avci, T. (2003). The effect of management commitment to service quality on employees’ affective and performance outcome. Journal of the Academy of Marketing Science, 31(3), 272–286. Boshoff, C., & Tait, M. (1996). Quality perceptions in the financial services sector: The potential impact of internal marketing. International Journal of Service Industry Management, 17(5), 5–31. Burgers, A., Ruyter, K., Keen, C., & Streukens, S. (2000). Customer expectation dimensions of voice-to-voice service encounters: A scale-development study. International Journal of Service Industry Management, 11(2), 142–161. Burmann, C., & Konig, V. (2011). Does internal brand management really drive brand commitment in shared-service call centers? Journal of Brand Management, 18(6), 374–393. Byrne, B. M. (2004). Testing for multi-group invariance using AMOS graphics: A road less travelled. Structural Equation Modelling, 11(2), 272–300. Chen, Y., Huang, H., Sung, S., & Huang, W. (2009). The role of fit on business performance: A profile deviation analysis. 2009 International Conference on New Trends in Information and Service Science (pp. 235–239). Chow, G. C. (1960). Test of equality between sets of coefficients in two linear equations. Econometrics, 28(3), 591–605. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple reg./cor. analysis for the behavioral sciences (3rd ed.). Mahwah, New Jersey: Lawrence Erlbaum Ass. Publishers. Culpepper, R. A., Gamble, J. E., & Blubaugh, M. G. (2004). Employee stock ownership plans and three-component commitment. Journal of Occupational and Organizational Psychology, 77(2), 155–166. Doty, H. D., Glick, W. H., & Huber, G. P. (1993). Fit equifinality and organizational effectiveness: A test of two configurational theories. Academy of Management Journal, 36(6), 1196–1250. Drazin, R., & Van de Ven, A. H. (1985). Alternative forms of fit in contingency theory. Administrative Science Quarterly, 30(4), 514–539. Fang, E., Palmatier, R. W., & Grewal, R. (2011). Effects of customer and innovation asset configuration strategies on firm performance. Journal of Marketing Research, 48(3), 587–602. Gustaffson, A. (2009). Customer satisfaction with service recovery. Journal of Business Research, 62(8), 1220–1222. Hult, G. T. M., Boyer, K. K., & Ketchen, D. J., Jr. (2007). Quality, operational logistics strategy, and repurchase intentions: A profile deviation analysis. Journal of Business Logistics, 28(2), 105–132. Jaccard, J., & Turrisi, R. (2003). Interaction effects in multiple regression (2nd ed.). Thousand Oaks, California: Sage Publications Inc.
1344
N. Malhotra et al. / Journal of Business Research 66 (2013) 1338–1344
Kabadayi, S., Eyuboglu, N., & Thomas, G. (2007). The performance implications of designing distribution channels to fit with strategy and environment. Journal of Marketing, 71(10), 195–211. Lindell, M. K., & Whitney, D. J. (2001). Accounting for common method variance in cross sectional research designs. Journal of Applied Psychology, 86, 114–121. Lings, I. N. (2004). Internal market orientation: Construct and consequences. Journal of Business Research, 57(4), 405–413. Malhotra, N., Budhwar, P., & Prowse, P. (2007). Linking rewards to commitment: An empirical investigation of four UK call centres. International Journal of Human Resource Management, 18(12), 2095–2127. Malhotra, N. K., Kim, S. S., & Patil, A. (2006). Common method variance in IS research: A comparison of alternative approaches and a reanalysis of past research. Management Science, 52(12), 1865–1883. Malhotra, N., & Mukherjee, A. (2003). Analysing the commitment – service quality relationship: A comparative study of retail banking call centres and branches. Journal of Marketing Management, 19(9/10), 941–997. Malhotra, N., & Mukherjee, A. (2004). The relative influence of organizational commitment and job satisfaction on service quality of customer-contact employees in banking call centres. Journal of Services Marketing, 18(3), 162–174. Meyer, J. P., Allen, N. J., & Smith, C. A. (1993). Commitment to organisations and occupations: extension and test of a three-component conceptualisation. Journal of Applied Psychology, 78(4), 538–551. Meyer, J. P., Stanley, D. J., Herscovitch, L., & Topolnytsky, L. (2002). Affective, continuance and normative commitment to the organization: A meta-analysis of antecedents, correlates and consequences. Journal of Vocational Behavior, 61(1), 20–52. Nunnally, J. (1978). Psychometric theory (2nd ed.). New York, New York: McGraw-Hill. Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1988). SERVQUAL: A multiple item scale for measuring consumer perceptions of service quality. Journal of Retailing, 64(1), 12–40.
