Articol Despre Personal

Journal of Business Research 66 (2013) 13381344

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, UKb Monash University, Clayton Building 11E, Room 272, Clayton Campus, Welington Rd., Clayton, 3168, Australiac Department of Marketing, School of Business, Montclair State University, Montclair, New Jersey, USAd Marketing Department, Aston University

The authors acknowledge the research support of K Corresponding author. Tel.: +44 121 2043151; fax:

E-mail addresses: [email protected] (N. [email protected] (F. Mavondo),[email protected] (A. Mukherjee), g.j.ho

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. Alldoi:10.1016/j.jbusres.2012.02.034

a b s t r a c t
a r t i c l e i n f o
Article history:Received 1 April 2011Received in revised form 1 September 2011Accepted 1 November 2011Available online 3 March 2012

Keywords:Profile deviation analysisConfiguration theoryBank branchesCall centersServiceQualityCommitment

Using a configuration theory approach, this paper conducts a comparative study between frontlineemployees in phone and face-to-face service encounters for a retail bank. The study compares the topperformers in service quality in relation to three components of organizational commitment and their demo-graphics by applying a profile deviation analysis. The results show that the profile deviation for face-to-faceemployees is significantly negative, while for call center employees nonsignificant. Although the study findsno significant differences in the three components of commitment, significant differences exist in the totalexperience and age of the best performers. Also, affective commitment dominates the profile of high per-formers, while poor service providers seem to exhibit a higher level of continuance commitment. Thisstudy 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 businessenvironment need to differentiate on service quality as a means ofachieving a competitive advantage, and frontline employees are cen-tral in determining this quality (Gustaffson, 2009). The frontline is thetouch point of the company; therefore, the service that frontline em-ployees provide is critical in developing customer relationships,gathering customer information, and in creating customer satisfac-tion, loyalty, and brand commitment (Burmann & Konig, 2011;Fang, Palmatier, & Grewal, 2011). Previous research indicates thatthe organizational commitment of frontline employees exerts astrong, positive influence on their service quality (e.g., Malhotra &Mukherjee, 2004; Vandenberghe et al., 2007). However, the actualcommitment-profile differences between high and low quality ser-vice performers have not been substantively researched.

yungwon Lee.+44 121 3334917.tra),

[email protected] (G. Hooley).

rights reserved.

Using a configuration theory approach, this study compares em-ployees in the two most difficult and important types of frontline ser-vice situations; face-to-face and telephone encounters. First, thisstudy identifies the profile of the top performers in each of the en-counters. Second, this study makes comparisons between the idealand non-ideal profiles to investigate whether deviations from theideal result in a decrease in service quality in each context. Any differ-ence this study finds among the profiles of ideal performers in thetwo contexts is useful to explore, especially for companies employingmulti-channel delivery, in order to recruit and manage frontline em-ployees effectively through appropriately designed internal market-ing strategies (see Lings, 2004; Wieseke, Ahearne, Lam, & VanDick,2009).

This study offers four contributions to the services marketing liter-ature. First, this study applies a profile deviation analysis with a basisin configurational theory, which is a methodological innovation be-cause profile deviation has rarely been used in marketing. Profiledeviation approaches provide significant advantages over traditionalapproaches such as regression analysis, slope analysis, and subgroupanalysis, particularly with individuals as the unit of analysis, in asses-sing fit in a way that is consistent with the multidimensional and ho-listic perspective of services marketing (Vorhies & Morgan, 2003).Previous research mainly applied the configuration theory and theprofile deviation approach to organizations as units of analysis to as-sess organizational performance (Chen, Huang, Sung, & Huang, 2009;Kabadayi, Eyuboglu, & Thomas, 2007). But, this study uses these con-cepts to predict individual behavior.
http://dx.doi.org/10.1016/j.jbusres.2012.02.034mailto:[email protected]:[email protected]:[email protected]:[email protected]://dx.doi.org/10.1016/j.jbusres.2012.02.034http://www.sciencedirect.com/science/journal/01482963

1339N. Malhotra et al. / Journal of Business Research 66 (2013) 13381344

Second, this study conducts an empirical test to see if deviationsfrom the ideal profile result in a decrease in service quality. In this re-spect, an important question is how different the top performers incall centers are from those in bank branches in face-to-face and tele-phone types of encounters.

