Service operations strategy, flexibility and performance in ... - CiteSeerX

The Emerald Research Register for this journal is available at http://www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/0144-3577.htm

Service operations strategy, ¯exibility and performance in engineering consulting ®rms Daniel Arias Aranda

Service operations strategy 1401

Departamento de OrganizacioÂn de Empresas, Universidad de Granada, Granada, Spain Keywords Service operations, Flexibility, Industrial performance, Spain Abstract In this paper, the relationship between operations strategy and performance through ¯exibility as a moderating variable is analysed within the service setting of engineering consulting ®rms in Spain. A framework for service strategy dimensions is suggested while manufacturing ¯exibility dimensions are applied to service operations considering necessary adaptations. A path analysis model is applied in order to enhance the understanding of interactions. This research proves that service operations strategy has a signi®cant positive and direct effect on service delivery performance. The fact that is especially relevant is that ¯exibility plays a more intense moderating role for ef®ciency performance measures than for customer satisfaction performance measures.

Introduction Operations strategy has been contemplated in the ®eld from a conceptual point of view and has dealt mainly with the content of strategy (Anderson et al., 1989; Skinner, 1978; Smith and Reece, 1999). All this research has con®gured the operations strategy concept as a pattern of decisions and actions driven to support the strategic goals established by the business unit. Operations should ®t the requirements of a business by striving for consistency between business capabilities and policies and business competitive advantages (Adam and Swamidass, 1989; Wheelwright, 1984). However, the scarcity of studies that analyse the effect of strategy on performance through its relationship with competitive priorities has become one of the great challenges in operations management research (Dangayash and Deshmukh, 2001; Sa®zadeh et al., 2000). Moreover, in service operations strategy, this panorama is still more meagre which leads to higher research opportunities (Nieto et al., 1999). Nowadays, the changing environments make ¯exibility one of the competitive priorities that most service ®rms have to deal with. The concept of ¯exibility of service processes suffers from the same confusion that overwhelmed manufacturing for years (Harvey et al., 1997). The nature and sources of manufacturing ¯exibility have been broadly studied, so comparison between manufacturing and services provides a good starting-point for the study of service ¯exibility. Flexibility is generally regarded as the ability to The author wishes to thank anonymous reviewers whose suggestions have been of inestimable help for improving the present paper.

International Journal of Operations & Production Management Vol. 23 No. 11, 2003 pp. 1401-1421 q MCB UP Limited 0144-3577 DOI 10.1108/01443570310501907

IJOPM 23,11

1402

respond or conform to new situations, being usually classi®ed as process, product or infrastructure related (Noori and Radford, 1995). Although the importance of ¯exibility has been broadly acknowledged, efforts to bring the existing theoretical research and practices together have not reached a substantial progress in order to evaluate both manufacturing and service ¯exibility (Suarez et al., 1996). In particular, the relationship between operations strategy and manufacturing ¯exibility has not received adequate attention at the time of making investment decisions in manufacturing technology (Adler, 1988; Arias-Aranda et al., 2001). Moreover, adequate recognition has been especially scarce in the implementation phase (Jaikumar, 1986). The relationship between operations strategy and ¯exibility has been understood in operations management (OM) literature in terms of environmental variability (Riley and Lockwood, 1997). The operations strategy intends to react to the changes that affect competition due to new industrial and social trends. The structural and infrastructural decisions that con®gure the operations strategy in services directly in¯uence the ¯exibility level of the delivery system (Bucki and Pesqueux, 2000). When the levels of ¯exibility in every dimension of the service delivery system support the implemented operation strategy, it directly affects operational performance (Spina et al., 1996). When analysing this relationship, traditional cost performance measures lack relevance in the service environment, as they do not re¯ect the customer-focus factors of quality, ¯exibility or satisfaction, which have become crucial to ®rm success (Neely et al., 1997). Performance measures, which capture and re¯ect service operations strategies, are required to examine whether non-cost-based measures cover the range of service operations practices (Ghalayini and Noble, 1996). According to these bases, the purpose of this paper is multiple and twofold: (1) to apply a manufacturing ¯exibility framework to service operations considering necessary adaptations to service industries’ nature; and (2) to enhance the understanding of interactions between service operations strategy and long and short-term performance measures by applying a path analysis model to the different dimensions de®ning both constructs in order to investigate the relationships between them. The empirical veri®cation of this research will be developed within the service setting of engineering consulting ®rms in Spain. Service operations strategy Issues regarding operations strategy content and process are often discussed in the current operations literature. The process of operations strategy is termed according to how strategic decisions are made in an organizational setting (Ho, 1996). De®nitions of strategy always mention enhancement of the ®rm’s

competitive position in the marketplace through resources building or positioning (Swink and Way, 1995). Most research focused primarily on the number of decision areas and goals of manufacturing in terms of performance criteria (Leong et al., 1990). Skinner (1969) identi®ed a set of decision areas through which manufacturing objectives are achieved. Later research proves a generally high agreement; even though many authors have developed different sets of manufacturing decision areas (Mills et al., 1998). However, traditionally OM literature has addressed service operations strategy to the development of the service delivery system in order to match the customer expectations with customer perceptions (Armistead, 1992). Models and frameworks have been suggested to explain this process through different services classi®cations and schemes (Johnston, 1994; Nayyar, 1992; Normann, 1984; Sampson, 1996; Schmenner, 1986). However, only a few studies analyse the differentiation and interaction between the different dimensions con®guring the service operations strategy. In this context, Arias-Aranda (2002) proposed a model based on the three basic operations strategies identi®ed in service literature according to the ®rm’s focus of activities. These basic operations strategies pursued by service industries are process, service or customer-oriented operations strategies (Berry and Parasuraman, 1997; Bowen and Youngdahl, 1998; Collier, 1994, 1996; Desatnik, 1994; Hart, 1995; Haynes and Du Vall, 1992; Johnston, 1994; Lusch et al., 1996; McCutcheon et al., 1994; Sampson, 1996; Tersine and Harvey, 1998). Arias-Aranda (2002) identi®ed nine structural and infrastructural decisions that lead to a determined service operations strategy. These are type of operations layout, PUSH/PULL orientation of the service delivery process, degree of process standardisation, number of different services offered, use of information technologies (cost reduction vs service improvement), back and front of®ce activities relationship, human resources specialisation, degree of customer participation, and new service design and development. Service ¯exibility Flexibility as a competitive goal still lacks a clear and accurate de®nition. In service ®rms, ¯exibility is a fuzzy concept, as not all service ¯exibility dimensions have been clearly determined. However, ¯exibility is generally accepted as a useful tool to improve the competitive position of the ®rm, especially when related to the process of decision-making in technologies  lvarez Gil, 1994; Jaikumar, 1986). adoption and implementation (Adler, 1988; A Slack (1987) analysed a sample of ten manufacturing ®rms concluding that managers’ vision of the ¯exibility concept is only partial and incomplete. Managers usually focus more on machines ¯exibility than on the overall system ¯exibility. Flexibility when focused only on technology implementation does not lead to a competitive advantage (Gupta and Somers, 1996). It is crucial to carefully plan and manage the whole system ¯exibility (Gustavsson, 1988)

