Journal of Business Research 66 (2013) 13381344

Contents lists available at SciVerse ScienceDirect

Journal of Business Research

Service quality of frontline employees: A prole 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:
Prole deviation analysis
Conguration theory
Bank branches
Call centers
Service
Quality
Commitment

a b s t r a c t
Using a conguration 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 prole deviation analysis. The results show that the prole deviation for face-to-face
employees is signicantly negative, while for call center employees nonsignicant. Although the study nds
no signicant differences in the three components of commitment, signicant differences exist in the total
experience and age of the best performers. Also, affective commitment dominates the prole of high performers, while poor service providers seem to exhibit a higher level of continuance commitment. This
study demonstrates the utility of prole 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 inuence on their service quality (e.g., Malhotra &
Mukherjee, 2004; Vandenberghe et al., 2007). However, the actual
commitment-prole 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: n.malhotra@aston.ac.uk (N. Malhotra),
Felix.Mavondo@buseco.monash.edu.au (F. Mavondo),
mukherjeeav@mail.montclair.edu (A. Mukherjee), g.j.hooley@aston.ac.uk (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 conguration theory approach, this study compares employees in the two most difcult and important types of frontline service situations; face-to-face and telephone encounters. First, this
study identies the prole of the top performers in each of the encounters. Second, this study makes comparisons between the ideal
and non-ideal proles to investigate whether deviations from the
ideal result in a decrease in service quality in each context. Any difference this study nds among the proles 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 prole deviation analysis with a basis
in congurational theory, which is a methodological innovation because prole deviation has rarely been used in marketing. Prole
deviation approaches provide signicant advantages over traditional
approaches such as regression analysis, slope analysis, and subgroup
analysis, particularly with individuals as the unit of analysis, in assessing t in a way that is consistent with the multidimensional and holistic perspective of services marketing (Vorhies & Morgan, 2003).
Previous research mainly applied the conguration theory and the
prole 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) 13381344

Second, this study conducts an empirical test to see if deviations

from the ideal prole 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, identication 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 prole 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 t by comparing face-to-face and
phone services.
This study integrates the service-quality literature with the
prole-deviation analysis literature to provide valuable new insights
into the theory and practice of service excellence, and is the rst to attempt to understand the role of conguration theory in designing internal marketing strategies.
2. Conguration theory
Conguration theory has been used by management over the last
two decades to assess complex, multidimensional phenomena implied in t 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 ndings.
Congurational theories posit higher effectiveness for employees
that resemble the ideal type that theory denes. The increased effectiveness comes from an internal consistency or t among the relevant structural and strategic factors (Doty, Glick, & Huber, 1993). Researchers who
treat congurations 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 congurations as categories, they predict that marginal
members of the category are as effective as the central members. On
the other hand, when treating congurations as ideal types, they predict
employees that marginally resemble the ideal conguration 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 dene the ideal type. Employees do not have to be classied 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 conguration

model as an interaction, selection, or systems approach to t
(Drazin & Van de Ven, 1985). The interaction approach is the basis
for many contingency theories that dene t 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 t assertions in congurational theories (Venkatraman & Prescott, 1990). Thus, this study
conceptualizes t 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 prole deviation and gestalt and denes t 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 t among factors is at a maximum in
those congurations. When applying t to employees, an opportunity
exists to provide a more holistic prole that leads to superior decisions in selecting and rewarding outstanding employees.
When considering t among multiple elements simultaneously
and examining the effects on outcomes, as this study does, conguration should be conceptualized and measured via prole deviation
(PD) analysis (Doty et al., 1993; Venkatraman, 1990; Vorhies &
Morgan, 2003). PD analysis views t as the degree to which a particular case (a customer contact employee, in this study) matches an
ideal prole (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 t 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 prole in achieving service
quality (Vorhies & Morgan, 2003; Zajac et al., 2000).
Employees whose proles 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 prole
(denition 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 proles
of service employees and the outcome of interest (i.e., service quality) is:
Proposition 1. The ideal type of employees will deliver signicantly
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 inuence 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) 13381344