Third, this study explores whether the form of commitment mat-ters in the commitment-service quality relation. Previously, amajority of the studies in marketing adopted one-dimensional ap-proach to organizational commitment while the implications of athree-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, identifi-cation with, and involvement in the organization; normative commit-ment (NC) denotes an employee's feelings of obligation to stay withthe organization; and continuance commitment (CC) is the commit-ment based on the costs that the employee associates with leavingthe organization. From the literature, AC seems to have more of a re-lation to service quality than NC and CC (Meyer, Stanley, Herscovitch,& Topolnytsky, 2002; Vandenberghe et al., 2007). Because not allforms of commitment necessarily have an association with high jobperformance, the profile of top service quality performers needs tobe understood with respect to the three components of commitment.

Fourth, this study also investigates whether the form of servicedelivery matters for performance fit by comparing face-to-face andphone services.

This study integrates the service-quality literature with theprofile-deviation analysis literature to provide valuable new insightsinto the theory and practice of service excellence, and is the first to at-tempt to understand the role of configuration theory in designing in-ternal marketing strategies.

2. Configuration theory

Configuration theory has been used by management over the lasttwo decades to assess complex, multidimensional phenomena im-plied in fit or congruence relations in ways that are more consistentwith the holistic framing of strategic management and marketingstrategy than traditional approaches like interactions or contingencytheory. Traditional approaches lack correspondence between verbaland statistical approaches when testing the theory. This lack of corre-spondence means that a weak link exists between theory buildingand theory testing that leads to inconsistent research findings.

Configurational theories posit higher effectiveness for employeesthat resemble the ideal type that theory defines. The increased effective-ness comes froman internal consistency or fit among the relevant struc-tural and strategic factors (Doty, Glick, &Huber, 1993). Researcherswhotreat configurations as categories rather than the ideal types fail to testthe core thesis of the theory. Many researchers have developed catego-ries of employees by using cluster analysis and then comparing themeans 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 marginalmembers of the category are as effective as the central members. Onthe other hand,when treating configurations as ideal types, they predictemployees that marginally resemble the ideal configuration as less ef-fective than those that closely resemble the ideal (Doty et al., 1993).

Modeling the ideal-type employee should begin by recognizing thatan ideal type is a theoretical construct, that is, a singular and discrete phe-nomenon rather than a nominal category. Hence, any empirical testshould involve a rich multivariate approach to define the ideal type. Em-ployees do not have to be classified into nominal groups because the cruxof thematter is the deviation of the employee from the ideal type. The de-viation measure (a Euclidian distance) can then be used to predict em-ployee effectiveness in a way consistent with theory.

Management researchers have conceptualized the configurationmodel as an interaction, selection, or systems approach to fit(Drazin & Van de Ven, 1985). The interaction approach is the basisfor many contingency theories that define fit as the statistical interac-tion of two variables (Schoonhoven, 1981), although population ecol-ogists adopt the selection approach to develop taxonomies. Neither ofthese approaches is consistent with the complex fit assertions in con-figurational theories (Venkatraman & Prescott, 1990). Thus, this studyconceptualizes fit through a systems approach, which is the most ap-propriate. Drazin and Van de Ven describe this approach as the mostcomplex and promising approach for future research. The systems ap-proach includes profile deviation and gestalt and defines fit in termsof consistency among multiple dimensions (see Kabadayi et al.,2007). Fit is high to the extent that an employee is similar to anideal type along many dimensions, and effectiveness is highest inthe ideal types because the fit among factors is at a maximum inthose configurations. When applying fit to employees, an opportunityexists to provide a more holistic profile that leads to superior deci-sions in selecting and rewarding outstanding employees.

When considering fit among multiple elements simultaneouslyand examining the effects on outcomes, as this study does, configura-tion 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 partic-ular case (a customer contact employee, in this study) matches anideal 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 thethree components of commitment and service quality as the degreeto which the commitment and demographic characteristics of a front-line employee differ from those of an ideal profile in achieving servicequality (Vorhies & Morgan, 2003; Zajac et al., 2000).