Service operations strategy 1403

IJOPM 23,11

1404

by creating a plant infrastructure that allows such system ¯exibility (Upton, 1995). The above multidimensional view of ¯exibility assumes that one of the key challenges to achieving a ¯exible operations system emerges when focusing and managing the different dimensions of ¯exibility (Gerwin, 1993). This is mainly due to the fact that ¯exibility is not accumulative; so more ¯exibility in different parts of the system does not imply a more ¯exible operations system. Changes in ¯exibility are of strategic nature, so they should not only involve process engineers, but also production and even business managers (Chambers, 1992). Operations strategy determines the level of uncertainty to be supported by the service delivery system by adapting the different ¯exibility dimensions to environmental changes (Chambers, 1992; Gerwin, 1993). In this context, ¯exibility in services involves the introduction of new designs and services into the service delivery system quickly, adjust capacity rapidly, customize services, handle changes in the service mix quickly and handle variations in customer delivery schedules (Suarez et al., 1996). Thus, improving the timing and quantity of resource allocations for performing a process to avoid employing human and material resources when they are not needed represents the core goal of service ¯exibility (Duclos et al., 1995). Under these assumptions, OM literature has focused on identifying different dimensions and types of ¯exibility (Brill and Mandelbaum, 1989; Browne et al., 1984; Carter, 1986; Chatterjee et al., 1984; Gerwin, 1987; Gupta and Goyal, 1992; Gupta and Somers, 1992; Hyun and Ahn, 1990; Son and Park, 1987). However, there is still no clear consensus about which are the exact costs and bene®ts of ¯exibility implementation according to every dimension. In many cases, ¯exibility bene®ts have been considered only in terms of machine programming and production models integration (Taymaz, 1989). In some other cases, these bene®ts are reduced to workforce versatility (Walton and Susman, 1987). Moreover, some authors maintain that ¯exibility is always bene®cial while some others af®rm that it can even be harmful in some cases (Cusumano, 1988; Gaimon and Singhal, 1992; Tombak, 1988). Such discussion reinforces the role of the multidimensional character of ¯exibility. In service industries, customer interaction and customization imply a high degree of ¯exibility according to the speci®city of services such as, for instance, the main input ¯ow is information, while material handling plays a less important role (Adler, 1988). Nowadays, customers are more demanding for integral services provided by the same ®rm (Vandermerwe, 1992). This fact will entail a new and different conception of service ¯exibility. In addition, process equipments in service industries are mainly based on information technologies (Gouillart and Sturdivant, 1994). Personnel ¯exibility and versatility must be directly considered as they perform a higher degree of the service delivery compared to machines in manufacturing industries (Harvey et al., 1997). Interconnection and interdependence between information

technology and personnel ¯exibility allow us to consider both as a whole (Hyun and Ahn, 1990). However, electronic data interchange (EDI) and information technologies (ITs) provide new market opportunities for service ®rms without high ¯exibility investment efforts. According to this idea, Harvey et al. (1997) suggested a model of ¯exibility in services in which technology controls variability reduction through the following dimensions: volume, time and place, needs and customer. Consequently, as ITs become widely used in services, reengineering processes are easier to accomplish in service delivery systems than in manufacturing systems (Hammer and Champy, 1993). In addition, service ®rms are able to design delivery systems in which customers are actively involved, becoming self-service delivery systems. This makes ¯exibility simpler and easier to implement considering that the customer develops certain activities of the service delivery system, especially when compared to the high costs of workers training for manufacturing ®rms (Bowen and Lawler, 1995). Performance measures Performance measures re¯ect how well the different competitive priorities ®t in the implemented operations strategy (Suarez et al., 1996). For service industries, effective performance measures related to operations strategies require a shift from measures that focus on manufacturing ef®ciency to those capturing the critical success factors related to customer initiated demands (Abernethy and Lillis, 1995). Traditional ®nancial accounting measures are short-term oriented and aggregate. Hence, they usually are not capable to support operations strategic priorities (Govindarajan, 1988; Simons, 1987). In addition, operations assumptions of standardisation and mass production in a stable environment inspire ®nancial performance measures (Kaplan, 1983). However, environmental conditions of manufacturing industries differ from those found in service industries, which consider ®nancial measures of operations performance less relevant (Brownell and Merchant, 1990). Empirical literature has determined in many studies the strong in¯uence that ¯exibility has on performance, especially for non-®nancial measures (Feigenbaum and Karnani, 1991; Jaikumar, 1986; Tombak and De Meyer, 1988). However, nowadays, service operations managers need relevant and speci®c feedback in order to take effective decisions in a changing environment (Chenhall, 1995). Non-®nancial measures contribute to enhance performance within these environments, as they deal with causes instead of effects. Hence, managers have an incentive to maximize performance on those activities on which their performance is measured (Hayes et al., 1988). As such, the dimensions of operations strategy need to be embedded in the performance measurement system (Sampson, 1996) in order to be enhanced. However, this argument does not intend to deny the role of ®nancial performance measures,