Schoonhoven, C. B. (1981). Problems with contingency theory: Testing assumptions hidden within the language of contingency "theory". Administrative Science Quarterly, 26(3), 349–377. Singh, J. (2000). Performance productivity and quality of frontline employees in service organizations. Journal of Marketing, 64, 15–34 (April). Smith, K. G., Guthrie, J. P., & Chen, M. J. (1989). Strategy, size and performance. Organization Studies, 10(1), 63–81. Van Herk, V. H., Poortinga, Y. H., & Verhallen, T. M. M. (2005). Equivalence of survey data: Relevance for international marketing. European Journal of Marketing, 39(3/4), 351–364. Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of the measurement invariance literature: Suggestions, practices, and recommendations for organizational research. Organizational Research Methods, 3(1), 4–70. Vandenberghe, C., Bentein, K., Michon, R., Chebat, J. C., Tremblay, M., & Fils, J. F. (2007). An examination of the role of perceived support and employee commitment in employee- customer encounters. Journal of Applied Psychology, 92(4), 1177–1187. Venkatraman, N. (1990). The concept of fit in strategy research: Toward verbal and statistical correspondence. Academy of Management Review, 27(1), 19–41. Venkatraman, N., & Prescott, J. E. (1990). Environment strategy co-alignment: An empirical test of its performance implications. Strategic Management Journal, 11, 1–23. Vorhies, D. W., & Morgan, N. A. (2003). Configuration theory assessment of marketing organization fit with business strategy and its relationship with marketing performance. Journal of Marketing, 11(1), 100–115. Wieseke, J., Ahearne, M., Lam, S. K., & VanDick, R. (2009). The role of leaders in internal marketing. Journal of Marketing, 73(1), 123–145. Zajac, E. J., Kraatz, M. S., & Bresser, R. K. F. (2000). Modeling the dynamics of strategic fit: A normative approach to strategic change. Strategic Management Journal, 21(4), 429–453.
Contents lists available at SciVerse ScienceDirect
Journal of Business Research
Service quality of frontline employees: A profile deviation analysis☆ Neeru Malhotra a,⁎, Felix Mavondo b, 1, Avinandan Mukherjee c, 2, Graham Hooley d, 3 a
Aston Business School, Aston University, Aston Triangle, Birmingham B4 7ET, UK Monash University, Clayton Building 11E, Room 272, Clayton Campus, Welington Rd., Clayton, 3168, Australia c Department of Marketing, School of Business, Montclair State University, Montclair, New Jersey, USA d Marketing Department, Aston University b
a r t i c l e
i n f o
Article history: Received 1 April 2011 Received in revised form 1 September 2011 Accepted 1 November 2011 Available online 3 March 2012 Keywords: Profile deviation analysis Configuration theory Bank branches Call centers Service Quality Commitment
a b s t r a c t Using a configuration theory approach, this paper conducts a comparative study between frontline employees in phone and face-to-face service encounters for a retail bank. The study compares the top performers in service quality in relation to three components of organizational commitment and their demographics by applying a profile deviation analysis. The results show that the profile deviation for face-to-face employees is significantly negative, while for call center employees nonsignificant. Although the study finds no significant differences in the three components of commitment, significant differences exist in the total experience and age of the best performers. Also, affective commitment dominates the profile of high performers, while poor service providers seem to exhibit a higher level of continuance commitment. This study demonstrates the utility of profile deviation approaches in designing internal marketing strategies. © 2012 Elsevier Inc. All rights reserved.
1. Introduction Organizations operating in today's highly competitive business environment need to differentiate on service quality as a means of achieving a competitive advantage, and frontline employees are central in determining this quality (Gustaffson, 2009). The frontline is the touch point of the company; therefore, the service that frontline employees provide is critical in developing customer relationships, gathering customer information, and in creating customer satisfaction, loyalty, and brand commitment (Burmann & Konig, 2011; Fang, Palmatier, & Grewal, 2011). Previous research indicates that the organizational commitment of frontline employees exerts a strong, positive influence on their service quality (e.g., Malhotra & Mukherjee, 2004; Vandenberghe et al., 2007). However, the actual commitment-profile differences between high and low quality service performers have not been substantively researched.
☆ The authors acknowledge the research support of Kyungwon Lee. ⁎ Corresponding author. Tel.: + 44 121 2043151; fax: + 44 121 3334917. E-mail addresses: [email protected] (N. Malhotra), [email protected] (F. Mavondo), [email protected] (A. Mukherjee), [email protected] (G. Hooley). 1 Tel.: + 61 3 990 59249. 2 Tel.: + 1 973 655 5126: fax: + 1 973 655 7673. 3 Tel.: + 44 (0) 121 204 4643. 0148-2963/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.jbusres.2012.02.034
Using a configuration theory approach, this study compares employees in the two most difficult and important types of frontline service situations; face-to-face and telephone encounters. First, this study identifies the profile of the top performers in each of the encounters. Second, this study makes comparisons between the ideal and non-ideal profiles to investigate whether deviations from the ideal result in a decrease in service quality in each context. Any difference this study finds among the profiles of ideal performers in the two contexts is useful to explore, especially for companies employing multi-channel delivery, in order to recruit and manage frontline employees effectively through appropriately designed internal marketing strategies (see Lings, 2004; Wieseke, Ahearne, Lam, & VanDick, 2009). This study offers four contributions to the services marketing literature. First, this study applies a profile deviation analysis with a basis in configurational theory, which is a methodological innovation because profile deviation has rarely been used in marketing. Profile deviation approaches provide significant advantages over traditional approaches such as regression analysis, slope analysis, and subgroup analysis, particularly with individuals as the unit of analysis, in assessing fit in a way that is consistent with the multidimensional and holistic perspective of services marketing (Vorhies & Morgan, 2003). Previous research mainly applied the configuration theory and the profile deviation approach to organizations as units of analysis to assess organizational performance (Chen, Huang, Sung, & Huang, 2009; Kabadayi, Eyuboglu, & Thomas, 2007). But, this study uses these concepts to predict individual behavior.