Employeeswhose profiles the innermost circle (A) represents are theideal type of employee who delivers the highest service quality to cus-tomers. Employees that inner circle (B) represents deliver lower qualitythan those in A but higher than those in C; and those in C, in turn, deliverbetter service quality than those in circle D (Fig. 2). The prediction of PDis that, as the Euclidian distance increases from the ideal, service qualitydeteriorates. Thus, the more a frontline employee is like the ideal profile(definition from AC, NC, and CC components and the total experienceand age demographics), the more superior the service quality is thatthe employee delivers. Hence, the key proposition relating the profilesof service employees and the outcome of interest (i.e., service quality) is:

Proposition 1. The ideal type of employees will deliver significantlysuperior service quality than non-ideal employees.

With the growth in multi-channel strategies among services, thetwo types of encounters involving frontline employees, face-to-faceand telephone are becoming increasingly crucial to manage. Howev-er, several differences exist in customer service delivery betweenface-to-face and telephone encounters that have a bearing on thetype 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 determi-nants of service quality; the customer plays a role in creating quality ser-vice through his/her behavior during the interaction, and people cancreate quality perceptions relating to the environmentwhere the servicetakes place (Burgers, Ruyter, Keen, & Streukens, 2000). On the otherhand, in phone encounters, the customer has less influence on servicequality. The service environment and tangibles are not part of the qualityperceptions; their judgement of service and quality comes purely fromintrinsic dimensions like reliability, responsiveness, assurance, and em-pathy (Boshoff & Tait, 1996; Burgers et al., 2000). Further, performancein branches involves less scripting and standardization as compared tocall centers, which provides an opportunity to the more able employees

Organizational Commitment

Affective Normative Continuance

Demographics Total experience Age

Employee profile fit with ideal profile

Deviation from an ideal employee profile leads to

inferior service quality

Service Quality

Fig. 1. A model of profile deviation for customer-contact employees.

1340 N. Malhotra et al. / Journal of Business Research 66 (2013) 13381344

to use their initiative, experience, and commitment to stand out. Keepingthis in mind, this research proposes:

Proposition 2. The ideal type of employees will differ significantlywith respect to their commitment profile and demographics in thetwo types of service encounters.

3. The methodology

3.1. Sample

This research comes from telephone call centers and branches of amajor retail bank in the United Kingdom. Because the perception isthat most financial products are high-involvement, complex, andlow on differentiation, the frontline employee, and not the service it-self, provides a source of differentiation and creates competitive ad-vantage (Burgers et al., 2000). Hence, this study measures servicequality from the viewpoint of the frontline employee.

Researchers mailed self-administered anonymous surveys to theheads of customer services and the regional branch managers whothen arranged for further distribution to their respective frontlineemployees. The researchers used stratified sampling to ensure thatno selection bias exists in the sample. They also provided self-addressed pre-paid envelopes along with the questionnaires, andthe employees returned the completed questionnaires directly toone of the researchers.

Six hundred and forty employees in the call centers and 300 em-ployees in the branches received the questionnaires. This distribution

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.

D

CB

Fig. 2. Profile deviation as concentric circles. A: The ideal type of employees who de-livers highest service quality to customers. B: Employees who deliver lower servicequality than A. C: Employees who deliver better service quality than D, but lower ser-vice quality than B. D: Employees who deliver lower service quality than C.

included all employees who satisfied the initial profile screening. Thisdistribution yielded 342 useable questionnaires from the call centersand 170 useable questionnaires from the branches, generating net re-sponse 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 organi-zational tenure of around 12.5 years. The profile of the respondents isrepresentative 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 orga-nizational tenure is around 3.5 years. The similarity between popula-tion and sample profiles ensures that no significant non-response biasexists.

3.2. Measuring instruments

In this study, frontline employees of the bank evaluated their ownperformance in terms of service quality on a shortened and adaptedversion of SERVQUAL (Parasuraman, Zeithaml, & Berry, 1988). Manystudies have effectively used employees perceptions of service deliv-ery (Babakus, Yavas, Karatepe, & Avci, 2003; Boshoff & Tait, 1996;Singh, 2000) in measuring performance. After extensive discussionswith managers of call centers and branches, researchers selected theappropriate items from the dimensions of SERVQUAL that were appli-cable in measuring service quality in both encounter scenarios. Inboth cases, they selected only those items that pertained specificallyto employee-related aspects of service quality (see Boshoff & Tait,1996).