Service operations strategy 1405

IJOPM 23,11

1406

but to demonstrate the limitations of ®nancial measures in the service industry context. Methodology In this research, we analyse empirically the relationships between service operations strategy and service delivery ¯exibility in order to know the in¯uence of this relationship on operations performance. As both constructs are multidimensional, we will also analyse the interrelationships among the different dimensions con®guring both constructs as well as the strength and magnitude of such relationships. Hence, the main hypotheses to be tested in this study are as follows. H1. The direct effect of operations strategy on ®nancial performance is higher than the indirect effect through ¯exibility as a mediating variable. H2. The direct effect of operations strategy on non-®nancial performance is lower than the indirect effect through ¯exibility as a mediating variable. These hypotheses can be disaggregated into a total of 36 subhypotheses. Each one of these subhypotheses refers to every direct and indirect effect (through ¯exibility) of the nine service strategy dimensions on the two different types of performance measures. Sample and the sampling procedure This study was conducted in the context of Engineering Consulting Firms in Spain. Three ®rm types (civil, industrial and environmental) were considered covering most activities of Engineering Consulting Firms. The sample is the same used by Arias-Aranda (2002). A questionnaire was used to obtain the data for the study. Initially, a copy of the questionnaire was sent to ten ®rms representing every turnover and activity group as a pre-test. They were requested not to answer the questionnaire, but to remark all doubts or possible mistakes detected. Only small syntactic changes were made, but none of the ®rms remarked dif®culties for concept understanding or misuse. Afterwards, the questionnaire was sent and collected from participating ®rms mainly via e-mail to the operations managers/executives or equivalent of the companies in the sample. The Spanish Association of Spanish Engineering Consulting Firms (Tecniberia) provided all the information about addresses and ®rm names. Initially, in order to attract the maximum number of participating ®rms, an e-mail was sent to all ®rms registered in Tecniberia asking for their participation while stressing the importance of the study. Finally, 129 ®rms with a turnover higher than 150.000 euros were selected as the target population. Twelve ®rms requested the questionnaire to be sent via

ordinary mail with a 100 per cent response rate. Non-respondents were contacted at least three times in order to invite them to participate in the study. Of these, usable data were collected from a total of 71 ®rms (55 per cent). The original language of the questionnaire was Spanish. Table I shows a description of the sample according to the ®ve turnover categories. Comparing the sample distribution with the sector as a whole, no signi®cant discrepancies were observed. Most of the ®rms’ turnover ranges from 300.000 to 3.000.000 euros (60 per cent approx. of the total sample). On the other hand, civil engineering ®rms represent the higher percentage of the sample (49 per cent) compared to 17 per cent of industrial engineering and 34 per cent of environmental engineering. Table II shows the turnover distribution of the ®rms according to the Spanish Ministry of Industry (1998).

Service operations strategy 1407

Measures The dimensions of operations strategy considered for the study are those of Arias-Aranda (2002). Questions de®ning such dimensions related to operations strategy are based on a ®ve-point Likert scale. Every one of the nine dimensions of operations strategy was clearly represented in differentiated blocks in the questionnaire. Control questions were included in order to verify the internal consistency of the questionnaire. For every dimension, a set of items was included in the questionnaire. For every item a Likert scale ranging from 1 (completely agree) to 5 (completely disagree) was used to measure the agreement of the operations managers/executives with such items. Cat.

Turnover (euros) Civil Firms Percentage

1 2 3 4 5

,

300.000 300.000-600.000 600.001-3.000.000 3.000.001-6.000.000 . 6.000.000 Total

7 11 11 3 3 35

20.0 31.4 31.4 8.6 8.6 100.0

Group of activity Industrial Firms Percentage 3 3 4 0 2 12

25.0 25.0 33.3 0.0 16.7 100.0

Environmental Firms Percentage 7 7 8 2 0 24

29.2 29.2 33.3 8.3 0.0 100.0

Source: Own processing

Turnover (euros) , 300.000 300.000-600.000 600.001-3.000.000 3.000.001-6.000.000 . 6.000.000 Percentage of ®rms 27.3 32.3 27.2 6 7.2 Source: Spanish Ministry of Industry (1998)

Table I. Sample distribution (turnover and group of activity)

Table II. Distribution in percentage of engineering consulting companies in Spain

IJOPM 23,11

1408

Partial indicators were developed in order to know the ®rms’ positioning for every operations strategy dimension. Such indicators combine the different items corresponding to each dimension in order to measure the ®rms’ trends. A global indicator was developed to measure operations strategy according to such trends taking into account that indicator’s rank should ¯ow between the values of 1 and 5 in order to be consistent with the Likert scale previously used (see Appendix). The ¯exibility dimensions of Ramasesh and Jayakumar (1991) were used as the main framework for the analysis of ¯exibility in services. According to the speci®c nature of service industries, some adaptations were included in the model in order to ensure the validation of the study. Some of these adaptations were suggested by the operations managers during the pre-test stage. A factor analysis was performed in order to test whether the items used to measure the different ¯exibility dimensions grouped consistently. Finally, seven ¯exibility dimensions were identi®ed through this analysis. They were named as expansion; distribution of information; routing; equipments and personnel; market; services and servuction[1], and process, programming and volume ¯exibility. Items con®guring each ¯exibility dimension were also measured through a ®ve-point Likert scale. In this case, the responders’ degree of agreement with each of the statements de®ning every ¯exibility dimension con®gures all items relating to system ¯exibility. Table III shows the ¯exibility dimensions and measures. Performance measures used in this study were classi®ed according to Abernethy and Lillis (1995). Respondents were asked to indicate the importance of each measure in the range ªof little or no importanceº to ªof utmost importanceº in a ®ve-point Likert scale. Operational measures were categorized as ®nancial/ef®ciency and customer satisfaction measures. Slight adaptations were made according to the pre-test in order to conform measures to engineering consulting ®rms. Respondents were asked to assign a level of relevance to each measure in a Likert scale (1-5). Measures are shown in Table IV. Inter-item analysis was used to check scales for internal consistency or reliability. Speci®cally, Cronbach’s reliability coef®cient (alpha) is calculated for each scale (dimension), as recommended by empirical research in operations by many researchers (Flynn et al., 1995; Smith and Reece, 1999; Swamidass and Newell, 1987). Cronbach’s alphas and trends for every dimension according to the indicator values are shown in Tables V and VI. Usually, a value of 0.7 in the Cronbach’s alpha is considered adequate in order to ensure reliability of the internal consistency of the questionnaire (Nunnally, 1978). However, a margin of 0.5-0.6 is generally considered adequate for exploratory work as well (Nunnally, 1978; Srinivasan, 1985). Construct validation is a process of demonstrating that an empirical measure corresponds to the conceptual de®nition of a construct (Schwab, 1980). Consequently, three types of validity can be established: nomological or theoretical validity, vertical

De®nition

Suggested measures

Ease with which capacity can be added Overall cost and time needed to add when needed capacity Upper limit on the amount of capacity expansion Distribution of Capability to distribute and share Ratio of the number of information information information through the service paths available in a system to the ¯exibility delivery system number of possible paths Increase in the performance of service delivery through the use of IT Routing ¯exibility Capability to use alternative processing Average number of ways in which a routes to deliver a service service can be delivered Routing entropy Decrease percentage in the throughput caused by breakdowns Cost of production lost due to rescheduling an urgent job Labour and Capability of personnel and machines Number of different operations equipment ¯exibility to perform different operations performed by machines and personnel weighted by the importance of tasks Cost of switching from one operation to another Extent of variation in the inputs that machines or workforce can handle Market ¯exibility Capability of the service delivery Cost of orders inattention system to adapt to market changes Cost of order delays

Expansion ¯exibility

Flexibility type

Gupta and Sommers (1992)

(continued)

Brill and Mandelbaum (1989) Gerwin (1987) and Sethi and Sethi (1990)

Chung and Chen (1990) and Kumar (1986)

Browne et al. (1984)

Chatterjee et al. (1984) and Newman et al. (1993)

Browne et al. (1984) and Carter (1986)

References

Service operations strategy 1409

Table III. Different types of service ¯exibility and some suggested measures

Table III.