N. Malhotra et al. / Journal of Business Research 66 (2013) 1338–1344
Second, this study conducts an empirical test to see if deviations from the ideal profile result in a decrease in service quality. In this respect, an important question is how different the top performers in call centers are from those in bank branches in face-to-face and telephone types of encounters. Third, this study explores whether the form of commitment matters in the commitment-service quality relation. Previously, a majority of the studies in marketing adopted one-dimensional approach to organizational commitment while the implications of a three-component model of commitment that comprises affective, normative, and continuance commitment, (Allen & Meyer, 1990) have remained largely unexplored (Culpepper, Gamble, & Blubaugh, 2004; Malhotra, Budhwar, & Prowse, 2007). Affective commitment (AC) is the extent of an employee's emotional attachment to, identification with, and involvement in the organization; normative commitment (NC) denotes an employee's feelings of obligation to stay with the organization; and continuance commitment (CC) is the commitment based on the costs that the employee associates with leaving the organization. From the literature, AC seems to have more of a relation to service quality than NC and CC (Meyer, Stanley, Herscovitch, & Topolnytsky, 2002; Vandenberghe et al., 2007). Because not all forms of commitment necessarily have an association with high job performance, the profile of top service quality performers needs to be understood with respect to the three components of commitment. Fourth, this study also investigates whether the form of service delivery matters for performance fit by comparing face-to-face and phone services. This study integrates the service-quality literature with the profile-deviation analysis literature to provide valuable new insights into the theory and practice of service excellence, and is the first to attempt to understand the role of configuration theory in designing internal marketing strategies. 2. Configuration theory Configuration theory has been used by management over the last two decades to assess complex, multidimensional phenomena implied in fit or congruence relations in ways that are more consistent with the holistic framing of strategic management and marketing strategy than traditional approaches like interactions or contingency theory. Traditional approaches lack correspondence between verbal and statistical approaches when testing the theory. This lack of correspondence means that a weak link exists between theory building and theory testing that leads to inconsistent research findings. Configurational theories posit higher effectiveness for employees that resemble the ideal type that theory defines. The increased effectiveness comes from an internal consistency or fit among the relevant structural and strategic factors (Doty, Glick, & Huber, 1993). Researchers who treat configurations as categories rather than the ideal types fail to test the core thesis of the theory. Many researchers have developed categories of employees by using cluster analysis and then comparing the means of the categories across effectiveness measures (Smith, Guthrie, & Chen, 1989). The appropriateness of this approach is questionable. When treating configurations as categories, they predict that marginal members of the category are as effective as the central members. On the other hand, when treating configurations as ideal types, they predict employees that marginally resemble the ideal configuration as less effective than those that closely resemble the ideal (Doty et al., 1993). Modeling the ideal-type employee should begin by recognizing that an ideal type is a theoretical construct, that is, a singular and discrete phenomenon rather than a nominal category. Hence, any empirical test should involve a rich multivariate approach to define the ideal type. Employees do not have to be classified into nominal groups because the crux of the matter is the deviation of the employee from the ideal type. The deviation measure (a Euclidian distance) can then be used to predict employee effectiveness in a way consistent with theory.
1339
Management researchers have conceptualized the configuration model as an interaction, selection, or systems approach to fit (Drazin & Van de Ven, 1985). The interaction approach is the basis for many contingency theories that define fit as the statistical interaction of two variables (Schoonhoven, 1981), although population ecologists adopt the selection approach to develop taxonomies. Neither of these approaches is consistent with the complex fit assertions in configurational theories (Venkatraman & Prescott, 1990). Thus, this study conceptualizes fit through a systems approach, which is the most appropriate. Drazin and Van de Ven describe this approach as the most complex and promising approach for future research. The systems approach includes profile deviation and gestalt and defines fit in terms of consistency among multiple dimensions (see Kabadayi et al., 2007). Fit is high to the extent that an employee is similar to an ideal type along many dimensions, and effectiveness is highest in the ideal types because the fit among factors is at a maximum in those configurations. When applying fit to employees, an opportunity exists to provide a more holistic profile that leads to superior decisions in selecting and rewarding outstanding employees. When considering fit among multiple elements simultaneously and examining the effects on outcomes, as this study does, configuration should be conceptualized and measured via profile deviation (PD) analysis (Doty et al., 1993; Venkatraman, 1990; Vorhies & Morgan, 2003). PD analysis views fit as the degree to which a particular case (a customer contact employee, in this study) matches an ideal profile (an optimal standing within a dataset) (Hult, Boyer, & Ketchen, 2007; Venkatraman, 1990; Zajac, Kraatz, & Bresser, 2000). The PD analysis in this study (Fig. 1) assesses the fit between the three components of commitment and service quality as the degree to which the commitment and demographic characteristics of a frontline employee differ from those of an ideal profile in achieving service quality (Vorhies & Morgan, 2003; Zajac et al., 2000). Employees whose profiles the innermost circle (A) represents are the ideal type of employee who delivers the highest service quality to customers. Employees that inner circle (B) represents deliver lower quality than those in A but higher than those in C; and those in C, in turn, deliver better service quality than those in circle D (Fig. 2). The prediction of PD is that, as the Euclidian distance increases from the ideal, service quality deteriorates. Thus, the more a frontline employee is like the ideal profile (definition from AC, NC, and CC components and the total experience and age demographics), the more superior the service quality is that the employee delivers. Hence, the key proposition relating the profiles of service employees and the outcome of interest (i.e., service quality) is: Proposition 1. The ideal type of employees will deliver significantly superior service quality than non-ideal employees. With the growth in multi-channel strategies among services, the two types of encounters involving frontline employees, face-to-face and telephone are becoming increasingly crucial to manage. However, several differences exist in customer service delivery between face-to-face and telephone encounters that have a bearing on the type of frontline employees suited for these encounters. In face-to-face encounters, both verbal and non-verbal behaviors (e.g., employee physical appearance and dress) are important determinants of service quality; the customer plays a role in creating quality service through his/her behavior during the interaction, and people can create quality perceptions relating to the environment where the service takes place (Burgers, Ruyter, Keen, & Streukens, 2000). On the other hand, in phone encounters, the customer has less influence on service quality. The service environment and tangibles are not part of the quality perceptions; their judgement of service and quality comes purely from intrinsic dimensions like reliability, responsiveness, assurance, and empathy (Boshoff & Tait, 1996; Burgers et al., 2000). Further, performance in branches involves less scripting and standardization as compared to call centers, which provides an opportunity to the more able employees
1340
N. Malhotra et al. / Journal of Business Research 66 (2013) 1338–1344
Organizational Commitment • Affective • Normative • Continuance
Employee profile fit with ideal profile Deviation from an ideal employee profile leads to inferior service quality
Service Quality
Demographics • Total experience • Age
Fig. 1. A model of profile deviation for customer-contact employees.