A three-component scale (Meyer, Allen, & Smith, 1993) measuresthe three components of organizational commitment. The scale (18items) has been extensively used by several researchers (Malhotra& Mukherjee, 2003, 2004) and has been well accepted for reliabilityand validity.

A five point, Likert-type scale measures all items, ranging fromstrongly disagree to strongly agree. Researchers also collected dataon the demographics of the respondents in terms of their age andtotal work experience.

4. Results

This study standardizes the data (mean-centered with a mean ofzero and a standard deviation of one) to remove the effects of differ-ent measurement units and potential multicollinearity (Jaccard &Turrisi, 2003). All the Chronbach's alphas are greater than 0.7 andtherefore are acceptable (Nunnally, 1978). An exploratory factoranalysis follows this standardization. The eight items that measureservice quality yield one factor, but the three components of commit-ment emerge as three distinct factors. The study runs a confirmatoryfactor analysis (CFA), and the measurement model exhibits strongpsychometric properties and acceptable fits (see Table 1). The studyalso tests common method variance (CMV). A correlation matrix of


the constructs in this study shows that the smallest correlation is be-tween AC and CC (0.002). This correlation becomes the conservativeestimate for CMV (Lindell & Whitney, 2001). Adjusting all correctionsfor this estimated value of CMV shows that all the correlations thatwere previously significant remain significant (Malhotra, Kim, &Patil, 2006). Thus, CMV is not a problem in this data. The extrapola-tion procedure that Armstrong and Overton (1977) suggest assessespotential non-response bias. When assessing the first quartile versusthe last quartile of the respondents in the sample group, no significantdifference exists on any of the four summated measures in the survey(i.e., service quality, AC, NC, and CC), indicating that the data are freefrom systematic difference or non-response bias.

4.1. Testing for measure equivalence

Measure equivalence of service quality is important to check be-cause respondents from the two settings could possibly interpret ser-vice quality differently. First, the study establishes the psychometricproperties of service quality in each sample separately. These proper-ties are acceptable (see Table 1). Using common service quality items,the researchers then perform multiple group comparison in AMOSversion 6. The first step is to test for configural equivalence that estab-lishes that both samples map the same measure, that is, they map thesame indicators and latent variables. Configural equivalence can notbe rejected (2=184.2, df=40, Cmin/df=4.6, RMSEA=0.08;IFI=0.91, CFI=0.91). Testing for metric equivalence, that is, con-straining the factor loadings to equality across the samples, showsthat measure equivalence at this level again can not be rejected be-cause the chi-square difference test is not significant (2=196.9,df=47, p>0.05). Imposing additional constraints of equality on in-tercepts leads to an acceptable, strong factorial equivalence(2=189.76, df=48, p>0.05). Support for the strong factorial

Table 1Instrument items and psychometric properties.

ITEM

Affective CommitmentI 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 CommitmentI 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 pI owe a great deal to my company.Continuance CommitmentRight 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 nI 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 workinOne of the few negative consequences of leaving this organization would be the scarcityService QualityWhen 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.

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; : Cro

equivalence indicates that any approach that specifies the modelwill not alter this finding (Vandenberg & Lance, 2000). Overall,these results suggest that support exists for measure equivalencethat allows all comparisons and the interpretation of the findings tobe 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 PDanalysis to test the predictions. The first step in the PD analysis is toidentify ideal employee-based profiles that can be used as bench-marks against which their fit can be examined (Doty et al., 1993;Vorhies & Morgan, 2003). To identify these profiles, the study exam-ines the frequencies of the outcome variable (service quality) coupledwith the general guidelines in PD studies of selecting about 10% of thecases to be included in the ideal profile (Venkatraman & Prescott,1990). The cutoff point becomes the top 10% of the employeeswhere a significant drop-off in service quality is apparent (resultingeach ideal profile comprising a range of 17 to 34 cases). One concernis that these tests might be self-fulfilling prophecies because onedataset provides both the ideal profiles and the rest of the data. Tobe clear, the development of the profiles comes from employees per-ceptions of their own service quality. The hypotheses in essence ex-amine the expectation that the closer the perceptions ofcommitment of a firm's employees, in general, align with those heldby the firm's best employees; the stronger the perception of theirown service quality.