Source: Adapted from Ramasesh and Jayakumar (1991)

Process, programming and volume ¯exibility

Suggested measures

Ratio of total output to setup costs Number of services delivered by year Time or cost to change from one service to another Option pricing approach Capability of a system to operate Expected percentage uptime during unattended for long periods second and third shifts Capability to produce different part Number of different parts that can be types without major effort produced Capability to operate at different levels Changeover costs between different of output known jobs within the current production plan Expected value of a portfolio of product for a given set of contingencies Ratio of average volume ¯uctuation to total capacity Stability of manufacturing costs over widely varying levels of production volume Smallest volumes for pro®table operation of the system

De®nition

Falkner (1986), Gerwin (1987) and Jaikumar (1986)

Browne et al. (1984)

Browne et al. (1984) Jaikumar (1986) and Son and Park (1987)

References

1410

Services and Capability of the system to add or servuction ¯exibility substitute services without major efforts

Flexibility type

IJOPM 23,11

Financial and ef®ciency measures

Customer satisfaction measures

Volume of services delivered Firm pro®t on net income Set up times

Number of customer claims Level of workers ef®ciency Frequency of errors in service delivery system New service development

Real costs ®t in relation to standard costs ROI Variation on prices of raw services and materials Percentage of use of service delivery capacity Percentage of no activity time Cost reduction due to quality increments

Expansion Distribution of information Routing Personnel and equipment Market Services and servuction Process, programming and volume

Ability of service customisation Level of customer satisfaction Level of workers cooperation at interdepartmental level

Financial performance Non-®nancial performance

Table IV. Performance measures

Cronbach’s alpha 0.9616 0.9325 0.8755 0.9329 0.8640 0.8676 0.5829

Source: Own processing

Performance dimension

1411

Number of services not ®nally delivered because of customer demand Time between service order and delivery

Source: Adapted from Abernethy and Lillis (1995)

Flexibility dimensions

Service operations strategy

Table V. Cronbach’s alphas for service ¯exibility dimensions

Cronbach’ alpha 0.8981 0.8455

Source: Own processing

validity, and horizontal or criterion-related validity. Hence, the measurement instrument establishes the basis for nomological or theoretical validity since all items are developed through an extensive review of the service operations strategy body of research. Factor analysis was used to check unidimensionality of scales, which provides evidence of a single latent construct (Flynn et al., 1995). Cronbach’s alpha values address vertical validity, which describes the extent to which a scale represents its construct. Evidence of criterion-related validity is presented through the Browne and Cudeck (1993) cross-validation

Table VI. Cronbach’s alphas for performance dimensions

IJOPM 23,11

1412

index for covariance structure modelling. Index value for this research is 0.642, which indicates a high probability that the model results are consistent with population parameters. Analysis and results According to the aim of our analysis, the technique of path analysis seems to be the most adequate to empirically verify the hypothesis. Such a technique estimates the magnitude of the linkages between the variables. Through these estimates, it is possible to provide information about the underlying causal processes (Asher, 1983). It also allows for distinguishing between direct and indirect effects in case there are compound paths. Our path analysis model is recursive as it assumes that there is no reciprocal causality in the form of causal feedback. For every path, a beta coef®cient is calculated. This coef®cient’s sign indicates the direction of the relationship (positive or negative) and its magnitude represents the strength of the interrelationship between the variables. The path coef®cients identify direct and indirect effects of each variable on the respective dependent variables. The total effect is simply the sum of the direct effect and all the indirect effects that occur through intervening variables. Tables VII and VIII report direct, indirect, total and

Direct

Table VII. Direct and indirect effects of operations strategy dimensions on ®nancial performance

Independent variable (Op. Strat. Dim.) Layout 0.465* PUSH orientation 0.598* Standardisation 0.578* Offered services 2 0.522* Use of IT 0.623* Back and front of®ce act. 0.528* HR specialisation 0.442* Customer participation 2 0.289 New services development 2 0.62* Dimensions of ¯exibility Expansion 2 0.136* Inf. distr. 0.098** Routing 0.127* Equip. and pers. 2 0.176* Market 2 0.153* Prod. and serv. 2 0.115* Proc., prog. and vol. 2 0.086** Standard R 2 F-ratio Notes: n = 71; *p , 0.01; **p , Source: Own processing

Dependent variable: ®nancial performance Indirect Total Spurious 0.312 0.291 0.324 2 0.280 0.354 0.383 0.372 2 0.165 2 0.247

0.05; ***p ,

0.886 35.164* 0.10;

r

0.777* 0.889* 0.902* 2 0.802* 0.977* 0.911* 0.814* 2 0.454495 2 0.867*

2 0.043** 2 0.017 2 0.083* 0.008 2 0.081* 2 0.059** 2 0.040 0.049** 0.077*

0.734* 0.872* 0.819* 2 0.794* 0.896* 0.852* 0.774* 2 0.405 2 0.79*

2 0.136* 0.098** 0.127* 2 0.176* 2 0.153* 2 0.115* 2 0.086**

0.041** 2 0.013 0.118* 0.081** 0.027 0.042** 0.179*

2 0.095* 0.085* 0.245* 2 0.095* 2 0.126* 2 0.073** 0.093*

Dependent variable: non-®nancial performance Direct Indirect Total Spurious Independent variable (Op. Strat. Dim.) Layout 2 0.486* PUSH orientation 2 0.365* Standardisation 2 0.405* Offered services 0.286* Use of IT 2 0.427* Back and front of®ce act. 2 0.437* HR specialisation 2 0.403* Customer participation 0.114** New services development 0.337* Dimensions of ¯exibility Expansion 0.186* Inf. distr. 2 0.073 Routing 0.146* Equip. and pers. 0.113* Market 0.115* Prod. and serv. 0.082* Proc., prog. and vol. 0.041* Standard R 2 F-ratio Notes: n = 71; *p , 0.01; **p , Source: Own processing