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 prole 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 signicantly
with respect to their commitment prole 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 nancial 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 stratied 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. Prole 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 satised the initial prole 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 prole of the respondents is
representative of the typical employee prole 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 proles ensures that no signicant 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 specically
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 ve 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 conrmatory
factor analysis (CFA), and the measurement model exhibits strong
psychometric properties and acceptable ts (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) 13381344

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 signicant remain signicant (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 rst quartile versus
the last quartile of the respondents in the sample group, no signicant
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 rst step is to test for congural equivalence that establishes that both samples map the same measure, that is, they map the
same indicators and latent variables. Congural equivalence can not
be 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, 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 signicant ( 2 = 196.9,
df = 47, p > 0.05). Imposing additional constraints of equality on intercepts leads to an acceptable, strong factorial equivalence
( 2 = 189.76, df = 48, p > 0.05). Support for the strong factorial

1341

equivalence indicates that any approach that species the model

will not alter this nding (Vandenberg & Lance, 2000). Overall,
these results suggest that support exists for measure equivalence
that allows all comparisons and the interpretation of the ndings to
be valid (Byrne, 2004; Van Herk, Poortinga, & Verhallen, 2005).
4.2. Testing the predictions of prole deviation
The study follows the literature on conguration theory and PD
analysis to test the predictions. The rst step in the PD analysis is to
identify ideal employee-based proles that can be used as benchmarks against which their t can be examined (Doty et al., 1993;
Vorhies & Morgan, 2003). To identify these proles, 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 prole (Venkatraman & Prescott,
1990). The cutoff point becomes the top 10% of the employees
where a signicant drop-off in service quality is apparent (resulting
each ideal prole comprising a range of 17 to 34 cases). One concern
is that these tests might be self-fullling prophecies because one
dataset provides both the ideal proles and the rest of the data. To
be clear, the development of the proles 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 rm's employees, in general, align with those held
by the rm'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 prole 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 rst 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 condence in my customers.

Factor
Loading

CALL CENTERS

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

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

1342

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

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 bj X sj X cj
;

j1

where MISALIGN = Degree of mist, 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
prole 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 proles
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 prole). 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 signicant, 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 proles produces stronger PD
coefcients 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 signicant and negative ( = 0.3, p b 0.001). This nding
supports the hypothesis that employees who deviate in their prole
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 nonsignicant
(p > 0.1) and the regression coefcient is very small ( = 0.01).
This nding provides further support for the robustness of the proposed model. An alternative nding 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
signicant.
Examining the prole 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 prole with relatively high AC, moderate NC, and low
CC. They have little experience relative to those in face-to-face encounters and are signicantly younger by comparison. Thus, the

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

Prole 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 signicant differences

exist in commitment among the best performers in call centers and
face-to-face contact. However, signicant differences exist in total experience (p b 0.001) and age (p b 0.001).
But, signicant differences exist in the commitment structures of
employees whose service quality perceptions rank among the bottom
10%. Of specic 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 signicantly 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 signicant 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 signicantly less than the AC of random employees in bank branches at p b 0.05. However, others display no
signicant differences.
5. Discussion
In both call centers and branches, similar proles 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
conrm that the continuance commitment dominates the proles in
both contexts for the bottom 10% of performers (Table 4). Thus, the
ndings indicate that the desirable commitment prole should comprise more affective commitment and less continuance commitment,
irrespective of the context.
The results also show that prole deviation is signicant in
branches but not in call centers (Table 2). This nding implies that
ideal employees deliver signicantly 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 signicant PD results
from branches in this study.
Although no differences exist in the commitment proles of top
performers, signicant 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) 13381344