To test for PD, the development of an empirically derived calibra-tion sample comes from the top 10% of employees rank-ordered byservice quality (Venkatraman & Prescott, 1990; Vorhies & Morgan,2003). The calibration sample uses the top 10% of employees showingthe highest service quality to identify the ideal profile in terms of the

BRANCHES CALL CENTERS

FactorLoading

C.R. AVE FactorLoading

C.R. AVE

0.83 0.83 0.52 0.82 0.83 0.500.54 0.68

0.79 0.68 0.790.81 0.76 0.810.80 0.79 0.800.59 0.58 0.59

0.89 0.88 0.55 0.78 0.77 0.510.65 0.510.88 0.810.87 0.800.78 0.62

eople in it. 0.58 0.400.63 0.45

0.84 0.83 0.48 0.86 0.85 0.500.72 0.610.55 0.64

ow. 0.64 0.740.79 0.73

g elsewhere. 0.72 0.82of available alternatives. 0.69 0.70

0.88 0.88 0.50 0.89 0.89 0.500.72 0.690.62 0.640.80 0.670.58 0.670.74 0.710.67 0.650.78 0.760.71 0.82

nbach's Alpha; CR: Construct Reliability; AVE: Average Variance Extracted


three commitment components and the demographics. Thus, thebenchmark for the comparisons of the rest of the sample is the cali-bration sample. The calculation of the Euclidian distance is accordingto the formula (Venkatraman & Prescott, 1990):

MISALIGN j1

bj XsjXcj 2

; 1

where MISALIGN=Degree of misfit, X are aspects of commitmentand demographics, and X-bar is the mean of the X for the calibrationsample (or the mean of the non-ideal 10% random sample).

A regression model tests for PD. Several advantages exist fromusing PD like the ability to add more variables without leading to in-creased demand on sample size; thus, yielding a more holistic de-scription of the ideal service employee. Developing a more holisticprofile can lead to superior employee selection and can be used for re-warding outstanding employees on some objective measure.

Alternative non-ideal models, where the average employees arethe benchmark for PD scores, also examine whether the ideal profilesprovide strong explanatory power. Thus, to further test the robust-ness of the model, this study uses a random sample of 10% of the em-ployees as another calibration sample (non-ideal profile). Thismethod tests whether deviation from a random sample of employeeshas performance implications. The resulting PD score substitutes forthe one from the top 10% of performers in order to re-estimate themodel. If the PD (random) is not significant, then this result providesfurther evidence of the robustness of the test and provides more sup-port for the research proposition in this paper. The results support theconcept that calibrating ideal employee profiles produces stronger PDcoefficients and greater explanatory power than the random bench-marks (see Chow, 1960; Cohen, Cohen, West, & Aiken, 2003).

The results in Table 2 indicate that PD for the face-to-face em-ployees is significant and negative (=0.3, pb0.001). This findingsupports the hypothesis that employees who deviate in their profilefrom the best performing employees deliver low levels of servicequality. The variance in the model is 8.5% for service quality. This per-centage might appear rather low but is consistent with comparablestudies (Vorhies & Morgan, 2003). The random PD is nonsignificant(p>0.1) and the regression coefficient is very small (=0.01).This finding provides further support for the robustness of the pro-posed model. An alternative finding might be interpreted as meaningthat deviation from any sample of 10% of the employees has an asso-ciation with poor service quality and is contrary to the theory. How-ever, the results in Table 2 for the call centers show that PD is notsignificant.

Examining the profile of the top employees for the face-to-face en-counters in branches shows that they have the following characteris-tics: moderate to high AC, moderate to high NC, and comparativelylow CC. They have a reasonable amount of experience and appear tobe middle aged at an average of 40. Females predominantly staffthis service delivery mode as noted in the demographics. On theother hand, for the call centers, the top performers have a similarcommitment profile with relatively high AC, moderate NC, and lowCC. They have little experience relative to those in face-to-face en-counters and are significantly younger by comparison. Thus, the

Table 2Profile deviation: service quality face-to-face (branches) & telephone (call centers).

Variable Face-to-Face ( Branches ) Telephone ( Call Centers )

10% Random 10% Random

Profile Deviation 0.31 (3.58***) 0.02 (0.15) 0.007 (0.13) 0.05 (0.60)R Sqr. 0.09 0 0.01 0.01Adj. R Sqr. 0.07 0.02 0.01 0.01F-ratio 6.44*** 0.03 1.17 0.90

***pb0.00.

results in Table 3 seem to suggest that no significant differencesexist in commitment among the best performers in call centers andface-to-face contact. However, significant differences exist in total ex-perience (pb0.001) and age (pb0.001).