0.05; ***p ,

2 0.425 2 0.431 2 0.463 0.337 2 0.481 2 0.509 2 0.524 0.280 0.311

r

2 0.911* 2 0.796* 2 0.868* 0.623* 2 0.908* 2 0.946* 2 0.927* 0.394** 0.648*

0.016 2 0.038 0.054 2 0.115 0.080 0.134 0.086 2 0.053 0.087

2 0.895* 2 0.835* 2 0.814* 0.508* 2 0.828* 2 0.812* 2 0.841* 0.341* 0.736*

0.196* 2 0.085 0.158* 0.127* 0.115* 0.095* 0.056*

0.009* 2 0.019** 0.011* 0.01* 0.014* 0.018* 0.022*

0.206* 2 0.104** 0.169* 0.137* 0.129* 0.113* 0.078*

0.932 46.415* 0.10

spurious effects of operations strategy dimensions on ®nancial and non-®nancial performance of the ®rm. To test H1 and H2, the effects of operations strategy dimensions and ¯exibility on ®nancial and non-®nancial performance were disaggregated into direct, indirect and spurious. The following results were obtained. . The direct effects of the operations strategy dimensions on ®nancial performance are positive and signi®cant ( p , 0.01) for all independent variables (direct effects range from 0.442 for human resources specialisation to 0.623 for the use of IT) except for offered services and new services development for which direct effects are negative. Customer participation is not signi®cant. . All effects of ¯exibility on ®nancial performance are negative and signi®cant (ranging from 2 0.086 for process, programming and volume ¯exibility ( p , 0.05) to 2 0.176 for equipments and personnel ¯exibility ( p , 0.01)) except for information distribution and routing ¯exibility for which effects are positive and signi®cant. Indirect effects of operations strategy dimensions on ®nancial performance through ¯exibility as the mediating variable are also high with values ranging from 2 0.1655 for customer participation to 0.384 for back and front of®ce activities.

Service operations strategy 1413

Table VIII. Direct and indirect effects of operations strategy dimensions on non-®nancial performance

IJOPM 23,11

.

1414

.

.

.

Direct effects of ¯exibility dimensions on ®nancial performance are negative and signi®cant (ranging from 2 0.086 for process, programming and volume ¯exibility ( p , 0.05) to 2 0.176 for equipments and personnel ¯exibility ( p , 0.01)) except for information distribution and routing ¯exibility for which effects are positive and signi®cant. Direct effects of operations strategy dimensions on non-®nancial performance are, on the other side, negative and signi®cant for all independent variables (values range from 2 0.365 for PUSH orientation to 2 0.486 for layout) with the exception of the services offered, customer participation and new services development for which direct effects are positive and signi®cant. All direct effects of ¯exibility on non-®nancial performance are positive and signi®cant ranging from 0.041 for process, programming and volume ¯exibility to 0.186 for expansion ¯exibility ( p , 0.01) except for information distribution for which direct effect is positive, but not signi®cant. Indirect effects of operations strategy dimensions on non-®nancial performance through ¯exibility as the mediating variable are also high with values ranging from 2 0.5246 for human resources specialisation to 0.337 for services offered. Direct effects of ¯exibility dimensions on non-®nancial performance are positive and signi®cant (ranging from 0.041 for process, programming and volume ¯exibility ( p , 0.01) to 0.186 for expansion ¯exibility ( p , 0.01)) except for information distribution ¯exibility for which effects are positive, but not signi®cant.

After performing the analysis, H1 is contrasted totally and positively, as all direct effects of the operations strategy dimensions on ®nancial performance are higher than all indirect effects through ¯exibility. H2 is contrasted partially and positively, as all direct effects of the operations strategy dimensions on non-®nancial performance are lower than all indirect effects through ¯exibility except for layout and new services development dimensions. Furthermore, the indirect effects of the operations strategy dimensions on ®nancial performance are mostly positive while such effects are negative for non-®nancial performance. However, this trend turns opposite for ¯exibility, whose incidence is mostly negative on ®nancial performance and positive on non-®nancial performance. The r coef®cients calculated let us know the goodness in the model ®t (Billings and Wroten, 1978). Spurious explanation of the model is obtained by comparing the r coef®cients with the total effects explained. In this model, there are only four values higher than 0.10 (10 per cent of spurious explanation). These are routing and process, programming and volume ¯exibility for ®nancial performance and services offered and back and front of®ce activities for non-®nancial performance.

Contributions and conclusions This research provides several contributions to the ®eld of service operations management for academicians and practitioners. First, based on the categories of structure and infrastructure decisions proposed by relevant service management literature, this study operationalizes the concept of service operations strategy. By doing so, new research opportunities broaden in order to improve and adapt this concept to different environmental conditions. Future operationalizations of the service operations strategy will require additional perspectives under a contingency analysis. In addition, heterogeneity of the service industries involves further attention according to competitive priorities. In this context, practitioners have a starting point to understand how functional decisions affect the different dimensions that con®gure the operations strategy in services. Second, a model of ¯exibility in service industries has been applied including some adaptations to the speci®c nature of service industries. This model is based on the classical manufacturing ¯exibility frameworks. This study examines this critical concept utilizing empirical methods within a ®eld-based setting. Still, some dimensions show low moderate reliability, so future research efforts may focus on improving and determining the exact internal and environmental factors affecting service ¯exibility according to speci®c sector characteristics. This can help practitioners in the process of choosing which dimensions of ¯exibility better support strategic priorities. Such knowledge is especially relevant when related to ®nancial decisions of IT investments and equipments as well as personnel training. Third, this research proves that service operations strategy has a signi®cant positive and direct effect on ®nancial performance. The results establish not only the existence of strong links between the service operations strategy and ®nancial performance, but also the magnitude of the impact of every operations strategy dimension on every single performance dimension. Nonetheless, the generalization of these ®ndings to other service industries cannot be guaranteed without cautiousness even though the robust statistical results for this relationship suggest that the ®ndings are quite reliable. Fourth, understanding the effects of operations strategy on performance is one of the major ®ndings of this study. The different strategy dimensions in¯uence ®nancial and non-®nancial performance in several manners. It is especially relevant to the fact that ¯exibility dimensions in¯uence positively on non-®nancial performance and negatively on the ®nancial performance. Hence, investments on ¯exibility lead to enhance the non-®nancial performance while negatively affecting the ®nancial performance. Therefore, practitioners can better establish priorities for competition according to the expected relationship between the operations strategy, ¯exibility and performance. Finally, this research has based all ®ndings on the empirical methods of analysis supported by the ®eld-based investigation to give support to concepts