1343

Table 3
Prole of top 10% employees. Table prole 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
Prole of bottom 10% employees. Table prole 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 signicant 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 prole 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 prole among
their employees. Organizations might be more prudent to foster affective commitment in their frontline employees as affective commitment dominates the commitment proles 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 prole
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 prole. 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 prole
needs to be encouraged and developed among frontline staff.
The results also offer some interesting insights for the recruitment
strategies of organizations as the proles of top performers in face-toface and telephone encounters differ signicantly 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 proles of the candidates in terms of their commitment
before nal 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 t as their ideal proles 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 proles. Also, the calibration sample can be taken at any level, which could improve the coefcient of determination, R 2. Some researchers prefer the top 5% while
others prefer the top 20%. These can be adjusted to t with the purpose of the exercise. Future research can also empirically compare
prole 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),
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 commitment 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 nancial services sector: The
potential impact of internal marketing. International Journal of Service Industry
Management, 17(5), 531.
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), 142161.
Burmann, C., & Konig, V. (2011). Does internal brand management really drive brand
commitment 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 road
less travelled. Structural Equation Modelling, 11(2), 272300.
Chen, Y., Huang, H., Sung, S., & Huang, W. (2009). The role of t on business performance: A prole deviation analysis. 2009 International Conference on New Trends
in Information and Service Science (pp. 235239).
Chow, G. C. (1960). Test of equality between sets of coefcients in two linear equations.
Econometrics, 28(3), 591605.
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), 155166.
Doty, H. D., Glick, W. H., & Huber, G. P. (1993). Fit equinality and organizational effectiveness: A test of two congurational theories. Academy of Management Journal,
36(6), 11961250.
Drazin, R., & Van de Ven, A. H. (1985). Alternative forms of t in contingency theory.
Administrative Science Quarterly, 30(4), 514539.
Fang, E., Palmatier, R. W., & Grewal, R. (2011). Effects of customer and innovation asset
conguration strategies on rm performance. Journal of Marketing Research, 48(3),
587602.
Gustaffson, A. (2009). Customer satisfaction with service recovery. Journal of Business
Research, 62(8), 12201222.
Hult, G. T. M., Boyer, K. K., & Ketchen, D. J., Jr. (2007). Quality, operational logistics strategy, and repurchase intentions: A prole deviation analysis. Journal of Business Logistics, 28(2), 105132.
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) 13381344

Kabadayi, S., Eyuboglu, N., & Thomas, G. (2007). The performance implications of designing distribution channels to t with strategy and environment. Journal of Marketing, 71(10), 195211.
Lindell, M. K., & Whitney, D. J. (2001). Accounting for common method variance in
cross sectional research designs. Journal of Applied Psychology, 86, 114121.
Lings, I. N. (2004). Internal market orientation: Construct and consequences. Journal of
Business Research, 57(4), 405413.
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), 20952127.
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), 18651883.
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), 941997.
Malhotra, N., & Mukherjee, A. (2004). The relative inuence of organizational commitment and job satisfaction on service quality of customer-contact employees in
banking call centres. Journal of Services Marketing, 18(3), 162174.
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), 538551.
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), 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 assumptions

hidden within the language of contingency "theory". Administrative Science Quarterly, 26(3), 349377.
Singh, J. (2000). Performance productivity and quality of frontline employees in service
organizations. Journal of Marketing, 64, 1534 (April).
Smith, K. G., Guthrie, J. P., & Chen, M. J. (1989). Strategy, size and performance. Organization Studies, 10(1), 6381.
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), 351364.
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), 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 in
employee- customer encounters. Journal of Applied Psychology, 92(4), 11771187.
Venkatraman, N. (1990). The concept of t in strategy research: Toward verbal and statistical correspondence. Academy of Management Review, 27(1), 1941.
Venkatraman, N., & Prescott, J. E. (1990). Environment strategy co-alignment: An empirical test of its performance implications. Strategic Management Journal, 11, 123.
Vorhies, D. W., & Morgan, N. A. (2003). Conguration theory assessment of marketing
organization t with business strategy and its relationship with marketing performance. Journal of Marketing, 11(1), 100115.
Wieseke, J., Ahearne, M., Lam, S. K., & VanDick, R. (2009). The role of leaders in internal
marketing. Journal of Marketing, 73(1), 123145.
Zajac, E. J., Kraatz, M. S., & Bresser, R. K. F. (2000). Modeling the dynamics of strategic
t: A normative approach to strategic change. Strategic Management Journal,
21(4), 429453.