But, significant differences exist in the commitment structures ofemployees whose service quality perceptions rank among the bottom10%. Of specific interest is the result that, although AC is the dominantdriver of the super achievers, CC seems to work most for the weakproviders. In both call centers as well as branches, AC levels drop sig-nificantly between the top and bottom 10% of employees, while CC in-creases. Even though the poor performers on service quality intend tostay with their organizations, service managers need to train, moti-vate, and assess them regularly so that they can be of some value totheir organizations.

4.3. T-test results

This study also conducts a t-test to see if significant differencesexist between the top, bottom, and random 10% of employees AC,NC, and CC across the sample groups. Only the AC of random em-ployees in call centers is significantly less than the AC of random em-ployees in bank branches at pb0.05. However, others display nosignificant differences.

5. Discussion

In both call centers and branches, similar profiles of top per-formers exist in terms of the three components of commitment. Af-fective commitment is the strongest; normative and continuancecommitment follow respectively. These results support the argu-ments presented in the literature (Allen & Meyer, 1990; Malhotra &Mukherjee, 2004; Meyer et al., 2002) that employees who value orga-nizational goals and identify with the organization are likely to per-form better than employees who merely stay under an obligation(normative) or a particular need (continuance). The results furtherconfirm that the continuance commitment dominates the profiles inboth contexts for the bottom 10% of performers (Table 4). Thus, thefindings indicate that the desirable commitment profile should com-prise more affective commitment and less continuance commitment,irrespective of the context.

The results also show that profile deviation is significant inbranches but not in call centers (Table 2). This finding implies thatideal employees deliver significantly superior service quality thannon-ideal employees in branches, but not in call centers. Possibly,this is because in call center environments, frontline work is very rou-tine and leads to excessive standardization that, along with intensivetraining and electronic monitoring, ensures minimum deviation instandards of performance. As compared to call centers, frontlinework in branches is less standardized, more satisfying, and socialdue to the nature of customer participation in these face-to-face en-counters. Good service quality delivery in face-to-face encountersplaces a great onus on employees discretionary efforts and demandsmuch more than just formally acquired skills. These discretionary ef-forts provide an opportunity to the more committed employees touse their initiative, experience, and ability in order to truly standout as better performers, as explained by the significant PD resultsfrom branches in this study.

Although no differences exist in the commitment profiles of topperformers, significant differences do exist in total experience(pb0.001) and age (pb0.001) between top performers in branchesand call centers. Also, the average organizational tenure in the caseof call centers is only 3.5 years as compared to 12.5 years in the caseof frontline employees working in the branches. Possibly, turnoverproblems continue to hit the call-center industry and thus limit ad-vancement opportunities. So, the call centers generally attract

Table 3Profile 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 3.45 3.23 0.94Normative commitment 3.06 3.03 0.02Continuance commitment 2.97 2.78 0.46Total experience 10.49 22.29 19.65***Age 31.53 40.86 12.87***

***pb0.001.

Table 4Profile 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 2.25 2.72 4.74*Normative commitment 2.53 2.69 0.39Continuance commitment 3.07 2.90 0.84Total experience 7.14 12.94 33.01***Age 29.25 30.20 61.30***

*pb0.05.***pb0.001.


younger people, which explains the significant differences in the twosamples regarding demographics.

6. Managerial implications and future research

As multi-channel service delivery strategies become more popu-lar, organizations need to better understand the right profile forfrontline employees that is suitable in different types of customer-contact encounters. The results of this study have implications for in-ternal marketing strategists on how to use various rewards and to de-velop programs that encourage the right commitment profile amongtheir employees. Organizations might be more prudent to foster af-fective commitment in their frontline employees as affective commit-ment dominates the commitment profiles of the top service qualityperformers in this study. Hence, internal marketing strategies shouldincorporate work variables like role clarity, autonomy, participationin decision making, training, feedback, and job satisfaction to encour-age employees affective commitment (Malhotra et al., 2007; Meyeret al., 2002).