Service operations strategy 1415

IJOPM 23,11

1416

and relationships that seem reasonable in order to provide empirical veri®cation. The speci®c methodology of this study and the results suggest that much of the conceptual efforts made in operations strategy for manufacturing may be applicable to service operations as well. However, the variables and dimensions used to con®gure this study have been adapted to the speci®c sector of the consulting engineering ®rm, which limits direct application of the model to other service industries. Further research efforts can help to determine to what extent the strategy, ¯exibility and performance relate in similar ways in heterogeneous service industries. Note 1. Even though service production and product-service are broadly used concepts we opted for the general denomination of services combined with the servuction term. References Abernethy, M.A. and Lillis, A.M. (1995), ªThe impact of manufacturing ¯exibility on management control system designº, Accounting, Organizations and Society, Vol. 20, pp. 241-58. Adam, E. Jr and Swamidass, P. (1989), ªAssessing operations management from a strategic perspectiveº, Journal of Management, Vol. 15 No. 2, pp. 181-203. Adler, P.S. (1988), ªManaging ¯exible automationº, California Management Review, Vol. 30 No. 2, pp. 34-56. Â Alvarez Gil, M.J. (1994), ªCapital budgeting and ¯exible manufacturingº, International Journal of Production Economics, Vol. 36, pp. 109-28. Anderson, J.C., Cleveland, G. and Schroeder, R.G. (1989), ªOperations strategy: a literature reviewº, Journal of Operations Management, Vol. 8 No. 2, pp. 1-26. Arias-Aranda, D. (2002), ªRelationship between operations strategy and size in engineering consulting ®rmsº, International Journal of Service Industry Management, Vol. 13 No. 3, pp. 263-85. Arias-Aranda, D., Minguela Rata, B. and RodrõÂguez Duarte, A. (2001), ªInnovation and ®rm size: an empirical study for Spanish engineering consulting companiesº, European Journal of Innovation Management, Vol. 4 No. 3. Armistead, C.G. (1992), ªIntroduction to service operationsº, in Voss, C., Armistead, C., Johnston, B. and Morris, B. (Eds), Operations Management in Service Industries and The Public Sector, Wiley, NJ. Asher, H.B. (1983), Causal Modeling, Sage Publications, Beverly Hills, CA. Berry, L.L. and Parasuraman, A. (1997), ªListening to the customer ± the concept of a service-quality information systemº, Sloan Management Review, Vol. 38 No. 3, pp. 65-76. Billings, R.S. and Wroten, S.P. (1978), ªUse of path analysis in industrial/organizational psychology: criticisms and suggestionsº, Journal of Applied Psychology, Vol. 63 No. 6, pp. 677-88. Bowen, D.E. and Lawler, E.E. III (1995), ªEmpowering service employeesº, Sloan Management Review, Summer. Bowen, D.E. and Youngdahl, W.E. (1998), ª`Lean’ service: in defense of a production-line approachº, International Journal of Service Industry Management, Vol. 9 No. 3, pp. 207-25.

Brill, P. and Mandelbaum, C. (1989), ªOn measures of ¯exibility in manufacturing systemsº, International Journal of Production Research, Vol. 27 No. 5, pp. 747-56. Browne, J., Dubois, K., Rathmill, S.P., Sethi, P. and Stecke, K.E. (1984), ªClassi®cation of ¯exible manufacturing systemsº, The FMS Magazine, Vol. 2 No. 1, pp. 747-56. Browne, M. and Cudeck, R. (1993), Alternative Ways of Assessing Model Fit, Kluwer Publishing, Dordrecht. Brownell, P. and Merchant, K.A. (1990), ªThe budgetary and performance in¯uences of product standarizations and manufacturing process automationº, Journal of Accounting Research, Vol. 2, pp. 388-97. Bucki, J. and Pesqueux, Y. (2000), ªFlexible workshop: about the concept of ¯exibilityº, International Journal of Agile Management Systems, Vol. 2 No. 1, pp. 62-70. Carter, M.F. (1986), ªDesigning ¯exibility into automated manufacturing systemsº, in Stecke, K.E. and Suri, R. (Eds), Proc. 2nd ORSA/RIMS Conference in Flexible Manufacturing Systems: Operations Research Model and Applications, Ann Arbor, Elsevier Science Publishers, Amsterdam, pp. 107-118. Chambers, S. (1992), ªFlexibility in the context of manufacturing strategyº, in Voss, C.A. (Ed.), Manufacturing Strategy. Process and Content, Chapman and Hall, London. Chatterjee, A., Cohen, M., Maxwell, W. and Miller, R. (1984), ªManufacturing ¯exibility: models and measurementsº, Proceedings of the 1st ORSA/TIMS Conference on FMS, Ann Arbor, MI, pp. 49-64. Chenhall, R. (1995), ªReliance on manufacturing performance measures, total quality management and organisational performance: an empirical investigationº, paper presented at University of New South Wales, Sydney. Chung, C.H. and Chen, I.J. (1990), ªManaging ¯exibility of ¯exible manufacturing systems for competitive edgeº, in Liberatore, M.J. (Ed.), The Selection and Evaluation of Advanced Manufacturing Technologies, Springer-Verlag, Berlin, pp. 280-305. Collier, D.A. (1994), The Service-Quality Solution: Using Service Management to Gain Competitive Advantage, Irwin, IL. Collier, D.A. (1996), ªTight and loose comprehensive customer contact (3C) plansº, Advances in Service Marketing and Management, Vol. 5, JAI Press, Greenwich, CT, pp. 153-68. Cusumano, M. (1988), ªShifting economies: craft production to the ¯exible factoryº, Working Paper 2012-88, Sloan School of Management, MIT, Cambridge, MA. Dangayash, G.S. and Deshmukh, S.G. (2001), ªManufacturing strategy: literature review and some issuesº, International Journal of Operations and Production Management, Vol. 21 No. 7, pp. 884-932. Desatnik, R. (1994), ªManaging customer service for the 21st centuryº, Journal for Quality and Participation, Vol. 17, pp. 30-6. Duclos, L.K., Siha, S.M. and Lummus, R.R. (1995), ªJIT in services: a review of current practices and future directions for researchº, International Journal of Service Industry Management, Vol. 6 No. 5, pp. 36-52. Falkner, C.H. (1986), ªFlexibility in manufacturing plantsº, Proceedings of the Second ORSA/TIMS Conference on Flexible Manufacturing Systems, Elsevier, New York, NY. Feigenbaum, A. and Karnani, A. (1991), ªOutput ¯exibility: a competitive advantage for small ®rmsº, Strategic Management Journal, Vol. 12, pp. 101-14. Flynn, B., Sakakibara, S., Schroeder, R., Bates, K. and Flynn, J. (1995), ªEmpirical research methods in operations managementº, Journal of Operations Management, Vol. 9 No. 2, pp. 250-84.