to use their initiative, experience, and commitment to stand out. Keeping this in mind, this research proposes: Proposition 2. The ideal type of employees will differ significantly with respect to their commitment profile and demographics in the two types of service encounters.
3. The methodology 3.1. Sample This research comes from telephone call centers and branches of a major retail bank in the United Kingdom. Because the perception is that most financial products are high-involvement, complex, and low on differentiation, the frontline employee, and not the service itself, provides a source of differentiation and creates competitive advantage (Burgers et al., 2000). Hence, this study measures service quality from the viewpoint of the frontline employee. Researchers mailed self-administered anonymous surveys to the heads of customer services and the regional branch managers who then arranged for further distribution to their respective frontline employees. The researchers used stratified sampling to ensure that no selection bias exists in the sample. They also provided selfaddressed pre-paid envelopes along with the questionnaires, and the employees returned the completed questionnaires directly to one of the researchers. Six hundred and forty employees in the call centers and 300 employees in the branches received the questionnaires. This distribution
D C B A
A: The ideal type of employees who delivers highest service quality to customers. B: Employees who deliver lower service quality than A. C: Employees who deliver better service quality than D, but lower service quality than B. D: Employees who deliver lower service quality than C. Fig. 2. Profile deviation as concentric circles. A: The ideal type of employees who delivers highest service quality to customers. B: Employees who deliver lower service quality than A. C: Employees who deliver better service quality than D, but lower service quality than B. D: Employees who deliver lower service quality than C.
included all employees who satisfied the initial profile screening. This distribution yielded 342 useable questionnaires from the call centers and 170 useable questionnaires from the branches, generating net response rates of 53.5% and 57% respectively. The employee sample in branches comprises 16% males and 84% females, with a mean age of around 36 years, and an average organizational tenure of around 12.5 years. The profile of the respondents is representative of the typical employee profile in branches. Similarly, the sample in call centers comprises 36% males and 64% females. The mean age of the employees is around 30 years. The average organizational tenure is around 3.5 years. The similarity between population and sample profiles ensures that no significant non-response bias exists. 3.2. Measuring instruments In this study, frontline employees of the bank evaluated their own performance in terms of service quality on a shortened and adapted version of SERVQUAL (Parasuraman, Zeithaml, & Berry, 1988). Many studies have effectively used employees’ perceptions of service delivery (Babakus, Yavas, Karatepe, & Avci, 2003; Boshoff & Tait, 1996; Singh, 2000) in measuring performance. After extensive discussions with managers of call centers and branches, researchers selected the appropriate items from the dimensions of SERVQUAL that were applicable in measuring service quality in both encounter scenarios. In both cases, they selected only those items that pertained specifically to employee-related aspects of service quality (see Boshoff & Tait, 1996). A three-component scale (Meyer, Allen, & Smith, 1993) measures the three components of organizational commitment. The scale (18 items) has been extensively used by several researchers (Malhotra & Mukherjee, 2003, 2004) and has been well accepted for reliability and validity. A five point, Likert-type scale measures all items, ranging from strongly disagree to strongly agree. Researchers also collected data on the demographics of the respondents in terms of their age and total work experience. 4. Results This study standardizes the data (mean-centered with a mean of zero and a standard deviation of one) to remove the effects of different measurement units and potential multicollinearity (Jaccard & Turrisi, 2003). All the Chronbach's alphas are greater than 0.7 and therefore are acceptable (Nunnally, 1978). An exploratory factor analysis follows this standardization. The eight items that measure service quality yield one factor, but the three components of commitment emerge as three distinct factors. The study runs a confirmatory factor analysis (CFA), and the measurement model exhibits strong psychometric properties and acceptable fits (see Table 1). The study also tests common method variance (CMV). A correlation matrix of
N. Malhotra et al. / Journal of Business Research 66 (2013) 1338–1344