Further, management needs to pay special attention to the profileof frontline staff dealing with customers in face-to-face encounters,because greater variations in service quality delivery exist dependingon their commitment and demographic profile. Hence, especially inface-to-face encounters where management has little control overthe behaviors of employees andwhere the discretionary effortmattersin delivering excellent service quality, the right commitment profileneeds to be encouraged and developed among frontline staff.

The results also offer some interesting insights for the recruitmentstrategies of organizations as the profiles of top performers in face-to-face and telephone encounters differ significantly in terms of age andexperience. Thus, organizations need to establish different selectioncriteria for these two types of encounters. A worthwhile investmentfor organizations would be to incorporate and develop suitable selec-tion mechanisms in their recruitment system that can decipher thepsychological profiles of the candidates in terms of their commitmentbefore final recruitment. Also, internal marketing strategies could bedevised to incorporate employee rotation between the two types ofencounters within the same organization according to employee-encounter fit as their ideal profiles demonstrate.

However, because this is a single bank study, the generalizabilityof results should be tested in the future. The study should be replicat-ed in other contexts using other indicators of performance, like prof-itability and customer satisfaction. A variety of ways exist in whichthe calibration sample can be drawn; for example, outstanding

service quality performers can be selected and their commitment orother variables computed to identify ideal profiles. Also, the calibra-tion sample can be taken at any level, which could improve the coef-ficient of determination, R2. Some researchers prefer the top 5% whileothers prefer the top 20%. These can be adjusted to fit with the pur-pose of the exercise. Future research can also empirically compareprofile deviation methods with existing statistical techniques.

References

Allen, N. J., & Meyer, J. P. (1990). The measurement and antecedents of affective, con-tinuance and normative commitment. Journal of Occupational Psychology, 63(1),118.

Armstrong, J. S., & Overton, T. S. (1977). Estimating non-response bias in mail surveys.Journal of Marketing Research, 14, 396402 (August).

Babakus, E., Yavas, U., Karatepe, O., & Avci, T. (2003). The effect of management com-mitment to service quality on employees affective and performance outcome.Journal of the Academy of Marketing Science, 31(3), 272286.

Boshoff, C., & Tait, M. (1996). Quality perceptions in the financial services sector: Thepotential impact of internal marketing. International Journal of Service IndustryManagement, 17(5), 531.

Burgers, A., Ruyter, K., Keen, C., & Streukens, S. (2000). Customer expectation dimen-sions of voice-to-voice service encounters: A scale-development study. Interna-tional Journal of Service Industry Management, 11(2), 142161.

Burmann, C., & Konig, V. (2011). Does internal brand management really drive brandcommitment in shared-service call centers? Journal of Brand Management, 18(6),374393.

Byrne, B. M. (2004). Testing for multi-group invariance using AMOS graphics: A roadless travelled. Structural Equation Modelling, 11(2), 272300.

Chen, Y., Huang, H., Sung, S., & Huang, W. (2009). The role of fit on business perfor-mance: A profile deviation analysis. 2009 International Conference on New Trendsin Information and Service Science (pp. 235239).

Chow, G. C. (1960). Test of equality between sets of coefficients in two linear equations.Econometrics, 28(3), 591605.

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple reg./cor. analysisfor the behavioral sciences (3rd ed.). Mahwah, New Jersey: Lawrence Erlbaum Ass.Publishers.

Culpepper, R. A., Gamble, J. E., & Blubaugh, M. G. (2004). Employee stock ownershipplans and three-component commitment. Journal of Occupational and Organiza-tional Psychology, 77(2), 155166.

Doty, H. D., Glick, W. H., & Huber, G. P. (1993). Fit equifinality and organizational effec-tiveness: A test of two configurational theories. Academy of Management Journal,36(6), 11961250.

Drazin, R., & Van de Ven, A. H. (1985). Alternative forms of fit in contingency theory.Administrative Science Quarterly, 30(4), 514539.

Fang, E., Palmatier, R. W., & Grewal, R. (2011). Effects of customer and innovation assetconfiguration strategies on firm performance. Journal of Marketing Research, 48(3),587602.

Gustaffson, A. (2009). Customer satisfaction with service recovery. Journal of BusinessResearch, 62(8), 12201222.