Service operations strategy 1417

IJOPM 23,11

Gaimon, C. and Singhal, V. (1992), ªFlexibility and the choice of manufacturing facilities under short product life cyclesº, European Journal of Operational Research, Vol. 60 No. 2, pp. 211-23. Gerwin, D. (1987), ªAn agenda for research on the ¯exibility of manufacturing processesº, International Journal of Operations & Productions Management, Vol. 7 No. 1, pp. 38-49.

1418

Gerwin, D. (1993), ªManufacturing ¯exibility: a strategic perspectiveº, Management Science, Vol. 39 No. 4, pp. 395-410. Ghalayini, A.M. and Noble, J.S. (1996), ªThe changing basis of performance measuresº, International Journal of Operations and Productions Management, Vol. 16 No. 8, pp. 63-80. Gouillart, F. and Sturdivant, F. (1994), ªSpend a day in the life of your customersº, Harvard Business Review, pp. 116-25. Govindarajan, V.A. (1988), ªA contingency approach to strategy implementation at the business-unit level: integrating administrative mechanisms with strategyº, Academy of Management Journal, Vol. 28, pp. 828-53. Gupta, Y.P. and Goyal, S. (1992), ªFlexibility tradeoffs in a ¯exible manufacturing system: a simulation studyº, International Journal of Production Research, Vol. 30 No. 3, pp. 527-57. Gupta, Y.P. and Somers, T.M. (1992), ªThe measurement of manufacturing ¯exibilityº, European Journal of Operational Research, Vol. 60, pp. 166-82. Gupta, Y.P. and Somers, T.M. (1996), ªBusiness strategy, manufacturing ¯exibility, and organizational performance relationships: a path analysis approachº, Production and Operations Management, Vol. 5 No. 3, pp. 204-31. Gustavsson, S. (1988), ªFlexibility and productivity in complex processesº, International Journal of Production Research, Vol. 22 No. 5, pp. 801-8. Hammer, M. and Champy, J. (1993), Reengineering the Corporation, HarperBusiness, New York, NY. Hart, C.W.L. (1995), ªMass customization: conceptual underpinnings, opportunities and limitsº, International Journal of Service Industry Management, Vol. 6 No. 2, pp. 36-45. Harvey, J., Lefebvre, L.A. and Lefebvre, E. (1997), ªFlexibility and technology in services: a conceptual modelº, International Journal of Operations & Production Management, Vol. 17 No. 21, pp. 29-45. Hayes, R.H., Wheelwright, S.C. and Clark, K. (1988), Dynamic Manufacturing: Creating the Learning Organization, Free Press, New York, NY. Haynes, R.M. and Du Vall, P.K. (1992), ªService quality management: a process control approachº, International Journal of Service Industry Management, Vol. 3 No. 1, pp. 14-24. Ho, C. (1996), ªA contingency theoretical model of manufacturing strategyº, International Journal of Operations & Production Management, Vol. 16 No. 5, pp. 74-98. Hyun, J. and Ahn, B. (1990), ªFlexibility revisited: review, unifying frameworks, and strategic implicationsº, mimeo, Korean Advanced Institute of Science and Technology, Seoul. Jaikumar, R. (1986), ªPostindustrial manufacturingº, Harvard Business Review, Vol. 64 No. 6, pp. 69-76. Johnston, R. (1994), ªOperations: from factory to service managementº, International Journal of Service Industry Management, Vol. 5 No. 1, pp. 49-63. Kaplan, R.S. (1983), ªMeasuring manufacturing performance: a new challenge for management accounting researchº, Accounting Review, Vol. 58, pp. 686-705.

Kumar, V. (1986), ªOn measurement of ¯exibility in ¯exible manufacturing systems: an information-theoretic approachº, Proceedings of the 2nd ORSA/TIMS Conference on Flexible Manufacturing Systems, Elsevier, New York, NY. Leong, G.K., Snyder, D.L. and Ward, P.T. (1990), ªResearch in the process and content of manufacturing strategyº, OMEGA International Journal of Management Science, Vol. 18 No. 2, pp. 109-22. Lusch, R., Boyt, T. and Schuler, D. (1996), ªEmployees as customers: the role of social controls and employee satisfaction in developing patronageº, Journal of Business Research, Vol. 35, pp. 239-52. McCutcheon, D.M., Raturi, A.S. and Meredith, J.R. (1994), ªThe customisation-responsiveness squeezeº, Sloan Management Review, Winter, pp. 89-99. Mills, J., Neely, A., Platts, K. and Gregory, M. (1998), ªManufacturing strategy: a pictorial representationº, International Journal of Operations & Production Management, Vol. 18 No. 11, pp. 1067-85. Nayyar, P.R. (1992), ªPerformance effects of three foci in service ®rmsº, Academy of Management Journal, Vol. 35 No. 5, pp. 985-1009. Neely, A., Richards, H., Mills, J., Platts, K. and Bourne, M. (1997), ªDesigning performance measures: a structural approachº, International Journal of Operations and Production Management, Vol. 17 No. 11, pp. 1131-52. Newman, W.R., Hanna, M. and Maffei, M.J. (1993), ªDealing with the uncertainties of manufacturing: ¯exiblity, buffers and integrationº, International Journal of Operations & Production Management, Vol. 13 No. 1, pp. 19-34. Nieto AntolõÂn, M., Arias Aranda, D., Minguela Rata, B. and RodrõÂguez Duarte, A. (1999), ªThe evolution of operations management contents: an analysis of the most relevant textbooksº, Industrial Management & Data Systems, Vol. 99 No. 7/8, pp. 345-53. Noori, H. and Radford, R. (1995), Production and Operations Management, McGraw-Hill, New York, NY. Normann, R. (1984), Service Management, Wiley, New York, NY. Nunnally, J. (1978), Psychometric Theory, McGraw-Hill, New York, NY. Ramasesh, R.V. and Jayakumar, M.D. (1991), ªMeasurement of manufacturing ¯exibility: a value based approachº, Journal of Operations Management, Vol. 10 No. 4, pp. 446-68. Riley, M. and Lockwood, A. (1997), ªStrategies and measurement for workforce ¯exibility: an application of functional ¯exibility in a service settingº, International Journal of Operations and Production Management, Vol. 17 No. 4, pp. 413-9. Sa®zadeh, M.H., Ritzman, L.P. and Mallick, D. (2000), ªRevisiting alternative theoretical paradigms in manufacturing strategyº, Production and Operations Management, Vol. 9 No. 2, pp. 111-27. Sampson, S.E. (1996), ªRami®cations of monitoring service quality through passively solicited customer feedbackº, Decision Sciences, Vol. 27 No. 4, pp. 601-21. Schmenner, R.W. (1986), ªHow can service businesses survive and prosper?º, Sloan Management Review, Vol. 27 No. 3. Schwab, D. (1980), ªConstruct validity in organizational behaviourº, Research in Organizational Behaviour, Vol. 2 No. 1, pp. 3-43. Sethi, A.K. and Sethi, S.P. (1990), ªFlexibility in manufacturingº, International Journal of Flexible Manufacturing Systems, Vol. 2 No. 4, pp. 289-328.