the constructs in this study shows that the smallest correlation is between AC and CC (0.002). This correlation becomes the conservative estimate for CMV (Lindell & Whitney, 2001). Adjusting all corrections for this estimated value of CMV shows that all the correlations that were previously significant remain significant (Malhotra, Kim, & Patil, 2006). Thus, CMV is not a problem in this data. The extrapolation procedure that Armstrong and Overton (1977) suggest assesses potential non-response bias. When assessing the first quartile versus the last quartile of the respondents in the sample group, no significant difference exists on any of the four summated measures in the survey (i.e., service quality, AC, NC, and CC), indicating that the data are free from systematic difference or non-response bias. 4.1. Testing for measure equivalence Measure equivalence of service quality is important to check because respondents from the two settings could possibly interpret service quality differently. First, the study establishes the psychometric properties of service quality in each sample separately. These properties are acceptable (see Table 1). Using common service quality items, the researchers then perform multiple group comparison in AMOS version 6. The first step is to test for configural equivalence that establishes that both samples map the same measure, that is, they map the same indicators and latent variables. Configural equivalence can not be rejected (χ 2 = 184.2, df = 40, Cmin/df = 4.6, RMSEA = 0.08; IFI = 0.91, CFI = 0.91). Testing for metric equivalence, that is, constraining the factor loadings to equality across the samples, shows that measure equivalence at this level again can not be rejected because the chi-square difference test is not significant (Δχ 2 = 196.9, Δdf = 47, p > 0.05). Imposing additional constraints of equality on intercepts leads to an acceptable, strong factorial equivalence (Δχ 2 = 189.76, Δdf = 48, p > 0.05). Support for the strong factorial
1341
equivalence indicates that any approach that specifies the model will not alter this finding (Vandenberg & Lance, 2000). Overall, these results suggest that support exists for measure equivalence that allows all comparisons and the interpretation of the findings to be valid (Byrne, 2004; Van Herk, Poortinga, & Verhallen, 2005). 4.2. Testing the predictions of profile deviation The study follows the literature on configuration theory and PD analysis to test the predictions. The first step in the PD analysis is to identify ideal employee-based profiles that can be used as benchmarks against which their fit can be examined (Doty et al., 1993; Vorhies & Morgan, 2003). To identify these profiles, the study examines the frequencies of the outcome variable (service quality) coupled with the general guidelines in PD studies of selecting about 10% of the cases to be included in the ideal profile (Venkatraman & Prescott, 1990). The cutoff point becomes the top 10% of the employees where a significant drop-off in service quality is apparent (resulting each ideal profile comprising a range of 17 to 34 cases). One concern is that these tests might be self-fulfilling prophecies because one dataset provides both the ideal profiles and the rest of the data. To be clear, the development of the profiles comes from employees’ perceptions of their own service quality. The hypotheses in essence examine the expectation that the closer the perceptions of commitment of a firm's employees, in general, align with those held by the firm's best employees; the stronger the perception of their own service quality. To test for PD, the development of an empirically derived calibration sample comes from the top 10% of employees rank-ordered by service quality (Venkatraman & Prescott, 1990; Vorhies & Morgan, 2003). The calibration sample uses the top 10% of employees showing the highest service quality to identify the ideal profile in terms of the
Table 1 Instrument items and psychometric properties. BRANCHES ITEM Affective Commitment I would be happy to spend the rest of my career with this organization. I really feel as if this organization's problems are my own.* I do not feel a strong sense of 'belonging' to my organization (R). I do not feel 'emotionally attached' to this organization (R). I do not feel like 'part of the family' at my organization (R). This organization has a great deal of personal meaning for me. Normative Commitment I don't feel any obligation to remain with my current employer (R). Even if were to my advantage, I don't feel right to leave my organization now. I would feel guilty if I left my organization now. This organization deserves my loyalty. I would not leave my organization right now because I have a sense of obligation to the people in it. I owe a great deal to my company. Continuance Commitment Right now, staying with my organization is a matter of necessity as much as desire. It would be very hard for me to leave my organization right now, even if I wanted to. Too much in my life would be disrupted if I decided I wanted to leave my organization now. I feel that I have too few options to consider leaving this organization. If I had not already put so much of myself into this organization, I might consider working elsewhere. One of the few negative consequences of leaving this organization would be the scarcity of available alternatives. Service Quality When I promise a customer that I will do something by a certain time, I do so. I perform the service right the first time. When problems occur, I give them all my attention in an effort to solve them speedily. I am never too busy to respond to the requests of my customers. I treat all customers courteously. I have the knowledge and ability to answer customers' questions. When a customer has a problem, I provide him/her with individual attention. My behavior instils confidence in my customers.
Factor Loading
CALL CENTERS
α
C.R.
AVE
0.83
0.83
0.52
Factor Loading
0.54
0.68
0.79 0.81 0.80 0.59
0.68 0.76 0.79 0.58 0.89
0.88
0.55
0.65 0.88 0.87 0.78 0.58 0.63
C.R.