Hult, G. T. M., Boyer, K. K., & Ketchen, D. J., Jr. (2007). Quality, operational logistics strat-egy, and repurchase intentions: A profile deviation analysis. Journal of Business Lo-gistics, 28(2), 105132.

Jaccard, J., & Turrisi, R. (2003). Interaction effects in multiple regression (2nd ed.). Thou-sand Oaks, California: Sage Publications Inc.


Kabadayi, S., Eyuboglu, N., & Thomas, G. (2007). The performance implications of de-signing distribution channels to fit with strategy and environment. Journal of Mar-keting, 71(10), 195211.

Lindell, M. K., & Whitney, D. J. (2001). Accounting for common method variance incross sectional research designs. Journal of Applied Psychology, 86, 114121.

Lings, I. N. (2004). Internal market orientation: Construct and consequences. Journal ofBusiness Research, 57(4), 405413.

Malhotra, N., Budhwar, P., & Prowse, P. (2007). Linking rewards to commitment: Anempirical investigation of four UK call centres. International Journal of Human Re-source Management, 18(12), 20952127.

Malhotra, N. K., Kim, S. S., & Patil, A. (2006). Commonmethod variance in IS research: Acomparison of alternative approaches and a reanalysis of past research. Manage-ment Science, 52(12), 18651883.

Malhotra, N., & Mukherjee, A. (2003). Analysing the commitment service quality re-lationship: A comparative study of retail banking call centres and branches. Journalof Marketing Management, 19(9/10), 941997.

Malhotra, N., & Mukherjee, A. (2004). The relative influence of organizational commit-ment and job satisfaction on service quality of customer-contact employees inbanking call centres. Journal of Services Marketing, 18(3), 162174.

Meyer, J. P., Allen, N. J., & Smith, C. A. (1993). Commitment to organisations and occu-pations: extension and test of a three-component conceptualisation. Journal of Ap-plied Psychology, 78(4), 538551.

Meyer, J. P., Stanley, D. J., Herscovitch, L., & Topolnytsky, L. (2002). Affective, continu-ance and normative commitment to the organization: A meta-analysis of anteced-ents, correlates and consequences. Journal of Vocational Behavior, 61(1), 2052.

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),1240.

Schoonhoven, C. B. (1981). Problems with contingency theory: Testing assumptionshidden within the language of contingency "theory". Administrative Science Quar-terly, 26(3), 349377.

Singh, J. (2000). Performance productivity and quality of frontline employees in serviceorganizations. Journal of Marketing, 64, 1534 (April).

Smith, K. G., Guthrie, J. P., & Chen, M. J. (1989). Strategy, size and performance. Organi-zation Studies, 10(1), 6381.

Van Herk, V. H., Poortinga, Y. H., & Verhallen, T. M. M. (2005). Equivalence of surveydata: Relevance for international marketing. European Journal of Marketing,39(3/4), 351364.

Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of the measurement in-variance literature: Suggestions, practices, and recommendations for organization-al research. Organizational Research Methods, 3(1), 470.

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 inemployee- customer encounters. Journal of Applied Psychology, 92(4), 11771187.

Venkatraman, N. (1990). The concept of fit in strategy research: Toward verbal and sta-tistical correspondence. Academy of Management Review, 27(1), 1941.

Venkatraman, N., & Prescott, J. E. (1990). Environment strategy co-alignment: An em-pirical test of its performance implications. Strategic Management Journal, 11, 123.

Vorhies, D. W., & Morgan, N. A. (2003). Configuration theory assessment of marketingorganization fit with business strategy and its relationship with marketing perfor-mance. Journal of Marketing, 11(1), 100115.

Wieseke, J., Ahearne, M., Lam, S. K., & VanDick, R. (2009). The role of leaders in internalmarketing. Journal of Marketing, 73(1), 123145.

Zajac, E. J., Kraatz, M. S., & Bresser, R. K. F. (2000). Modeling the dynamics of strategicfit: A normative approach to strategic change. Strategic Management Journal,21(4), 429453.
Service quality of frontline employees: A profile deviation analysis1. Introduction2. Configuration theory3. The methodology3.1. Sample3.2. Measuring instruments4. Results4.1. Testing for measure equivalence4.2. Testing the predictions of profile deviation4.3. T-test results5. Discussion6. Managerial implications and future researchReferences

Articol Despre Personal

Documents

Transcript of Articol Despre Personal