Service operations strategy 1419

IJOPM 23,11

Simons, R. (1987), ªAccounting control systems and business strategy: an empirical analysisº, Accounting, Organizations and Society, Vol. 12, pp. 357-74. Skinner, W. (1969), ªManufacturing ± missing link in corporate strategyº, Harvard Business Review, May-June, pp. 136-45. Skinner, W. (1978), Manufacturing in the Corporate Strategy, Wiley, New York, NY.

1420

Slack, N. (1987), ªThe ¯exibility of manufacturing systemsº, International Journal of Operations & Production Management, Vol. 7 No. 4, pp. 35-45. Smith, T.M. and Reece, J.S. (1999), ªThe relationship of strategy, ®t, productivity and business performance in a services settingº, Journal of Operations Management, Vol. 17, pp. 145-61. Son, Y.K. and Park, C.S. (1987), ªEconomic measure of productivity, quality and ¯exibility in advanced manufacturing systemsº, Journal of Manufacturing Systems, Vol. 6 No. 3, pp. 193-206. Spanish Ministry of Industry (1998), ªEl sector de la IngenierõÂa Civil en EspanÄaº, Ed. Ministerio de Fomento. Spina, G., Bartezzaghi, E., Bent, A., Cagliano, R., Draaijer, D. and Boer, H. (1996), ªStrategically ¯exible production: the multi-focused manufacturing paradigmº, International Journal of Operations & Production Management, Vol. 16 No. 11, pp. 20-41. Srinivasan, A. (1985), ªAlternative measure of system effectiveness: associations and implicationsº, MIS Quarterly, Vol. 9 No. 3, pp. 243-53. Suarez, F.F., Cusumano, M.A. and Fine, C.H. (1996), ªAn empirical study of manufacturing ¯exibility in printed circuit board assemblyº, Operations Research, Vol. 44 No. 1, pp. 223-40. Swamidass, P.M. and Newell, W.T. (1987), ªManufacturing strategy, environmental uncertainty and performance: a path analytic modelº, Management Science, Vol. 33 No. 4, pp. 509-24. Swink, M. and Way, M.H. (1995), ªManufacturing strategy: propositions, current research, renewed directionsº, International Journal of Operations & Production Management, Vol. 15 No. 7, pp. 4-26. Taymaz, E. (1989), ªTypes of ¯exibility in a single machine production systemº, International Journal of Production Research, Vol. 27 No. 11, pp. 1891-9. Tersine, R. and Harvey, M. (1998), ªGlobal customerization of markets has arrived!º, European Management Journal, Vol. 16 No. 1, pp. 79-90. Tombak, M.M. (1988), ªThe importance of ¯exibility in manufacturingº, Wharton/PIMS Research Center Working Paper, No. 89-002, Philadelphia, PA. Tombak, M. and De Meyer, A. (1988), ªFlexibility and FMS: an empirical analysisº, IEEE Transactions on Engineering Management, Vol. 35, pp. 101-7. Upton, D.M. (1995), ªWhat really makes factories ¯exible?º, Harvard Business Review, pp. 74-84. Vandermerwe, S. (1992), ªThe power is in the service because the value is in the resultº, in Lovelock, C.H. (Ed.), Managing Services, 2nd ed., Prentice-Hall, Englewood Cliffs, NJ, pp. 298-309. Walton, R. and Susman, G. (1987), ªPeople policies for the new machinesº, Harvard Business Review, Vol. 66 No. 2, pp. 98-106. Wheelwright, S.C. (1984), ªManufacturing strategy: de®ning the missing linkº, Strategic Management Journal, Vol. 5, pp. 77-91.

Appendix The indicator was designed as follows: h± P ² ­ ± ²­ i Pd Pb ­Pd ­ b 5 +1 i= a A in 2 i= c Ain + ­ i= c A in 2 5 i= a Ain ­ ­ ­ ­ i ± P ² ± ² Ebn = h­ P P P ­ b ­­ d ­ d + ­ i= c Ain 2 5 bi= a Ain ­ +1 ­5 i= a Ain 2 i= c A in ­

where Ebn is the indicator; Ain is the score obtained in question i of block n in the questionnaire. Rank [a,b ] represents questions scoring towards one of the trends in each block. Rank [c,d ] represents questions scoring towards opposite extremes of rank [a,b ] in each block. Hence, Á ! d b X X Ain 2 5 Ain i= c

i= a

represents the smallest reachable value, considering that one ®rm scores the highest (score 5) in all questions for one of the trends and the smallest (score 1) in all questions of the opposite trend. On the other hand, Á ! b d X X 5 Ain 2 Ain i= a

i= c

represents the smallest reachable value for a ®rm positioned at one extreme, scoring the smallest (score 1) and the highest (score 5) for the opposite trends. Once the extremes and possible intermediate values have been obtained, the indicator transforms this rank in a scale from 0 to 5 by adding to the value obtained, the smallest reachable value plus one. The value obtained is ®nally divided by the highest reachable value adding the smallest value plus one in order to make the scale positive. Finally, the value obtained is multiplied by ®ve to ®t it to the 1 - 5 scale.

Service operations strategy 1421