AVE
0.82
0.83
0.50
0.79 0.81 0.80 0.59 0.78
0.77
0.51
0.86
0.85
0.50
0.89
0.89
0.50
0.51 0.81 0.80 0.62 0.40 0.45 0.84
0.83
0.48
0.72 0.55 0.64 0.79 0.72 0.69
0.61 0.64 0.74 0.73 0.82 0.70 0.88
0.72 0.62 0.80 0.58 0.74 0.67 0.78 0.71
α
0.88
0.50 0.69 0.64 0.67 0.67 0.71 0.65 0.76 0.82
CFA Fit Statistics: IFI = 0.91; CFI = 0.91, TLI = 0.90; RMSEA = 0.04; CMIN/DF = 1.95 (R) = reverse-coded item; *deleted from further analysis due to poor factor loading; α: Cronbach's Alpha; CR: Construct Reliability; AVE: Average Variance Extracted
1342
N. Malhotra et al. / Journal of Business Research 66 (2013) 1338–1344
three commitment components and the demographics. Thus, the benchmark for the comparisons of the rest of the sample is the calibration sample. The calculation of the Euclidian distance is according to the formula (Venkatraman & Prescott, 1990): 2 MISALIGN ¼ ∑ bj X sj −X cj ;
ð1Þ
j¼1
where MISALIGN = Degree of misfit, X are aspects of commitment and demographics, and X-bar is the mean of the X for the calibration sample (or the mean of the non-ideal 10% random sample). A regression model tests for PD. Several advantages exist from using PD like the ability to add more variables without leading to increased demand on sample size; thus, yielding a more holistic description of the ideal service employee. Developing a more holistic profile can lead to superior employee selection and can be used for rewarding outstanding employees on some objective measure. Alternative non-ideal models, where the average employees are the benchmark for PD scores, also examine whether the ideal profiles provide strong explanatory power. Thus, to further test the robustness of the model, this study uses a random sample of 10% of the employees as another calibration sample (non-ideal profile). This method tests whether deviation from a random sample of employees has performance implications. The resulting PD score substitutes for the one from the top 10% of performers in order to re-estimate the model. If the PD (random) is not significant, then this result provides further evidence of the robustness of the test and provides more support for the research proposition in this paper. The results support the concept that calibrating ideal employee profiles produces stronger PD coefficients and greater explanatory power than the random benchmarks (see Chow, 1960; Cohen, Cohen, West, & Aiken, 2003). The results in Table 2 indicate that PD for the face-to-face employees is significant and negative (β = − 0.3, p b 0.001). This finding supports the hypothesis that employees who deviate in their profile from the best performing employees deliver low levels of service quality. The variance in the model is 8.5% for service quality. This percentage might appear rather low but is consistent with comparable studies (Vorhies & Morgan, 2003). The random PD is nonsignificant (p > 0.1) and the regression coefficient is very small (β = −0.01). This finding provides further support for the robustness of the proposed model. An alternative finding might be interpreted as meaning that deviation from any sample of 10% of the employees has an association with poor service quality and is contrary to the theory. However, the results in Table 2 for the call centers show that PD is not significant. Examining the profile of the top employees for the face-to-face encounters in branches shows that they have the following characteristics: moderate to high AC, moderate to high NC, and comparatively low CC. They have a reasonable amount of experience and appear to be middle aged at an average of 40. Females predominantly staff this service delivery mode as noted in the demographics. On the other hand, for the call centers, the top performers have a similar commitment profile with relatively high AC, moderate NC, and low CC. They have little experience relative to those in face-to-face encounters and are significantly younger by comparison. Thus, the
Table 2 Profile deviation: service quality – face-to-face (branches) & telephone (call centers). Variable
Profile Deviation R Sqr. Adj. R Sqr. F-ratio ***p b 0.00.
Face-to-Face ( Branches )
Telephone ( Call Centers )
10%
Random
10%
− 0.31 (− 3.58***) 0.09 0.07 6.44***
− 0.02 (− 0.15) 0 − 0.02 0.03
0.007 (0.13) 0.05 (0.60) 0.01 0.01 0.01 − 0.01 1.17 0.90
Random
results in Table 3 seem to suggest that no significant differences exist in commitment among the best performers in call centers and face-to-face contact. However, significant differences exist in total experience (p b 0.001) and age (p b 0.001). But, significant differences exist in the commitment structures of employees whose service quality perceptions rank among the bottom 10%. Of specific interest is the result that, although AC is the dominant driver of the super achievers, CC seems to work most for the weak providers. In both call centers as well as branches, AC levels drop significantly between the top and bottom 10% of employees, while CC increases. Even though the poor performers on service quality intend to stay with their organizations, service managers need to train, motivate, and assess them regularly so that they can be of some value to their organizations. 4.3. T-test results This study also conducts a t-test to see if significant differences exist between the top, bottom, and random 10% of employees’ AC, NC, and CC across the sample groups. Only the AC of random employees in call centers is significantly less than the AC of random employees in bank branches at p b 0.05. However, others display no significant differences. 5. Discussion In both call centers and branches, similar profiles of top performers exist in terms of the three components of commitment. Affective commitment is the strongest; normative and continuance commitment follow respectively. These results support the arguments presented in the literature (Allen & Meyer, 1990; Malhotra & Mukherjee, 2004; Meyer et al., 2002) that employees who value organizational goals and identify with the organization are likely to perform better than employees who merely stay under an obligation (normative) or a particular need (continuance). The results further confirm that the continuance commitment dominates the profiles in both contexts for the bottom 10% of performers (Table 4). Thus, the findings indicate that the desirable commitment profile should comprise more affective commitment and less continuance commitment, irrespective of the context. The results also show that profile deviation is significant in branches but not in call centers (Table 2). This finding implies that ideal employees deliver significantly superior service quality than non-ideal employees in branches, but not in call centers. Possibly, this is because in call center environments, frontline work is very routine and leads to excessive standardization that, along with intensive training and electronic monitoring, ensures minimum deviation in standards of performance. As compared to call centers, frontline work in branches is less standardized, more satisfying, and social due to the nature of customer participation in these face-to-face encounters. Good service quality delivery in face-to-face encounters places a great onus on employees’ discretionary efforts and demands much more than just formally acquired skills. These discretionary efforts provide an opportunity to the more committed employees to use their initiative, experience, and ability in order to truly stand out as better performers, as explained by the significant PD results from branches in this study. Although no differences exist in the commitment profiles of top performers, significant differences do exist in total experience (p b 0.001) and age (p b 0.001) between top performers in branches and call centers. Also, the average organizational tenure in the case of call centers is only 3.5 years as compared to 12.5 years in the case of frontline employees working in the branches. Possibly, turnover problems continue to hit the call-center industry and thus limit advancement opportunities. So, the call centers generally attract
N. Malhotra et al. / Journal of Business Research 66 (2013) 1338–1344
1343
Table 3 Profile of top 10% employees. Table profile 10% - mean value between telephone and face-to-face encounters. Variable
Telephone (Call Centers) N = 34
Face-to-Face (Branches) N = 17
F-Ratio
Affective commitment Normative commitment Continuance commitment Total experience Age
3.45 3.06 2.97 10.49 31.53
3.23 3.03 2.78 22.29 40.86
0.94 0.02 0.46 19.65*** 12.87***
***p b 0.001.
Table 4 Profile of bottom 10% employees. Table profile 10% - mean value between telephone and face-to-face encounters. Variable
Telephone (Call Centers) N = 34
Face-to-Face (Branches) N = 17
F-Ratio
Affective commitment Normative commitment Continuance commitment Total experience Age
2.25 2.53 3.07 7.14 29.25
2.72 2.69 2.90 12.94 30.20
4.74* 0.39 0.84 33.01*** 61.30***
*p b 0.05. ***p b 0.001.
younger people, which explains the significant differences in the two samples regarding demographics.
6. Managerial implications and future research As multi-channel service delivery strategies become more popular, organizations need to better understand the right profile for frontline employees that is suitable in different types of customercontact encounters. The results of this study have implications for internal marketing strategists on how to use various rewards and to develop programs that encourage the right commitment profile among their employees. Organizations might be more prudent to foster affective commitment in their frontline employees as affective commitment dominates the commitment profiles of the top service quality performers in this study. Hence, internal marketing strategies should incorporate work variables like role clarity, autonomy, participation in decision making, training, feedback, and job satisfaction to encourage employees’ affective commitment (Malhotra et al., 2007; Meyer et al., 2002). Further, management needs to pay special attention to the profile of frontline staff dealing with customers in face-to-face encounters, because greater variations in service quality delivery exist depending on their commitment and demographic profile. Hence, especially in face-to-face encounters where management has little control over the behaviors of employees and where the discretionary effort matters in delivering excellent service quality, the right commitment profile needs to be encouraged and developed among frontline staff. The results also offer some interesting insights for the recruitment strategies of organizations as the profiles of top performers in face-toface and telephone encounters differ significantly in terms of age and experience. Thus, organizations need to establish different selection criteria for these two types of encounters. A worthwhile investment for organizations would be to incorporate and develop suitable selection mechanisms in their recruitment system that can decipher the psychological profiles of the candidates in terms of their commitment before final recruitment. Also, internal marketing strategies could be devised to incorporate employee rotation between the two types of encounters within the same organization according to employeeencounter fit as their ideal profiles demonstrate. However, because this is a single bank study, the generalizability of results should be tested in the future. The study should be replicated in other contexts using other indicators of performance, like profitability and customer satisfaction. A variety of ways exist in which the calibration sample can be drawn; for example, outstanding
service quality performers can be selected and their commitment or other variables computed to identify ideal profiles. Also, the calibration sample can be taken at any level, which could improve the coefficient of determination, R 2. Some researchers prefer the top 5% while others prefer the top 20%. These can be adjusted to fit with the purpose of the exercise. Future research can also empirically compare profile deviation methods with existing statistical techniques. References Allen, N. J., & Meyer, J. P. (1990). The measurement and antecedents of affective, continuance and normative commitment. Journal of Occupational Psychology, 63(1), 1–18. Armstrong, J. S., & Overton, T. S. (1977). Estimating non-response bias in mail surveys. Journal of Marketing Research, 14, 396–402 (August). Babakus, E., Yavas, U., Karatepe, O., & Avci, T. (2003). The effect of management commitment to service quality on employees’ affective and performance outcome. Journal of the Academy of Marketing Science, 31(3), 272–286. Boshoff, C., & Tait, M. (1996). Quality perceptions in the financial services sector: The potential impact of internal marketing. International Journal of Service Industry Management, 17(5), 5–31. Burgers, A., Ruyter, K., Keen, C., & Streukens, S. (2000). Customer expectation dimensions of voice-to-voice service encounters: A scale-development study. International Journal of Service Industry Management, 11(2), 142–161. Burmann, C., & Konig, V. (2011). Does internal brand management really drive brand commitment in shared-service call centers? Journal of Brand Management, 18(6), 374–393. Byrne, B. M. (2004). Testing for multi-group invariance using AMOS graphics: A road less travelled. Structural Equation Modelling, 11(2), 272–300. Chen, Y., Huang, H., Sung, S., & Huang, W. (2009). The role of fit on business performance: A profile deviation analysis. 2009 International Conference on New Trends in Information and Service Science (pp. 235–239). Chow, G. C. (1960). Test of equality between sets of coefficients in two linear equations. Econometrics, 28(3), 591–605. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple reg./cor. analysis for the behavioral sciences (3rd ed.). Mahwah, New Jersey: Lawrence Erlbaum Ass. Publishers. Culpepper, R. A., Gamble, J. E., & Blubaugh, M. G. (2004). Employee stock ownership plans and three-component commitment. Journal of Occupational and Organizational Psychology, 77(2), 155–166. Doty, H. D., Glick, W. H., & Huber, G. P. (1993). Fit equifinality and organizational effectiveness: A test of two configurational theories. Academy of Management Journal, 36(6), 1196–1250. Drazin, R., & Van de Ven, A. H. (1985). Alternative forms of fit in contingency theory. Administrative Science Quarterly, 30(4), 514–539. Fang, E., Palmatier, R. W., & Grewal, R. (2011). Effects of customer and innovation asset configuration strategies on firm performance. Journal of Marketing Research, 48(3), 587–602. Gustaffson, A. (2009). Customer satisfaction with service recovery. Journal of Business Research, 62(8), 1220–1222. Hult, G. T. M., Boyer, K. K., & Ketchen, D. J., Jr. (2007). Quality, operational logistics strategy, and repurchase intentions: A profile deviation analysis. Journal of Business Logistics, 28(2), 105–132. Jaccard, J., & Turrisi, R. (2003). Interaction effects in multiple regression (2nd ed.). Thousand Oaks, California: Sage Publications Inc.
1344
N. Malhotra et al. / Journal of Business Research 66 (2013) 1338–1344
Kabadayi, S., Eyuboglu, N., & Thomas, G. (2007). The performance implications of designing distribution channels to fit with strategy and environment. Journal of Marketing, 71(10), 195–211. Lindell, M. K., & Whitney, D. J. (2001). Accounting for common method variance in cross sectional research designs. Journal of Applied Psychology, 86, 114–121. Lings, I. N. (2004). Internal market orientation: Construct and consequences. Journal of Business Research, 57(4), 405–413. Malhotra, N., Budhwar, P., & Prowse, P. (2007). Linking rewards to commitment: An empirical investigation of four UK call centres. International Journal of Human Resource Management, 18(12), 2095–2127. Malhotra, N. K., Kim, S. S., & Patil, A. (2006). Common method variance in IS research: A comparison of alternative approaches and a reanalysis of past research. Management Science, 52(12), 1865–1883. Malhotra, N., & Mukherjee, A. (2003). Analysing the commitment – service quality relationship: A comparative study of retail banking call centres and branches. Journal of Marketing Management, 19(9/10), 941–997. Malhotra, N., & Mukherjee, A. (2004). The relative influence of organizational commitment and job satisfaction on service quality of customer-contact employees in banking call centres. Journal of Services Marketing, 18(3), 162–174. Meyer, J. P., Allen, N. J., & Smith, C. A. (1993). Commitment to organisations and occupations: extension and test of a three-component conceptualisation. Journal of Applied Psychology, 78(4), 538–551. Meyer, J. P., Stanley, D. J., Herscovitch, L., & Topolnytsky, L. (2002). Affective, continuance and normative commitment to the organization: A meta-analysis of antecedents, correlates and consequences. Journal of Vocational Behavior, 61(1), 20–52. Nunnally, J. (1978). Psychometric theory (2nd ed.). New York, New York: McGraw-Hill. Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1988). SERVQUAL: A multiple item scale for measuring consumer perceptions of service quality. Journal of Retailing, 64(1), 12–40.
Schoonhoven, C. B. (1981). Problems with contingency theory: Testing assumptions hidden within the language of contingency "theory". Administrative Science Quarterly, 26(3), 349–377. Singh, J. (2000). Performance productivity and quality of frontline employees in service organizations. Journal of Marketing, 64, 15–34 (April). Smith, K. G., Guthrie, J. P., & Chen, M. J. (1989). Strategy, size and performance. Organization Studies, 10(1), 63–81. Van Herk, V. H., Poortinga, Y. H., & Verhallen, T. M. M. (2005). Equivalence of survey data: Relevance for international marketing. European Journal of Marketing, 39(3/4), 351–364. Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of the measurement invariance literature: Suggestions, practices, and recommendations for organizational research. Organizational Research Methods, 3(1), 4–70. Vandenberghe, C., Bentein, K., Michon, R., Chebat, J. C., Tremblay, M., & Fils, J. F. (2007). An examination of the role of perceived support and employee commitment in employee- customer encounters. Journal of Applied Psychology, 92(4), 1177–1187. Venkatraman, N. (1990). The concept of fit in strategy research: Toward verbal and statistical correspondence. Academy of Management Review, 27(1), 19–41. Venkatraman, N., & Prescott, J. E. (1990). Environment strategy co-alignment: An empirical test of its performance implications. Strategic Management Journal, 11, 1–23. Vorhies, D. W., & Morgan, N. A. (2003). Configuration theory assessment of marketing organization fit with business strategy and its relationship with marketing performance. Journal of Marketing, 11(1), 100–115. Wieseke, J., Ahearne, M., Lam, S. K., & VanDick, R. (2009). The role of leaders in internal marketing. Journal of Marketing, 73(1), 123–145. Zajac, E. J., Kraatz, M. S., & Bresser, R. K. F. (2000). Modeling the dynamics of strategic fit: A normative approach to strategic change. Strategic Management Journal, 21(4), 429–453.